Decentralization is linked with local government systems and has been practiced in Tanzania since colonial era. Thus, there is a plethora of analyses on the subject in the literature [15] - [23] .

Practically, there have been at least four major initiatives for reforming local government system in order to improve social services delivery in Tanzania [24] . The first attempt to implement decentralization in Tanzania can be traced back to the colonial period when the British Colonial rule introduced local governments. The British introduced local government as part of the broader strategy to govern their colonies through a system called indirect rule [25] .

After independence in 1961, the newly independent government inherited the British Colonial structures and integrated them into the national system of administration [26] . The local authorities experienced troubled times during the post-independence era, especially after the constitutional changes that introduced one party system in 1965 [25] . Later on the enactment of the Decentralization of the Government Administration (Interim Provisions) Act of 1972 abolished all local government authorities and centralized power and authorities in central governments [25] [26] . The District Development Councils and Regional Development Councils replaced the LGAs.

The economic crises in the late 1970s and 1980s led to rapid decline in essential services [27] , prompting the government to re-introduce local governments in 1982. In 1984, the newly established LGAs were constitutionally recognized through the amendments of the constitution vide Act No. 15 of 1984; this was major development as the LGAs cannot be abolished without tabling a motion in the Parliament [25] .

Due to the failure of local governments to perform the mandated functions and responsibilities, the government attempted decentralization reform that involved promotion of devolution [28] . The local government agenda [28] provided the vision for the future local government system and the policy paper on local government reform [14] became the official guiding policy document regarding local government reforms and decentralization by devolution (D by D) [25] [26] . The aim was to speed up political, financial and administrative accountability at district level, improve transparency in the local government transactions and bring public services closer to the grassroots levels [14] [29] . The next sections present the historical development of the provision of the two services this is done to contextualize the research problem.

The Health Sector Reform (HSR) has also involved decentralization of health care provision within the sector, with the aim of promoting greater stakeholders’ participation and choice, as well as improving efficiency in service delivery [30] . The health reforms have been defined as the institutionalized changes in the way the health services (curative, preventive, promotive and rehabilitative) are provided and financed [29] . As the move to achieve “Global Health for all by the year 2000” proclaimed in the Alma Ata Declaration of 1978, Tanzania government in collaboration with WHO and UNICEF sponsored 1987 Bamako Initiative and institutionalized decentralization within the health care service delivery [29] .

The initiative called for the decentralization of the public health systems, strengthening community participation in the health, and cost recovery programmes [30] . In the wake of this broader reforms Tanzania government through the ministry of health adopted health sector reforms in the early 1990s. These reforms resulted into significant organizational, managerial and financial changes to health care planning and delivery [29] .

As the result, the delivery of health services is now organized at three administrative levels; at the top is the Ministry of Health, Community Development, Gender and Elderly (MoHCDGE) which is charged with the development of the sector policies, regulatory framework, monitoring and evaluation in collaboration with the President’s Office-Regional Administration and Local Government (PO-RALG). At the regional level, the regional medical officer is the member of regional secretariat, responsible for the provision of supervisory and technical backstopping to LGAs and at the local level are the district councils [31] .

The HSRs and LGRs give mandate to district councils to manage district hospitals, health centers and dispensaries using subversions from the central government and locally generated sources [9] .

Even though the importance of portable water to human life is universally understood, domestic water requirements were not on top of priority of the global social and political agenda until mid-1970s [32] . To be specific it was during the UN water conference, in Mar del Plata, Argentina, in 1977. The conference come up with the proposal that 1981-90 be designated as the International Water Supply and Sanitation Decade, with the aim of delivering water related services for 100% of the world’s population [32] [33] . Although the decades’ ambitious target to increase access to water and sanitation proved impossible to achieve, for the first time, water and sanitation appeared as top priority in development agenda [33] .

A decade later, the UN comes up with the Millennium Development Goals (MDGs), which adopted only one of the objectives of the decade: “Target 7(c) which seeks to halve by 2015 the population of people without access to safe drinking water”. Moreover, later on, in 2002 the target was extended to include sanitation [32] . Moreover, the 2005-2015 was declared “International Decade for Action: Water for life, and subsequently, the year 2008 was declared the International year of Sanitation (ibid).

Despite these global political efforts, the data indicates that only moderate achievements have been registered globally to date, and still huge inequalities appears when comparing access to water services between rural and urban and across the countries [33] . As [34] report, Sub Saharan Africa will only reach the MDG water target by 2040 and still, some 400 million people living in Sub Saharan Africa will be left without access to safe water.

It is against this backdrop that the Government introduced a new National Water Policy (NAWAPO) in 1991. Since then, Tanzania has been facing a transition from a socialist economy-based on the principle of “free water for all”―to a more liberal economy where cost recovery is inevitable [35] . Today, more than 45 years later, the situation shows little improvement as providing safe water and improved sanitation while reducing the existing service coverage gap between rural and urban areas remains a herculean task. A new policy framework has been recently developed, to define the appropriate strategy to achieve ambitious national sector-related targets. The NAWAPO was revised in 2002, the National Water Sector Development Strategy (NWSDS) was formulated in 2004, and the Government launched in 2005 a new National Strategy for Growth and Reduction of Poverty (NSGRP), embarking on a process of “decentralization by devolution”, with control over water service delivery moving to local governments [36] .

Finally, a comprehensive Water Sector Development Programme (WSDP) has been developed to put water-related policies in a functional framework, and it is to be implemented through three national programmes, covering the sector areas of major concern: Rural Water Supply and Sanitation (RWSS), Urban Water Supply and Sewerage (UWSS), and Water Resource Management (ibid). In executing this strategy in Tanzania, stakeholders remain the same but are assuming different commitments. For example, the governments’ new role is that of policy and guidelines formulation, coordination, monitoring and regulations [36] . Similarly, in accordance with the principle of “decentralization by devolution,” the management and coordination of day-to-day activities move to the local authorities, and since the delivery of water services is based on demand driven approach, community participation at different levels should be guaranteed [35] .

In Tanzania, both health and water sectors are guided by their sector policies, but the overall direction of the health and water issues is being guided by the National Strategy for Growth and Reduction of Poverty (NSGRP) popularly known in Swahili as “Mkakati wa Kukuza Uchumi na Kupunguza Umasikini Tanzania” (MKUKUTA). These reforms are being implemented under the broader local government reform programme (LGRP) which commenced on 1996 to speed up political, financial and administrative accountability at district level [29] . The programme also intends to improve transparency in the local government transactions and bring public services closer to the grassroots levels [14] .

Several studies attempted to analyze the implementation of Decentralization and Service delivery in Tanzania. To name but a few they include the following: [37] - [46] . Some of these studies were comparative correlations. Other studies evaluated the extent to which the reform programme was undertaken. This study attempted to use Principal-Agent Theory as well as users and providers’ experiences as the main sources of information.

The study involved 208 respondents, where 127 were individual respondents to the survey questionnaires on health and water services delivery, while 51 participated in the in-depth interviews. Specifically, information was collected from key policy makers and planners at the national level and health and water officials from the council level and the service users. The respondents were purposively selected because they were directly involved in decision-making on implementation of decentralization at the central level, and respondents from the LGAs were selected because of their directly involvement in actual implementation of the decentralized health and water services at the district level.

At the national level, the study involved the coordinator of Regional Secretariat who was in-charge of four ministries directly responsible for implementation of decentralization policy in the health and water sectors: the then MoHSW, PMO-RALG, MoW and Ministry of Finance (MOF). At the district level, 33 key informants who were directly involved in implementation and supervision of decentralized health and water services delivery were purposively interviewed in Kinondoni and 17 key informants were interviewed in Ulanga.

The study also administered questionnaires on health services to 20 and 13 respondents in Kinondoni and Ulanga, respectively. To get the perspective of users of water services, the researcher administered 49 questionnaires in Ulanga and 45 in Kinondoni Municipality, the respondents were randomly selected. The two FGDs in Ulanga were composed of 20 participants, whereas in Kinondoni the FGD was composed of 10 participants.

This article makes use of a research data from two case councils (Kinondoni and Ulanga) which were used for PhD study of the author in 2014-2015. Findings from interviews and FGDs were transcribed; those from the questioners were cleaned, verified and checked for quality and consistency, sorted, and organized carefully. The interview guides were labeled to identify code numbers of interviewees, which enabled the identification number for the informants.

Data analysis was based on themes or perspectives that were prepared. Thematic analysis approach involved reading through the transcribed text of each interview and identifying responses relevant to the main questions asked. The coding process was used to develop a small number of categories. Codes with similar concepts were grouped together to form category. Quantitative data obtained from the households’ surveys were coded and entered into the Statistical Packages for Social Science (IBM SPSS Statistics 19) to make them amenable for statistical analysis, summary frequencies were run at the end of data entry exercise to check for accuracy and completeness.

Descriptive statistics were done to understand the relationships between two cases. Triangulation of the methods yielded simultaneous interpretations, which were complimented by the qualitative responses by means of quotations.

Four forms of participation facilitated examination of participation in decentralized health and water service delivery in this study. They include attendance in public meetings and other health related activities, participation in water and health governance structures, health and water services facilities’ autonomy, roles and responsibilities of service users and support from the government and other stakeholders.

The most important way through which community members are involved in the planning and management of the health services is through their participation in the meetings and committees [55] . Surveyed respondents in Kinondoni MC and Ulanga DC reported that they were participating in the health policies implementation through attending public meetings, receiving feedback on services, promotion services, attending presentation about health behaviors, and family visits. Specifically, in the studied districts, 14.8 percent of the respondents reported to attend public meetings in Kinondoni MC and 100 per cent of the respondents attended meetings in Ulanga DC, 14.8 percent of the respondents reported to attend seminars on health behaviour in Kinondoni MC and 84.6 percent of the respondents in Ulanga DC had similar opinions.

Moreover, the study findings indicated that community participation in implementation of the health policies was also done through health official visits to families in villages and streets as reported by 13.1 percent of the respondents in Kinondoni MC and 84.6 percent of the respondents in Ulanga DC. Delivering health education for children at schools was also a form of participation as reported by 16.4 percent of all respondents in Kinondoni MC and 92.3 percent of the respondents in Ulanga DC. Giving feedback to the communities on the health services delivered was another form of participation as reported by 15.6 percent of the respondents in Kinondoni MC and 84.6 percent of the respondents in Ulanga DC.

Community participation was also through promotion of health services, whereby it was reported by 15.6 percent of the respondents in Kinondoni MC and 84.6 percent of the respondents in Ulanga DC participated. The findings further revealed that community members were involved in assessment of their health needs, whereby 9.8 percent of the respondents reported to participate in Kinondoni and 14.8 percent of the respondents reported to participate in the assessments of their needs in Ulanga DC. Community participation at this level is “practical and active specific”. The findings from Ulanga and Kinondoni districts are summarized in Table 2.

The above findings from Kinondoni and Ulanga are supported with the findings from an in-depth interview from one key informant in Ulanga:

Opportunities and Obstacles to Development (O & OD) have improved participation of the communities in various Local Government plans, it gives them an opportunity to see the opportunities and obstacles and how do they and government contribute in the development. In Ulanga, for example if we want to build a health Centre (HC) we give the community the proposed plan and what they should contribute as beneficiaries (Key Informant₁ Ulanga DC February 2014).

Source: field data (2014).

The above findings are in sharp contrast with [29] conclusion that “decentralization in whatever form does not automatically provide space for community, the assumption that devolution to local governments promotes transparency, accountability and community participation is far from reality”. This study finding posits that decentralization reforms have the potential to create spaces for community participation in the delivery of the services.

Tanzania has developed governance structures for participation in the health services [56] . For example, users are involved in the implementation of the policies through planning and management of the health services delivery [55] .

The Village Health Committees (VHC) and Health Service Boards (HSB) facilitate user participation. These are organs working very close with local governments in primary health care delivery. In the studied areas, it was reported that health facilities’ boards and health committees were present in both councils, as indicated by 100 percent of the respondents in Kinondoni MC and 92.3 percent of the respondents in Ulanga DC respectively. It was further reported that boards and committees met every month 50.0 percent of the respondents in Kinondoni MC and 61.5 percent of the respondents in Ulanga DC.

However, on the composition of the board/committee, the responses were as follows; 95.0 percent of the respondents mentioned the health staffs as committee and boards members in Kinondoni, whereas 76.9 percent of the respondents reported the same in Ulanga DC. Moreover, study respondents indicated that there was minimal community participation in the health boards and committees as reported by 5.0 percent of all respondents in Kinondoni MC and none 0 percent of the respondents in Ulanga DC. The above information is summarized in Figure 1.

The health facility service boards are established under prevailing laws, which include Local Government Act of 1998, sections 76, 78 and 44 and 45 for District and Urban Councils [56] . The amended Act number 6 of 1996 gives legal mandate for LGAs to establish these boards. Findings from this study revealed that health facility boards were present in some facilities in Ulanga DC and Kinondoni MC. However, out of six interviewed health facility board members in Kinondoni Municipal Council, only two had university degrees. Moreover, all members were fairly experienced with 3 to 6 years in the boards. In Ulanga District Council, for example, the highest education level of the health facilities’ board member was a certificate of secondary education. Yet, board members’ education level is very important in implementing their responsibilities as placed up on them through various guidelines [56] .

Examination of findings from this study in light of Principal Agent Theory reveals that the decision space provided for governance and rules in the health sector is “narrow” as establishment, size, composition and tenure of the board members are centrally determined by Acts. Thus, community participation at this level is “passive participation”.

In both councils, community members revealed that they initiated construction of water facilities. Specifically, in Kinondoni MC, 40.8 percent of the respondents said that community members were responsible for initiation of construction of water facilities and in Ulanga DC, 22.2 percent of the respondents reported the same. At the household level, 10.2 percent of the respondents in Kinondoni MC said husbands and 2.2 percent of respondents in Ulanga DC said husbands were responsible for the construction. Moreover, 4.1 percent of the respondents in Kinondoni MC pointed out that wives initiated the construction of the facilities and 0 percent of the respondents mentioned wives in Ulanga DC. Yet, regarding women’s central roles on family decision regarding water, all water projects should have treated women as “valued customers” [8] . Sadly, in most African countries, women are not represented in decision-making organs regarding water management and service delivery and thus, as the results severely hindering the sustainability of the projects (ibid.). Furthermore, 44.9 percent of the respondents mentioned “others” to be responsible for construction of water facilities in Kinondoni MC, whereas 75.6 percent of the mentioned “others” in

Source: Field data (2014).

Ulanga DC. “Others” could be the Government, Non-Governmental Organizations (NGOs) or development partners.

Following the [7] typology of participation, community participation at this level is “Practical active-specific” in terms of labour. These findings corroborate with a demand-responsive decentralized water service delivery study in central Java in Indonesia by [57] who concluded, “Only if users were directly involved in service design and selection, were services likely to match users’ preference.” Inherently in this quotation is that informed participation by the local community will make the project more sustainable than the decision of the leaders alone. The findings are summarized in Table 3.

According to the [36] water user groups are the most important in the provision of the service. To ensure sustainability and professionalization of Water User Groups, the new Act of 2009 calls for registration of former water user groups with their respective councils to acquire the status of Community Owned Water Supply Organization (COWSO). The registered water user committees will automatically become COWSO [36] .

Once WUG is registered, it will automatically stop reporting to the village governments and acquire large autonomy to participate in water service delivery [36] . Findings from this study revealed that in Ulanga DC, only one Water User Group from Uponera ward had applied for registration and Lupiro ward had approved one WUG constitution at ward level. During the time of undertaking this study, the district registrar had already been appointed in Ulanga DC (UDC, 2013). In Kinondoni MC none of the Water User Groups was registered by the time of undertaking this study. When the respondents were asked about when they will start registering the Water User Groups in Kinondoni MC, one key informant had this to say:

“We have received a letter from Municipal Director instructing us to register all Water User Groups the latest by July, and we have, in turn issued guidelines for registration of all WUGs in Kinondoni by 2015” (Key Informant₂ Kinondoni, January, 2014).

Viewing findings of this study in the limelight of Principal Agency Theory water facilities in Ulanga DC and Kinondoni MC have “moderate decision space.” This is because majority of Water User Groups in both districts were not registered and therefore, they lacked full autonomy to deliver their full mandate.

Contributing to initial capital investment is the form of activity-specific participation [8] . Therefore, water users can contribute towards construction of water sources either in term of money or labour [58] .

In both studied councils, it was reported that communities contributed in construction of water facilities, specifically, 88.9 percent of the respondens in Ulanga DC and 49.0 percent of the respondents in Kinondoni MC. However, 51 percent of the respondents did not contribute in Kinondoni MC, whereas in Ulanga DC 11.1 percent of the respondents did not contribute.

Failure to contribute may be partly attributed to poverty level and high illiteracy rate among the beneficiaries [59] . One key informant from Ulanga DC had this to say to support such position.

“Lack of enough community members’ participation in cost-sharing is also attributed to their little understanding of water policy, which clearly stipulates that water is a commodity. Majority especially in rural areas are of the view that water is the gift from god! Why should they then contribute to access a gift? This is the big challenge in implementing water strategies in Ulanga” (Key Informant₃ Ulanga, January 2014).

The obvious inference from these findings is that communities in Ulanga district would have contributed more in the construction of water facilities if the water policy would have been known to them. Following the ladder of participation; community participation at this level is “practical and active-specific” in terms of money and labour. The findings are shown in Figure 2.

The study respondents indicated that community members participated in both councils. Specifically, 86.7 percent of the respondents in Ulanga DC and 61.2 percent of the respondents in Kinondoni MC were involved in the management of water facilities. At the household level, more men participated as indicated by 18.4 percent of the respondents than women as indicated by 10.2 percent of the respondents in Kinondoni MC, whereas in Ulanga DC men and women’s participation was equal as indicated by 2.2 percent of all respondents. Furthermore, 8.9 percent of the respondents in Ulanga DC and 10.2 percent of the respondents in Kinondoni MC mentioned “others” to be instrumental in the maintenance and management of the water facilities.

Respondents further indicated that management and maintenance of water facilities were done through routine inspection, routine cleaning and treatment of water sources. Specifically, respondents reported that they participated in routine cleaning, and distribution per council was as follows; 51.1 percent of the respondents in Ulanga DC and 38.8 percent in Kinondoni MC. Respondents also reported to participated in routine inspection in both councils, specifically, 28.0 percent of the respondents in Kinondoni MC and 15.6 percent of the respondents in Ulanga DC.

Similarly respondents participated in water treatment in the studied districts and the distribution per council was 6.1 percent of the respondents in Kinondoni MC and 24.4 percent of the respondents in Ulanga DC, and 18.1 percent of the respondents mentioned other means in both councils. Following [7] community participation at this level was “practical and active specific.”

Specifically, 6.1 percent of the respondents in Kinondoni MC reported that the government supported construction of the facilities, whereas in Ulanga DC, 6.7 percent of the respondents reported the same. Moreover, 10.2 percent of the respondents said NGOs supported construction in Kinondoni MC and 2.2 percent of the respondents reported the same in Ulanga DC.

The findings further indicated that 8.2 percent of the respondents mentioned local community in Kinondoni MC and 4.4 percent of the respondents in Ulanga DC reported the same. Majority of the respondents mentioned donors to be the major supporter in the construction of water supply systems as indicated by 86.7

Source: field data (2014).

percent of the respondents in Ulanga DC and 75.5 percent of the respondents in Kinondoni MC. Viewing the findings through [7] community participation at this level was “consultative-participation.” The findings are summarized in Table 4.