Beyond inclusion: practical lessons on striving for gender and disability transformational changes in WASH systems in Cambodia and Timor-Leste

Globally, 3.6 billion people lack access to safely managed sanitation services at home, and 2 billion people lack access to safely managed drinking water at home (WHO/UNICEF 2021). Despite progress towards increased access to water, sanitation and hygiene (WASH), WHO/UNICEF's Joint Monitoring Programme (2021) found inequalities in access to WASH based on geography, wealth quintiles and other factors. As Cambodia and Timor-Leste drive progress towards the Sustainable Development Goal 6 by strengthening WASH systems and services (WHO/UNICEF 2021), it is critical that they include, benefit and empower women and girls in all their diversity, and that they do the same for people with disabilities (Overseas Development Institute/WaterAid 2017).

There is a deep understanding of the importance of gender equality in efforts to strengthen WASH systems (Dery et al. 2020; MacArthur et al. 2020) and an increasing awareness of the importance of disability inclusion (Mactaggart et al. 2018). However, the literature on these intersecting issues is limited, and there is a shortage of studies that focus on the importance of integrating them. To support women's empowerment and gender equality, WASH system interventions must contribute to the transformation of political, economic and social systems.

This paper examines the changes to gender equality, disability and social inclusion (GEDSI) which occurred in three programmes designed to strengthen WASH systems. These findings are presented in the following three programmatic cross-country case studies:

• Shifting gender norms in rural community and households (Timor-Leste).

• Building women's WASH leadership (Cambodia).

• Government-led disability-inclusive WASH systems (Cambodia).

The authors analyse GEDSI outcomes across these programmes by drawing on the following: WaterAid's outcome and performance data obtained from its monitoring of GEDSI changes; qualitative studies of WASH system interventions; key informant interviews with participants; face-to-face questionnaires and field observations by WaterAid staff. The analysis tracks changes in behaviours among WaterAid's partners including the water user committees called Grupu Maneja Fasilidade (GMFs) and communities in Timor-Leste, and provincial and district government staff in Cambodia.

What emerges is a set of practical solutions for improving gender equality and disability rights through WASH system efforts in Cambodia and Timor-Leste. These findings are useful to WASH practitioners wishing to better understand empowerment outcomes and what works within a system-strengthening approach.

Evidence shows that poor access to WASH disproportionately affects marginalised groups, poorer households and women. Women and girls experience challenges in relation to WASH due to their practical needs such as the need for safe access to WASH services and the need to manage menstrual bleeding (Fisher et al. 2017). In addition, their strategic gender interests are affected by shifts in power, social structures and status (Moser 1989; Kabeer 2002; MacArthur 2020). Despite this, women are often left out of governance systems and decision-making processes on WASH at all levels – from households to national policy and agenda setting (Dery et al. 2020).

People living with disabilities make up 15% of the global population (WHO 2010) and 80% live in low- and middle-income countries (WHO 2010). People with disabilities experience disproportionate challenges in accessing WASH (Mactaggart et al. 2018). Barriers range from physical ones such as inaccessible toilet facilities in homes and schools to social issues such as the risk of violence, discrimination or harassment when utilising services (Groce et al. 2011). People with disabilities are often left out of decision-making processes and leadership positions due to stigma and discrimination, or because they have not had opportunities to develop technical knowledge or skills (UNICEF 2016).

Systems thinking and practice for WASH has evolved significantly in the past decade, largely in response to concerns about the sustainability of WASH services. The literature on WASH systems describes programmatic approaches that aim to bring together diverse perspectives in order to address ‘wicked problems’ – problems for which there is no easy answer, and often different start and end points based on the context – encountered in WASH service provision (Neely 2019). The literature also presents systems thinking as a means of understanding and addressing the ways in which a multitude of factors affect the sustainability of WASH services (Valcourt et al. 2020).

The broader academic literature on influencing systems refers to the need to bound a system, understand it, assess how its component parts interact and then find ‘levers of change’ (Foster-Fishman et al. 2007) or leverage points (Meadows 1999), which can be used to influence the system. Meadows (1999) describes leverage points that include system goals; the structure of information flows; the distribution of power and decision-making; norms, attitudes, values and beliefs; and the mindset or paradigm out of which the system arises. There is limited literature on GEDSI in WASH systems, despite many of these leverage points being very relevant to transformational gender and disability programming. Power, norms and attitudes, decision-making and mindsets are highly gendered in any context and are likely to reinforce inequalities (Moser 1989; Kabeer 2005; Cornwall 2016). Gender or disability, if they are referenced in the WASH system literature at all, tend to be treated as two among many ‘factors’ (Valcourt et al. 2020) affecting how a system functions, or as one potential ‘leverage point’ for change, such as improved school WASH leading to improved educational outcomes for girls (Grant & Willetts 2019). Grant & Willetts (2019) identify gender equality and inclusion as one aspect of the mindset and paradigm out of which a system arises, and give an example:

The empowerment literature, particularly in relation to gender equality, explores empowerment as an ongoing process, which enhances a person's ability to make decisions or strengthen their agency (Kabeer 2005). It also examines notions of power and decision-making (ISF-UTS 2019; Dery et al. 2020). Empowerment is seen as arising from a combination of a person having the capacity to make choices and a shifting of structural opportunities around them, so that those choices can be actioned (Cornwall 2016). As Rowlands (1997) describes it, empowerment is an iterative process, which is also about collective action and shifts in power relations away from ‘power over’ towards ‘power within’ (self-representation), ‘power with’ (social capital) and ‘power to’ (agency). It is a process in which all four power shifts are required. Carrard et al. (2013) drew on Moser's framework (Moser 1989) to describe how shifting the social dynamics around gender equality can occur within the personal dimension and/or the relationship dimension. Changes in the personal level include changes related to the roles of women and men, and changes to attitudes or self-perceptions. Changes in the relationship level include women and men changing the ways in which they relate with one another (Moser 1989; Carrard et al. 2013). Dery et al.’s (2020) review of the WASH research on empowerment found the following five common interrelated dimensions of empowerment: access to information; participation; capacity building; leadership and decision-making. The authors noted that empowerment is both a cause and an outcome of successful gender-sensitive WASH programmes. It has also been recognised that gender equality work in WASH can result in backlash, resistance and negative outcomes, which may cause harm if the gender work is not done sensitively and does not employ strong monitoring and mitigation strategies (Willetts et al. 2014). As there is limited literature on the empowerment of people with disabilities in WASH, for the purposes of this paper, empowerment literature related to gender is also applied to analyse empowerment outcomes for people with disabilities.

In this paper, we analyse findings related to gender equality and disability outcomes across three WASH system-strengthening programmes (WaterAid 2017, 2020). An internal review team was established. It consisted of three field-level equality and inclusion practitioners, who are based in Timor-Leste and Cambodia, and lead programming with partners in the community, and of two technical leads (system strengthening and equity and inclusion – E&I). The team established the following criteria to select three programme examples from Cambodia and/or Timor-Leste:

• Technical focus: Gender and/or disability as a core focus of programme outcomes.

• Strong evidence: Qualitative monitoring evidence which had been collected throughout or following implementation which demonstrates changes to inclusion or empowerment.

• WASH system levels: Represent a cross-section of the different ‘levels’ at which we seek to strengthen the WASH system and influence transformative/empowering GEDSI outcomes: household; community; sub-national and national.

Of the eight programmes reviewed, the following three were considered to have met the core criteria through a review team workshop (see Table 1 summary):

• Gender equality changes in households and communities (Timor-Leste).

• Women's leadership in the sub-national WASH (Cambodia).

• Disability-inclusive sub-national WASH system building (Cambodia).

The evidence used in this paper has been taken from WaterAid's programme-monitoring systems and processes for each programme. These evidence sources are designed on a programme level to respond to specific programme outcomes and assumptions, and are intended to respond to the contextual GEDSI issues and desired influence of the programmes. This means that while the monitoring evidence and methods for evidence gathering are based on a common framework, the Equality, Inclusion and Rights Framework (Figure 1), the methods and evidence type vary between the three programmes. See the evidence type and source in Table 1 (WaterAid, 2021).

Once these programmes were selected, the review team applied WaterAid's gender equality framework to reflect on and analyse the programmes using the following learning questions:

• What inclusion or empowerment outcomes were achieved in each programme? What is the strength of the evidence?

• What system-strengthening approaches may have contributed to those outcomes?

• What were the common and divergent themes and what do the findings mean for practice?

The results presented in this review come from WaterAid's ongoing performance monitoring of GEDSI changes within WASH system programmes. The results track changes in behaviours among WaterAid's partners including the GMFs and communities in Timor-Leste, and provincial and district government staff in Cambodia. To monitor, evaluate and learn from GEDSI changes in programming, WaterAid applies outcome mapping methods. Using outcome mapping, a WASH system programme can target a behavioural change, and it can aim to measure the contribution the change has made to an intended outcome, while recognising that other actors in the local system may also have contributed to the observed change (Earl et al. 2001). This is of particular relevance to gender and disability changes, where specific behavioural changes related to challenging norms or power can be mapped. Changed behaviours can indicate changes to the underlying mindsets, understandings and norms of individuals and groups.

To map empowerment outcomes, WaterAid developed and later revised an Equality, Inclusion and Rights Framework (2021). WaterAid's framework drew together gender empowerment theory (Moser 1989; Carrard et al. 2013; Inter-agency Gender Working Group (IGWG) 2017), to create four ‘levels’ of equality and inclusion outcomes – harmful, inclusive, empowering and transformative (see Figure 1). It aims to guide how we work and what outcomes we want to achieve in WASH system-strengthening efforts at levels ranging from the individual and the household to policy and government actors (see Figure 2) (WaterAid 2017, 2020).

In the results section, we analyse the evidence and show what inclusion or empowerment outcomes were achieved in each programme, and what system-strengthening approaches may have contributed to those changes/outcomes. These are summarised in Table 2.

Since 2016, WaterAid has led participatory gender dialogue sessions within the rural Community Action Planning for Sanitation and Hygiene (PAKSI in Tetun language), a localised version of the Community-Led Total Sanitation (CLTS) which uses disgust rather than shame, to trigger community behaviour change and structured community action planning. The approach aims to motivate households to share WASH labour more equally between women and men. The sessions facilitate dialogue about gender norms, roles and rights in support of achieving WASH outcomes. A range of facilitator tools make up five modules, which local partners and women's rights organisations implement. The tools are implemented at strategic moments alongside WASH activities that establish community water user groups – GMF. While the dialogue sessions do not explicitly encourage greater women's involvement in the GMFs, they seek to influence the societal norms that would otherwise limit gender equity in the community-level system for the management of WASH services.

To ascertain what gender equality changes had occurred in communities, a qualitative study (Grant & Megaw 2019) examined the strategic and practical gender equality changes experienced by men and women at the community and household levels. One hundred and seventy-two people across nine communities participated. As documented in Table 2, the review found that the qualitative study's findings aligned with WaterAid's ‘Empowering’ outcomes across household and community change domains of the rural Timor-Leste's WASH system.

Changes in WASH responsibilities were reported at the household level, as men and women were taking on new roles. Overall, both women and men reported ‘a lot of change’ with respect to WASH-related roles. Eight communities reported men and women being more willing to share household tasks. Men, in particular, reported that the modules gave them a newfound understanding of the work that women did in the home.

Participants also reported changes in gender roles at the household and community levels. Eight communities reported that women had become more involved in household-level decision-making. Nine communities reported a rise in the status of women, with more opportunities to fill leadership roles. Nine communities also reported that women had increased access to opportunities for decision-making.

Women in six communities reported that men supported them to participate more in village and sub-village meetings. A women's group discussion noted that because of the gender sessions ‘we are confident to participate in trainings and meetings’ and that this improved equality in access to the WASH and health information. Also, it was reported that women were taking up new roles in the GMF structure, and in technical and construction roles in WASH projects.

The research also indicated that more women now shared responsibility with men for controlling resources (such as financial and property resources) and/or had control of household finances and livestock. As stated in one women's group discussion: ‘Through recent training and activities, women can also make decisions about when to build pipes, and we can also become managers of technical finance in the community’.

The review found a lack of evidence of the avoidance of ‘harmful’ outcomes as articulated in WaterAid's framework. The qualitative study reported some limitations: the need to go deeper into identifying gender equality changes; unintended consequences were not measured such as backlash or increasing women's work burdens and prevailing social norms made change slow and difficult.

In Cambodia, WaterAid's gender framework informed a Gender Power Analysis and a participatory process to identify barriers and outcomes for equality in the WASH system. Both processes identified limited women's participation and representation in WASH sector decision-making and limited support to women from their male colleagues and supervisors as key barriers to equity in the WASH system. In response, WaterAid developed a Women's Leadership Training Program in Kampong Chhnang to strengthen women WASH professionals' networks and leadership confidence. Twenty women from different government WASH-related departments participated in the programme which aimed to improve leadership skills and confidence on WASH topics and brought women leaders together in a peer support model. The programme facilitated a series of six sessions over a period of 6 months (December 2019–July 2020).

The programme's contribution to outcomes and changes were recorded from:

• Participants' self-reflections during sessions (recorded by facilitators).

• Formal feedback at the end of each session to inform training rounds.

• Facilitators' reflections back to the WaterAid team.

• Interviews with participants led by WaterAid at the end of the programme.

• Follow-up interviews with 11 participants 1 year after the training.

As documented in Table 2, the review found that data from the programme's real-time monitoring and follow-up interviews aligned with WaterAid's ‘inclusive’ outcome. A range of inclusive outcomes were documented across individual people connected to the sub-national level of Kampong Chhnang's WASH system.

During the course, participating women self-reported that they had increased capacity and more confidence to express opinions around male colleagues. They felt they had increased their leadership skills and improved their public-speaking abilities. Some women reported at the later stages of the training that they had greater support from their supervisors and some even reported promotions (although this may not be attributable to the programme).

In an interview conducted immediately after she attended the leadership course (June 2020), one participant, who is a Deputy District Governor, reported having the confidence to take on a bigger role in the effective functioning of an inclusive WASH Committee. She reported that since taking part in WaterAid's leadership programme, she confidently leads meetings with relevant people to make collective decisions on WASH. She now led the quarterly and monthly district WASH Committee meetings with members and commune councillors to check and review the progress of the WASH plan implementation:

Real-time learning data from facilitators' reflection notes documented changes in women's confidence, knowledge and skills. One facilitator reported that at the end of the programme, a female participant said: ‘My confidence increased after the leadership training’. Another facilitator reflected following the training that rather than the theoretical study, there needed to be practical learning which met the needs of women and their own critical consciousness ‘especially building more on Power with, Power Within, and Power To’ (WaterAid real-time monitoring, November 2019).

Some participants reported that improvements to their knowledge and skills helped empower other women they worked with in the community. One facilitator noted that the sessions were ‘building a group of supportive women’. However, another facilitator reflected that there had been limited progress in building a strong collaborative platform.

One end-of-training reflection noted that the course was a pathway that was ‘important in [enabling] transformation [for] women to move up [to] the next level of their leadership and capacity at workplaces’. However, other real-time data revealed challenges to the ways in which women applied leadership skills within their organisations. One facilitator reflected that an overall learning from delivering the programme was that there is a lack of pathways for career progression and role models in workplaces and that this ‘really restricts women's efforts to develop facilitative leadership style and use of their leadership potential no matter how much they try to break this glass [ceiling]’.

Another facilitator's reflection highlighted the barriers created by existing leadership styles. The facilitator reported that half of the female participants rated their organisation's leadership style as ‘dictatorial’ and the majority of the participants reported that their preference was for ‘facilitative’ leadership.

One year after the training, WaterAid undertook follow-up qualitative interviews using the Kirkpatrick Model (Kirkpatrick & Kirkpatrick 2016) to assess how the programme had contributed to gender equity in WASH decision-making. The Kirkpatrick Model was used to design interview questions based on the desired changes to behaviours and outcomes identified in the training programme plan. All training participants were invited to be interviewed with 11 of the 16 participants agreeing to an interview. All participants reported that there had been at least some change in their knowledge or their capacity, or that they had put their learning into practice. Many women reported using what they learnt to advocate for women's participation in the community level and formal meetings. One woman reported advocating for men to become champions for women.

The most common change reported was an improved confidence to speak up and share ideas. Most (8/11) participants reported that their participation in decision-making had changed. This included being more involved in decision-making, raising their voices and being heard in meetings, and positive improvements in participation generally.

Most women (9/11) reported they had observed other participants being strong role models, for example, by contributing ideas in meetings and fulfilling their roles. However, fewer (4/11) participants reported that other women had changed their participation in decision-making. Where it was reported, observable changes included an increased confidence to speak and/or contribute to decision-making. Few participants also reported making progress against the personal goals they had set in the training. Only one respondent reported their goal had been achieved and four reported making some progress.

Participants reported a wide range of unexpected results from the training, including changes in WASH knowledge and mobilisation of resources for WASH. Several women (4/11) observed that their increased confidence had noticeably improved their community WASH promotion work which was helping to strengthen community WASH knowledge. Two of these respondents especially noted that conveying WASH knowledge to women in the community was fundamentally about supporting women to understand their rights and to improve their decision-making role in their family. One participant noted their improved confidence to collaborate with commune leaders was helping to mobilise both commune budget for WASH and also increase philanthropic contributions from better-off community members.

The review of these findings also found a lack of evidence of avoiding ‘harmful’ outcomes as articulated in WaterAid's framework. Three women reported that there are ongoing barriers to women's ability to contribute, such as women not being promoted or listened to, and an inability for women to decide things in meetings.

To address the limited operationalisation of Cambodia's National Inclusive WASH Guidelines, since 2018, WaterAid has partnered with the Rolea Bier District Administration and Organisations for People with Disabilities (OPDs) to create an inclusive WASH learning hub to showcase practical implementation of the guidelines.

The inclusive learning hub adopted a range of approaches including:

• OPD-led orientations on the guidelines.

• Participatory trainings.

• Accessibility audits at homes and public institutions.

These efforts aimed to build the Rolea Bier District WASH Committee's understanding of the experiences and barriers encountered by people with disabilities in their WASH access.

The review analysed monitoring data gathered through real-time documentation of changes referred to as ‘Inclusive’ WASH outcomes under WaterAid's Framework, as captured in Table 2. The data that were analysed came from WaterAid staff field-monitoring reports, workshop notes, partners' reports and anecdotes shared by district government partners in online learning forums. Collectively, the evidence from these sources shows what action government counterparts began taking to increase the inclusion of people with a disability.

Evidence from the monitoring reports shows the following:

• In late 2018, WaterAid and the District WASH Committee used field trips to showcase existing accessible infrastructure as a method to inspire and trigger change. The monitoring report states, ‘After training the participant created their own plan to build toilet for all in their own district. They will target three communes, three pagodas, and their district office’. (WaterAid real-time monitoring, October 2018)

• Three months later (February 2019), a (non-disability focused) system-strengthening workshop found ‘evidence that the recent E&I training has influenced people's perceptions in Kampong Chhnang. When the district WASH committees were asked to draw their vision for their district in 2020 most of the toilets drawn were accessible [to people with disabilities]’. (WaterAid real-time monitoring, February 2019)

• In November 2019, the District WASH Committee learning on accessibility standards had led them to conduct an inclusive CLTS process. There was greater diversity of participants in the process, including people with disabilities, older people, pregnant women and women from women-led households; the field note reflected that this was a change from the past when only a few older people and no people with disabilities joined.

• Throughout 2020, there was evidence of a growing commitment on the part of the District WASH Committee to address disability-inclusive WASH. The Committee organised training on inclusive WASH standards, inclusive principles and construction monitoring to other government staff, religious leaders and toilet construction businesses (February–March 2020). One real-time monitoring report (November 2020) captures how the influencing components worked: ‘They also visit each commune to follow up and encourage all commune offices to have accessible toilets at their community especially at public space.…This is a role model, and hope [it will] have a positive influence on other communes in Kampong Chhnang’ (WaterAid real-time monitoring, November 2020). As a result, three commune councillors (Kork Banteay, Krang Lvea and Pong Ro) used their own budgets to build accessible toilets with technical support from the District WASH Committee.

• In March 2021, the Rolea Bier District WASH Committee had led two inclusive WASH activities by themselves, demonstrating that the Committee members had taken on the responsibility of implementing their own inclusive WASH plans based on their willingness and capacity. However, one challenge reported by this note was that the Committee had limited engagement of marginalised groups in its activities.

Observed changes in the mindsets and priorities of the district administration as they became the inclusive WASH champions appear to have led to public investment in inclusive WASH facilities at public offices, healthcare centres and pagodas, and the mainstreaming of inclusive principles in WASH planning and community activities. Other districts have approached Rolea Bier or the Rolea Bier District Administration to learn from their example.

The review found that two of the three GEDSI-related WASH system initiatives might have contributed to inclusive outcomes; one might have contributed to empowering outcomes and as yet none demonstrated evidence suggesting transformative outcomes. The review also found limited evidence of initiatives, ensuring that there were no harmful outcomes. The key themes and their implications for practice that emerged from the review were as follows.

Both the Timor-Leste gender dialogue sessions and the women's leadership development programme in Cambodia sought to empower women within the WASH system by building two dimensions of empowerment: ‘power within’ (whereby women recognise their own power) and ‘power with’ (whereby collective action is built to achieve social change) (Cornwall 2016).

The analysis of gender module sessions in Timor-Leste found evidence of changes to women's strategic interests (Moser 1989) including confidence to participate in meetings; to contribute more to community decisions about WASH and to participate more actively in the GMFs that were formed to manage the water systems. At the household level, the evidence demonstrating a shift towards men taking on more WASH responsibilities in households represents a change in the demands placed on women (Moser 1989). This reflects progress towards WaterAid's programming outcomes targeted through the approach of dialogues between men and women. Since women have traditionally been expected to take on the role of WASH work in the home, the evidence from eight communities reporting men and women being more willing to share household tasks is a positive change in social dynamics (Carrard et al. 2013).

The qualitative data showed that broader community members wanted to support women to speak up and find practical ways to enable women to participate more. This is an example of collective power (Batliwala 2007) and shows evidence suggesting that the gender dialogue approach led communities to adopt a ‘power with’ approach (Rowlands 1997).

A review of the data of Cambodia's leadership programme gathered from women and facilitators during and immediately after the course, as well as 1 year later, revealed that women felt more confident in themselves. For example, the leadership courses focused on public speaking and self-esteem, and women participants and the facilitators reported that their confidence to speak out and raise their voices in the decision-making processes had increased. This reflects a positive change of ‘power within’ (Rowlands 1997). It also aligns with the notion that empowerment is a process of building self-esteem (Cornwall 2016). The programme's empowerment changes demonstrate ‘Capacity Building’, one key component of Dery et al.’s (2020) empowerment outcomes in WASH.

Both the Timor-Leste gender dialogue sessions and the inclusive WASH learning hub in Cambodia sought to address systemic barriers within the WASH system's enabling environment – norms, attitudes and entrenched decision-making structures – which limit empowerment.

The evidence from Timor-Leste demonstrates a contribution to changing men's attitudes and actions, which is a critical part of empowerment work. As Sen (1997) describes it, the prevailing social norms also have to shift in order for women to be able to act on their critical consciousness, which requires space and an enabling environment for this change to be supported. The review found an attitudinal change reflected empowerment outcomes (Cornwall 2016) as evidenced by the qualitative study's documented changes of men's attitudes and actions, such as supporting women in village meetings, sharing household decision-making and having a better understanding of the gendered vision of labour.

The review found in Cambodia changes in disability-inclusive practice were not empowering of people with disabilities themselves. Rather, the changes were to the political commitment of district authorities and to the actions taken by existing leaders to make WASH more inclusive. So, the approach did not shift power relationships; however, it could be seen as an example of ‘power with’ (Rowlands 1997) if these efforts are viewed as part of collective action (Batliwala 2007). The evidence of the District WASH Committee independently training others on inclusive WASH standards and advocating for greater prioritisation of inclusive WASH services by other WASH actors may indicate the beginning of collective support among WASH service providers for people with disabilities.

The review also found evidence in Rolea Bier of ‘mindset changes’ among WASH actors (Grant & Willetts 2019; Neely 2019) including changing attitudes towards people with disabilities and responding to their specific WASH needs. This was evidenced by the provincial and district governments regularly meeting with the ODP and building the capacity of marginalised groups, especially persons with disabilities. They did this by listening to their voices, conducting public forums, and visiting places where marginalised groups would otherwise not have been listened to at all.

The Timor-Leste and Cambodia case studies demonstrate that even where empowerment interventions intended to change gender attitudes at a community level (Timor-Leste) or among one cohort (women leaders in Cambodia), such change is slow due to prevailing social norms. Persistent cultural and societal norms (such as attitudes towards women) must be addressed in order to achieve more empowering outcomes for women and people with disabilities in decision-making.

The review team identified that the key to moving from inclusive to empowering was a holistic approach which focused on strengthening the confidence and self-belief of women and/or people with disabilities at an individual level while simultaneously tackling the systemic barriers preventing their empowerment (Cornwall 2016). Among the three case studies, the Timor-Leste gender dialogue sessions were found to be contributing to empowering outcomes, while the women's leadership and inclusive learning hub in Cambodia were contributing to inclusive outcomes.

The Timor-Leste gender dialogue sessions shifted women's understandings of their ability to participate in decision-making and also challenged wider community attitudes to gender norms. This combination led to changes in gender relations and roles.

The two interventions in Cambodia did not attempt to achieve empowerment outcomes with such a holistic approach and consequently did not achieve meaningful empowerment outcomes for women or people with disabilities.

Women who participated in the women's leadership training in Cambodia reported feeling more confident, and having improved knowledge within themselves, but have not quite reached empowerment achievements such as exercising leadership or greater power (Kabeer 2005). Some of the participants even named ongoing barriers in their environment which meant that their confidence did not translate to more meaningful participation in WASH decision-making. Barriers such as women not being promoted, or listened to, or having access to decision-making authority, are all examples of limiting paradigmatic factors (Foster-Fishman et al. 2007; Cornwall 2016) and relational issues that are part of the women's local WASH system. This shows that in order to create meaningful change and personal transformation, women's leadership efforts in WASH systems need to maximise power to and power with (Rowlands 1997). The leadership programme addressed only ‘power within’. Batliwala (2007) describes collective action as being essential to the process of empowerment. The leadership programme promoted the creation of role models of women who shared ideas. However, without the engagement of male champions this did not enable women's role to extend to decision-making or collective action between men and women.

Conversely, the Rolea Bier inclusive WASH learning hub changed attitudes towards people with a disability among WASH decision-makers. However, while local ODPs were involved in building knowledge and awareness, the limited leadership and ownership by people living with disabilities of the ongoing roll-out of the inclusive WASH guidelines in the district missed opportunities for meaningful participation (Dery et al. 2020) and true collective action with a shift in power relations from ‘power over’ to ‘power with’ (Rowlands 1997). There was little evidence throughout the real-time learning data to demonstrate a shift in power within and power to (Rowlands 1997). The four requirements from Dery et al. (2020) – participation, capacity building, leadership and decision-making – were not met. The review found there was little evidence of self-empowerment of people with disabilities due to the focus of the activities being on the WASH systems and actors rather than on individuals or groups themselves.

These case studies reinforce the framing in the literature of empowerment as a process which requires both shifts in consciousness and changes in culturally embedded normative beliefs and ideas (Cornwall 2016).

The review of outcomes across the three case studies found changes at different levels of WASH systems. It highlights that integrating gender, disability and social inclusion outcomes within system-strengthening programming provides useful opportunities and leverage points for addressing systemic barriers to inclusion and empowerment.

Building commitment for gender equity and disability inclusion among those in positions of power and authority within the WASH system can contribute to shifting norms and behaviours, as demonstrated by the Rolea Bier district decision-makers regarding disability-inclusive WASH. Changing fundamental aspects of the system such as norms, mindsets and paradigms, especially at the societal level, is known to be difficult (Meadows 1999; Foster-Fishman et al. 2007). Working with active change agents (Meadows 1999) to change how system stakeholders see the world, in this case by encouraging visions of more inclusive services, leads to changed practices and behaviours and eventually alters the system context (Foster-Fishman et al. 2007), both for WASH and for other public services. Early evidence that this is occurring in Rolea Bier includes the District WASH Committee influencing communes, pagodas and other public facilities to the extent that they utilised their own budgets to build accessible toilets, which is itself a positive outcome for inclusive services. Notwithstanding the limited leadership by people with disabilities themselves, the District WASH Committee's increasing engagement of diverse stakeholders and perspectives to solve the multiple problems (financing, accessibility and inclusive participation) facing WASH services in their local system shows the district authorities are beginning to demonstrate more participatory leadership (Dery et al. 2020) and use their ‘power to’ and ‘power over’ to advocate for more inclusive practice among other actors in the system.

While prevailing norms may change slowly, community gender dialogues in Timor-Leste demonstrate that there can be key moments of WASH system changes when introducing equity, and inclusion initiatives can have rapid impacts on the paradigms within which WASH systems function (Meadows 1999). In the Timor-Leste example, the creation of GMFs for managing water infrastructure provided opportunities for community collective action (Rowlands 1997) and new power relations within the new decision-making structures (Meadows 1999). An initial shift in social norms occurred through facilitating community dialogues. This process contributed to the emergence of women role models and demonstrated women's proficiency in decision-making and resource management. It is an example of ‘insert[ing] people with the new paradigm in places of public visibility and power’ (Meadows 1999, p. 18). It is also an example of gender empowerment as both a cause and an outcome of successful gender-sensitive programming (Dery et al. 2020) and of the creation of positive feedback loops that enable women's more visible leadership in WASH to reinforce new social norms (Meadows 1999).

This review did not find sufficient evidence to assess whether greater gender equity or disability inclusion can create a stronger WASH system. In both Timor-Leste and Cambodia, positions of formal power and authority in WASH systems continue to be dominated at all levels (household, community, district and national) by able-bodied men. It would be useful for future research to consider whether WASH decision-making, planning and service provision that is more equitable, participatory and representative of service users, performs more strongly on system metrics such as service access, accountability, equality, transparency and sustainability.

As well as the programmatic limitations mentioned in sections ‘Results’ and ‘Discussions’ (such as limited monitoring of potentially harmful consequences of gender programming and the need to deepen the monitoring of gender equality changes), the authors note two key limitations of the study design and methodology:

• 1.

  The sources of data in all three case studies did not include baseline or data collected from control groups. Therefore, changes observed cannot be fully attributed to WaterAid's programming, but as a contribution towards social changes. These findings need to be considered as one potential explanation of how the change has happened and is not representative of all empowerment outcomes in WASH systems.

• 2.

  Only the qualitative evidence of the gender equality changes in households and communities in Timor-Leste was collected by independent reviewers. The evidence from the programmes in Cambodia may be affected by social desirability bias by participants in their responses to the women's leadership interviews. The real-time observations made by the WaterAid staff in documenting the inclusive WASH learning hub may have been biased towards particular programme targets and objectives. However, this bias was mitigated through triangulation. In Cambodia, the data were collected on a broader set of outcome questions. The monitoring strategies aimed to measure broader outcomes and it was a participant-led approach therefore some potential self-reported bias was mitigated. For example, the follow-up surveys encouraged participants to reflect generally on what changes they had observed and we later analysed based on what outcomes we sought to harvest evidence of.

These limitations mean the study's analysis and results are not representative of all contexts and WASH systems’ outcomes. There is a limited ability to apply these findings to other settings, programmes and monitoring systems. The difficulties in attributing change, as well as the analysis and interpretation of findings, being internally led means the findings need to be interpreted from a practitioner lens. The authors are not suggesting that these results represent a complete narrative of how our work has shifted norms in WASH systems. The authors sought to understand the extent to which our programming has been effective against our own objectives/targets. In future, we recommend practitioner-led research to measure impact having quality baseline data to improve rigour and improve triangulation to limit social desirability bias.

The following recommendations are for practitioners addressing GEDSI in WASH system-strengthening and based on the review team's findings of limitations and areas for improvement in the above case studies:

• WASH system-strengthening interventions can leverage strategic entry points to address systemic inequalities

While WASH systems’ thinking evolved in response to the sustainability challenges of WASH services, it has great potential to reorientate prevailing power structures and harmful attitudes. The leverage points and factors which system-strengthening interventions seek to influence – power, norms and attitudes, decision-making and mindsets – are highly gendered in any context and are likely to reinforce inequalities (Cornwall 2016). Applying a GEDSI lens such as WaterAid's ‘Equality and Inclusion Framework’ when bounding, understanding and assessing the WASH system, and designing interventions, can reduce inequalities and move from inclusive outcomes towards transformative GEDSI outcomes.

• To achieve transformative change, GEDSI interventions need to shift individuals' self-perceptions and also address attitudes and norms within the broader WASH system.

Among the three case studies, the most empowering outcomes were demonstrated when interventions raised critical consciousness and engagement (Batliwala 2007) and normative beliefs (Cornwall 2016). Increasing women's confidence and skill-building are not sufficient for empowerment if women still face cultural barriers and negative attitudes. Similarly, changing attitudes among WASH system leaders is not sufficient for empowerment if it is not complemented by empowering participation and collective action. Changing four dimensions of power – ‘power within’, ‘power with’, ‘power to’ and ‘power over’ – is necessary for meaningful empowerment outcomes.

• Employ a variety of methods to generate collective evidence to monitor deeper empowerment changes.

While the evidence in the qualitative study of Timor-Leste's gender dialogues was strong, one qualitative study finding was that programmatic monitoring needed to go deeper to identify empowerment changes. For example, there was a lack of indicators in WaterAid's programme monitoring to measure decision-making changes. The review team noted that, had the qualitative study not been done, regular programme monitoring would not have documented changes to decision-making. In the Cambodia examples, systematically monitoring indicators could have improved tracking of changing attitudes and marginalised groups participation and thus informed programme adaptations.

Overall, the review found that examining empowerment outcomes through analysis of real-time monitoring and learning data from Cambodia related to both disability and gender programming was an effective method to understand small changes over time. In contrast, as a ‘point in time’ study, the Timor qualitative study provided deeper insights which can be used to reorientate programming. The review team recommends that a mixture of both methods would be best to analyse and understand empowerment changes across the levels of WaterAid's E&I framework.

• Actively monitor and respond to harmful outcomes in GEDSI and system-strengthening programming.

While inclusive or empowering outcomes were reported in both Cambodia and Timor-Leste case studies, a common gap was that the ‘harmful’ outcome articulated in WaterAid's framework was not strongly mapped and tracked; therefore, the ‘harmful’ outcomes of WaterAid's framework may not have been fully avoided. The literature shows that resistance and backlash are common in gender equality work (Willetts et al. 2014), and it is strongly recommended that all the gender and WASH programming in system strengthening has practical tools to map, monitor and mitigate any unintended harm.

A review of three field approaches in Timor-Leste and Cambodia demonstrates the evidence of WASH system-strengthening initiatives providing an opportunity to move beyond inclusive outcomes towards achieving empowering outcomes. The review found that a system approach to WASH which integrates GEDSI can help to address the systemic normative and attitudinal barriers which prevent equality and inclusion. To achieve empowering outcomes, WASH system-strengthening actors must seek to shift the self-belief of women and people with disabilities while at the same time addressing the wider norms and attitudes to GEDSI which may present structural barriers to empowerment.

Funding for WaterAid to undertake the programmes reviewed in this paper was provided by the Australian Government's Department of Foreign Affairs and Trade/Australian Aid through the Civil Society WASH Fund and Water for Women Fund (Timor-Leste) and the H&M Foundation through the SusWASH Program (Cambodia). The authors wish to acknowledge the community members and participants who contributed their knowledge and information to this study, as well as the contribution of the ISF-UTS who led the gender dialogue review in Timor-Leste (Grant & Megaw 2019).

All relevant data are included in the paper or its Supplementary Information.