A de jure study of social accountability for water and sanitation services in Uganda

Despite intense focus on the improvement of access to water and sanitation services over the past decade, achieving universal coverage by 2030 (one of the Sustainable Development Goals (SDGs)) will require a quadrupling of current rates of progress in safely managed drinking water, safely managed sanitation, and basic hygiene services. An increasingly challenging aspect of attempts to deliver this goal in low-income countries is the provision of effective post-installation or post-intervention support. For example, while most rural water supply and wastewater management systems are funded and constructed by donors and governments they are left in the hands of users like community-based management groups for maintenance (Horvey & Mukongo 2020). Limited post-construction monitoring and poor quality installations lead to frequent breakdowns and undesirable second-order impacts (e.g. water leakage due to limited infrastructure maintenance not only affects water quality but also water security) with field data suggesting that failure rates can be as high as 15% of installed units after the first year of installation and 25% after the fourth year of installation (Banks & Furey 2016). Poorly planned and financed maintenance programmes (Nagpal et al. 2019), corruption (Pusok 2016), poor governance (Samuel et al. 2021), and lack of both capacity and standards (Oduro-Kwarteng et al. 2015) are frequently cited as contributing factors to the fragility of water and sanitation interventions.

Debate surrounding corrective adjustments to improve the durability of water and sanitation interventions has increasingly engaged with concepts of social accountability (Asís et al. 2009; Jiménez et al. 2018). There are varying opinions on how broadly or narrowly social accountability can be defined. Some have emphasised the role of citizens in demanding greater accountability (Baez-Camargo & Jacobs 2013), while others have emphasised the steps taken by the state to fulfil its obligations (Joshi & Houtzager 2012). Alternative perspectives that infer citizen-led engagement while downplaying the role of civil society-led engagement (e.g. Halloran 2016) are too nuanced to be a useful point of departure and our adopted characterisation (broadly aligned with that proposed by both Grandvoinnet et al. 2015; UNDP/UNICEF 2015) acknowledges that both citizen action and state action enhance social accountability. For the purposes of the study reported below, we define social accountability as ‘Actions taken by either citizens or the state which seek to hold state actors accountable for their (collective) commitments’.

Social accountability, as a desirable trait of social improvement initiatives and as a useful metric is increasingly being used to reinforce conventional forms of accountability such as political checks and balances and financial auditing systems. It has been the subject of growing interest from donors and non-governmental organisations (NGOs) as they attempt to impose greater responsibility and answerability to the poor (World Vision International 2019). Generally, social accountability mechanisms can increase citizen empowerment as they allow vulnerable and disadvantaged groups to gain access to information about their rights and entitlements. Second, they can improve governance processes and outcomes, as demonstrated through increased state effectiveness triggered by the ability of the poor to go beyond mere protest to also engage with politicians and bureaucrats in a more systematic and organised manner. Finally, they can improve development effectiveness realised through citizen empowerment, promoting dialogue between policy-makers, citizens, and service providers, and providing incentives for improved performance (Malena et al. 2004).

It is only relatively recently that the importance of providing an effective enabling environment within which both citizens and the state can formalise accountability expectations and shape delivery processes has been highlighted within a development context (Fox 2015). The embedding of suitable principles and mechanisms in legal instruments and government policies is one way in which accountability can be promoted and effected. It also provides a set of reference points which citizens, advocacy groups, and social justice groups can use to pursue failing administrations through the courts (Herrera & Mayka 2020). Turning ambition and pledges into prescribed functions and processes, if accompanied by transparency, constitutes a model for deeper social engagement. The lack of constructive participatory frameworks, whereby stakeholders are given a voice and informed about developments, is problematic and has been shown to generate destructive outcomes and sustained political frustrations (Kanyane et al. 2020). Federal and regional authorities are thereby restoring and renewing part of a wider social contract which calls for the peoples' representatives to be answerable for their actions. While, as we shall see, the form of obligations can range from providing opportunity (you can do this) to requiring compliance (you will do this), the intention is commonly virtuous.

A variety of social accountability tools have been used in development contexts with those that directly involve citizens in allocation, disbursement, monitoring, and evaluation of public resources having proved particularly effective (Malena et al. 2004). However, it has been noted that success in one context does not guarantee similar effectiveness in another (Joshi & Houtzager 2012) and that the deployment of multiple tools in an integrated cycle could yield better results (Janmejay & Parmesh 2003). Generally, the success of most social accountability approaches has been hindered by poor understanding of the contextual suitability of a tool (Gaventa & McGee 2013) and the limited responsiveness of service providers (Joshi & Houtzager 2012).

The recent history of social accountability initiatives in the world's poorest regions has seen the targeting of high priority challenges such as health and education as well as the development of new concepts and mechanisms reported through the academic literature. Water and sanitation as a target for new developments has, in one respect, been left behind. Data from the Scopus database between 1997 and 2019 suggest that for every contribution on social accountability in the water and sanitation sector there are over 15 on education or health.

In the context of water and sanitation services provision, evidence-based studies have indicated a range of benefits resulting from social accountability as a route to good service delivery and enhanced water security (Malena et al. 2004) for both already empowered and marginalised groups (Holland & Schatz 2016). The significant role which trust plays in mediating social accountability has recently been exposed in the context of water provision, demonstrating how it is partly a natural outcome of long-term relationships between customers and service providers (commercial or public) and partly an outcome of the ease with which customers can communicate with providers and are empowered to influence decision-making (Munro & Kweka 2021). In a study with interesting parallels to the work reported below, Dhungana et al. (2021) found that social accountability tools provide a platform for water users to participate and deliberate on issues related to the execution of water and sanitation schemes, thereby enhancing the legitimacy of interventions. Their observations that (i) the concept of deliberation and downward accountability does not necessarily match with local power relationships and cultural norms and (ii) the pursuit of accountability through formal procedural mechanisms can be elusive where the capacity for judgment and sanctioning are weak, offer a useful counterpoint to a de jure-focused assessment covering regulatory and legal expectations. The dynamic between de jure and de facto measures and processes is a relatively uncharted dimension of studies into accountability. Where it has been directly addressed, primarily with respect to institutional or organisational accountability, we note that there has been a tendency to view the relationship as causal (de jure requirements triggering de facto interventions) rather than co-evolutionary. So, while the connections between the two forms are often recognised and characterised in detail (e.g. Mechkova et al. 2019) the dependencies are, in our view, overplayed to the detriment of a more nuanced understanding.

Previous work on the levels of accountability in low-income country water and sanitation contexts has found evidence of concerning differences between the availability and effectiveness of accountability in urban and rural settings (Jiménez et al. 2018). The often-limited capacity of service providers in rural contexts is retarding progress towards SDG realisation and incentivizing the development of new service delivery frameworks. Interestingly, this relative deficit of accountability in rural environments has been independently observed in other sectors such as health and education (King 2015).

In what follows we investigate the extent to which social accountability principles are reflected in the laws, regulations, policies, and other statutory provisions of a low-income country which has made good governance a political imperative. In doing so we expect to be able to draw conclusions about the breadth and force of such commitments and provide an interpretive frame for exploring potential reasons for evidenced cases of poor and strong social accountability. The contribution will also provide an evidence base against which future studies of practical accountability enhancement initiatives might be benchmarked.

Despite possessing vast water resources and land, Uganda is a country that struggles to provide adequate water supply and sanitation services for its citizens. Currently, Uganda's water supply coverage stands at 79% in urban areas and 69% in rural areas with some districts achieving only 35% coverage (Ministry of Water & Environment (MWE) 2019). Currently, safely managed sanitation is provided for 7.1% of the population in rural areas and 42.8% in urban areas (MWE 2019). Several types of institutional actors have some responsibility for Uganda's water and sanitation sector including the following: Ministry of Water and Environment (MWE), Ministry of Health (MoH), Ministry of Education and Sports (MoES), Ministry of Local Government (MoLG), National Water and Sewerage Corporation (NWSC), National Environment Management Authority (NEMA), Local governments (urban and rural), Community-Based Organisations (CBOs), and NGOs. Urban areas mostly rely on piped water supply and sewerage services provided by the NWSC. Smaller towns are provided for by regional umbrella authorities such as those serving the southwest, the Karamoja district, and the central area. Rural areas rely on point water sources such as shallow wells and spring wells whose operation and management are entrusted to community water user committees under the local government that works with government ministries (MWLE 1999; MWE 2019).

Progress in extending water supply services has been generally minimal. Over the 6 financial years to 2018, coverage has increased by only 5% in rural areas and 6% in urban areas (MWE 2019). Operation and maintenance for rural water supply has been greatly hindered as the current community-based approaches are yielding limited results (MWE 2019) and water and sanitation has been repeatedly allocated relatively low budgets compared with other sectors (MWE 2019). Progress on sanitation access has been particularly hampered by a non-legally binding memorandum of understanding signed in 2001 between MWE, MoES, and MoH. This has not only limited accountability and fragmented responsibility but has also reduced sector funding (MWE 2019).

While many of the root causes of fragile water and sanitation interventions have been observed in Uganda (Kwiringira et al. 2021), the need for good accountability and good governance has been a publicly highlighted priority in the state over recent years, it being one of the few in Africa to place combatting corruption high on the development agenda (Jacobson et al. 2010). Anti-corruption and good governance strategies have been set at national level under the MWE including establishment of a Good Governance Working Group (MWE 2016) and complaint handling mechanisms through a regulatory status revised clients charter (MWE 2018a, 2018b). At the community level, several social accountability initiatives have been implemented including ‘Baraza’ public meeting programmes (Kyohairwe 2014), score card reports (Care 2013), and crowd-sourcing ICT initiatives (Grossman et al. 2018; Van Campenhout et al. 2018).

Provisions to report irregular, illegal or corrupt practices have been made available via The Whistle Blowers Act (2010). In the midst of all these initiatives, several accountability challenges have been reported within the country's governance framework such as the lack of appropriate institutional arrangements, overlapping of roles and responsibilities at national level, poor coordination among state actors, non-professional staff at national and district level, limited district engagement in project procurement processes, and limited public participation in decision-making (Civil Society Budget Advocacy Group 2015; MWE 2019).

Our agenda for enquiry has analogues in, and is complementary to, the UN global analysis and assessment of sanitation and drinking water (WHO 2015). The research approach comprised a content analysis of documents containing national legislation, policies, and plans. The assumption underpinning the value of the evidence base being that a commitment to social accountability would leave a clear trail of obligations and responsibilities in governance principles and mechanisms. This search for de jure evidence provided a focus on officially sanctioned functions and processes. Documents (all available in English and mostly accessed via websites hosted by Ugandan ministries and the Uganda Legal Information Institute) were selected for inclusion in the study if they contained measures of direct relevance to water or sanitation services provision (see Table 1).

Document analysis was conducted using a modified form of the national de jure accountability framework developed by Water Witness International (Hepworth et al. 2020). We selected the following four accountability aspects for analysis: (i) appropriate rules, norms, processes, and standards are in place, (ii) responsibilities and duties are clearly assigned, (iii) performance of duties is monitored, reviewed, disclosed, and explained, and (iv) corrective measures, incentives, or sanctions are imposed where necessary and are effective. We evaluated these aspects across 21 key functions in the water and sanitation sector (Table 2).

Between one and three reflective questions were used to assess the extent to which each accountability aspect underpins the key functions. So, for example, in the case of the aspect ‘roles, processes, and standards’, one of the questions (relevant to all the key functions) was ‘Are the rules, processes and standards for delivery of this function clearly spelled out?’ A score between 0 and 2 was allocated to each answer with a score of 0 indicating that no evidence could be identified in the target document, a score of 1 indicating some evidence, and a score of 2 indicating strong evidence. Where no evidence was found this typically meant that a particular aspect was not covered at all in the target document. Score allocation was undertaken by one of the authors with a small sample (10%) being independently cross-checked by a second author as part of a QA protocol. The analysis generated 231 data points in total.

We report the results of the document analysis by accountability aspect as described above. Cell numbers in the tables below indicate the number of key functions for which there was strong, weak, or no evidence with respect to the question (A, B) and context (urban/rural). Column totals indicate the sum of an evidence class across questions and the profile of totals thereby provides a measure of how credible the de jure commitments to accountability are for a particular aspect across a collection of relevant legal statutes.

Overall, we can see significant evidence for the provision of rules, processes, and standards in the legislation (Table 3). Representative areas of weakness for this accountability aspect in rural contexts include the following: a lack of detail on how future demand for services is to be monitored, lack of clarity around whether districts are to include service levels and targets within their plans, and no provision being made for social impact assessment. For urban contexts, we see poor definition around complaint handling, a focus on safe water resources management rather than safe water supply, lack of compliance requirements for emergency sewage discharges, and a lack of detail as to the service levels provided by water authorities.

Multiple legal instruments have been used to set out requirements and expectations across both urban and rural contexts. An obvious area of weakness, however, based on the metrics used in this study, is the co-development and review of regulatory interventions with affected stakeholders. We also note that functions such as the collection of water fees and charges in rural areas have no transparency safeguards in place to prevent mishandling of funds and that the participation of stakeholders in the formulation of rules and procedures is weak, despite the existence of policy committees consisting of national and district representatives with a duty to advise on policy. Policy committees are required to provide advice ‘upon request’ by the minister according to the Water Act (1997). This would imply that, if not requested, the policy committee may not participate in the scoping and wording of legislative material.

Evidence of clearly laid out responsibilities and duties was also generally strong across both urban and rural related legislative tools (Table 4). The few areas of weakness exposed largely related to lack of clarity and process. For example, the regulations for setting water charges are specified within the Water Action Plan, 1995 and Water National Policy 1999 but these hold no legal status, and it is therefore unclear whether water charges are to be approved by an official or not. The legislation does not provide clear procedures on how complaints, incidents and problems are to be managed by regional and local councils and the Public Health Act (1935) does not prescribe standards for sanitation provision in institutions like hospitals and schools.

Here again, there is substantial evidence for effective legislative support. Responsibilities and duties are effectively and clearly described across a large number of legal measures – although slightly biased towards urban compared with rural contexts. A significant weakness in accountability in urban areas is the overlapping functions relating to the setting of fees and charges. According to the Water act 1997, it is required that charges are set by water authorities upon approval by the minister while according to the National Water and Sewerage Cooperation Act of 1995, the NWSC is given the power to set fees and charges for its services. This is a confusing mandate which could compromise the pursuit of accountability.

Performance monitoring is somewhat sparingly covered in the documents reviewed, for both urban and rural contexts (Table 5). Taken across both questions asked of the documentation, there was no evidence of significant commitments to accountability across the majority of key functions. Those functions designated as having weak evidence for their implementation were typically described as optional actions rather than obligatory. In both urban and rural areas, some key functions are not monitored. For example, it's not clear how complaints, incident management and the activities of authorities such as NWSC and the rural executive committees are to be monitored and reported on.

Rules and regulations typically describe a functional performance envelope but fall short of specifying how such standards will be audited or how stakeholders might be engaged as part of the process. For example, according to the Water Act (1997), an authorised person may enter land for the purpose of inspecting works or use of water and an environmental inspector may inspect effluent quality through sampling. However, the Act does not detail which particular standards are to be checked. It is only for private sewers where the Water Act 1997 specifies that the inspector will check the design and operation of the installed system. There are also familiar weaknesses with respect to stakeholder participation where, for the great majority of key functions, stakeholders are not engaged in the monitoring process and thus transparency is not provided. In urban areas, monitoring for water authorities is conducted by an inspector who in turn reports to the minister. So, information is only shared between the inspector, the minister and the respective water and sewerage authority. Similarly, in rural areas, it is the district council or village executive committees (involves 10 elected members as per the LGA Act, 1997) that are involved in monitoring without much emphasis on engagement of community members. However, it is worth noting that several newer regulations (some of which are yet to come into force such as the National Environment Act) do provide for public consultation, for example during environment audits.

A much more mixed picture emerges from the data on corrective measures, incentives, and sanction, again across both urban and rural contexts. References to these functions are in evidence in multiple documents although there is a clear weakness in the provision for community and stakeholder ability to initiate enquiries and investigations (Table 6). Of those that were existent some are contextually relevant to incentivise alignment while others were weak. For example, according to the Uganda Water Action Plan of 1995, it is recommended that pollution fines should be set in a way that discourages practice. However, prescribed monetary fines for illegal wastewater discharge set at 10 million Ugandan shillings (about US$280 at the current exchange rate) seem light, especially in comparison with the capital investment budgets of large water service companies. The same problem applies to fines given to local government when they fail to properly account for the use of public funds – 5 million Ugandan shillings.

Similarly, very light sanctions are used in the case of night soil containment in both rural and urban areas. Regulation 18 of the Public Health Act 1935 requires a fine of not more than 100 Ugandan shillings (about 3 US Cents) to be imposed for not repairing buildings with defective latrines. Communities do have the ability to trigger investigations and sanctions through making complaints and whistle-blowing but these provisions are embedded in anti-corruption legislation such as the Inspectorate of Government Act (IoGA) 2002 and the Whistle Blowers Act 2010. This is one of many examples where the possibility of sanctions when complaints are made is indicated but where no clear complaint mechanism or process is described. Another example can be found in the Environmental Impact Assessment Regulation 32 (3) which indicates that the public may petition the executive director to cause investigation but details no specific procedural mechanism.

In the results described above, it is important to look at where there is evidence rather than where there is not, as a proportion of the document assemblage will, because of its remit and focus, not register as providing evidence for some of the metrics. With this caveat in mind, while this study has identified legislative provision across all four measures (although with minor differences between urban and rural contexts), it has also identified some key social accountability challenges within the water and sanitation sector. For example, limited integrity in delivery of some of the key functions, limited stakeholder engagement in the formulation of rules, processes and standards, compromised performance monitoring by government officials most especially for national level functions, limited community engagement during performance monitoring, and non-existent or inadequate sanctions for some key functions. Previous studies have reviewed the country's water accountability through assessing policies and conducting stakeholder interviews but rarely are there reported challenges related to the content and substance of legislation and policies. Most contributions to these debates have been framed by policy implementation challenges relating to a limited scope of water and sanitation features (e.g. Jacobson et al. 2010; Holvoet et al. 2016; Nicol & Odinga 2016). This study diverges from the dominant policy narrative to capture a wider range of social accountability challenges across different water and sanitation functions with stark implications for both legislators and communities.

The limited integrity in legislation implementation that we identify is partly a result of a lack of anti-corruption standards for some high-risk functions (like collection of fees), thereby confirming the findings of several studies that have assessed water accountability challenges in rural water supply (Kanyesigye et al. 2004; Jacobson et al. 2010). Operation and maintenance activities have been partly challenged by a lack of trust by community members towards water user committees and this has limited the collection of operation and maintenance funds. Additionally, dependency on the non-legally binding Memorandum of Understanding (Ministry of Water and Environment 2019) between the three ministries confirms findings from CSBAG (2015) that the practical implementation of the MoU has been affected by the lack of a clear accountability mechanism, limited coordination among ministries, and limited commitment of funds by those ministries. The lack of any legally binding responsibilities along with challenges in coordination and resource commitments has arguably compromised the intended impact of the MoU which has not been translated into improvements in sanitation and hygiene standards (IRC 2018). The Ugandan authorities have recently recognised this weakness, calling for a less fragmented institutional home for sanitation within the government (MWE 2019).

Challenges have also been reported (e.g. by Frederick 2012) with respect to performance monitoring where district officials do not satisfactorily monitor service providers and consequently poor quality fittings have been used which compromise the reliability and resilience of water supply. Our findings diverge somewhat from Fredrick's comments as performance monitoring is more likely to be conducted by ministry inspectors that monitor services managed at national level like water abstraction and water pollution. Any relevant duties and responsibilities mandated to be carried out by district councils would be driven by the Local Government Act, 1997. With this, it is evident that accountability issues are not only limited to gaps within legislation but also implementation. This accountability issue is partly the reason why most water sources fail at the early stages of installation and operation as indicated by Banks & Furey (2016). On the other hand, effective performance has been supported by the establishment of technical support units mandated to support implementation of water and sanitation activities at district level for rural water supply (Ministry of Water and Environment 2018a, 2018b). The existence of these units is however not explicitly provided for within the legislation.

Corruption by way of bribery has been reported when awarding contracts for private contractors and private operators (Jacobson et al. 2010). It is therefore not surprising that sometimes local contracts are tendered to firms that are non-specialists, or inexperienced in the water sector. District water engineers also report that they are not involved in contract review processes for rural water supply (Frederick 2012). These phenomena align with our study results where integrity may be compromised by ministries at national level because the benchmark criteria for selection of water authorities or private water operators is not clearly indicated within the Water Act. It is worth noting that the MWE currently has an existing Good Governance Sub-Sector Working Group mandated to formulate measures to promote and monitor transparency, accountability and good governance and currently indicators such as customer satisfaction are being monitored (Ministry of Water and Environment 2020).

Reluctance in complaint handling by sub county officials has been reported as a significant problem in Ugandan and other African contexts (Smet et al. 2010; Walusansa 2017). Our findings suggest that such ineffectiveness might happen where legislation makes no provision for monitoring complaints in rural areas and fails to provide clear complaint handling mechanisms to trigger sanctions. Corruption complaints made to the Inspectorate General of Government in Uganda are seen to follow a lengthy process and local governments have been advised to create complaint handling mechanisms for those that can be solved at local level (Segawa 2019). Outside of legislation, the MWE has a client charter that specifies standards to be delivered with complaint procedures (MWE 2018a, 2018b). However, the awareness and ability of rural residents or the poor to access and make use of this charter may be questioned.

The majority of roles and responsibilities relating to water and sanitation services are clearly spelled out in the surveyed documents with only one case where duties overlap causing a confusing mandate. This is related to the approval of fees and charges where it is unclear whether such charges require governmental approval or not. This finding aligns well with previous studies, but goes further to identify overlapping responsibilities in policy setting. The ‘de jure’ nature of the study reported above can only partially reveal the difference between mandate and obligation. In choosing to focus on accountability commitments and mechanisms enshrined in law we must be wary of assuming that such mechanisms necessarily bestow a right on behalf of citizens or an obligation by the state. They could equally, for perfectly valid reasons, be provided on a conditional or discretionary basis.

The analysis of legislation reported above unpacks a number of concerning social accountability gaps within statutes, such as lack of complaint mechanisms; limited stakeholder engagement when passing laws; weak performance monitoring; and a lack of integrity in delivery of some water and sanitation functions. Evidence from other sources suggests that these gaps are partially filled by local accountability initiatives, some of which are shaped by legislation and regulations which do not take water and sanitation services as their primary focus. For example, technical support has been decentralised to help districts to fulfil their responsibilities, while at national level accountability structures like the good governance working group has been formed to strengthen accountability initiatives.

We noted in the introductory sections that Ugandan government pronouncements on, and commitments to, social accountability have been strong and, in many respects, meaningful. This study has revealed some weaknesses in legal provision for such mechanisms, particularly with respect to stakeholder and community involvement. However, there are accountability initiatives which are active on the ground (e.g. technical support units, client charter for complaint handling) but not specified within legislation. We would argue that while such initiatives are often effective, their status and longevity will remain vulnerable where not underpinned by legally binding requirements. Researchers might use the prevalence and effectiveness of informal social accountability initiatives to further study the role and relevance of de jure social accountability provision. In particular, the ways in which legally sanctioned and local or citizen-initiated accountability processes interact would provide useful insights into how such a blend of the formal and informal constitutes a workable and effective model for social accountability.

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