The limits and impact of communication and context in implementing social interventions in a pandemic: Ghana's free water policy revisited

The study examined the implementation of the free water intervention initiated by the government of Ghana during the peak of the Covid-19 pandemic. The study deployed a case study design within the qualitative research method. A total of 22 respondents were interviewed for the study. Findings from the study revealed that the implementation suffered some challenges which mostly stemmed from both the suddenness of the intervention and the complex context of the water sector in Ghana. The abruptness of the intervention resulted in poor communication and little or no consultation between the presidency and the implementer. The other major setback was the withdrawal of services by many tanker drivers from the supply of free water to designated Covid-19 reservoirs due to delays in reimbursement by the government. Furthermore, there was no effective monitoring mechanism to ensure complete adherence to the policy directive by both household vendors and landlords. Generally, the objective of the intervention to provide free water to all was largely ineffective due to the challenges highlighted above.


Introduction
COVID-19 has drawn global attention to the critical role of water in managing infectious outbreaks (Smiley et al., 2020). Global leaders tacitly agreed to a need for a massive, coordinated and nationally-led attack on the pandemic. Besides the World Health Organisation's (WHO) non-pharmaceutical measures, governments across the globe rolled out various social interventions, the most famous of these interventions the was free provision of water to help mitigate the adverse effects of the crisis. In Africa, eleven (11) governments rolled out some form of free or discounted water policies, deferred payment of water bills, or suspended any disconnection of water supply for some months (Root, 2020). The free provisioning of water became globally critical for two main reasons. First, to help facilitate the World Health Organization's (WHO's) non-pharmaceutical measures of frequent hand-washing and maintenance of personal hygiene and the second reason was to cushion citizens and reduce the impact of the crisis (Duti, 2020;Root, 2020;Smiley et al., 2020;USAID, 2020USAID, , 2021. The decision of the government of Ghana to provide free water sought to achieve a common objective, primarily, to expand access and to assuage the suffering of citizens from the socio-economic hardship of the pandemic. However, there are concerns that the admirable intentions behind this beautiful initiative may not be achieved due to the complexity of factors surrounding water delivery in Ghana. First, the overall problem of water insecurity in Ghana (as in many developing countries) poses a hindrance to the successful implementation of the policy. Before the outbreak of the pandemic, about 5.5 million Ghanaians lived without access to clean water (Water Aid Global, 2021). The COVID-19 Pandemic has compounded this problem and has further deepened and entrenched the already existing inequities in the Water Sanitation and Hygiene sector. Consequently, the chances of already excluded groups benefiting from the free water policy are near impossible largely because they live in informal settlements in poor urban, peri-urban and rural communities. This study sets out to examine the limit and impact of Ghana's COVID-19 free water intervention. Specifically, the study seeks to examine the implementation of the intervention to ascertain its effectiveness or otherwise. The study makes a three-fold contribution to literature. Firstly, it makes a general contribution to the bourgeoning empirical studies on governments' policy responses to COVID-19. Secondly, it fills the void in the literature on the impact of the myriad COVID-19 interventions by the government with a specific focus on the free water policy. Finally, the study makes a contribution to the general literature on implementation science with a focus on the relevance of communication and context in the implementation process.

The context of water delivery in Ghana
The Ghana Water Company Limited (GWCL) was the leading sector agency charged with the implementation of the government's free water programme. It was supposed to coordinate the activities of other public sector institutions as well as private water delivery systems in the provision of free water services to Ghanaians within the stipulated time. Smiley et' al. (2020) also contend that the type of water source (public, private, formal or informal), the spatial context (urban, rural, or periurban), and a wide range of institutional factors in a country will determine the success or failure of water policies. Understanding the geography of water delivery in Ghana is therefore crucial to the successful and effective implementation of the government's free water policy.
For this study, we group water delivery systems into two broad categories-formal service providers and informal service providers. The formal service providers are represented by institutions formally established and recognized by the state to provide water services. In the context of Ghana, the formal sector comprises the Ghana Water Company Limited (GWCL) and the Community Water and Sanitation Agency (CWSA). These institutions deliver services through individual and collective connections, as well as home delivery. The informal service providers are represented by private water providers or vendors such as mobile water providers, standpipes and water kiosks. These informal service providers complement the formal in the sense that, they operate in areas where the network of the latter cannot reach (Rachmadyanto et al., 2016).
In the many areas where formal water providers' networks cannot reach, communities and private individuals put in place mechanisms to respond to their water demands thereby creating a niche for informal water providers. They are 'informal' in the sense that, they resort to informal processes in the management and provision of water (Rachmadyanto et al., 2016). Four approaches are used by informal water providers to meet water demand. The first of these is the standpipe which may be publicly owned by the District Assembly or community or privately owned. They are usually made of one or more taps, and connected to the public utility provider's main network. Consumers access to water from these taps at a cost (Kariuki & Schwartz, 2005). It is the major source of water for households in most poor urban and peri-urban communities. A study conducted by the Ministry of Water Resources, Works and Housing and Ministry of Local Government and Rural Development revealed that 52% out of 90% of households who reported using water from the GWCL, access water from either public or private standpipes, while only 38% have direct piped water connections to their houses (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). The second approach is the water porters. These are usually individuals and small companies who use various means of transportation such as water tankers, push-carts, wooden yokes or tricycles (motor king) to distribute water to consumers. Water porters are mostly secondary providers and buy water from a collective point-a public utility provider, a standpipe, water vendors, or even unknown sources-and sell to consumers and communities who normally have no direct connections to the formal utility provider (GWCL). They normally form associations to cater for their welfare needs (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012;Rachmadyanto et al., 2016). Water kiosks are the third approach used by informal providers to deliver water. This mostly includes individuals or groups who store water in tanks and sell it to people within their neighbourhood at a profit. They may be directly connected to the main utility provider's network (legally or illegally), construct their hand-dug wells or boreholes (self-supply); or may get the water from water porters. Water kiosks are a crucial component of the water supply chain, and mostly serve the poorer sections of the community. Consumers may directly buy water from them, or rely on water porters-who buy water from water kiosks for onward retailing on request (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). The fourth approach is household resale. This is where households that have a direct piped connection to the public utility network (GWCL) provide water to neighbours without direct piped connections. Such households foot the monthly bill to the public utility provider and resell the water to neighbours for profit. However, as Rachmadyanto et al. (2016) noted, one's ability to access water from this source depends on the willingness of the seller to sell water to them. Arguably, the sector makes an invaluable contribution to water management and delivery.
Both the formal and informal water sectors are confronted with formidable challenges in terms of efficient and effective water provision. The informal sector in particular is confronted with poor water quality. The quality of water supplied under this model is questionable as a result of many factors including the structure for collecting, treating, transporting, storing and distributing water. Although the quality of water at treatment points is normally good, it gets contaminated during transportation. Again, the use of the same tanker trucks in the supply of drinking water and raw water for construction presents additional quality challenges to those who rely on such services for drinking water (Kariuki & Schwartz, 2005;Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). Unfortunately, there is an inadequate quality control mechanism in place to check the quality of water supply by the informal sector. Furthermore, the primary motive of the informal sector is profit maximization. The Ministry of Water Resources, Works and Housing & Ministry of Local Government and Rural Development (2012) identified that some residents paid 10-20 times the approved tariff to access water from these sources. Kariuki and Schwartz (2005) also make a common revelation that communities and individuals who depended on these sources paid between 3 and 20 times higher than the normal price. As Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development (2012) noted, even though the PURC regulates the price at which tankers purchase water from GWCL, it does not regulate how much they sell to consumers. The poor therefore end up paying more for water than the rich.
The sector is also riddled with unreliable water supply for a couple of reasons. The first of these is undefined operational hours. Unlike the formal sector, these sources have no definite schedules for the provision of water. Even if such schedules exist, they are not under any obligation to always ensure frequent water supply. Consumers therefore can only have access to water only when it is made available by providers or when providers deem it fit. For instance, Kariuki and Schwartz (2005) identify limited operational hours as one challenge associated with this sector. Again, uninterrupted supply also depends on a provider's ability to maintain their facilities and pay their utility bills. Lastly, unlike the public utility providers, informal water supply systems and their workforce are not obligated to operate under emergency conditions (Stoler et al., 2020, p. 2). USAID (2021) identified that some (17%) operators reported temporarily stopping operation during the COVID-19 pandemic.

Challenges facing the water sector in Ghana
The sector is bedevilled with institutional and structural capacity deficiencies. Daily supply deficit, service failures and intermittent water supply continue to hamper the WASH sector. The GWCL's daily production capacity is unable to meet the daily demand of consumers, resulting in shortfalls. As a result, water is rationed in the majority of urban areas. For 2012, against a daily demand of 113 mm 3 , GWCL was only able to produce 0.708 mm 3 (62.65%) per day. Again, frequent service failures and intermittent and unreliable water supply continue to be frequent occurrences. As of 2012, less than 30% of households with a pipe connection had a water supply every day, while 35% obtained water for two days each week or less (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). Despite an appreciable improvement in urban water supply over the years-from 62% in 2010 to 77% coverage in 2022 (GWCL, 2022; Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012), -a substantial part of the population still do not have access to clean and sustainable water supply forcing them to resort to other sources (informal water systems) for water supply, which may be unclean, unreliable, or simply unavailable.
Studies point to a couple of factors underlying these capacity challenges. On one hand, rapid development and changing demographics keep overstretching WASH systems and causing systemic failures. On the other hand, these systems are unable to adapt to the ever-growing demand as a result of financial constraints, obsolete infrastructure, lack of requisite human resources, unreliable power supply etc. Moreover, excessive political controls and manipulations affect short-term policies and tariff setting; influence the appointment and determination of the tenure of office of CEOs and board members; and stifle the autonomy of the sector, thereby posing serious administrative challenges (Egyin, 2011;Manu, 2015 (2012), significant overlaps exist in the functions of the sector's three main sub-sector institutions -GWCL, CWSA, and WRC. For instance, it is unclear which institution(s) is/are responsible for peri-urban and small-town water service delivery.
With regards to regulations, sector institutions have either been ineffective or compliance has largely not been encouraging. Even though the Public Utilities Regulatory Commission (PURC) is responsible for regulating the urban water sector-setting water rates, establishing regulatory guidelines and protecting the interests of consumers etc.-the institution has been ineffective over the years. As Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development (2012) noted, the PURC lacks the requisite equipment and logistics for monitoring urban water quality, and therefore, relies on GWCL facilities for water quality testing. Secondly, in the rural water sub-sector, there are no or inadequate systems in place for monitoring water quality at point sources beyond the initial water quality tests that are conducted before handing over facilities to communities (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). Thirdly, it is unclear which of the sector institutions-PURC, GWCL, CWSA, WRC-is responsible for regulating the informal water sector and how effective such institution(s) has/have been. As Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development (2012) identified, the PURC used to regulate the price at which the tankers purchase water from GWCL but not how much they sell to consumers (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). As a result, there is a poor quality of water supply especially in the informal sector, with high prices and unreliable services.
Poor institutional linkages among the sector's key players-WRC, GWCL, CWSA, DAs, the Traditional Authorities and the community at large-are all governance issues facing the WASH sector in the country.
Production and supply of clean water is a capital-intensive venture which entails colossal capital investment (Manu, 2015). Unfortunately, the water sector faces huge financial constraints, shackling its ability to deliver effective and efficient services. Even though a chunk of the water sector's revenue is supposed to come from the government of Ghana, the government's contributions to the sector over the years have been substantially low, "with variations between budgeted and actual allocations reaching as high as 95% in some instances" (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). Aside from government transfers, the sector also depends on donor funding and investments for effective service delivery and administration. However, such assistance has been insignificant and insufficient over the years. As Manu (2015) mentions, donor funding has been mired by undue delays, hindering the effective and efficient implementation of water projects. Moreover, corruption, moribund internal revenue mobilization systems, illegal connections, and payment defaults from customers among others pose formidable financial challenges to the sector (Egyin, 2011;Manu, 2015 Poor development planning and the haphazard nature of building and infrastructure in Ghana, most especially in peri-urban and poor communities also pose serious challenges to the public utilities providers (GWCL) in ascertaining accurate data. Most poor localities either fall outside the formal planning zone of the authorities or lack a service network. The data available are therefore mere estimates by GWCL (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). Secondly, households in poor localities are often unable to get connections because They live in areas that are far from the network; lack land titles, which is a prerequisite to be eligible for a connection; and cannot afford the high upfront connection fees (Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development, 2012). Even when they have connections, because poor consumers generally lack "voice" and find it difficult to lobby for service improvements, they have poor and less maintained networks, experience perennial service interruptions, and have longer response times to service calls.

Situating Ghana's Covid-19 free water intervention in the implementation theory
The task of implementation establishes a link that allows the goals of public policies to be achieved as outcomes of governmental activity. This includes the creation of a "policy delivery system" in which specific means are designed and pursued with the expectation of arriving at particular ends (Grindle, 1980). Another definition which is quite intriguing and very relevant to this paper is that provided by Ripley and Franklin (1986) in which implementation is defined as what happens after laws are passed authorising a programme, a policy, a benefit or some kind of tangible output. In respect of the Covid-19 free water intervention, the government authorised the provision of free water which provides some form of benefit to the affected population. It is therefore important to investigate what happened after the authorization was made. Clearly, this demands a detailed examination of the implementation process, the effectiveness, the challenges and even the factors that facilitated the implementation. The science of implementation is arguably the most protracted area of study in the social sciences. There are simply too many theories, models and frameworks making it difficult for researchers to decide on one model to guide their research. A recent study by Nilsen (2015) identified three predominant aims of the use of theories, models and frameworks in implementation science: (1) Describing and/or guiding the process of translating research into practice, (2) understanding and/or explaining what influences implementation outcomes and (3) evaluating implementation. The second aim can further be disaggregated into determinant frameworks, classic theories and implementation theories. Therefore, there are five categories of theoretical approaches in implementation science as follows: (1) Process models, (2) determinant models, (3) classic theories (4) implementation theories and evaluation theories.
This study is situated within the determinants models which fall under theories that seek to understand and/or explain what influences implementation outcomes. Determinant frameworks highlight various types of determinants that have been found to influence implementation outcomes. These determinants or explanatory factors are made up of individual barriers or facilitators that have an impact on implementation outcomes. Some of these factors include the nature of decisions made at the design or formulation stage, the kinds of objectives that have been specified for it and how the goals have been stated. The content of the policy can also affect the implementation process. For instance, if public policy seeks to introduce change in social, political and economic relationships, they generally stimulate considerable opposition from those whose interests are threatened (Grindle, 1980). Mazmanian and Sabatier (1983) also provided six checklists for effective policy implementation: Clarity and consistency of policy objectives, a programme based on a valid causal theory, adequate structuring of the implementation process, commitment to programme goals by officials, interest groups and executives and legislature; and finally, there should not be detrimental changes in the socioeconomic framework conditions. The sixth point is consistent with Grindle's (1980) discussion on the potential impact of the policy content on implementation.
The context of implementation is critical to successful policy implementation. The context is generally explained as the conditions or surroundings in which something occurs, it is an analytical unit that is higher than phenomena directly under investigation (Nilsen, 2015). For instance, the current study examines the Covid-19 free water policy of the government of Ghana but the implementation of the intervention is carried out in an environment which transcends the free water intervention. Bottom uppers in particular, argue that successful policy implementation must consider the interaction of policy and local context: Micro-level institutional settings, local actors, and social differences among target communities (race, gender, language, culture, class, citizenship status, etc.) in conjunction concerning resources, access, power, and social capital (Stachowiak et al., 2016). The issue of context cannot be overemphasised in the analysis of the Covid-19 free water intervention. This is due to the peculiar nature of water provisioning in Ghana where informal water suppliers played a crucial role in the success or otherwise of the programme.

Methodology
The study deployed a case study design within the qualitative research method. The design was deployed because the study sought to conduct a detailed examination of the Covid-19 free water intervention in Ghana. According to Schoch (2020), a case study allows for a detailed analysis of a particular event, situation, organization or social unit. It is often used when a researcher wants to better understand or explain about 'how' or 'why' a phenomenon within a particular context happened. The research instrument used for the collection of primary data was the interview. Interviews were conducted from December 2021 to January 2022 and some additional interviews were conducted in March 2022. A total of 22 respondents were interviewed for the study. Two main reasons accounted for the choice of a sample size of 22. According to Schoch (2020), sample sizes in case studies are typically small. This is supported by the works of Creswell (2013cited in Creswell, 2014) that there is no specific number of participants to include in a qualitative study. However, he holds that the sample size used depends on the qualitative design being used. For instance, the narrative would normally include one or two individuals; phenomenology would typically be from three to ten; grounded theory can have between twenty to thirty participants and case studies include about four to five cases (Creswell, 2014). Indeed, there are researches where single participants were used for a case study research. For example, a longitudinal study of a CEO who is highly psychopathic is a plausible example of a single-case approach being groundbreaking and informative (Boddy, 2015). Furthermore, Ministry of Water Resources Works and Housing & Ministry of Local Government and Rural Development (2012) recommended 15-30 interviews for case studies and according to Boddy (2016), any qualitative sample size over 30 becomes unwieldy to administer and analyse. Another basis for the choice of a sample size of 22 hinged on the idea of saturation. It is argued that a researcher can stop collecting data when the categories (or themes) are saturated. That is when gathering fresh data no longer sparks new insights or reveals new properties (Charmaz, 2006in Creswell, 2014Guest et al., 2006). Consequently, the sample size of 22 deployed in this study is considered large enough to support the conclusions drawn from the analysis. These respondents comprised officials from the Ghana Water Company Limited -GWCL (2), customers of GWCL (10) and private/informal water distributors (10). The customers of GWCL are divided into two groups of customers, those who are directly connected to the GWCL pipelines and those who are not connected to the pipelines and therefore source water from private providers. The informal water providers are also put under four distinct groups -(1) household providersthese providers are connected to GWCL pipelines and in turn, sell water from their homes to the general public, (2) standpipes -This constitutes standpipes that belong to the government but are managed by private individuals on behalf of the government and (3) water Porters with focus on water tanker providers these are private individuals who use water tanker trucks to supply water to areas outside the coverage area of GWCL. (4) Water kiosks -this mostly includes individuals or groups who store water in tanks and sell it to people within their neighbourhood at a profit. They may be directly connected to the main utility provider's network (legally or illegally), construct their hand-dug wells or boreholes (self-supply); or may get the water from water porters.
The respondents were purposively drawn across four regions in Ghana, Winneba (Central Region), (Kumasi) Ashanti region, Techiman (Bono East Region) and Tamale (Northern Region). The four regions were selected due to the proximity of the researchers to these locations as well as accessibility of respondents. Another reason for the choice of the four regions was to reflect the three geographical zones of Ghanathe coastal belt, forest belt and savannah belt. According to Scheib (2003 cited in Schoch, 2020), a site can be selected because of easy access. It is also informative to mention that the study was a single case (Ghana) with a relatively homogenous population and as such any qualitative sample size that is more than 30 in in-depth interviews could be considered large and would require justification (Boddy, 2016). 20 of the interviews were conducted in person and two were conducted via phone. Documentary analysis was also heavily relied upon as a form of primary data. Largely, news items were analysed as a form of triangulation for the study.
The data gathered from the interviews were analysed thematically. The raw data were first prepared and organisedthis involves transcribing the interviews, typing field notes, and sorting and arranging the data into different types depending on the sources of information. The data was carefully studied and coded. A combination of emerging and predetermined codes were used. The report was written in the form of a narrative which made use of rich language and text to convey the findings of the study. The analysis also involved an interpretation of the findings, ultimately gleaning lessons from the findings. Furthermore, the analysis was done along with the discussion of the findings where the findings were discussed in the context of the relevant existing literature.
One major limitation of the study was the fact that the researchers did not have an opportunity to interview the staff of the implementer (GWCL) at its headquarters. Access to the headquarters staff would have helped explain the discrepancies in the implementation of the programme at the local level.

Swift decisions during pandemics and the relevance of communication
Goal clarity and communication should go hand in hand with government decisions and implementation efforts. The effects of Covid-19 took governments across the globe aback. As a result, many government interventions were hastily handed out from the top down without any form of consultation with the institutions charged with their implementation. Officials from the GWCL, the main institution responsible for the implementation of the free water policy during the height of Covid-19 indicated they were not consulted before the introduction of the intervention. There was also an initial lack of clarity regarding who constituted the beneficiaries of the intervention. The intention to have all Ghanaians enjoy the intervention could not materialise due to the absence of prior communications or consultation between the presidency and the GWCL. Officials from GWCL captured the lack of clarity and prior communications between the implementers and the policymaker as follows: "No, the communication was not done. It was after the president's announcement that the top hierarchy was informed. The views and input of frontline staff were not sought" (GWCL official 2, March 2022).

"You see like I said we have codes for different categories of users/consumers. So when the President announced the directive, he did not specify who gets what free, so it created confusion. So we had to draw his attention, and the Minister for Water Resources came to explain and to specify who pays and who gets it for free. Even with that, it was difficult to convince people to pay" (GWCL official 1, February 2022).
Furthermore, initial communications from the headquarters of the GWCL suggested that customers who are indebted to the company and those who had their homes disconnected from the public provider will not enjoy the free water supply announced by the president. The CEO of GWCL, Clifford Braimah is reported to have made a statement that: "The people who owe us, we expect that they pay before they get free water. Somebody said those of them who have been disconnected, we have to connect them so that they will benefit from the three months [free water service]. I said the President is not telling me to give free water to those who are not connected because the president is not paying for January or February. And so those owing will have to pay," (Graphic Online, 2020).
The statement by the CEO was later clarified that customers who owed the company would enjoy the free water supply. However, those who were already disconnected would have to pay to be reconnected to the GWCL pipelines. In other words, households that were disconnected from the public pipeline were exempted from the government's free water policy and were compelled to source water from private providers. This finding is in line with Fielmua and Mengba (2022) that urban households that could not settle their outstanding water bills before the implementation of Ghana's free water policy, as well as rural households who depend on point sources for water supply did not benefit from the intervention. The finding is also in line with Gbedemah (2022) who reported that many people in Somanya in the Eastern Region of Ghana depended on private water vendors and paid for the water services they received during the Covid-19 period.
Indeed, the back-and-forth communication from the quarters of the GWCL is a clear indication of a lack of prior correspondence between the government and the implementer in the issuance of the policy directive. It also indicated a lack of clarity on the goals of the intervention. Consequently, not all household consumers enjoyed the free water intervention as announced by the president. Similarly, those who used water for commercial purposes including small businesses such as car washing operators, and restaurants among others were charged during the free water intervention. The lack of clarity in the goals as well as the nature of the decision, and the fact that the decision was taken in haste fall within the inhibiting factors that affect successful implementation as discussed (Grindle, 1980;Mazmanian & Sabatier, 1983).

The prospects and limits of context on implementation
A discussion of the context of water service delivery in Ghana ultimately includes the structure of the water sector and the resources of the public provider. Water delivery in Ghana as mentioned earlier can be grouped into formal service providers and informal service providers. The formal service providers include the Ghana Water Company Limited (GWCL) and the Community Water and Sanitation Agency (CWSA). The GWCL is strictly in charge of managing water services delivery in urban areas on one hand. The CWSA on the other hand is charged with rural water services. This analysis largely draws on the implementation of the intervention in the urban and peri-urban areas. Many households in the urban and peri-urban areas are not connected to the pipelines of the GWCL. These households mostly rely on informal service providers represented by private water vendors such as mobile water providers, standpipes and water kiosks. This affirms Smiley et al.'s (2020) study that most households in informal settlements in Ghana lack piped and metered water supply, hence, they access water from water tankers and other private vendors whose services were not reliable at the height of Covid-19. Further, according to the Ghana Living Standards Report by the Ghana Statistical Service (GSS, 2019), close to 3.9 million households in the country are off the national grid of pipe-borne water. The complex context along with the nuances it presents affected the fluid implementation of the free water intervention in varying ways.
The presence of neighbourhoods outside the catchment area of the public provider posed the challenge of how to rope these neighbourhoods into the free water intervention. The GWCL had to devise appropriate means to provide water to these areas. Consequently, water tanks (labelled Covid-19 tanks) were placed in areas that were not connected to the public provider's pipelines and made arrangements with some private providers (water tanker drivers) to fill these water tanks as and when necessary. These tanker drivers before the policy intervention sold water to these unconnected neighbourhoods. However, with the commencement of the free water policy, they were tasked to supply water to these neighbourhoods for free for onward reimbursement from the government. This is captured in the excerpt below.
"The GWCL people engaged us to supply water to the new sites and the areas beyond their catchment zones so that they reimburse us later. That was the arrangement they made with us" (Water tanker driver 2, February 2022) As evinced from the excerpt above, the government made some efforts to provide free water to areas that were not connected to government pipelines. However, the presence of a vast area that is still not connected to the public provider's pipelines made it difficult for the government to provide Covid-19 water tanks to all the unconnected areas. This meant the exclusion of some neighbourhoods from enjoying the free water supply by the government. An official from the GWCL office in Tamale Metropolis reported the inability of the GWCL to supply water to all the unconnected areas as follows: "We were not able to cover all the areas. The tanks were mounted in areas that were connected to GWCL pipelines but were not getting water supply" (GWCL official, March 2022).
This suggests that many neighbourhoods that were not yet connected to GWCL pipelines were not provided with free water. Furthermore, a water tanker driver from the Winneba area captured the inability of the public provider to supply water to all unconnected areas in the statement below: "So like I said, those Covid-19 tanks were mostly provided in villages around Winneba. So the villagers mostly accessed them for free. But apart from that, if you are in say Winneba town where there is no Covid-19 tank and you need water, then you have to pay for it. That is how we worked" (Water tanker driver 6, February 2022).
Unequal access to the free water provided by the government during the pandemic is supported by some studies. Nkrumaha et al. (2021) reported that access problems, completely or partially excluded about 5 million households in Ghana from the government's free water intervention during Covid-19. A related study by Gbedemah et al. (2022) on the coping strategies for water accessibility during the Covid-19 pandemic in Somanya established that most people did not benefit from Ghana's free water policy intervention due to lack of in-dwelling pipe-borne water facilities.
Another dimension of the intervention had to do with its impact on the livelihoods of private providers and how that in turn affected the general implementation of the policy. Many private vendors who hitherto the intervention sold water to neighbourhoods that are not connected to GWCL pipelines were engaged by the government to deliver water to Covid-19 water tanks mounted in some of the unconnected areas for reimbursement by the government. Thus, the successful implementation of the free intervention in the unconnected areas largely depended on the cooperation and participation of the water tanker drivers. interviews with the drivers indicated that the government reached out to them to participate in the programme and many of them indeed joined the scheme scheme. A good number of tanker drivers reported that they participated in the programme up to some point and withdrew their services due to delays in reimbursement on the part of the government. Some of the drivers opted not to participate in the programme because of the anticipated delays in reimbursements. The delays in reimbursements affected their operations and livelihoods in different ways. Those who withdrew their services did so because their daily operations depended on the regular flow of income as they needed money to fuel their vehicles daily to move from hydrant stations to the communities they served. Some of them also complained that the communities they operated in did not have any public hydrant; as a result, they relied on private hydrant stations which came with the additional cost of paying for the water itself. In addition to the preceding points, it was mentioned that the trucks belonged to truck owners, who expected to receive daily sales from tanker drivers. The fact that there were delays in reimbursement meant that the drivers had to struggle to fuel their trucks and pay private hydrant owners in areas there were no public hydrants. Furthermore, tanker owners did not have the luxury of waiting on the government to pay at the end of the month as they had families to feed daily. Essentially, the delays in reimbursement brought some form of hardship on both the water tanker drivers and the tanker owners, leading to the discontinuation of the provision of free water to the general public by many water tanker drivers. The drivers captured the delays and their attendant effects on their livelihood in the excerpts below: "… Those GWCL people here don't understand the nature of our work. They only told us to supply the water for free so they pay us later without thinking about what the car owner will say. Besides, how do I fuel my car? Don't forget I must also get something at the end of the day to feed my family. All these things were not considered. So if you just tell me to supply water for free and wait till God knows when before you pay me, you know there will be problems" (Water tanker driver 5, December 2021).
Some water tanker drivers decided not to participate in the free water supply at all. One of them mentioned prior experiences where they worked for the government with promises of reimbursement which never materialized to date.
"Yes, as I said earlier, the GWCL people told us to come to the office to register if only we wanted to participate. But like I said earlier, I declined. I didn't believe it; and I am talking from experience. During the construction of the Bronkyempem Assembly Hall, the assembly engaged some of us to supply them water for construction, and that they would reimburse us later. We also agreed and supplied. My brother, as I am talking to you, I haven't received a dime. They still owe me. I went there so many times but they are still yet to pay me. So I will be a fool to be tricked for the second time. Besides, per the nature of our work, we need money in order to be able to function. So if I supply you and you delay payment for more than three days, you are collapsing my business. I cannot function without money for fuel and even money to buy the water" (Water tanker driver 1, December 2021).
An interesting observation from the interviews was that almost all the drivers including those who participated in the government free water programme engaged in a form of hybrid water delivery during the free water intervention. The drivers who registered to deliver free water continued to sell water to areas that were not under the provider's service areas. Another category of private vendors that cannot be ignored is the household vendorsthese are individuals whose homes are connected to the public provider's pipelines and they sell water to their neighbours who do not have their homes connected. Borehole operators also fall under this category. While there was no arrangement to cater for the household vendors (People who sold water from their homes), the same cannot be said about borehole operators. Generally, household vendors were asked not to sell water during the period. As to whether they were billed or not, there were contradictory reports from both the vendors and GWCL operators. One household vendor mentioned that they were asked to provide water for free and in exchange, she was not billed throughout the free water intervention. Others also claimed they were billed and for that matter, they were forced to sell water during the period.
"No all they said was my bills are free but the government didn't make any arrangement with us. So our water was free. So if they said we won't pay bill and people too were homes without going to work, so we sold for only Koko sellers because they are going to use it for commercial purpose so when they come to fetch we sold it to them because one can go about three of four barrels, but those who fetch for household use like cooking and bathing were not fetching so much" (Household vendor 3, December 2021).
One thing that was consistent across all vendors was the fact that they had no form of arrangement with the government in terms of financial support during the free water supply. On the contrary, borehole operators were paid to provide free water during the intervention in the northern region.
"In Damango, borehole operators were registered and the government paid them for their services" (GWCL official 2, January 2022).
Even among the borehole operators, there were discrepancies in how the government implemented the intervention. One GWCL official in Winneba stated that the government did not deal with borehole operators. The following are his words: "We don't deal with boreholes. If someone has his borehole, you cannot dictate to them how they should operate" (GWCL official 1, January 2022).
The quotes above indicate there were discrepancies in the implementation of the free water programme in respect of household vendors. It is therefore not surprising that many of them decided to sell water during the period despite the warnings from the government. These findings are in sync with many studies conducted on the government's free water programme. As observed from the analysis and supported by extant literature, some water tanker drivers temporarily stopped operating under the free water directive due to the inability to bear initial operational expenses (USAID, 2021, p. 6). Braimah et al. (2018) found that water tanker drivers cover daily operational expenses on maintenance, gasoline, and daily income, among others. Drawing on their fear to wallow in this cost, some tanker drivers were reluctant to engage in the supply of the free water programme during the pandemic. Smiley et al. (2020) noted that tanker drivers may be unwilling to supply households with the expected quantity of free water. Furthermore, Amankwaa and Ampratwum (2020) found that the free water intervention decimated the revenue of water service providers such as the GWCL, CWSA, and other private water providers due to the government's inability to defray its debts to water service providers. Consequently, the authors concluded that inadequate financial and operational functionality of water service providers to deliver adequate services contributed to the relapse in the implementation of the free water intervention. The USAID also reported that reimbursements by the Government of Ghana to the operators were minimal. While 67 per cent of water systems reported receiving some compensation, the amount received was only a small fraction (4-16 per cent) of the amount requested (USAID, 2021, p. iii).
The final dimension of free water supply was how the intervention was implemented in respect of household consumers. The intervention worked smoothly for households that were directly connected to the GWCL. They had full access to the government's free water intervention. However, some nuances were observed with the households that were connected to the public pipelines. A distinction can be made between those who lived in their own homes and those who are renters/tenants. Homeowners fully benefitted from the intervention while the same cannot be argued for tenants. While some tenants were lucky enough to have been spared from paying water bills by their landlords, other tenants were not so lucky. Their landlords continued to take water charges from them throughout the intervention. The excerpt below captures the predicament of some tenants during the interventions: "The Answer is a yes and no. This is because even though we heard about it on the radio and TV, our landlord continued to take money from us, at a discount I must add. So somehow, I can say we enjoyed the free water policy but not completely. Before the policy, I was paying GHC 13.00 per month, but when the free water initiative was announced, we were paying half of the amount-At times GHC 7.00 and at times GHC 6.00" (Respondent, December 2021).
While the government decided to provide free water during the height of Covid-19, some landlords continued to charge their tenants but at a reduced rate. This indicates the absence of a mechanism for checking compliance on the part of vendors and households. The noncompliance of the landlords' was also made possible by the complex context of the water sector in Ghana. People who do not have their homes connected to the GWCL pipelines and did not also have the Covid-19 water reservoirs in their neighbourhoods appear to have been worst affected. Unlike tenants who were exploited by their landlords but enjoyed some discounts, those who had no connectivity in their homes and also had no Covid-19 reservoirs paid for water throughout the free water intervention. The above scenario affirms the conclusions of Duti (2020) that the free water policy only resolved the issue of affordability for those who have access to portable drinking water, it does nothing for the many vulnerable rural and peri-urban communities who do not have direct piped connections and effective and reliable water tanker services. Furthermore, the fact that water tanker drivers, household vendors and landlords continued to sell water during the period is indicative of the ineffectiveness of the implementation process. This also supports the findings of USAID (2021) that about 10% of small piped systems surveyed in Ghana continued to apply tariffs to customers, while 84% of piped water consumers in both national and district surveys reported paying some amount for water. The challenges the free water intervention faced was largely that of context. This underscores the relevance of paying considerable attention to the local context in both policy formulation and implementation. According to (Stachowiak et al., 2016) successful policy implementation must consider the interaction of policy and local context: Micro-level institutional settings, local actors, and social differences among target communities (race, gender, language, culture, class, citizenship status, etc.) in conjunction with the resources, access, power, and social capital.

Conclusion
The study set out to examine the effectiveness of the implementation of the free water intervention initiated by the government of Ghana during the height of the Covid-19 pandemic. The art of implementation is a complex process and not always straightforward as implementers would desire. The implementation of the free water intervention was no doubt going to be a challenging task due to the complexities of Ghana's water context. The challenges of the implementation stemmed from both the hasty manner in which the intervention was initiated and the complex context of the water sector in Ghana. The study found that the abruptness of the decision resulted in poor communication and little consultation between the government and the implementer. The lack of communication and consultation between the two did not make it possible for the implementers to provide the government with the requisite information on the water sector to make the right allocation of resources. This generally led to a lack of clarity on the goals of the intervention which in turn led to a series of contradictory communication about who benefits from the free water policy.
The other major setback of the implementation stemmed from the complex context of the water sector in Ghana. The first of the contextual challenges emerged from the fact that many homes and neighbourhoods are not connected to the public provider's pipelines and therefore rely on informal providers for their water. The implementer and government made efforts to ensure that groups/individuals that are not connected to the public pipelines were catered for during the intervention. However, the strategy adopted by the government to include these disconnected groups met some setbacks. Many of the water tanker drivers contracted to supply water to designated areas abrogated their contracts a few weeks into the intervention. Many reasons were cited for the discontinuation of supply to the designated areas. Paramount among them was the delay in the reimbursement of water tanker drivers. The delays affected their operations and livelihoods in many ways. For instance, drivers needed money to fuel their trucks on daily basis to enable them to move from the hydrant to the areas they were serving. Drivers also indicated that some of the areas they served did not have public hydrants, which meant they had to pay before they could fill their tanks at private hydrants. Another difficulty brought about by the delay in payments had to do with the inability of water tanker drivers to pay water tanker owners on daily basis. Most of the drivers are not owners of the water tankers they drive. They drive other people's trucks and they are expected to pay truck owners daily. However, the nature of the agreement with the government meant that they will not be able to pay the truck owners daily. The inability to pay truck owners on daily basis means an inability to honour the agreement they have with the truck owners which could lead to the termination of their appointment. Moreover, the drivers depend on daily sales for their livelihoods. Thus, any prolonged delay in reimbursement will adversely affect their livelihoods and that of truck owners. Consequently, many of the drivers abandoned the contract to deliver water to areas that are not connected to public pipelines. The implication of this is that many areas that are not connected to public pipelines-mostly peri-urban, rural or poor neighbourhoods-were partially or totally left out of the free water intervention. Sadly, many inhabitants in these communities are among the low-income groups in the country. Thus, the poor, whose livelihoods were greatly affected by the pandemic, were further excluded from benefiting from the government's social intervention. Ironically, the rich who live in urban and wealthy neighbourhoods were the ultimate beneficiaries of the intervention.
Another challenge suffered by the implementation was the absence of an effective monitoring mechanism to ensure complete adherence to the policy directive by both household vendors and landlords. The study found that consumers who are not directly connected to the public pipelines and those who are tenants were extorted by household vendors and landlords respectively. Some household vendors continued to sell water throughout the period because that is their only source of income and since the government did not make any arrangement to support them during the intervention, they were compelled to sell water despite the free water intervention. Some landlords also continued to charge their tenants for water rates. In most cases, they discounted the water charges for their tenants when they (landlords) were not paying any bills to the government.
Generally, the objective of the intervention to provide a free water supply to all was largely ineffective due to the challenges highlighted above. Though the relevance of goal clarity, communication and the local context of the policy environment has received a lot of attention in the implementation literature, they continue to substantially affect implementation. This is an indication that policymakers do not fully accord these factors the needed attention in the policy-making process, which inevitably negatively affects implementation.