2.1 Complex Challenges

In the United States and Europe, the study of complex challenges began in the 1970s, when they were characterized as “wicked problems” (Reference Rittel and WebberRittel & Webber, 1973). These challenges are recognizable by their seemingly contradictory requirements, with complex interdependencies that take significant time and sustained effort even to define. Reference Rittel and WebberRittel and Webber (1973) transformed the thinking with the idea that a formulation of these kinds of problems was, necessarily, the solution to these problems, because the solution creation is what leads to definition. The leadership literature describes complex challenges as “adaptive” – because of their complexity, stakeholders may not only perceive the solutions differently but may even have difficulty agreeing on the problem (Reference HeifetzHeifetz, 1994).

By contrast, Reference Rittel and WebberRittel and Webber (1973) identified “tame” challenges as those that a manager, who had the right education and competencies, could understand and solve through a formulaic process. The leadership literature calls these “technical” challenges, those that groups or a technical community would perceive and tend to design a solution the same way (Reference HeifetzHeifetz, 1994).

According to Rittel and Webber and Heifetz, understanding complex challenges comes through a deep knowledge of context, and the context is used to give the problem scope and to understand what solutions are possible. Solutions are best identified according to individual and group interests, values, and ideologies through a process involving multiple parties who are equipped, interested, and able to create the solutions.

2.2 Cross-sector Collaboration

The study of complex challenges evolved into the study of cross-sector collaboration. This new field began to develop rapidly in 2006, aided by an important literature review by Bryson, Crosby, and Stone that coalesced the fragmentary evidence from many disciplines into a picture catapulting the research funding and interest at the municipal, state, and federal levels in the United States. They defined cross-sector collaboration as “the linking or sharing of information, resources, activities, and capabilities by organizations in two or more sectors to achieve jointly an outcome that could not be achieved by organizations in one sector separately” (Reference Bryson, Crosby and StoneBryson, Crosby, & Stone, 2006, p. 44). They structured their review around a number of “propositions” that they constructed based on their own research, weaving the nascent evidence from multiple fields into a picture that was accessible to both researchers and practitioners.

In 2015, this team published an updated review explaining the evolution of the field and how this research and practice, although challenging, vastly improved the way that governments – and other collaborating partners – respond to public challenges in developed countries (Reference Bryson, Crosby and StoneBryson, Crosby, & Stone, 2015). They identified seven holistic theoretical frameworks created in the prior ten years and honed important concepts, such as design, strategic management, and governance, that had come about during that time. Elaborating on the theme that diverse entities must collaborate effectively to impact and ultimately resolve complex challenges, the review identified a number of important areas for future research focus.

Even though their review specifically excluded developing-country evidence, looking at the developed-country progress offers new avenues for thinking about how to implement the research and practice of cross-sector collaboration in developing countries. However, even with such a comprehensive and inspiring review as a starting point, when this project began, it was difficult to see how the results could be applied in Ghana. For example, one influential case study, used to advance the theory and practice, involved a $1.1 billion demonstration project to reduce congestion on an urban transportation corridor in Minneapolis, Minnesota (Reference Bryson, Crosby, Stone, Saunoi-Sandgren and ImbodenBryson et al., 2011b). This funding implies a level of infrastructure and human resources that does not exist in developing countries. This resource disparity explains why it is challenging to apply the collaboration literature in developing countries and points to, perhaps, why development industry norms have not yet evolved to incorporate collaboration best practices.

2.3 Development as Usual

Debate among critics and proponents of international development funding has been focused on whether, or the extent to which, international aid funding and development programs should exist (Reference Flint and zu NatrupFlint & zu Natrup, 2019). Critical works such as Damisa Moyo’s Dead Aid demand an end to aid, arguing that it exacerbates poverty (Reference MoyoMoyo, 2010). Academic Jeffrey Sachs champions the other side of the debate in The End of Poverty: Economic Possibilities for Our Time – that aid can transform developing economies (Reference SachsSachs, 2006). Some argue development programs should exist but adapt, taking into account evidence from social capital theory (Reference Woolcock and NarayanWoolcock & Narayan, 2000). Others advocate reducing overall aid and point to alternatives like social entrepreneurship and civic innovation (Reference FowlerFowler, 2000). The debate continues, with general agreement that current aid-funded programs do not achieve desired results, and vast improvements are needed if these systems are to end poverty (Reference EasterlyEasterly, 2008). This project started with the premise that international development programs lack efficacy and fail to become sustainable because international development funding mechanisms and projects do not take into account the evidence base on how to resolve complex challenges.

While international development models call for work across sectors, development agencies – and therefore their grantees, including researchers and practitioners – have not yet adapted to incorporating best practices from cross-sector collaboration literature and cases. Virtually every development funding agency operates with a sectoral approach, requiring proposals around predetermined issues and preferencing predesigned projects with established metrics. These factors fail to account for the human relationships (Reference EybenEyben, 2010) and conflicts that must be worked through to resolve complex challenges. Thus, complex challenges are not – and cannot be – addressed effectively.

Few, if any, funders offer the flexibility of a stakeholder-driven, strategic approach that is suitable for actually resolving complex challenges. Compounding this issue, funders hinder the time-consuming process of creating sustainable solutions by imposing short timelines for reporting results (Reference Airing and TeegardenAiring & Teegarden, 2012). Thus, when development researchers or practitioners try to address complex challenges and at the same time attempt to complete their predefined program of work within a standard three- or five-year funding cycle, they run into the issue of deadlines. How would one manage the relationships and the necessary conflict that arise in the resolution of these challenges? It is not possible. Instead, the researchers and practitioners try to “tame” the “wicked” problems (to use Rittel’s and Webber’s language), defaulting to outdated evidence as they treat these problems as standard projects. This approach is designed to fail.

Alternatively, they may try to conduct evidence-based cross-sector collaboration, but on a shortened timeline, thus resorting to practices that are rigid, lack rigor, and misapply or do not use current evidence. In the implementation literature, this is called “rival framing”; for purposes of this discussion, it means that performance measurements and donor reports distort the activity – the work of collaboration – to comply with desired new collaboration norms (Reference KritzKritz, 2017). In one glaring example, dozens of articles touted the efficacy of a major international collaboration with the laudable goal of treating and preventing HIV/AIDS in Botswana. Yet, when local stakeholders were later interviewed, researchers discovered that the Dutch project leader worked extensively in South Africa but lacked cultural competence in Botswana. The top-down approach did not readily incorporate practices that would work locally, and project results were prioritized over relationships. It took several years for the project to achieve some “mutuality” so that local government leaders believed they were valued partners (Reference Ramiah and ReichRamiah & Reich, 2006). These issues were compounded by enormous external pressure due to short funding timelines.

Cross-sector collaboration research is needed to improve these sorts of international development programs. A more robust, contextually relevant evidence base could identify collectivist cultures that are more suited to collaboration, available resources, personnel, government structures, key jobs for international and donor organizations, and many other opportunities. This research would strengthen relationships and ultimately improve program results. This project was developed in part to address these issues.

2.4 Cross-sector Collaboration Evidence in Developing Countries

This research began with a literature review of the developing-country cross-sector collaboration evidence. The goals of the review were to find out how collaboration research was conducted in developing countries and to identify collaboration interventions that could be used as a model for Old Fadama.

The literature search began with health literature, which is robust and particularly well organized. A broadly constructed search identified 20,000 articles mentioning collaboration-related topics in their title or abstract. However, only 165 articles contained data on how cross-sector collaboration was implemented, and just a handful of those articles included rigorous study of the collaboration itself (Reference KritzKritz, 2017). Because of the lack of evidence, the review then incorporated an extensive hand search in a number of other fields, yielding little additional data. Thus, the most important outcome from the review was to illustrate how little rigorous collaboration process research has been done in developing countries.

Nonetheless, the results, described in Figure 4 were interesting in that they explained how contextual factors differed in developing countries, while generally mirroring the developed-country theory. Thus, the results related to how to construct collaboration process were organized around the Bryson, Crosby, and Stone team’s themes of design, strategic management, and governance (Reference Bryson, Crosby and StoneBryson, Crosby, & Stone, 2006, Reference Bryson, Crosby and Stone2015; Reference Stone, Crosby, Bryson, Cornforth and BrownStone, Crosby, & Bryson, 2013).

Figure 4 Systematic review model

Analyzing the results of the review combined theoretical understanding and empirical evidence and focused on explaining the relationship between the context, mechanisms, and outcomes (Reference Pawson, Greenhalgh, Harvey and WalshePawson et al., 2005). Regarding how to conduct the collaboration research, the small number of research studies that were deemed to be rigorous had several commonalities that were used to construct this project (Reference KritzKritz, 2017). The studies (1) used mixed methodology including interviews, focus groups, questionnaires, and observation; (2) incorporated a high degree of contextual complexity into the research design; (3) involved multi-level participation in the research, including nongovernmental organizations, grassroots community organizing, and government offices; and (4) captured mechanisms, the often-unstated emotional reactions activated by the cross-sector collaboration intervention (Reference Ahmed and AliAhmed & Ali, 2006; Reference Campbell, Nair and MaimaneCampbell, Nair, & Maimane, 2007; Reference Manning and RoesslerManning & Roessler, 2014; Reference Pridmore, Carr-Hill and Amuyunzu-NyamongoPridmore et al., 2015; Reference Sanchez, Perez and CruzSanchez et al., 2009). Because of the limited developing-country evidence, both the developed-country evidence base and the systematic review were used to establish the previous evidence base and the collaboration principles that undergirded the PAR process.

2.5 Collaboration Principles

Rittel and Webber explain how “the analyst’s ‘world view’ is the strongest determining factor in explaining a discrepancy and, therefore, in resolving a wicked problem” (Reference Rittel and WebberRittel & Webber, 1973, p. 166). With that said, it is important to understand the “world view” – the collaboration principles – that undergirded this research. At the beginning, these were not clearly articulated ideas but rather areas where a stakeholder-driven approach, informed by the evidence, with data collected based on the stakeholders’ decision making, might advance the existing evidence. Some of these principles did not work (and are not discussed in this Element). Those described in Sections 2.5.2–2.5.8 (see Figure 5) became part of the PAR intervention results (see Figure 10).

2.5.2 Creating Shared Understanding

Cross-sector collaboration creates shared understanding of complex challenges (Reference Bryson, Crosby and StoneBryson, Crosby, & Stone, 2015), but it is not an easy task to create shared understanding in a high-conflict environment such as Old Fadama. This context requires a conflict-sensitive approach (Reference AndersonAnderson, 1999; World Health Organization, 2019). Conflict can be value neutral and is important in that it surfaces issues that need to be addressed for those in conflict to begin to collaborate (Reference CarpenterCarpenter, 2019). However, these conflicts need to be handled with care to achieve solutions (Reference BinghamBingham, 2009). The PAR interviews were used in part to infuse the process with references to local culture and values regarding the importance of working together, respecting others, and seeking understanding across divisions. The research team used the PAR process to build on these norms and create a shared understanding of Old Fadama’s challenges. To do so, the facilitator focused on the stakeholders’ shared interests rather than their “positions,” by identifying each stakeholder’s basic human and organizational needs and focusing attention on doing the best “for Mother Ghana” (see Figure 6). Thus, the PAR process was used to create a value-neutral understanding of the conflict between the government and the community, which provided the opportunity to develop a shared understanding of the challenges.

Figure 6 Old Fadama community meeting

These norms have further developed and include both a shared terminology and a value system based on mutual respect and understanding (Reference Thomson and PerryThomson & Perry, 2006). As the norms grew, new stakeholders were able to more rapidly expand their perspective, understand the PAR process, share their ideas, and see these ideas incorporated, because of the shared understanding of and language around Old Fadama’s challenges.

3.1 Exploratory Research to Identify Stakeholders and Community

This section explains in further detail the initial stakeholder identification described in Section 1. The research director began by identifying, in a purposive process, possible research participants who could work together to identify a challenge. Catholic sisters seemed to be the nongovernmental actors best-known in Ghana for making consistent, long-term commitments to poor and marginalized communities. The National Catholic Health Service identified a purposive sample of four sisters from congregations known for their work with the poor, who might want to participate. Congregations are religious organizations of sisters dedicated to social service in the world, rather than a monastic or cloistered way of life. There has been a concerted push in international development to engage faith organizations (Reference Duff and BuckinghamDuff & Buckingham, 2015). However, they are nearly absent from the collaboration literature, so it was unclear if the sisters would enroll (Reference KritzKritz, 2017). The research director interviewed them, and Sr. Matilda Sorkpor, HDR (the “lead sister”) agreed to participate and chose Old Fadama as a place she wanted to work but was hesitant to enter because the challenge was so great. In 2018–2019, she educated and enrolled several additional sisters from different congregations.

The municipal government seemed to be a logical partner because Old Fadama was an urban slum. When the research director contacted Boateng, the director of public health of the Accra Metropolitan Assembly (the AMA, the mayor’s office), he responded with enthusiasm and relief:

Little is documented about effective government partnership in the scholarly literature, so it was unclear what to expect from government participation (Reference Barnes, Brown and HarmanBarnes, Brown, & Harman, 2016). Through interviews, representatives of the Department of Public Health shared that their typical efforts combatted cholera in other areas but did not work in Old Fadama. The evidence suggests that sectoral failure is an antecedent to cross-sector collaboration, and a facilitating organization can assist collaboration formation (Reference Bryson, Crosby and StoneBryson, Crosby, & Stone, 2006). Consistent with that evidence, and because of the research focus and Catholic sisters’ engagement, Boateng and others from the Department of Public Health became enthusiastic partners. They supported community involvement and provided public and environmental health research to inform stakeholders’ decision making. Later, when the research participants undertook a latrine project, the Department of Public Health coordinated between AMA offices that scoped the latrines, developed plans, and provided resources for permits, beginning to change the city’s policy around Old Fadama slum improvement.

As described in Section 1, the Old Fadama community was represented by Imoro Toyibu, a resident who had served for years as its secretary and has recently been hired as the municipal government’s public health officer for Old Fadama. After the initial interviews, he identified several research participants, leaders from Old Fadama community organizations. These included the Kayayei Youth Association, a community association of women and girls locally known as “kayayei,” head porters who carry goods (typically balanced on their heads) in markets and sell products on the streets throughout Accra; and the Old Fadama Youth Development Association (OFADA), the Old Fadama community association governed by the chiefs (community elders, see Figure 8) of the sixteen tribes residing in Old Fadama. OFADA did not include the chiefs of the two warring tribes that many believed were responsible for much of the organized violence in the slum. However, the association represented the majority of Old Fadama residents. Those familiar with community-based approaches know that the idea of a community may “create the illusion that people in a particular location, neighborhood, or ethnic group, are necessarily cooperative, caring, and inclusive. The reality may be very different, as power differentials in gender, race, and class relations may result in exclusion, and threaten the apparent cohesiveness of the group in question” (Reference Mathie and CunninghamMathie & Cunningham, 2003, p. 475). Consistent with the evidence, Old Fadama was not homogeneous, and decision making and negotiation often took place along tribal lines (Reference PallerPaller, 2014). A number of other community leaders and members also enrolled as research participants to offer their input on strategy and projects.

Figure 8 Old Fadama community elder

Two National Catholic Health Service (NCHS) officers served as the Ghana-based facilitation and research team. NCHS is a local organization, an association of Catholic health institutions that provides research and planning services and coordinates with government health services. The director of NCHS provided early guidance on collaboration with sisters and identified a staff member, Peter Batsa, to collect data and serve as the collaboration’s facilitator. The facilitation literature describes this vital role as creating consultative meetings and platforms for discussion to build relationships and accountability between differently resourced organizations with different capacities (Reference KritzKritz, 2017). A number of case studies in the literature detail important facilitation skills and responsibilities. These include research (Reference Campbell, Nair and MaimaneCampbell, Nair, & Maimane, 2007; Reference Pridmore, Carr-Hill and Amuyunzu-NyamongoPridmore et al., 2015; Reference Sanchez, Perez and PérezSanchez et al., 2005), as well as catalyzing (Reference Saadé, Bateman and BendahmaneSaadé, Bateman, & Bendahmane, 2001), linking (Reference Kielmann, Datye, Pradhan and RanganKielmann et al., 2014), bridging (Reference Manning and RoesslerManning & Roessler, 2014), brokering (Reference Sablah, Klopp and SteinbergSablah et al., 2012) or serving as an intermediary (Reference Murthy, Frieden, Yazdani and HreshikeshMurthy et al., 2001; Reference Probandari, Utarini, Lindholm and HurtigProbandari et al., 2011), coordinating (Reference Brooke-Sumner, Lund and PetersenBrooke-Sumner, Lund, & Petersen, 2016; Reference Thaennin, Visuthismajarn and SutheravutThaennin, Visuthismajarn, & Sutheravut, 2012), convening (Reference Li, Huikuri, Zhang and ChenLi et al., 2015), and facilitating (Reference Manning and RoesslerManning & Roessler, 2014; Reference Rangan, Juvekar and RasalpurkarRangan et al., 2004; Reference WessellsWessells, 2015).

Due to the complexity of Old Fadama’s problems, all of these responsibilities were deemed essential. As the collaboration grew, a Catholic sister, Sr. Rita Ann Kusi, HDR, joined the research team to serve in the important role of community liaison. The data later showed that her presence conveyed the message that this was a nonpolitical, public interest effort, thus building trust in the community and differentiating this project from efforts to sway voters in the lead-up to the contentious 2016 national elections. Later in the project, she received a grant to build a block of latrines in partnership with the municipal government and with input from the Old Fadama community.

Each research participant informed and sought input from their own important constituencies – stakeholders within their own organizations as well as others. As examples, Boateng consulted the leadership of the Ga State, a politically powerful Accra group with a historic ownership interest in the Old Fadama land. NCHS and the sisters consulted the Ghana Catholic Bishops Conference and the Office of the Metropolitan Archbishop. These offices represent the hierarchy of the Catholic Church and supported the sisters’ work in a coordinated manner. In time, some constituents became stakeholders. For example, as the project progressed, Boateng liaised with the Ministry of Inner-City and Zongo Development and the Ghana Health Service, government ministries that offered their input and later became stakeholders when the community and municipal priorities aligned with their offices’ national planning goals. Figure 9 describes the research participants and their stakeholders (Ministry of Health, Archdiocese of Ghana, Ga State) who joined the collaboration between 2015, when it began, and 2017, the end of the concept phase.

Figure 9 Stakeholder diagram

3.2 Development of the Participatory Action Research (PAR) Intervention Components: Interviews, Focus Group Discussions (FGD), and a Community Survey

As described, the stakeholders began by identifying a challenge that they would like to try to address through cross-sector collaboration, using their own resources and applying for funding as needed. There were three PAR phases: (1) key informant interviews with Catholic sisters identified a location for the study, to which the municipal Department of Public Health agreed; (2) focus group discussions (FGDs) set community priorities, aligned them with the priorities of government planning, and created strategies and projects; and (3) a community survey increased community member participation and further defined project goals and measurement.

Participatory action research couples knowledge generation – such as would occur in traditional research – with an additional component: a process to create or support organizational action and change (Reference Cornwall and JewkesCornwall & Jewkes, 1995; Reference Greenwood and LevinGreenwood & Levin, 1998). PAR was identified as a methodology that could meet the stakeholders’ needs in a complex environment, including a rapidly evolving urban slum, a diffuse city agency, and multiple congregations of Catholic sisters that were taught to plan strategically and work together as they made the decision to move from better-served rural areas to an urban slum. PAR incorporates local priorities, processes, and perspectives (Reference Cornwall and JewkesCornwall & Jewkes, 1995). A PAR process involves researchers and participants working together to define the problem and formulate context-specific research questions and solutions. The research team and participants worked together to create the PAR process. They made joint decisions to establish the research agenda; collect and analyze data on stakeholders’ opinions; identify priorities, strategies, and projects; and incorporate the resulting knowledge into the PAR process.

These steps transformed each participating organizations’ practices in comparison with the way they worked in the past, and the way other similar organizations worked. The research team incorporated a holistic understanding of the context of each of the stakeholder organizations, including their structures and history with cross-sector collaboration, cultures, organizational norms, and strategies. Through the use of qualitative methods, the research team was able to understand the stakeholders’ organizations’ practices within their own contexts, why cross-sector collaboration was the tool they chose to employ, and where evidence could inform their decision making (Reference Cornwall and JewkesCornwall & Jewkes, 1995; Reference Miles, Huberman and SaldanaMiles, Huberman, & Saldana, 2014).

Throughout the process, the research team used purposive sampling to select research participants based on their perspective and role (Reference StringerStringer, 1999), and intensity sampling based on the understanding of current information and the need to fill remaining gaps (Reference PattonPatton, 2015). The research team took participant observation notes at all interviews and focus groups, as well as during the survey and meetings. Most focus groups were audio recorded. The focus groups allowed the research team to observe stakeholder interactions within the context of their own organization and between organizations. As noted in the Figure 7, the terms “research participants” and “participants” are used to describe those who formally participated in data collection to inform the study. The term “stakeholders” is used to describe groups including participants as well as others who did not participate in formal data collection, usually because the process was believed to have achieved saturation. The methodology of the project included continuous data collection, and the research team remained nimble, faced with constantly shifting contextual dynamics that are common in an unstable, rapidly developing urban slum. Consistent with the evidence base, as a result, the process was reflexive, flexible, and iterative (Reference Cornwall and JewkesCornwall & Jewkes, 1995). Ghana is an Anglophone country, so the process was conducted in English except where otherwise noted.

3.3 Piloting and Results of the Intervention Components

The research director suggested initial steps, with significant input by Sr. Matilda and the Department of Public Health. Reflecting PAR principles developed around priority setting, throughout the process the research team trained the stakeholders on the relevant collaboration evidence, supported their decision making, and collected data from the stakeholders to feed back into the process (Reference Patten, Mitton and DonaldsonPatten, Mitton, & Donaldson, 2006). Each phase resulted in decision making by consensus, meaning that all stakeholders came to agreement on the decision. Thus, when this Element describes how “the collaboration” took an action, that means research participants in a full FGD or smaller FGD reached a consensus-based decision, decided on a follow-up action, and designated one or more research participants to take that action on behalf of all participants in the collaboration. The results of each PAR phase are organized by the two main themes that emerged: process management of the cross-sector collaboration and strategy and project development based on the stakeholders’ priorities. There were three phases to the PAR process.

4.1 Process Results

The result of this concept phase is the PAR intervention described as a flowchart in Figure 10. The research team employed grounded theory to consolidate the collaboration principles, interviews, FGDs, and survey. The intervention includes outputs, outcomes, and impact described in Section 4.2.2: areas of focus that the research participants sought or occurred through the PAR process.

Figure 10 Participatory action research intervention flowchart

4.2 Strategy and Project Results

As described, the collaboration identified five community priorities based on the community chiefs’ and leaders’ definition of their own needs: sanitation, addressing community violence, supporting vulnerable kayayei, solid waste management, and building a clinic. The needs were triangulated with the government priorities, and the stakeholders created strategies and then projects to address each priority at its root cause. This section focuses on the sanitation strategy and the latrine project, and how they informed the PAR intervention impact.

4.2.1 Latrine Strategy and Project Results

Latrines were exciting for the community leaders because, as one of them said in FGD 2, “cholera is killing us today.” Two aspects of latrine installation impacted the broader cross-sector collaboration process: (1) the creation of policy change and (2) the stakeholders’ planning for latrine sustainability. As described, from the beginning of the Old Fadama project, participants from all sectors – government, nongovernmental organizations, community, and citizens – were interested in creating sustainable solutions. They explored different strategies to address the sanitation priority. After discussing portable latrines, it became clear that installing permanent public latrines was the most sustainable option. However, as noted, there was a city policy against upgrading Old Fadama. In September 2015, the facilitator and community liaison, at the request of the collaboration, entered into negotiations with the AMA to change city policy. Following the community survey, the collaboration filed an application for permits to install public latrines and bathhouses, formalizing the negotiation.

The permits required planning approvals from multiple AMA offices that reviewed the collaboration’s latrine proposal. The approvals had the gradual effect of changing AMA policy to allow slum upgrading in Old Fadama. For five months, the director of public health, facilitator, and community liaison contacted various city offices in turn, seeking approvals on the latrine and bathhouse permits. Each office provided input and then directed them to a different office for a new approval. It appeared that no office wanted to give the final approval for the plan, as no one wanted responsibility for finalizing such a change to city policy.

Although intense and frustrating, this protracted negotiation served the important purpose of improving the longer-term planning for sustainability (Reference KritzKritz, 2018). As the latrine proposal circulated from office to office receiving different levels of approval, stakeholders were called upon to identify and analyze key barriers to sustainability. For example, the government identified that Old Fadama was not included in city planning because the slum developed on city land that was set aside as a floodplain. This meant, as noted earlier, that resources needed to be reallocated for the planning. In addition, the gap in city planning in the past meant that there were no roads to accommodate sanitation waste pickup trucks, so they would not be able to enter the slum due to access issues. However, the community leaders explained how the city’s partial demolition of the slum – which destroyed latrines and created a newly cleared area – presented an opportunity by creating space for the new sanitation facility, and also opening up an access road.

The Catholic sisters, known for their long-term attention to communities in need, identified the importance of doing “what the community wants,” meaning that the project needed to respond to community requests if it were to ultimately achieve sustainability. The sisters have a long-held value of creating charitable networks to support communities in need. Thus, when funding was needed for latrine installation, the Catholic sisters were able to use their charitable networks to raise foundation funds to supplement the research funding that was reallocated to latrine installation.

The facilitation and research team from NCHS determined that community leaders from multiple tribes regularly engaged in destroying one anothers’ infrastructure. For example, political leaders from multiple tribes had used the headquarters of the community association for years. However, after an election, it had been destroyed in an act of supposed political vigilantism to keep another political party from using the facility. This cycle of violence was attributed to political retaliation, but just as often these kinds of actions seemed to be a guise for other issues. Thus, latrine signage about the collaboration helped insulate the latrines from destruction (see Figure 11).

Figure 11 Latrine block signage

Video 3

Tribal leader involvement.

www.cambridge.org/Kritz_2nd_edition

And finally, through the analysis required to receive approval from each government office, all stakeholders identified the absence of a maintenance culture in the government. Their implementation planning addressed this gap through a community survey that bolstered community leaders’ knowledge of the cross-sector collaboration process, coaching them on how to take responsibility for various aspects of latrine management. The city donated the land, approved infrastructure investment, and finally entered a long-term public-private partnership with the stakeholders – represented by Sr. Matilda and Batsa as key implementing partners – to manage the latrines.

Video 4

Building sustainable latrines

Video available at www.cambridge.org/Kritz_2nd_edition

At the same time, other outcomes of the PAR process continued to develop, unbeknownst to the research team and participants. During the time that the research team shuttled from AMA office to office seeking the necessary permits, AMA employees discussed the proposal with their colleagues in Accra’s sanitation business. As the business community took notice, the latrine project became a lively topic of conversation. The concept attracted the interest of multiple sanitation businesses that viewed the strategy as workable. They recognized a business opportunity.

The December 2016 national elections led to a new political administration with a specific policy and new ministry supporting slum improvement. Early in 2017, new AMA government officers expressed their gratitude for the collaboration’s willingness to take on the challenge of sanitation in Old Fadama and immediately approved the stakeholders’ latrine permits, formalizing the long-awaited policy change for Old Fadama slum improvement.

Having previously assessed the collaboration’s proposal, the local sanitation businesses observed the city policy change and wanted to participate. This is where the research team discovered some surprising outcomes. As the cross-sector collaboration began latrine construction, other sanitation businesses, aware of the strategy, adopted the plans for latrine installation: on their own initiative and with their own resources, they began latrine construction in other areas of the slum. To date, six additional latrine and bathhouse blocks have been installed by local businesses (see Figure 12), investing their own resources and providing sanitation coverage in new areas of Old Fadama, providing another path toward local sustainability.

Figure 12 Latrines installed by local sanitation business

Figure 13 Kayayei at work

The strategic prioritization and planning process created a workable plan that external organizations eagerly adopted and began to implement on their own, to the benefit of the community. This was a very positive, yet unintended result for a community that had been so long frustrated with short-term projects led by internationally funded foreign investigators or collaborators. The previous projects resulted in time and resources taken from community members – but “nothing changed.” Now all stakeholders could point to their own project results and community empowerment, having engaged multiple AMA offices and the community chiefs as well as survey participants in site selection and management. Over time, a local sanitation company was engaged to manage the latrines. While the latrine block is not yet a sustainable enterprise, the research participants believed that local sanitation businesses created a path toward sustainability on the broader issue of sanitation in Old Fadama.

5.1 Phase I–II: Key Informant Interview, Focus Group Discussion, and Process Management Results

In 2017, through a purposive process, stakeholders identified a new research participant and implementation partner, Positive Action for Porter Girls, a community-based kayayei project in the Madina community, to work alongside the Kayayei Youth Association in Old Fadama. Creating programming was challenging for a number of reasons. Kayayei migrate back and forth between Old Fadama and their home communities, so programs begun in either location are frequently not completed. Kayayei endure the same challenging conditions as the rest of the community; however, their gender and their isolation from traditional family networks make them uniquely vulnerable. Due to the long hours they spend working, their children are often left unsupervised and cluster in groups around the slum.

Replicating the PAR intervention, the research team and participants conducted a series of interviews and FGDs to learn about the priorities of a purposive sample of one hundred kayayei identified by the Kayayei Youth Association and Positive Action for Porter Girls. The team learned that these women came to Accra with several priorities: health-care services, education for their children, and better economic opportunities.

Two FGDs were held with the sisters, the Kayayei Youth Association, Positive Action for Porter Girls, and technical officers from NCHS to explore possibilities for a strategy for addressing the kayayei community’s challenges based on their priorities. The research participants identified three priorities and the facilitator and community liaison triangulated these with government priorities. Areas of shared priority were health (treating and preventing chronic and acute health conditions); economic (skill building and credit access); and education (education and social welfare of the children of kayayei). Stakeholders identified trafficking as an emerging issue.

In January 2018, three Catholic sisters (including Sr. Rita, the former community liaison) became facilitators. Coached by Batsa, they began to replicate the PAR intervention around the priorities to create new strategies to support the kayayei. The research team coached these new facilitators in an ongoing way, similar to the concept phase, on how to collect data to feed back into the PAR process and inform stakeholder decision making. These new facilitators and research participants – supported by the research team – began by identifying additional new stakeholders and inviting them to join the new cross-sector collaborations. At the outset, the research team taught new stakeholders how policy change occurred in the past and encouraged them to pursue this avenue.

In 2018–2019, stakeholders launched four projects in urban slums in Accra and Cape Coast, and eight projects in rural communities in northern Ghana.

While the education, economic, and trafficking collaborations achieved similar government participation, the health sector in Ghana is larger and yielded more robust data on the health collaboration so it is reported here.

5.2 Health Project and Policy Results

The stakeholders’ strategy, informed by concepts of social justice, was to provide comprehensive health services for each member of the kayayei community based on needs they identified. Aggregating the numbers of beneficiaries was challenging due to the community-driven design, which led to variability in services provided. The intervention, funded by this research project for 300 kayayei beneficiaries, included:

1. 1. Conduct health talks educating the community about the need for screening and treatment;

2. 2. Conduct community-based basic health screening and vaccination;

3. 3. Identify treatment sites and connect kayayei to treatment for chronic and acute conditions diagnosed during the screenings;

4. 4. Accompany kayayei to follow up care, using a trust fund supported by this project to pay for treatment as needed;

5. 5. Enroll or re-enroll kayayei in insurance, as needed.

Many of the original one hundred kayayei key informants from 2017 migrated back to northern Ghana. So the sisters, Kayayei Youth Association and Positive Action for Porter Girls used a purposive process to identify new beneficiaries. Sr. Angelina Gerharz, SSpS, served as this project’s facilitator, training and supporting officers from the Positive Action for Porter Girls. Nurses from the Catholic Health Guild, a new stakeholder, provided volunteer medical services. The project goals were immediately expanded to serve 940 kayayei when the local government polyclinic became a stakeholder, agreeing to provide test kits, reduced-cost vaccines, and screening and treatment materials.

The National AIDS Control and National TB Control offices and National Health Insurance Scheme (NHIS) were contacted, and they saw an opportunity to provide services to a population they were unable to reach. The facilitators were able to explain, based on the project’s prior results, how policy change had happened before, and why it would be effective. These entities modified their policies in order to cost-share the setup for the screening. As the project expanded, government policy change included multiple steps described below. Due to health policy change, the collaboration expanded the number of beneficiaries from 940 to 1,534 kayayei.

The National Health Insurance Scheme (NHIS) was similarly invited to attend the screening and cost-share. Their financial policy was to engage in community-based enrollment when one hundred or more people could be enrolled or re-enrolled. The NHIS is a health agency and NHIS funds are put to work in the health sector. NHIS was enthusiastic about the volume of beneficiaries. However, funding for vouchers for indigent people to receive health insurance is administered by the Department of Social Welfare (DSW). DSW did not have funding to conduct community outreach. So, the collaboration paid for DSW to attend community health screenings and identify kayayei who could not afford health insurance. DSW screened these individuals and provided vouchers, which NHIS processed and the collaboration distributed to the kayayei in follow-up visits. Thus, rather than enrolling 300 kayayei, 1,789 indigent kayayei were enrolled in national health insurance in three locations – Old Fadama, Madina, and Ashanti Region.

Educated by the prior results of the intervention and coached how to make policy change, five government health agencies and the department of social welfare made policy changes including:

• budget allocation and reallocation

• service delivery adaptation: moving to provide services in the community itself

• modifying data collection processes accordingly

• working with collaboration stakeholders to incorporate the collaboration into government processes, creating accountability for staff follow-up

• increasing efficiency by deputizing stakeholders (Catholic Health Guild nurses, National Catholic Health Service research staff, and community-based organizations) to collect health and social services data.

The process replication continued, with new funding and a new target of 750 beneficiaries. By mid-2022, more than 10,000 beneficiaries received health services that they requested, thanks to the collaboration. Figure 14 visualizes the health stakeholders at the end of 2022.

Figure 14 Health stakeholder diagram

In addition to direct services, community-based health provision provided lasting and widespread benefits regarding the stakeholders’ 2015–2018 priority to build a clinic in Old Fadama. At that time, community residents felt unwelcome at nearby health facilities. As one community resident stated, “they think we are dirty and violent, and they don’t want us there.” So, the stakeholders created health access by conducting health programs in the community, accompanying beneficiaries on their follow-up visits to the polyclinic, and providing insurance and trust fund support where needed. This health intervention received substantial government investment, additional funds from a local charitable organization and later, international philanthropy. Thus, in 2019, when community leaders revisited the issue of building a clinic, the community itself deemed it unnecessary, reducing duplication and waste. This removed a major sticking point in the relationship between the community and the municipal government.

5.3 Network Analysis

As the PAR intervention created new collaborations, it was not possible to collect data from each group of stakeholders in the same exhaustive detail as during 2015–2017, leading to the idea of network analysis. In the United States, systematic research examined collaborative network activities among public organizations, making important contributions to the understanding of governance. Robert Agranoff’s Managing Within Networks: Adding Value to Public Organizations (2007) provided a significant early step. He examined diverse sectors and employed grounded theory to describe and inform our understanding of how these interorganizational networks emerge, operate, and affect public problems. As public administration network analysis evolved, the field continued to focus on networks as a strategy to resolve public problems. The research focus continued to be placed on evaluating or describing networks in order to understand typology and effect (e.g., Provan & Milward, 1995).

More recently, intervention research from public health was picked up by network analysts in public administration, who called for the field to expand upon the more evaluative or descriptive approach described earlier (Reference Whetsell, Siciliano, Witkowski and LeibleinWhetsell et al., 2020). Siciliano and Whetsell’s scholarship harmonized the evidence, calling for public administration network analysis to incorporate the public health evidence on intervention, manipulating network mechanisms in order to achieve better network configurations to impact public problems (2021).

In 2015, the research director had a similar approach to this project, seeking to demonstrate how cross-sector collaboration evidence from developed countries could impact public problems in a developing country. To do that, the PAR process harmonized the public health evidence on intervention design with the robust cross-sector collaboration evidence from public administration. By 2020, the collaborations grew. The research team looked at network analysis as a way to visualize collaboration expansion, and understand why some collaborations were more successful than others. As Reference Scott and UlibarriScott and Ulibarri (2019) point out, public administration studies focusing on institutional elements but lacking explicit network analysis can be categorized as “implicit network analysis.” The research team took this perspective, looking retroactively at what developed in order to understand what factors were of interest to the stakeholders, in order to plan for network analysis as a potential tool to measure progress.

In 2018–2020, the new organizations and participants in the collaborations were straightforward, although the collaborations developed so rapidly it was challenging for the research team to keep track of all the new stakeholders. The research team employed grounded theory and similar methods to Agranoff’s, moving beyond the social network evidence base into “network analysis” (2007, pp. 6–7) in order to incorporate not only the structure and operations of human networks but also issues of representation, formalization, and organizational effort. As reported earlier, the research team first reanalyzed the 2015–2017 collaboration, and validated and refined these results (which are reported in this second edition) with data from the PAR intervention replication in 2018–2022. The unit of analysis was each collaboration. Like Agranoff’s subjects, the Ghana collaborations were substantively diverse including sanitation, microfinance, skills training, education, and health. Collaborations were both chartered (formally established through legal means) and nonchartered (not formalized under a legal construct) entities. Similar to Agranoff’s methods, the research team’s preparation for network analysis relied upon observation, document coding, and data from semi-structured interviews and focus groups.

The more challenging aspect of the research was understanding the mechanisms that caused the collaborations to develop, expand, and provide services to vulnerable populations. In particular, the research director was interested in understanding the relationship change that precipitated cross-sector collaboration – the more appropriate configuration of stakeholders, such as Siciliano and Whetsell posited (2021). One key finding from Agranoff’s work is that the value networks add to public management is context specific, often indirect, and not necessarily in keeping with traditional notions of outcomes. Measuring intangible and unpredictable outcomes is challenging (Peters, Gonsamo, & Molla, 2011). Measurements must be used to inform and build practice, improve efficacy, and contribute to openness and honestly through learning; this often takes place on a relational level (Taylor & Soal, 2010). These intangible benefits were explored to explain the pathway that was created from individual transformation to organizational transformation, when these new government stakeholders readily reallocated resources toward vulnerable populations.

Recent groundbreaking work on the psychology of democracy (Reference MoghaddamMoghaddam, 2016) was helpful to explain stakeholders’ individual transformation as they developed new skills working together. Moghaddam’s study of the psychology of democracy built on a career studying intergroup relations, informed by living and researching the psychological foundations of democracy and dictatorship in countries with a variety of governance systems. He adopts a cultural psychology framework, which looks at human psychology through a rich holistic perspective on the relationship between an individual and their culture (see, e.g., Shweder, 1991). The research team felt this holistic approach was required to explore individual transformation in the context of this cross-sector collaboration intervention.

Moghaddam outlined ten critical convictions or behavior styles that guide the actions of a democratic citizen, represented in a circle diagram (Figure 15) because more than one can develop in the same time in the same person (2016, p. 51).

Figure 15 Ten convictions of the democratic citizen

Analyzing the data through the lens of this diagram, it appears that the research team coached the stakeholders to develop skills for democratic actualization as follows:

• Stakeholders were reminded that they could be wrong through revisiting their prior individual approaches to Old Fadama that did not work. They were coached to actively listen to other stakeholders who had different perspectives on why they, too, failed in Old Fadama, and to be open to these different viewpoints.

• This approach led the stakeholders to critically question everything that they believed about Old Fadama in the past.

• They were coached to revise their opinions when faced with new evidence. As the research team repeatedly sought information and opinions from different sources, analyzed this data and shared the results in a continuous process, stakeholders realized they could learn from those who are different from them. At times, the facilitator coached them to reach consensus which meant that they had to seek to understand those who were different.

• Ghana is a religious culture, and meetings frequently begin and end with Christian and Muslim prayers. The sisters’ prayers for each stakeholder to “help our brothers and sisters who are less fortunate” encouraged the stakeholders to be actively open to the new experience of the collaboration. As they developed shared understanding, stakeholders expressed that earlier feelings of being alone and overwhelmed at the challenges presented by Old Fadama were replaced by feelings of camaraderie that came from addressing the challenges together. An example of a mechanism that was expressed repeatedly by all stakeholders at the individual level was transformation from not knowing one another, and often feeling at odds with organizational positions of other sectors, to feelings of friendship. These feelings were encouraged and supported by the Catholic sisters who, at the outset of the project, decided among themselves to try to build positive relationships between the stakeholders.

• This encouraged the municipal government to decide to create new experiences with the Old Fadama community, through designing strategies to impact community challenges.

• The idea that there are principles of right and wrong was brought to the forefront by engaging the sisters as moral leaders. They encouraged the stakeholders to actively seek experiences of higher value, such as deciding to install permanent latrines because they would be more sustainable and that is what the community had requested, even though this policy change was a more challenging route.

This exploration of principles from the psychology of democracy helped to explain how the PAR intervention was successful at building the stakeholders’ skills to engage in collaboration and therefore participate in democratic governance, and the individual transformation the stakeholders experienced as a result of this activity. In 2018–2022, individual transformation created a welcoming environment for new collaboration stakeholders. The original stakeholders expressed kindness, enthusiasm, friendship, camaraderie, and gratitude. Over the years, these sentiments became norms that new government stakeholders adopted upon joining the collaboration. This shift will become the basis for network analysis in the next phase.