Research Paper Pathways for securing government commitment for activities of collaborative approaches

Decentralization in many African countries makes local governments responsible for water, sanitation, and hygiene (WASH) service provision. Yet service provision is complex and requires concerted inputs from many actors, thus local governments are increasingly collaborating with internal and external stakeholders. These ‘ collaborative approaches ’ , which bring together diverse stakeholders to solve complex problems facing service delivery, have been found to strengthen service provision when they obtain support and commitment from local government decision makers. However, the ways to obtain this commitment remain understudied. This cross-case comparative study investigated how collaborative approaches in Ethiopia and Uganda obtained government commitment for their activities. Results draw upon experiences of 13 cases where suf ﬁ cient commitment either was or was not gained. Deep case knowledge of the research team, aided by Qualitative Comparative Analysis, revealed that the collaborative must (a) be seen as legitimate and credible, (b) align activities within government mandates, and (c) demonstrate the value of the activities to decision-makers. Together this strategy provides a reliable way for collaborative approaches to secure government commitment for their activities. Notably, the results also show that commitment can be obtained even amidst turnover and political dynamics.


INTRODUCTION
overnight. Yet, other reasons for a lack of commitment fall on the shoulders of the collaborators themselves, who did not prove their work to be more valuable than the countless other priorities of decision-makers. In reality, the changes sought by collaborative approaches often take many years to come to fruition, meaning that their activities must also compete with activities that produce more immediate results or other 'lower-hanging fruit'. Nonetheless, there are many collaborative approaches that have successfully built government commitment.
To better understand what enables government commitment in collaborative approaches, we systematically analyze combinations of key factors in 13 instances where government commitment was either gained or not gained, using Qualitative Comparative Analysis. As a result, this study presents practical and evidence-based strategies for collaborative approaches.
This manuscript first covers the current state of literature, outlining how collaborative approaches need government 'uptake of recommendations' to make progress, how 'uptake' is comprised of commitment followed by action, and finally, as 'commitment' is similar to the oft-studied concept of 'political will', we then review factors that have been found to secure political will. After tying this back to the idea of 'commitment' in an overview of the research context and cases, we delve into our methodology, followed by a presentation and discussion of the results and implications.

Collaborative approaches rely on government uptake
By their nature, collaborative approaches are highly intertwined with government, who are the responsible authorities for ensuring service delivery takes place (Uganda Legal Information Institute 1997). Pugel et al. (2021) found that progress on goals of the collaborative group hinges on government actors taking up and acting on the group's recommendations. Furthermore, they documented how this 'uptake' can waver and change over time: 'Implementers must continually maintain and regain government [uptake] over the long lifetime of a collaborative approach, especially in contexts affected by political dynamics, shifting priorities, and turnover ' (2021, p. 10). Implementers and funders of collaborative approaches, they argued, must make purposeful, concerted, and regular efforts to gain the support and commitment of relevant government decision makers. These findings expanded existing knowledge of the role of government in collaborative approaches, which fail to acknowledge the importance of government uptake of recommendations of the collaborative. Previously, theory has only focused on the importance of government endorsements (Lasker et al. 2001) and involvement of government either as leaders (Ansell & Gash 2007) or as members (Gray & Purdy 2018).
Notably, the lack of published knowledge on government uptake may partially be due to the fact that virtually all theories and frameworks related to collaborative approaches (e.g. Ansell & Gash 2007;Gray & Purdy 2018) have been based on cases in high-income countries and non-development contexts, which experience very different governing regimes (Pugel et al. 2020). Because existing theories insufficiently cover government uptake, strategies for how collaboratives can gain and maintain government uptake for their activities have likewise not been investigated in literature but are particularly relevant to the WASH sector.

Uptake consists of commitment and action
Gaining government uptake of recommendations from collaborative approaches requires first commitment and then action, where action relies on access to funding mechanisms to follow through on those commitments. This study focuses on the first step: commitment to an activity. The idea of commitment is highly related to the oft-studied concept of 'political will'. As such, we will refer to the concept of 'political will' when reviewing literature, but shift back to 'commitment' when discussing the WASH context and collaborative approaches.
Political will is defined by Post et al. (2010, p. 659) as 'The extent of committed support among key decision makers for a particular policy solution to a particular problem'. In WASH work at sub-national levels, political will is a matter of the government committing and taking responsibility for policy implementation. The phrase is often used to encapsulate the complicated processes, incentives, and politics that are deeply embedded in WASH systems. The concept of political will remains nebulous as it can be simultaneously blamed for why change efforts fail (Tsinda et al. 2021) and praised as a driver for why they succeed (Eliasson 2013).
To overcome this obscurity, significant efforts have been made to understand the inner workings and drivers of political will. Multiple factors have been found to influence political will, including: the perception that the problem exists and is urgent (Kingdon & Stano 1984;Shiffman 2003); the strategic framing of issues to be seen as legitimate and valuable (Roche et al. 2018); the legitimacy of those proposing the policy or recommendation (Roche et al. 2018;Bogenschneider & Corbett 2021); the existence of similar ongoing efforts (Pierson 2000); and perhaps most important and most challenging in reality, the overarching interests and priorities of decision makers (Fritzen 2005;Bogenschneider & Corbett 2021).
Notably, when applied to the context of collaborative approaches, these theories have one main limitation. Despite the wealth of theory on drivers of political will and government commitment, no theories have focused specifically on how collaborative approaches can obtain government commitment for their activities. We posit that the nature of collaborative approaches in WASH mean that their strategies for securing commitment may be different from strategies identified by existing theory. For example, it may be easier for them to build upon ongoing initiatives or address existing priorities because they involve decision-makers and technical government officials in the collaborative process, and often these decision-makers have already agreed on some program goals in advance. At the same time, it may be more challenging for them to contribute to ongoing efforts or address existing priorities due to the frequent turnover and shifting priorities in some WASH contexts. We presume that donor power dynamics may also play a role, where decision-makers may be more receptive to a donor's priority if it means additional funding could be secured for their district or town. Ultimately, in light of the many ways that commitment can be gained more broadly, there remains a need to investigate how collaborative approaches specifically can secure government commitment for their activities.

Research question and context
The current study answers the question: What combinations of factors allow collaborative approaches in WASH to build government commitment for their recommendations? We investigate this by looking at instances where commitment was secured, and instances where it was not, by collaborative approaches seeking to strengthen WASH service delivery in Ethiopia and Uganda.
Five of the authors of this study are field experts from international non-governmental organizations that either managed or facilitated one of five separate collaborative approaches in Ethiopia or Uganda. Field experts' direct roles in the collaborative approaches meant they were the most knowledgeable about the work but also heavily invested in its success, thus, the remaining authors were university partners who conducted data analysis in a way that ensured consistency and rigor across cases while also allowing field experts to be closely involved in identifying factors and interpreting results.
From 2017 to 2021, these collaborative approaches regularly convened a diverse array of local stakeholders at the district or town level to identify solutions critical to improving water or sanitation services in their district or town. At the start of the project, all collaboratives obtained government permission to operate and willingness to participate. All collaboratives were originally managed by international non-governmental organizations, with leadership and close involvement of relevant government entities. Stakeholders included decision makers and other government agencies and municipalities from water, sanitation, health, education, finance, and/or environment sectors; civil society organizations; private sector; academic institutions; and community representatives. Over time, some collaboratives passed leadership roles such as facilitation, setting agendas, inviting participants, and securing funding over to key government officials.
All collaborative approaches were broadly seeking to improve sanitation or water services, but the specific activities they focused on to improve services varied. Each of these five collaborative groups of local stakeholders, or 'collaboratives', had two or three focus areas (for a total of 13 across all five collaboratives). For example, a town-level collaborative in Ethiopia had two task teams focused on improving sanitation service delivery in the town, with one focused on building consensus on a fecal sludge dumping site and the other focused on standardizing management of shared latrine facilities. Some focus areas gained more traction and commitment from government decision makers than others. Thus, the 13 focus areas, originating from the five collaboratives, were considered separate cases for this study.
Throughout this paper, the following terms will be used: each 'case' will represent one focus area, a specific topic defined by a collaborative, a group of local stakeholders working together as members of the collaborative approach. Each focus area could comprise a few distinct activities. For example, the first case A1 was the focus area of maintenance in Collaborative A, which included a set of activities including setting up a spare parts supply shop and formalizing water user associations for maintenance of rural water infrastructure in the Collaborative A district (Table 1).

METHODS
This study used an action-research approach to data collection and analysis, where members of the research team were themselves the implementers of the collaborative approaches (McTaggart 1991). A deeply embedded research team enabled continuous access to rich context-specific case knowledge throughout the entire program. Additionally, over 2,500 pages of documentation were collected regularly over the course of 4 years of the program (from March 2017 to March 2021). Documentation included program proposals, signed memorandums of understanding or terms of references, technical assessments, meeting reports, quarterly program reporting, field visit reports, and facilitator diaries. In addition, interviews were conducted with members and facilitators of the collaborative and these interviews were transcribed. All documentation was qualitatively coded as a part of a larger study to build deep case knowledge among all researchers. The research protocol was reviewed by the University of Colorado Institutional Review Board under Protocol #19-0207.
Fuzzy-set Qualitative Comparative Analysis (fsQCA) was then used to conduct a structured cross-case comparison across 13 cases to investigate combinations of factors that contributed to commitment. FsQCA is a cross-case comparative method that provides more structure and replicability that traditional cross-case comparative methods, allowing the research team to investigate all 13 cases simultaneously. FsQCA uses set logic and case knowledge to investigate causal relations, thus making it a hybrid between variable-oriented and case-oriented analytical methods. The authors direct the reader to foundational documents for additional information related to QCA broadly (Ragin 1987) and QCA in the WASH sector specifically (Kaminsky & Jordan 2017). In brief, the QCA process entails selecting an outcome of interest, identifying factors theorized to contribute to that outcome, evaluating the extent to which each of the factors and outcomes are present in each case, developing categorizations to represent variation, and then following a highly structured comparative process to identify cooccurrence of all factor combinations and the outcome. With the aid of the software, instances of co-occurrence can be further interpreted using Boolean minimization and set theory so that the most influential relationships are identified. Each of these four main steps ( Figure 1) are highly iterative and based deeply in case knowledge, for example the factors and their categories were re-defined many times to best represent this set of cases. These four steps are outlined in detail in the following sections.
Step one: select outcome We define government commitment as 'Significant committed support among key decision makers for a particular activity or set of activities recommended by a collaborative', adapting the definition of political will from Post et al. (2010). Significant committed support in our cases were meaningful actions that would be expected to make a difference in the implementation of the case. Examples of this include passing policies or resolutions, implementing policies or major activities, or contributing financially or in-kind to implementation. Significant commitment was demonstrated in D1, for example, when the town deputy mayor called together a technical committee to identify a site location for the fecal sludge disposal facility. Similarly, significant commitment was demonstrated in A1 when the district water office head used part of the revolving fund to secure spare parts for a water utility that the collaborative had helped set up. If too few decision makers demonstrated commitment or an otherwise insufficient amount of commitment was secured, the case was considered to have not secured significant government commitment. For example, more commitment was needed from the C2 case, where even though the town administration was aware of the need for improved latrine management, there was a lack of follow-up and technical support to the latrine managers.
Data were collected about the extent of commitment gained, then within Step 3, each case expert on the research team, using evidence from process documentation, assessed the relative strength of commitment on a 4-point scale from 0 (low commitment) to 1 (strong commitment). Cases that gained significant commitment but at a slower rate than expected were assigned a strength of 0.66, while a strength of 0.33 was assigned to cases that had some commitment but still required more to have a meaningful impact. Full definitions for each increment on the 4-point scale were iteratively developed to best fit these cases and were defined based on set theory, as is standard in QCA. See Supplementary material for all categorization definitions.
Overall, eight of the 13 cases were found to have gained significant government commitment. The five remaining cases still required greater levels of commitment to have a meaningful impact on their implementation.
Step two: identify factors Five factors were selected for investigation, supported by both case knowledge and theory: Demonstrated Value, Existing Priority, Ongoing Initiatives, Aligned with Clear Mandates, and Legitimacy of the Collaborative. These factors were identified through multiple structured discussions by the research team, which includes the facilitators and experts. These discussions lasted from April to August 2021, followed by a review of documentation and theory to synthesize factors into those hypothesized to play the most influential role in gaining and maintaining government uptake.
Decision makers prioritizing the activity have been cited throughout literature as an underlying force driving commitment and political will (Kingdon & Stano 1984;Fritzen 2005;Bogenschneider & Corbett 2021). In our cases, this manifested in different ways, where some activities were Existing Priorities before the collaborative formed. Regardless of whether it was an existing priority or not, some collaboratives made concerted efforts to convince decision makers of the activity's urgency, value, and/or impact, here called Demonstrated Value. This value was demonstrated sometimes through advocacy, data and evidence, or both. Path dependency theory tells us that activities with ongoing efforts are more likely to gain government commitment because they allow for policy continuity (Pierson 2000). This aligns with our cases as some had the presence of Ongoing Initiatives, where momentum and ideas were already flowing about the topic. In other cases, collaboratives ensured their activities Aligned with Clear Mandates of government entities and made substantive efforts to convince decision makers that supporting the activities would help fulfill their mandates.
Finally, policymaking literature argues that credibility of the intermediary entity who is proposing policies heavily influences commitment by policymakers (Bogenschneider & Corbett 2021). This aligns with our cases as significant efforts were made to build the Legitimacy of the Collaborative, ensuring that decision makers saw the collaborative itself as a legitimate entity. This legitimacy was gained by the collaborative using a variety of strategies including consistent messaging and continuous updating, gaining permission from higher levels of government first, involving decision makers in problem identification, formal agreement through memorandums of understanding, and the conveners demonstrating long-term commitment ( Table 2).
Step three: assess factor and outcome presence and score cases This study then investigated how all of these factors -Existing Priorities, Demonstrated Value, Ongoing Initiatives, Aligned with Mandates, and Legitimacy of the Collaborativeworked alone or together in combination to contribute to collaboratives building government commitment for their focus areas. Using the data collected, alongside deep case knowledge, the extent to which each factor and outcome was present in each case was systematically scored by the research team through a process called indirect calibration (see Basurto & Speer 2012). Rather than 'scores', indirect calibration assigns 'set categorizations' (Fully Present, Partially Present, Partially Absent, Fully Absent) to qualitative statements in line with set theory, case knowledge, and theoretical knowledge (Ragin 1987). An example of qualitative statements assigned to set categorizations for the factor Aligned with Clear Mandates is given in Table 3. Set categorizations all factors are provided in the Supplementary material. Set categorizations were assigned systematically based on documentation and case knowledge of the members of the research team with closest knowledge of the cases. Once each case expert assigned categorizations based on case knowledge and backed by documentation, the first author worked with all individuals to ensure consistency and to re-define and refine set categorizations as needed. Iterative refinement used set logic to emphasize the differences that occurred between cases when variation occurred. Once all scores were assigned for all factors and all cases, the result was a data matrix reflecting the extent to which each of the five factors were in the set of each of the 13 cases (Table 4).
Step four: systematically investigate influential relationships Using the software fsQCA 3.0, we systematically compared and quantified the set relationships between the outcome and all possible combinations of the factors that varied across cases. All cases were used for each comparison, for example when investigating contributors to Commitment, cases that did not gain sufficient Commitment were used as counterfactuals. 1 One example of a set relationship comparison is as follows: how often did cases with both Demonstrated Value and Ongoing

Existing Priority
Decision makers prioritized the activity before the collaborative formed, but they had not necessarily taken action on it.

Demonstrated Value
The collaborative convinced decision makers of the activity's value, impact, and/or urgency.
Ongoing Initiatives Initiatives such as legislation or programs heavily related to the activity topic were already underway when the collaborative formed.

Aligned with Clear Mandates
Activities clearly fit within the mandate of a government entity.

Legitimacy of the Collaborative
The collaborative proposing activities are seen as a credible, legitimate, and trustworthy entity.
Initiatives fully present also have Commitment fully present, and how often did cases that did not have Demonstrated Value nor Ongoing Initiatives not gain Commitment? Partial presence and partial absence were also investigated but not listed in the example. Case knowledge guides the entire process, supplemented by two validity measures. The validity measures represent two slightly different ways of looking at the set relationship, one from the perspective of the factors and one from the perspective of the outcome. 'Consistency' focuses on the factors. Higher consistency values (.0.80) mean that most of the times that combination of factors was present, the outcome was present. 'Coverage' instead looks from the outcome's perspective. Higher

Fully Present
The activities in the focus area clearly fit within an existing mandate of the relevant sector offices, with clear roles and responsibilities.

Partially Present
The activities in the focus area fit within a few overlapping government mandates and thus are not entirely clear.

Partially Absent
The specific activity is not the responsibility of the local government at the scale the collaborative is operating, but it is still somewhat related to one of their mandates.

Fully Absent
The specific activity is not the responsibility of the local government such as if it was the responsibility of a nongovernment entity, such as private sector, non-government organizations, or communities.

RESULTS AND DISCUSSION
This investigation focused on how different combinations of factors contribute to building government commitment, and the results identify the most influential combinations that can reliably build commitment. First, before presenting the results of fsQCA, we discuss the only factor that was present in all cases, regardless of whether they gained commitment or not: Legitimacy of the Collaborative. Its presence in all cases is evidence of its fundamental importance, but also prevented us from investigating it in the cross-case analysis due to the lack of variance. After discussing Legitimacy of the Collaborative, we present a single combination of factors, or 'pathway', that was found to be the most influential driver of commitment in the cases studied. We interweave the results with discussions of the implications of these results and advice for collaborative approaches who seek to build government commitment.

Legitimacy of the collaborative is foundational
All five collaboratives included in this study were seen as legitimate and credible entities. The collaboratives involved decision makers either directly or indirectly in identifying the problem areas and would regularly update them on progress and challenges. Transparency and consistent messaging were prioritized during discussions with decision makers. All collaboratives had formal Memorandums of Understanding, and many had Terms of Reference documents detailing roles and responsibilities. The conveners of the collaborative had good reputations, demonstrated long-term commitment to the locality, and gained permission from higher levels of government before starting activities. This factor, while critical, could not be further included in the analysis because the QCA methodology cannot investigate factors that do not vary across cases. In addition, as some of the researchers of this study managed and/or facilitated the collaboratives, there are biases preventing us from fully investigating perceptions of legitimacy. This opens up an area for future research to more deeply investigate, specifically regarding the ways in which collaboratives are seen as legitimate by different entities and stakeholders.
Aligning with mandates and demonstrating value create a pathway for commitment One combination of factors that allowed cases to gain government commitment clearly emerged from the comparative analysis ( Figure 2). Collaboratives that are seen as legitimate and credible can reliably gain government commitment on their activities when the activities are Aligned with Clear Mandates and the collaborative Demonstrated Value of the activities. This pathway proved to be highly reliable, in that every instance where these two factors were present, commitment was gained (consistency ¼ 1.0) and this accounted for nearly all cases that gained strong commitment (coverage ¼ 0.88). Thus, corrected Proof these two factors together had a stronger influence on government commitment than the other factors of Ongoing Initiatives and Existing Priorities.

Align activities with clear mandates
Collaboratives working toward improving public services are, by definition, seeking to work in domains where government is the legally responsible entity, thus their efforts should align with the mandates of government. This idea is particularly wellestablished in the Ethiopia and Uganda WASH sectors, for example, one of the pillars of the Ethiopia ONE WASH program is alignment with government mandates (The Federal Democratic Republic of Ethiopia 2019). In line with this sentiment, the results of this study show that in order for collaborative approaches to gain commitment from decision-makers, they must align their activities within and in support of existing mandates. For example, in 2017, a Collaborative A was formed in one district within a larger Zone in Ethiopia, comprised of stakeholders from the water, health, finance, and administrative offices of the District and Zone as well as non-governmental organizations, local universities, and community representatives.
Over the course of a few meetings, they identified key desired outcomes of maintenance, monitoring, and master planning as leverage points to strengthen rural water services in the district. They knew that maintenance and master planning fell within the mandate of the District Water Office while most monitoring activities unfold at a larger scale and thus are predominantly the responsibility of the Zone Water Department. Thus, as activities were planned and it came time for implementation, the District Water Office Head was targeted for the maintenance and master planning activities while the Zone Water Department Head was targeted for the monitoring activities. We do not mean to suggest that the issues faced by collaboratives can fit neatly within a single entity's mandate. On the contrary, these problems are complex, often either spanning the mandates of multiple entities or stuck in between misaligned mandates. For this reason, they require cross-sectoral collaboration to solve. Since government entities may exhibit fragmentation or are otherwise not 'joined up', strategic framing issues in different ways helps to clearly communicate to different public actors both why and how the issue is aligned with their portfolio of responsibilities.
If existing mandates are not clear, collaboratives need to work together with local government to clarify mandates and establish clearer roles and responsibilities within the country's legal framework. Unclear mandates were a challenge for the collaboratives focused on small-town sanitation, where before the collaborative formed, the reality on the ground was that government actors would push responsibilities of fecal sludge disposal site preparation to one another: the municipality would push the responsibility to the utility, the utility to the health office, and the health office to the environmental protection office. The mandate was clearly the responsibility of the town government as a whole, certainly, but it was unclear which town department or office held that mandate. Both collaboratives that sought to establish fecal sludge dumping sites in their respective towns faced this issue (D1, C1), and as a result they could not simply design the activities to fit within existing mandates. Instead, both collaboratives made significant efforts to clarify mandates for the fecal sludge dumping site and achieved consensus on clear roles and responsibilities for each office that played a role.
Understanding existing mandates is critical for implementers of collaborative approaches. Government decision makers are not a monolith or a single entity whose commitment must be gained. Different government entities and decisionmaking bodies need to be targeted for different types of activities. Effectively, it matters who the decision makers are. In our Ethiopia small-town sanitation collaboratives (C, D), the decision makers are the town administration with inputs from the town offices. In the Ethiopia rural water collaboratives (A, B), the decision makers were the District Administration advised by the district WASH sector offices. For some activities, both District-level and Zonal-level officials were the decisionmaking bodies (A2, B2) while other activities fell within the decision-making power of woreda-level officials only (A1, A3, B1, B3). Finally, in the Uganda rural water collaborative (E), decision makers for one desired outcome were the District Council (E1), the Chief Administrative Officer for another (E2), and technical staff for the final (E3). Thus, conveners of collaborative approaches, and the collaboratives themselves, need to have a strong working knowledge of local politics and decisionmaking dynamics before embarking on their work. Tools such as the political economy analysis (Kempster & Hueso 2015) can be used to build this knowledge.

Demonstrate value of activities through evidence-based advocacy
Our results showed that value demonstration is a further necessary step. Governments, and particular local governments, have myriad responsibilities and competing priorities. Therefore, to get their buy-in for a collaborative's focus areas, it was imperative that the collaboratives showed decision makers how the activities will make a significant impact.
One way collaboratives demonstrated value was by using comprehensive systems assessments to reveal leverage points where single WASH interventions could lead to a wider change in the performance of local systems. These technical results were followed with advocacy to convince decision makers of the urgency, impact, and value of the activities. In addition, relationship building and trust between the government and the collaborative gave greater weight to the issues raised in such assessments.
Demonstrating the value of one desired outcome can come at the cost of demonstrating the value of other desired outcomes, as was shown in the D Collaborative. In D1 and D2, a comprehensive town-wide assessment of sanitation was conducted, revealing that 80% of the town's fecal sludge was being unsafely disposed into the groundwater and nearby river. This affected the lives and livelihoods of the towns' citizens, but the decision makers were unaware that such a large portion of the fecal sludge was unsafely disposed. The assessment revealed that a primary reason that fecal sludge was unsafely disposed of was that there was no designated disposal facility and thus most of the waste collected by vacuum trucks was released into the river. While the assessment also revealed that there were inadequate management practices for communal and public latrines, it also showed that this was a far less pressing issue than the lack of a proper disposal facility. The results were presented to the decision makers, which clearly proved the value of the disposal site (D1), making it a top priority. At the same time, the D Collaborative's second desired outcome, communal latrine management (D2), was seen as a lower priority item in comparison.

Commitment can be gained amidst turnover and shifting priorities
Commitment can also be influenced by a variety of events and factors outside of the control of collaborative approaches. Rampant turnover, shifting political dynamics, and unpredictable events such as natural disasters can quickly change decision-makers' priorities and challenge efforts to secure commitment for a particular issue (Naiga et al. 2015;Pugel et al. 2021). Many of the collaboratives included in this study faced turnover or shifting priorities over the course of the 4 years of study, and yet many still were able to secure commitment from relevant decision makers. This ability to secure commitment amidst turnover and shifting priorities is reflected in the results, where the two factors related to existing efforts, Existing Priorities of decision makers and Ongoing Initiatives, were not necessary to achieve government commitment, and thus dropped out of the final pathway. In the fsQCA process, this occurs when factors are present in some cases that gained commitment and absent in others. This means that the role they played in driving commitment was relatively low compared to ensuring activities are Aligned with Clear Mandates and Demonstrating Value. These findings do not entirely align with existing literature on decision-making, which emphasizes the importance of ongoing initiatives and existing priorities. This inconsistency is likely because many WASH service delivery systems in Ethiopia and Uganda are challenged by weak governance. When institutions and governance structures are weak, WASH systems over-rely on individual commitments and relationshipsand, when paired with turnover and other political dynamics, this can create contexts where priorities can suddenly and unpredictably shift. It is thus reassuring that cases can obtain government commitment even amidst these conditions.

Commitment does not guarantee action
Collaborative approaches rely on uptake from government decision makers in order to make progress (Pugel et al. 2021). We investigate what contributes to one component of 'government uptake', government commitment. The other component, having the capacity to follow through on commitments, is an equally important component of uptake. Follow-through on commitments requires having implementing capacity, including adequate funding, skills, staffing, and other capacities (van Steenbergen & Kumsa 2015); however, low capacity and inadequate financing mechanisms remain the reality of many WASH contexts.

CONCLUSION
This study is timely for two reasons. First, collaborative approaches are increasingly applied in the WASH sector because they are seen as well-suited to addressing complex problems that by nature cannot be solved by any single entity. Second, recent research has identified that government commitment is critical for progress of collaborative approaches. Thus arises the need for well-founded strategies for how collaborative approaches can gain and maintain government commitment. This study filled this gap by comparing 13 instances where significant commitment either was or was not gained by collaborative approaches, and systematically investigating all possible combinations of five key factors that could contribute to government commitment using Qualitative Comparative Analysis.
The results revealed that collaboratives must be seen as legitimate and credible entities that decision makers can trust. Collaboratives must then align the activities within clear government mandates and demonstrate value, impact, and urgency of the activities to the most relevant decision-making bodies. We advise all conveners of collaborative approaches to ensure that they have a working knowledge of local decision-making dynamics before they start their programs. This strategy, quite notably, does not rely on existing efforts or priorities, meaning that commitment can be gained amidst turnover and political dynamics. Finally, it is critical to note that obtaining commitment does not guarantee action. Follow-through on commitments relies on government entities having the capacities and financing mechanisms available to catalyze action.