Exit Strategies for the Resettlement of Drought Prone Populations: Technical Report

EXIT STRATEGIES FOR RESETTLEMENT OF DROUGHT PRONE POPULATIONS

PROJECT TECHNICAL REPORT Simon Batchelor Kevin McKemey Nigel Scott Department for International Development 94 Victoria Street LONDON SW1E 5JL Contract No: R7136 April 2000

152 Cumberland Road Reading RG1 3JY UK

phone: 44 (0)118 926 7039 fax: 44 (0)118 935 1374 email: [email protected]

Gamos Ltd Final Technical Report- “Exit Strategies” R7136

© Gamos Ltd 2000

This document is an output from a project funded by the UK Department for International Development (DFID) for the benefit of developing countries. The views expressed are not necessaril y those of the DFID.

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page i

Executive Summary.

Introdu ction Throughout the world welldrilli ng in drought prone areas creates unnatural settlements which have a high dependency on the water supply. A number of agencies in Africa have been working with displaced populations (through conflict or environmental factors) to restart communities in home areas or in new areas. These programmes have involved creating water points which seem to be magnets for returnees. The communities become dependent on the new water supply and are vulnerable to its breakdown. This creates a challenge for leaving locally sustainable systems. Handover of maintenance from the external agency to local communities and local government is often difficult and it is at this point that much of the gains of the agency can be undone. Agencies use a variety of participatory approaches, vill age level maintenance structures, standardised pumps to fit government recommendations, organisation of spares supply and training of local government teams. This report presents the work of an interagency project funded by the British Government (DFID) which compares and contrasts recent variations on these approaches. The project is undertaking a thematic comparison of three case studies in Africa. This is for the practical objective of identifying common elements of a successful exit strategy for resettlement programmes that have introduced new water supplies. Organisation Care World Relief Concern

Universal Country Mozambique Mozambique Malawi Province/Area Inhambane Gaza Central and

Southern No of wells 170 163 172 Start date 1993 1993 1992 Average depth 60m 63m 53m Average dynamic water level 50m 46m 35m Each partner started with an emergency response to water needs, and evolved during the following years to a more “developmental” approach. The gradual shift on all programmes has been:- • a growing emphasis on community level pump maintenance, • increasing levels of cost recovery and • greater involvement of the private sector. These differences of entry (pre-installation) and exit (post-installation) strategies have allowed the research project to investigate relative importance of various elements in the strategies in creating local sustainabili ty. “ Sustainabili ty is the capacity to maintain service and b enefits both at the community and agency levels even after external agencies (i.e. NGOs) “ special assistance” (managerial, financial and technical) have been ph ased ou t.” (Evaluation o f the HESP/VLOM Programme, Concern Universal 1996)

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page ii

Methodo logy In each case about 15 vill age communities were visited. For each community there were both focus groups discussions and individual interviews resulting in a total of 305 useable responses (average of 7 per community). Interviews were kept as open discussions allowing the communities to define the important factors alongside the framework created from the literature. The data from the pump user and key informant interviews, and observations were coded and entered into computer software for analysis. Given the nature of the data, non-parametric tests were applied to identify possible relationships, i.e. Mann-Whitney test for significant differences, and Spearman’s correlations. These relationships were then used to highlight those important links to pump availabili ty and ownership. These relationships were further investigated using the qualitative data gathered during the interviews.

Results Results are presented based on an analysis of the impact of key social contextual and community organisational issues (Vill age level Operation and Maintenance variables) on the sustainable availabili ty of water from the borehole. The time a pump is broken (downtime) is the cumulative effect of the possible delays. The data is used to investigate each delay and state whether over the whole sample it is a defining element. In summary, the data suggests that where a local community has been mobili sed to repair their own pump, the downtime is reduced (compared to a more central approach) but the frequency of breakage increases. Thus the total reliabili ty of the pump is not necessarily increased by localisation of the repair process. Using the results to investigate social factors, confidence in local technical competence is found to have the strongest relationship with the sustainabili ty of the system. If there is competence in the locality (either within the community or nearby, possibly a private area mechanic) then the reliabili ty of the pump improves (minimal downtimes and low frequency of breakage). Participation of the wider community and organisation both contribute to competence but are not the defining factors. Good technical training is necessary to create sustainable systems. Another defining element to the process of repairing the pump is the availabili ty of spares. The agencies concerned have been developing a private enterprise supply chain. The combination of the private chain and the spares originally supplied during the emergencies has meant that spares availabili ty is not a significant problem for the communities. The results show that this is not a significant delaying factor in the repair process. However, this may not represent the longer term. Private suppliers do not seem to be making sufficient profit from the spares to motivate them and are currently selli ng spares as a community service. This may or may not be sufficient for the long term. The report encourages agencies to allow retailers a greater handling charge on the small i tems of pumps. Having placed greater emphasis on local technical competence, the reports draws attention to the fact that social mobili sation is valuable in itself as a prelude to community problem solving and future development activities. The literature also

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page iii

suggests that it has considerable value for long term health benefits. The results show that increased social mobili sation tends to lead to greater care of the wellhead, which may be an indicator of hygiene awareness. The results show that rules and sanctions set by a water committee do not encourage care of the wellhead. Those communities with less rules actually show a greater care and ownership for their pump. The results seem to indicate that dialogue and understanding are more important than legislation. The results demonstrate that even the simplest of involvements in installation of a pump, such as provision of materials or working on the road, generates a greater sense of ownership than otherwise. Even during rapid emergency programmes, involvement of the community should be made a priority. However, the results also suggest that some elements of an exit strategy can go some way to making up a poor entry strategy. An emphasis on technical training alongside training on management, finance and health, can bring a community to ownership of the pump. Competence will be formed by technical knowledge, which in turn is influenced by the length of training, the level of training and the number of trained mechanics. “Length of training – frequency of breakage and downtime are both reduced with longer training. More importantly; “ emergency pumps” which have had longer training tend to have less frequent breakdowns and shorter downtimes than “ development” pumps. Ownership was increased with the longer training for emergency pumps and the ownership levels were almost the same for emergency pumps with more than 1 week training and development pumps.” This statement is made of the sample of Afridev pumps only (Volanta pumps excluded) Key Findings:- There is a need to balance the training given to include suff icient technical focus. Building the capacity of a committee on subjects such as community mobilisation and mechanisms of community management enhances long term development. However, there is no substitute for technical training. Technical competence influences the sustainability of the pump. There is also no substitute for gett ing the community involved early on. Even the simplest of involvement, such as clearing roads or providing materials, enhances the future sense of ownership. To be inclusive in the installation process does not necessarily takes a lot of time and it should be possible for even emergency programmes to be mindful of exit during the entry process. External agencies as part of their exit strategies should set in place mechanisms for ongoing support. There should be a recognition that communities will need assistance for complex (albeit infrequent) second level repairs. Building the capacity of the private sector or local water institutions to be able to provide this technical support is part of a realistic exit strategy. Community mobilisation for problem solving, including Village Level Operation & Maintenance of the water supply, is a long term process. Communities need long term encouragement to maintain the motivation and develop the skills required for development. Social mobilisation of the communities may (will l ikely) come from local institutions other than water departments staffed by engineers (eg health and Community services). Local institutions able to provide ongoing social

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page iv

mobilisation should be identified and enabled to include the water system in their programmes.

Exit strategies The report gives a number of strategies addressing noted exit issues, based on the framework of concepts developed from the literature and partner interviews. The Table in Section 7 of the main report presents the options that are available to an external agency considering exiting an area. In addition, the report raises the question of funding exit strategies and of creating a supply chain for spares. Regarding funding and policy:- The case studies suggest that both funding and implementing agencies working with water supplies and/or having integrated programmes on water, sanitation and health, should not abandon water points that were the outcome of an emergency or resettlement programme. They should work with the local authorities and local communities on a planned exit strategy to turn the water points into sustainable supplies owned and managed by the local population. Exit strategies need not cost a lot of money although the case studies seem to indicate that to stimulate effective ownership of a water supply and to put in place the technical competency required to locally manage a borehole, costs about the same amount as the original hardware of the water supply. Emergency water supplies are often put in by agencies with considerable sums of money at their immediate disposal (funds raised to save human life in response to the emergency). Development agencies often have a lot less money for disposal. Emergency agencies could put aside an amount equal to the original cost of putting in the hardware for a longer term planned exit strategy. Regarding the supply chain:- private suppliers should be given the opportunity to make a profit by placing a reasonable handling charge on small i tems, and not being undermined by free gifts from local government or agencies. This supply of spares should not need to be subsidised at any stage of the long term process either as an incentive to the suppliers or to the distributors. The research indicates that communities can raise finance for spares if they are committed to the water supply. Private suppliers could also undertake the more complex repairs and become “area mechanics” with good technical competence providing ongoing technical support to the community. Training by the agency can be offered during the exit process perhaps for a nominal training fee to ensure commitment by the new businesses.

Recommendations The report makes the following recommendations

• Agencies that have been involved with water supplies during emergency or resettlement programmes should have planned exit strategies regardless of their entry strategies. A planned exit strategy can enhance sustainabili ty.

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page v

• Donor agencies should be prepared to fund follow up work after an emergency or resettlement programme. Strategic inputs to a planned exit strategy can be a cost effective mechanism for creating sustainable water assets in an area.

• Donor agencies should create policies which link departments concerned with

emergency work with those for longer term development so that it is clear which department should fund the necessary exit strategies.

• Agencies should note that a key factor in creating a sustainable water supply is

to have adequate technical and management competence within the locality. Competence has a greater influence on the total availabili ty of water than the details of community organisation and wider community participation. Limited budget or resources for an exit strategies should be used to enhance technical competence especially if traditional forms of community organisation are effective.

• Agencies wishing to exit an area must acknowledge that some ongoing support

is required. Ongoing technical support is required for diff icult technical repairs and ongoing institutional support is required to encourage ongoing social mobili sation. Those responsible for water are not necessarily the best local authority or agency for the social mobili sation and other departments such as health or community development should be considered. Agencies should make strategic links and build capacity to leave some form of ongoing support. The private sector is a strong possibili ty in present day Africa for both the supply chain and some form of area mechanic.

• Exit strategies for sustainabili ty should include the following practical

elements:- �

Social profili ng – to know in advance of potential conflicts within the community. �Work with existing structures if they exist. Do not attempt to introduce new democracy – let the vill age community suggest the most appropriate way of selecting people for training, committees and care-taking �Training of the community selected committee in technical, management, financial and health skill s with sufficient emphasis on technical skill s to ensure a competent repair process. �Training on the financial mechanisms should leave the community to decide if regular payments is the best mechanism for their community operation and maintenance. (Collecting quotas at the time of breakdown does not tend to delay repairs). �There is a need to save towards replacement of the system – strategic connections between the community and the financial service sector may need to be made, this will vary based on the country context. �Training on management should not be dogmatic on the need for rules and sanctions. Rules can discourage ownership by the general community. Training should include discussion of the most appropriate mechanisms to manage the pump. �Training of as many people as possible (within a community and within budget) in those skill s.

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page vi

�Training of both men and women in all the skill s. (Do not reinforce stereotypes) �Work with the government and other agencies to ensure a supply chain of spares. Preferably the supply chain should result in spares being available to the community from a local retailer. �Capacity building of local government or private enterprise, to form a group that can undertake more difficult repairs on an ongoing basis after the agency has exited. �Work with local government other than just the Water Department. Health and Community Development (Youth, Gender) workers often have more resources (time and capacity) to visit communities regularly. Agencies should encourage local government workers to encourage communities to sustain their motivation.

Recommended immediate actions • The current Guidance notes circulating in DFID do not seem to address the

overlap between an emergency WS&S programmes, resettlement or rehabili tation, and longer term development WS&S programmes. Guidance notes based on this research should be disseminated to complement and support existing literature.

• If possible, review DFID policies and guidelines surrounding the transition from emergency WS&S programmes to longer term development. This may require a workshop/consultation with relevant stakeholders to discuss the findings and suggest practical administrative changes.

• Guidance notes targeted at different institutions should be generated based on the information in this report, eg a guidance note for national governments experiencing an emergency and for NGOs working with WS for resettlement and development

The research has identified the important factors that make up a good exit strategy that leaves the water point as sustainable as possible within the country context. While there is no substitute to a good participatory entry strategy, this gives some hope for retaining the assets of an emergency or resettlement programme beyond the life of the programme and presence of the external agency.

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page vii

Acknowledgements This research and action project has been a joint effort of many people and agencies. The main partners were Care, Concern Universal, World Relief and Gamos. Although many individuals deserve mention, it is not possible to list everyone who contributed. Accordingly, we would like to record here the input of the Care team (Mozambique) led by Patrick Sayer and Fernando Pilil ao, the Concern Universal team (Malawi) led by Samson Hailu, Bruck Wolde and Robert Kampala, and the World Relief team (Mozambique) led by Galen Carey and Johnson Nghatsane. Special thanks to Juma of World Relief. The many people from the communities who gave up their time to respond to our question deserve thanks. We are grateful to the Governments of Mozambique and Malawi for their co-operation, particularly the local officials who gave us considerable time in the field, Francis Devison who travelled from Malawi to Maputo for the workshop, and Alberto José Cumbane.

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page viii

Glossary Afridev VLOM pump suitable for manufacture in developing countries. availabili ty the probabili ty that equipment will be in operating condition Care Care (NGO) CU Concern Universal (NGO) CWSP Community Water and Sanitation Project DFID Department for International Development, UK Down time period for which the pump is out of service when it does break down FAO Food and Agriculture Organisation of the United Nations Frequency of breakage

Alternative representation of mean time before failure

GOM Government of Mozambique Grameen Bank

Provides credit to the poorest of the poor in rural Bangladesh without any collateral

HESP Hygiene Education and Sanitation Promotion (Malawi) HSA Health Surveill ance Assistant (Malawi) Mk Malawi currency Mt Mozambique currency MW Mann-Whitney statistical test Naira NGO Non Governmental Organisation O&M Operation and maintenance PNA Politica Nacional de Aguas, National Policy for Water of

Mozambique PRA Participatory rural appraisal PRA Participatory Rural Appraisal SCF Save the Children Fund SCF Save the Children Fund (NGO) UNDP United Nations Development Programme UNICEF United nations Childrens Fund VLOM Vill age level Operation and Maintenance Volanta Deepwell pump developed in the Netherlands. WEDC Water, Engineering and Development Centre, at Loughborough

University, UK WES Water and Environmental Sanitation WHO World Health Organisation WMA Water Monitoring Assistant (Malawi) WR World Relief WTP willi ngness to pay

Gamos Ltd Final Technical Report- “Exit Strategies” R7136 Page 1

1 INTRODUCTION........................................................................................................5

1.1 OBJECTIVES OF RESEARCH ......................................................................................5 1.2 SUSTAINABILITY .....................................................................................................5 1.3 THE PROBLEM .........................................................................................................5 1.4 RESEARCH RESPONSE..............................................................................................6 1.5 METHODOLOGY OVERVIEW .....................................................................................7

2 DESCRIPTIONS OF PROGRAMM ES......................................................................8

2.1 NATIONAL CONTEXT - MOZAMBIQUE.......................................................................8 2.1.1 World Relief ................................................................................................... 8 2.1.2 Care Mozambique ..........................................................................................9

2.2 NATIONAL CONTEXT – MALAWI ............................................................................11 2.2.1 Concern Universal........................................................................................12

2.3 GENERAL COMMENTS............................................................................................13

3 BACKGROUND FOR RESEARCH.........................................................................15

3.1 GENERAL LITERATURE SURVEY .............................................................................15 3.1.1 Operation and maintenance (VLOM)............................................................15 3.1.2 VLOM assumptions ......................................................................................16 3.1.3 Community Participation..............................................................................18 3.1.4 Water Committees and Management.............................................................19 3.1.5 The role of women........................................................................................21 3.1.6 Training .......................................................................................................22 3.1.7 Links to local government, private suppliers and community ........................23 3.1.8 Technology Choices......................................................................................24 3.1.9 Emergency Rehabilit ation Environments......................................................25

3.2 DFID GUIDANCE MANUAL ...................................................................................27 3.2.1 Participation ................................................................................................27 3.2.2 Cost recovery – a sign of sustainabilit y?.......................................................28 3.2.3 Private suppliers...........................................................................................28 3.2.4 Sustaining community based operation and maintenance..............................29 3.2.5 Agency links................................................................................................. 30 3.2.6 Technology................................................................................................... 30 3.2.7 Measuring sustainabilit y?.............................................................................31

4 METHODOLOGY .................................................................................................... 32

4.1 BASIS FOR THE INVESTIGATION..............................................................................32 4.1.1 Data gathering.............................................................................................32 4.1.2 Analysis........................................................................................................32 4.1.3 Verification................................................................................................... 32

4.2 SAMPLE ................................................................................................................33 4.2.1 Sample Size .................................................................................................. 33 4.2.2 Sampling Process.........................................................................................33 4.2.3 Description of acquired sample.................................................................... 34

5 FINDINGS..................................................................................................................37

5.1 THE CONCEPTUAL CONSTRUCT .............................................................................37


5.2 VERIFICATION BY THE PARTNER AGENCIES.............................................................40 5.2.1 Key Issues.....................................................................................................40 5.2.2 The question of community involvement ........................................................40 5.2.3 The question of finance.................................................................................41 5.2.4 The question of training................................................................................41 5.2.5 The question of technology............................................................................42 5.2.6 Agency constraints........................................................................................43 5.2.7 Partner suggestions for data framework .......................................................44

5.3 VERIFICATION FROM CONFERENCE DELEGATES .....................................................45 5.4 SUSTAINABILITY ISSUES FOR FRAMEWORK OF RESEARCH.......................................45

6 RESULT S AND ANALYSIS.....................................................................................50

6.1 GENERAL DISCUSSION ...........................................................................................50 6.2 PROGRAMME OUTCOME - PUMP RELIABILITY .........................................................50 6.3 BUILDING THE PICTURE ON AVERAGE DOWNTIME. ..................................................53

6.3.1 Diagnosis......................................................................................................53 6.3.2 Accessing the mechanic ................................................................................53 6.3.3 Funding the repair ........................................................................................54 6.3.4 Mechanic capabilit y......................................................................................54 6.3.5 Spares availabilit y ........................................................................................55 6.3.6 Other factors highlighted by the data............................................................57

6.4 FREQUENCY OF BREAKDOWNS...............................................................................59 6.4.1 Reliance on mechanics..................................................................................60 6.4.2 Use of the pump............................................................................................61 6.4.3 Initial involvement ........................................................................................61 6.4.4 Involvement and Ownership..........................................................................62

6.5 SOCIAL FACTORS...................................................................................................63 6.5.1 Section introduction......................................................................................63 6.5.2 Competence.................................................................................................. 65 6.5.3 Organisation................................................................................................. 66 6.5.4 Participation of the wider community ...........................................................68 6.5.5 Section conclusion........................................................................................69

6.6 OUTCOME INDICATORS FOR HEALTH AND HYGIENE................................................69 6.7 PROGRAMME HISTORY - THE INFLUENCE OF AN EMERGENCY CONTEXT .................72 6.8 PROGRAMME DESIGN - CAN TRAINING MAKE UP FOR A WEAK ENTRY STRATEGY?...74

7 EXIT STRATEGIES – A DISCUSSION...................................................................76

7.1 IS EXIT ACTUALLY POSSIBLE?.................................................................................76 7.2 ESSENTIAL ELEMENTS OF AN EXIT STRATEGY .........................................................77

7.2.1 Participation................................................................................................. 77 7.2.2 Organisation................................................................................................. 78 7.2.3 Training (Education) ....................................................................................79 7.2.4 Context .........................................................................................................80 7.2.5 Optimal use of water .....................................................................................81 7.2.6 Settlement .....................................................................................................81 7.2.7 Environmental Sustainabilit y ........................................................................81

7.3 FUNDING OF EXIT STRATEGIES ...............................................................................81 7.3.1 Who funds the exit?.......................................................................................81 7.3.2 How much to spend on an exit strategy? .......................................................82 7.3.3 Supply of Spares...........................................................................................83


8 CONCLUSIONS AND RECOMM ENDATIONS.....................................................89

8.1 RECOMMENDATIONS:-...........................................................................................91 8.1.1 Recommended immediate actions ................................................................. 92

9 REFERENCES ..........................................................................................................94

10 APPENDIX MAPS OF MALAWI AND MOZAMBIQUE.................................. 97

11 APPENDIX - LOGICAL FRAMEWORK OF THE RESEARCH ......................99

12 APPENDIX RESPONSES FROM WEDC CONFERENCE...............................100

12.1 BOREHOLES AND PROGRAMMES ...........................................................................100 12.2 KEY ISSUES..........................................................................................................101

12.2.1 Key activities leading to local sustainabilit y ................................................102 12.2.2 Principal blockages.....................................................................................103

12.3 SUMMARY OF KEY ISSUES....................................................................................104 12.4 EMERGENCY INSTALLATIONS ...............................................................................105 12.5 REASONS GIVEN ...................................................................................................105

13 APPENDIX VALID PERCENTAGES OF VARIABLES..................................106

14 APPENDIX - DATA ANALYSIS – WELL PERFORMANCE INDICATORS.116

14.1 SUMMARY TABLE ................................................................................................116


1 Introdu ction

1.1 Objectives of Research “To identify elements of a successful exit strategy for water programmes with resettlement rural populations in Southern Africa.”

The research project considers three case studies in southern Africa of water provision for large rural resettlement populations in drought prone areas. It compares and contrasts the approaches investigating the elements of a successful handover to local government and communities. This work is intended to enhance programmes working with vulnerable resettlement communities in southern Africa. The risk of an inadequate exit strategy is that any gains in quality of life made during the programme are threatened by failure of the water supply once the external agency has withdrawn. The project result is intended to be a clear analysis of the dangers and opportunities presented to agencies who complete their programmes and withdraw from an area handing over to local communities and local government. The project has attempted to identify strengths and weaknesses of various approaches. In particular it has looked at capacity building of the local communities, resources and local government. The intention of this output report is that it may be used throughout the world with water projects and programmes that deal with displacement and resettlement as a “ lessons learned” document.

1.2 Sustainabili ty “ Sustainabili ty is the capacity to maintain service and b enefits both at the community and agency levels even after external agencies (i.e. NGOs) “ special assistance” (managerial, financial and technical) have been ph ased ou t.” (Evaluation o f the HESP/VLOM Programme, Concern Universal 1996)

1.3 The problem Throughout the world borehole drilli ng in drought prone areas creates unnatural settlements which have a high dependency on the water supply. This is true particularly of southern Africa where there have been a number of large resettlement programmes. Agencies have been working with displaced (through conflict or environmental factors) populations to restart communities in their home areas or in new areas. These programmes have involved creating new water points or rehabili tating derelict ones. These new water points are sometimes magnets for returnees. The communities become very dependent on the new water supply and are vulnerable to its breakdown. The overall population densities are low and the areas of the programme sometimes vast (e.g. Mozambique World Relief programme 400 by 200 miles). This causes a tremendous challenge for creating sustainable maintenance systems. Handover of maintenance from the NGO to local communities and local government is often difficult and it is at this point that much of the gains of the agency can be undone. Reductions in Malaria, Bilharzias, increases in general child health, increases in food security, even some income generation activities are all dependent on the water supply.


In “development” activities agencies (mainly) install the pump after a participatory approach, they mobili se the people and create vill age level maintenance structures. This entry strategy is seen as critical to gaining the ownership of the people of the new water point. This mobili sation of the community is complemented by other technical and policy factors. The agencies often use standardised pumps to fit government recommendations, organise the flow of spares and train local government teams. However. even with this apparent coherency of (developmental) programme approach, end of programme evaluations often question the sustainabili ty of the water system, and point to the vulnerabili ty of the communities to the future. They note that communities vary in their stated ownership and capacity for caring for the pump. In emergencies or (rapid) rehabili tation, the participatory approaches are sometimes (often) compromised. The urgency of helping populations to have an acceptable water supply causes agencies to take rapid and often centralised decisions. These emergency water points therefore have even less chance of survival because of ownership and support issues resulting from non-participatory entry strategies. This research considers the difference between water points that were installed with rapid entry and those that have been undertaken in the same region with a more participatory entry. The case studies involve three programmes that started as emergency programmes and have evolved into development programmes. . The logical framework for the project is given in appendix 1

1.4 Research Respon se The basic methodology concerned:-

• thematic evaluation of three water programmes in Southern Africa • analysis of common factors which have led to effective handover • workshop attendants endorsement of common factors

This included an initial li terature review, three project visits including focus group discussions and a workshop of the major players in case studies. In addition, attendance at the WEDC Water Conference was used as an opportunity to hold wider discussions with practitioners. The project undertook a thematic comparison of three case studies in Southern Africa. This was for the practical objective of identifying common elements of a successful exit strategy for resettlement programmes that have introduced new water supplies particularly those dependent on deep bore systems. The three programmes have each respected government standards and introduced water pumps according to the dictated specifications. They have trained the community in maintenance procedures, organised the communities to lobby local government for spares, created precedence of good maintenance and ensured as far as is possible access to parts etc. However there are some variants in the approach. The three programmes involved in the research project are:

• World Relief, Gaza province, Mozambique. • Care, Inhambane Province, Mozambique


• Concern Universal, Malawi Organisation Care World Relief Concern

Universal Country Mozambique Mozambique Malawi Province/Area Inhambane Gaza Central and

Southern No of wells 170 163 172 Start date 1993 1993 1992 Average depth 60m 63m 53m Average dynamic water level 50m 46m 35m A map can be found in the Appendices. Due to the nature of the research and the central question, this report does not contain a comparison of the partner agencies. The differences of approach vary between agencies but also within agencies over time and between districts. An agency may have a well motivated team in one location and a less motivated team in another. Similarly they may have applied a particular entry and exit strategy during a two year programme and then modified their strategies for the subsequent years. During the analysis, data was checked for agency differences to eliminate elements that might be agency specific. However, the research group took a decision not to discuss those agency specifics.

1.5 Methodo logy overview The research project visited two partners in Mozambique and one in Malawi in 1999. In each case about 15 vill age communities were visited. For each community there were both focus groups discussions and individual interviews resulting in a total of 305 useable responses (average of 7 per community). Interviews were kept as open discussions allowing the communities to define the important factors. A more detailed description of the respondents profile and the construct is given in section 4. The data from the pump user and key informant interviews, and observations were coded and entered into computer software for analysis. Given the nature of the data, non-parametric tests were applied to identify possible relationships, i.e. Mann-Whitney test for differences, and Spearman’s correlations. These relationships were then used to highlight those important links to pump availabili ty and ownership. These relationships were further investigated using the qualitative data gathered during the interviews.


2 Descriptions of programmes

2.1 National context - Mozambique During the emergency phase (1992 – 1995), the responsibili ty for maintenance and most construction of rural water supplies lay with the Government of Mozambique (GOM). The 1997 evaluation for Care states “supply users regarded the systems as belonging to government and expected government to look after them and repair them, a role the GOM did not have the resources to fulfil. The result was a system which relied heavily on outside funding, with little or no ownership or financial contribution from consumers, and which quickly fell into disrepair.” In 1995 the GOM adopted a new policy (PNA) which “accentuates not only the sustainabili ty triangle, but also the organisational triangle with the consumers at the apex.” This plan emphasises the role of participation by the users at all stages, and the role of the private sector in improving and maintaining water supplies. Both Care and World Relief have attempted to follow the national water policy, although Care has been taking a greater lead in the development community responding to the PNA plan. It has not been easy to make the transition to community managed systems and the response varies from area to area depending on local authorities. Care in particular has taken the lead with pre-payment for pump-heads by communities and “stopping of free distribution of spare parts which were undermining the fragile but growing private sector spare parts stocking and sale”. Given this evolution of responses with Mozambique both World Relief and Care have experience of an emergency response to the need for water supplies and of making a transition to a more “developmental” response In Mozambique the official standardised pumps are the Afridev for up to 45m and the Volanta for over 45m. Both are well known pumps. Details of the Afridev and Volanta can be found in an appendix. The Volanta has some technical diff iculties associated with it (eg cracking of pipe joints). World Relief persevered with the Volanta modifying the way it was installed to try to address the technical shortfalls. With the combination of the Volantas poor performance, plus the fact that the spares are relatively expensive and it would be difficult to set up a sustainable supply, Care decided that it was best to install Afridev pumps even to 60m. They reasoned that regular breakdowns of a less expensive pump tend to be better maintained than those with few but costly repairs.

2.1.1 World Relief World Relief’s Emergency Water Supply Programme was a response to the worsening drought situation along the Limpopo Corridor in the province of Gaza during 1992. The programme was designed to complement the governments rural water supply programme for Gaza and was an attempt to rapidly meet the needs of the population by providing access to clean water through the drilli ng of boreholes and installation of handpumps in accordance with government guidelines. Between November 1992 and July 1993 World Relief subcontracted a drilli ng company, drill ed 41 boreholes and installed 29 handpumps in two phases. This represents a success rate in what was considered an unfavourable area for groundwater development of over 70%. The average direct cost of the borehole during this phase was $6,750. The average total cost per borehole with pump was $16,041.


The response of World Relief took into account the shifting needs of the population. During the emergency major settlements had developed along the Limpopo valley corridor. By strategic placement of water supplies at the more remote areas, on the heights of the valley sides, there was a planned intention to facili tate resettlement over a greater area. The report for this stage states :- “Although emergency in nature, one of World Relief’s objectives was to ensure sustainabili ty of the water supply in the long term. This was achieved through training of members of the community in pump management and maintenance and creating awareness within the community of the pump and other water related issues.” However, in practice during this phase trainings were short and the pump repair relied on an area mechanic. Following that initial programme, World Relief undertook a number of amendments and extensions to its programmes, such that by 1995 it had installed 167 handpumps in the area. For most of this period the training focussed on selected individuals who would maintain pumps throughout the localised area. In addition there was a vill age pump attendant for basic pump supervision. During an evaluation in 1997, it was noted “The maintenance team are repairing on average 20 to 23 pumps per month. One of the reasons for this is the excess use most are receiving. Many of the communities have grown rapidly since the pumps were installed.” We should note in these statements the reliance on a mobile team and the pressure on the boreholes. The 1997 evaluation continues, “There is a great deal of concern regarding the future of this maintenance programme.” Since that report was written World Relief have closed their water programmes. 1997 and 1998 were years spent retraining vill agers, and attempting to create a sustainable supply of spares. This research project occurred at a time when World Relief was in the final process of withdrawing as an agency from supporting repairs and maintenance. It was having discussions with local government about the equipment it had and whether all equipment should be handed over to the local authorities. However it continued to be in the area for other development activities.

2.1.2 Care Mozambique Care started an emergency water supply project in Inhambane province in 1993. As with World Relief, Care made a rapid response to a drought stricken population who were also suffering the after effects of civil war. In that first phase approximately twenty functioning handpumps were installed. A second phase started in January 1995 called a Community Water and Sanitation Project (CWSP). “This phase was planned to be for three years and has concentrated on development issues, particularly those relating to sustainabili ty and encouragement of demand responsive approaches” . The 1997 evaluation continues – “The move from emergency to development is seldom easy, as it requires the change of attitude of all involved, at community, project and government levels, and thus evolves gradually, rather than being an instantaneous transformation. During this evolution, CARE has encountered and overcome several problems which result from being an organisation at the forefront of such a change….”


The emergency proposals state that there would be community involvement during the siting of the pump, and that the community would be trained in VLOM. Inhambane was one of the first provinces to begin to recover from the war, and as a result communities in this region were better established. Even during the emergency phase, CARE was working with relatively stable communities. During the CWSP phase, a high demand for water and lack of alternative sources combined with the stabili ty of the population to present favourable conditions for demand driven responses. By 1997 there were 400 wells and boreholes in the four districts where Care works. Of these approximately 170 have been constructed or rehabili tated through the CARE projects. The direct costs of a borehole in the 95 to 97 period were on average $9,000 per community. In addition, if the total project costs are taken into account (facili tation and overhead) this figure rose considerably to give high overall costs. These costs include all the social mobili sation such as organising and facili tating prior contribution, post installation hygiene and sanitation education, etc. As stated in the national context section, Care took a lead in implementing the PNA National water policy. This was not an easy task. New communities who wanted pumps were asked to purchase the pumphead from a local supplier. This “prior contribution” before drilli ng started was intended to stimulate ownership within the community for the pump. A secondary feature of this approach was to ensure the community made a connection with the local supplier and therefore knew where to get spares. This system was supported by training programmes for VLOM (although these had been in place before). It is interesting to note that the 1997 evaluation comments “Fund raising is well accepted by the communities if not always by the local politicians. I t has helped strengthen a feeling of ownership…..” .


CARE Inhambane Project, Mozambique CARE’s Afridev project in Inhambane has monitored pump performance since 1993. Many of the pumps have been installed to depths which are over 45m and serve more than 100 user families. In-country manufacture of pumps by a Dutch company has been established, with third party inspection, and the sale of spares has been commercialised since 1996. Originally, communities only had to pay for routine preventive maintenance with the government or NGO providing major repairs for free. These were needed frequently given the large number of users and installation depths. In 1996 CARE began transferring full maintenance responsibility to communities by training ‘maintenance and repair groups’ (MRGs) at village level. These were to deal with all repairs including the removal and repair of rising mains, which sometimes cracked. By December 1997, 97% of the 135 handpumps in the project were functioning and had an average downtime of less than 10 days. 77% of breakdowns were repaired wholly or partly by communities, 19% needed some external assistance and 4% were done by local government workshops or CARE. Lessons from the Project • Communities had the capacity to pay for routine maintenance and repair, but it was uncertain whether they could pay for major repairs such as replacement of the entire rising main. The involvement of the private sector helped in making unsubsidised spares available at acceptable prices. • A backstopping system (a ‘second tier’) of local mechanics was recommended in addition to MRGs to support communities in those activities requiring more confidence and skill such as replacing the rising main, or fishing for dropped pipes and rods. • A suitable technical solution for deep installations was needed due to a big increase in breakdowns at pump depths over 45m, of which pipe joint failure was a common cause. (Obiols and Baumann, 1998, quoted from Colin 1999)

2.2 National context – Malawi In 1992/93 an estimated 2 milli on Malawians experienced severe water shortages as a result of drought. A survey by Save the Children (SCF) during the 92/93 drought showed that 40 to 60% of the boreholes provided were out of use at any given time due to inadequate capacity for maintenance by the governments repair teams, lack of spares and lack of transportation. It was SCF that first initiated district wide community based management training for beneficiary communities in Malawi. The training covered technical operation and maintenance skill s at community level, leadership skill s, mechanisms for fund raising and financial management, and hygiene education and sanitation promotion. 1993/94 was a year of more drought in Malawi and some agencies continued to make a mixed response. 1988 to 1993 also saw the provision of water points to Mozambique refugees hosted by the Government of Malawi in refugee impacted districts. After the repatriation of the Mozambiquans, a total of 900 water points remained for use by Malawians in all the refugee impacted districts. UNHCR discontinued its centralised care and maintenance teams for the water points, and again about 40-50% were out of order at any given time thereafter. The Afridev is the standard pump for boreholes up to 45m in Malawi. On shallow holes the Naira is installed.


Introdu cing VLOM in Malawi: The Karong a Experience The Karonga project provided Afridev pumps to a population of 60,000; subsequently, four years were devoted to motivating and preparing the community for VLOM. It thereby served to test whether, given the right support, a community can manage their own water supplies using the VLOM approach. In each village a water and health committee was established in addition to one committee per pump with at least two trained attendants for repair and maintenance. High-demand spares were supplied by a local wholesaler to village shops, while the government’s role was restricted to monitoring and policy-making. The intention was that even if the government did not fulfill its role, the pumps could still be maintained. When the project began the Afridev had not been fully tested and some parts had to be changed on all pumps after installation. Appropriate tools had also not been designed. These problems were eventually resolved and in 1997, two years after all project support had ended, a study found that:

� 95% of the handpumps and boreholes were working, with 75% working “well”; � communities repaired their handpumps, even replacing pump rods and rising main sections. However, they did no preventive maintenance or repairs to aprons and headworks;

� half of the pumps had either no or only one attendant. In most cases, at least one of the four essential tools had been lost or was never issued;

� most village water and health committees were defunct but pump committees remained active;

� communities had small amounts of cash on hand for maintenance but did not buy spares in advance of a breakdown. No regular contributions were made to maintenance funds; and

� village shops no longer sold spares, as there was too little business. Town-based wholesalers, however, still sold them. From Colin 1999

2.2.1 Concern Universal The combined forces of UNICEF, National Water Department and Concern Universal (CU) equipped Chikwawa District with 100 functioning boreholes through rehabili tation and new boreholes. While there was some consultation with communities, it was in 1994 that CU worked with UNICEF, the Water Department, Ministry of Health and Ministry of Women and Children Affairs and Community Services, to undertake a two year programme with two key training components. Hygiene Education and Sanitation Promotion (HESP) and Vill age Level Operation and Maintenance (VLOM). In 1996 CU obtained funding from Department for International Development (DFID) UK to undertake a Water and Environmental Sanitation (WES) project. Its goal was to mitigate the recurrent effects of the droughts and improve health of targeted communities. It was to provide sustainable water supply systems and promote improve sanitation and hygiene practices. This work continued to be in partnership with the government of Malawi in Dedza, Ntcheu and Chikwawa districts. This programme had a planned component of community mobili sation. Participatory Rural Appraisal (PRA) techniques were used to involve the community in the planning and implementation of the project. For this two year project 1996/98 CU installed or rehabili tated successfully approximately 138 boreholes and pumps. CU used a Eureka small portable drilli ng rig for much of its work. This innovation can be found described in “Introduction of Low Cost Rotary Drilli ng to Malawi: Specifically the Eureka Port-a-Rig” Concern Universal, BOND 1998. The paper notes that -“Given the present coverage and technologies available, Malawi urgently needs to


find a means of accelerating it’s population’s access to clean water and adequate sanitation facili ties. Investigations are necessary to find the technologies that are quickest and most cost effective for installi ng both water and sanitation facili ties." The main advantage of the rig is “Low Cost. The nominal cost per hole using the Port-a-rig currently stands at approximately £1,200 or MK36,000. This is a very considerable saving over the use of commercial contractors, where total construction costs exceed MK100,000. The Port-a-rig can replace a good number of holes traditionally drill ed by commercial contractors.”

2.3 General comments As we can see the programmes are similar and have a similar evolution from an emergency to development approach. However, there are also differences in the details of the approaches. Between them and over the years, the agencies have undertaken, or have been involved with communities who have interacted with other agencies, such that the case studies cover a wide variety of approaches and factors. For instance:- On the number of mechanics trained:- WR had mechanics during their early programme but was now discouraging them in preference for VLOM. However, Care has come full circle and now sees that there is need for some competence at an area level to undertake the more difficult repairs. Concern Universal in Chikwawa trained the local grocery owners to stock spares and to understand how to do the second level repairs. Care in their strategy are not training single mechanics but a group who are able to do other things such as lobby local government. The committee size and constitution varies from agency to agency. The words that one agency uses to distinguish roles are caretakers or committee. Caretakers do the work of maintenance and repair. For Care the executive committee do not do the work, they decide that the pump needs repairing and collect the money, an operational committee or maintenance group do the actual repairs, another group are for cleaning etc. Care currently encourages 12 in the committee with 6 doing the fund raising and 6 doing the O&M. CU in Malawi has 10 people who all tend to do the same thing, although only 3 are officially trained in mechanics. The role of government varies. In Malawi the work was co-operative with the Government and the Health Surveill ance Assistant (HSA) have become a vital component of the ongoing support to the VLOM system . The role of the HSA in the Malawian system is important. They relate to the Ministry of Health as opposed to the Ministry of Water. The HSAs have responsibili ty for only 5 or 6 vill ages each where as the official water mechanic from the Ministry has about 100 vill ages in his (her) responsibili ty. The HSA makes proactive visits the vill age regularly while the Water Monitoring Assistant (WMA) tends only to respond to requests for a visit. Country/cultural differences occur . The committee have to feedback the expenses to the community in Mozambique, while in Malawi they seem to rely much more on the natural meetings that generally occur rather than have special “pump” meetings. In Mozambique the agencies encourage a more planned public accountabili ty. So while in Malawi, CU has training sessions on financial management, the feedback is more integrated with vill age life e.g. funerals, weddings, general gatherings; they also use the church and school. In Mozambique if they need a meeting then the person by the pump tells the


people to come. Regarding motivation and payment, the person by the pump gets some benefits – free water, labour for their fields – and notes who is getting lot of water etc. Some differences between the countries are minor such as the details of tracking water payments. In Malawi the communities have cards for who has or has not paid and the cards were designed by the community. In Mozambique they also use the card system. They have a community member sit by the pump and collect payment, sign the card and/or list. Community differences:- For some variables, such as storage of money for spares, the situation changes for every community. In Mozambique some communities have treasurers who keep the money. To store the money they sometimes (but not often) use the bank. Alternatively they put it with a man who has a lot of animals so they can reclaim the animals if something goes wrong or they spread it among a few people so that if someone runs away they only lose a small part of it. It is not necessary to document all the different details of the systems covered by the research project. The preferred options are covered in the results section.


3 Background for Research This section explores the background for the research project. It makes a general li terature survey. However in order to ensure relevance of the research to current practice a section on The DFID Guidance manual for water supply and sanitation Programmes (WS&S) is presented. This is a special case in the literature taken to ensure the research results are of maximum use within DFID and its partners.

3.1 General li terature survey This review of literature on rural water supplies concentrates on issues of sustainabili ty of water supplies and water supplies installed during emergency conditions. The principal topics currently under debate in the literature relate to the management of water supplies. Issues of participation influence project planning, management structures and maintenance provision. The involvement of women is an aspect of participation that has received great emphasis, but is subject to assessment and scrutiny. Training is flagged as necessary in order to enable communities to participate effectively. It is now widely recognised that technology can be a critical factor in the sustainabili ty of water projects, and it appears to be widely accepted that beneficiaries should be expected to pay for water services. This section presents some of the arguments in recent literature on what appear to be the most important topic areas.

3.1.1 Operation and maintenance (VLOM) The review starts with VLOM for Rural Water Supply: Lessons from Experience (Colin 1999). This gives a useful overview of the factors involved in vill age level communal water supplies. The term VLOM, originally meaning Vill age Level Operation and Maintenance, was developed in the 1980s as a technological concept relating specifically to handpumps for rural water supply. A typical approach would include the following components (Noppen, 1996; White, 1996).

• Project staff formed user committees in each vill age prior to surveying and construction, and subsequently trained them.

• The committees organised community labour and contributions to capital costs. After construction they would co-ordinate operation and maintenance including fund-raising.

• Local leaders and committee members might sign a contract with the project agency specifying the responsibili ties of each; there might also be a handover ceremony.

• Project staff trained designated users in repair and maintenance and provided basic tools, possibly including an initial supply of spares.

• The project organised the supply of spares either via a government department or, ideally, through local shops.

It was clear to Colin (1999) that handpumps, including some VLOM designs, despite their many advantages, were not living up to earlier expectations. What appeared to be lacking


was the abili ty of both governments and user communities to maintain them; this was especially true in Africa. Affordabili ty of both capital and repair costs was also cited as a problem. Even where communities did attempt to take care of their pumps, this largely involved carrying out repairs after a breakdown; preventive maintenance was hardly ever undertaken (Skinner,1996). Colin 1999 notes that both the provision and maintenance of many new generation handpumps has proved expensive, especially where spare parts have to be imported. The most important factor affecting maintenance costs is the number of pumps covered by the maintenance system (Skinner, 1996). He suggested that in sparsely populated areas where the density of handpumps is low, the unit costs for setting up a maintenance scheme are high. Thus, many vill agers in such areas cannot afford to contribute significantly to the upkeep of their pump. This also raises the question of who will replace the pumps at the end of their working life. These problems are exacerbated by a lack of standardisation in some cases, and a more general failure to develop the practice of preventive maintenance which saves costs in the long run. Spare parts distribution has also been noted as a problem, especially in areas of low population (and hence pump) density. Unless there is a critical mass of pumps in a district, the market for spares is so small that no commercial operator is likely to have an incentive to import and stock them. This again leads to dependency on government or, more commonly, on donors.

3.1.2 VLOM assumptions Colin 1999 goes on to say that there are three assumptions in the VLOM concept that may not be true.

• The user community will be able and willi ng to maintain communal handpumps • Government will be able to provide an enabling environment to support VLOM • Communal handpumps will be able to meet most rural water supply needs

Colin 1999 notes:- “Handpumps are often ‘handed over’ to communities in the belief that true ownership is also transferred. This has proved to be a gross under-estimate of what it takes to instill a communal sense of responsibili ty; in reality, communities rarely accept the ownership of communal facilities. The test of ownership comes when the community are expected to pay for maintenance or repairs; often they do not, especially when they have previously been led to believe that government would be responsible for such services. “A community’s confidence in its abili ty to maintain pumps also needs careful nurturing. Frequent breakdowns, especially those requiring major works, do not help and may give the community no confidence in the pump itself. “VLOM systems sometimes fail due to the lack of basic technical skill s in sparsely populated rural areas; again this is a major problem in Africa. “Mobili sation and training for operation and maintenance usually receive low priority in project design. Many projects ignore or grossly underestimate the time needed for development of VLOM; project durations and targets are based around the time taken for hardware implementation, with relatively little time devoted to training caretakers during handpump installation.


“Where VLOM has been adopted it is usually the result of a heavy donor input of expertise and resources. “Supporting VLOM may also be no cheaper to government than a centralised maintenance system. Recent experience indicates that any savings from ending the repair service will be offset (initially at least) by the cost of training and mobili sation activities, which on the basis of NGO experience may take several years. “Too often, the external support agency hands over the support of VLOM to government (formally or implicitly) just as it hands over pumps to communities, without any preparation for the task. Unfortunately, most government departments do not at present have the vision, the resources or the management and extension skill s to promote and support VLOM effectively. “The adoption of VLOM usually marks a clear departure from existing custom and practice in both community and government and therefore has implications for the management and orientation of government services as a whole. It is diff icult to introduce VLOM or community management while the broad policy environment is still based on traditional notions of government provision. Colin 1999 concludes that “While the handpump has its merits, there is now considerable doubt that VLOM in its true sense is possible. In the African context there is little evidence that communities are able or willi ng, on their own, to support handpump programmes in the long-term. If handpumps are to be used, therefore, reliance on VLOM alone should be avoided. It may be more appropriate to develop the concept of shared responsibili ty for maintenance under a two or three-tier system with local caretakers responsible for simple tasks and more skill ed district-based teams, supported by government institutions, dealing with the more complex problems. (emphasis ours)

Lessons from Noppen 1996 VLOM is not a technical but a community concept; technology is only one part of what is needed to make VLOM a reality. Training for operation and maintenance should be part of project design, not an afterthought. It will be a long, slow process but will l ead to improved sustainabili ty. There is also a need to build confidence in the community concerning their ability to manage their handpumps. Operation and maintenance structures should be designed in accordance with the strengths and weaknesses of the community and supporting institutions.� Where local skill s are minimal only one type of handpump should be offered, to simplify maintenance tasks. This also increases the chances of a commercial market for spares.

� The improvement of traditional sources such as springs and wells should also be investigated as an alternative to handpumps.

� Extension staff are the key link between project agency and community and can create or destroy trust and confidence. It is very important to ensure that they are properly trained for their role.

� A distinction should be made between high and low demand spares (fast- and slow-moving spares). The former should be available through vill age shops, the latter via bigger, town-based dealers. (Noppen,1996, quoted in Colin 1999, emphasis ours)


Given the above overview on Operation and Maintenance let us now consider more details of Community Participation, Water Committees and management, the role of women, training, and links to local government, private suppliers and community.

3.1.3 Community Participation

3.1.3.1 Par ticipation enshr ined even in emergencies Smout and Parry-Jones (1999) articulate the commonly held view that it is important to involve community members in the planning and implementation of projects as a means of engendering the sense of ownership needed to provide motivation to effectively operate and maintain systems. They refer to projects where failure to achieve this involvement has led to unsustainable programmes. Even in a sudden emergency context Chalinder (1994) urges that, despite lack of time to gather information, these principles of participation and ownership used in development programmes apply equally strongly, and every effort should be made to involve the beneficiary community. However, Chalinder also acknowledges the difficulty of this in practice. “Experience has shown that in emergencies decisions are often taken on the basis of very little information. This can sometimes be justified on the grounds of the need to take measures to save lives. However, it can also mean that decisions which will have a major impact on the population being assisted are taken at a centralised level. The widely recognised direct benefits of community participation in water projects relate to information (e.g. local knowledge of water supplies and understanding of socio-cultural structures), communication (e.g. local contacts and skill s for mobili sation and capacity building), and ownership (FAO 1995),. Community participation has been seen by donor agencies as a means of enhancing effectiveness of programmes. Other benefits associated with community participation include;

• It is a vehicle for decentralisation, capacity building and empowerment • Improved sustainabili ty of projects through beneficiary commitment • Spinoffs in other projects due to improved capacity • It provides an effective learning mechanism for better project design.

Manikutty (1998) that there were five aspects of community participation that should be considered when evaluating participation:

• Integration of community participation in planning • Role of intermediary agencies • Mechanisms for interfacing with community • Design of project organization for facili tating participation • Learning mechanisms.

However Smout and Parry-Jones (1999) also describe the complexity of meaningful community involvement. One of their examples is an integrated programme in Nepal, aimed at improving health through water initiatives, education and horticulture. In this case the emphasis on an imposed agenda from the external agency inhibited community ownership. At the same time the dominance of a powerful leader in one vill age rendered a


democratic committee structure unworkable, but his sympathetic attitude meant that the programme worked well, showing the need for flexibili ty in approach. Here we begin to see that although community participation is a desirable state for longer term development, particularly the softer elements such as education and hygiene initiatives, participation may not directly correlate with “effectiveness” when effectiveness is discussed purely in terms of pump performance. It is important to be clear about what community participation means in a given project, and how it is to be integrated into the project.

3.1.3.2 Local Knowledge & learning mechanisms Brett (1996) draws a distinction between two forms of participation:

• Using co-operative organisations to deliver services • Enabling local people to influence the structure and management of interventions.

Arguments in favour of co-operative organisations are based on potential increases in efficiency that can be achieved by giving people a stake in organisations. The effectiveness of co-operatives is restricted by ignorance and by self interest – especially where money is concerned. In this case, management structures can only work if they have leadership to make informed decisions, and there is a comprehensive system of rules and sanctions to prevent opportunism. Given this statement, Brett goes on to say that people should only be allowed to participate in decision making where they have appropriate information and skill . A broad control base is only appropriate where there is low complexity of activities involved, with no need for expertise. Co-operatives tend to succeed where there is a high degree of solidarity amongst people. This can be brought about by an external threat, as will often be the case in an emergency situation, e.g. violence or drought. On the other hand, participatory systems will not be appropriate where conflicts of interest are so severe that they cannot be resolved through agreement. This might be the case where indigenous peoples are in conflict (e.g. Rwanda), or where displaced people settle amongst existing communities. Where conflicts are strong, participative structures maximise opportunities for self seeking behaviour. As with any grouping, contributions of members will be different, depending on commitment and abili ty. Attention needs to be paid to maintaining the motivation of active members e.g. payment, status (see Appiah (1999) below). Once the motivation of a capable member has been eroded, it commonly gives way to corruption and abuse of position.

3.1.4 Water Committees and Management The main mechanism most agencies use for ensuring that a sustainable water supply is operated and maintained by a community, is the Water Committee. This may be an existing group or structure that has its attention turn to the water supply or its mandate extended to include the water supply. It may also be a completely new grouping of representatives from the community possibly through a democratic process.


Manikutty (1998) notes that where water committees were weak, they proved ineffective in what could have been a useful role of coordinating related project elements within the overall programme. Factors contributing to this weakness were the fact that they were set up some time after the water points were installed, and they were given an ill defined role. In his case study in India both construction and maintenance was the responsibili ty of the government water board. He goes on to say that empowerment and enthusiasm of members of committees was achieved where vill agers were involved at all stages:

• objectives and details of the project were explained during planning • committees were formed and were active in locating water points, and later in

carrying out minor maintenance work; • the community was involved in training.

In one project with a high degree of community participation, the facili ties were handed over to the local committees, but the issue of how participation would be sustained thereafter was not addressed, and indeed it did not continue. Appiah (1999) looked at water committees in Ghana. He found that the mechanisms for appointing water committees is often irrelevant, as chiefs and powerful leaders manipulate procedures behind the scenes. He also points out that if effective community organisations already exist (e.g. a vill age improvement committee), then there is no need to create another committee; indeed this can be counterproductive by giving rise to friction. The motivation of water committees presents ongoing problems. Appiah suggests several approaches including issuing members with T-shirts, gum boots, etc., awarding certificates for good service, ongoing education, visits from external agents. All of these require a low level but ongoing input from external agents and assume that a water agency has not actually exited from the area. This reinforces the benefits associated with establishing links with local government organisations which may be able to provide this level of support once an NGO intervention ends. Payment of staff is important in a number of ways. Firstly, it is critical to ensure that members are motivated. They are unlikely to take on the responsibili ty and work load of management (particularly over along period of time) without tangible benefit; this becomes especially true when committees see that water supplies can generate revenue (e.g. for maintenance). Secondly, women’s non-economic status is improved if they earn income (Regmi and Fawcett (1999)), so payment for their services to committees can contribute to wider development objectives of empowerment of women. Appiah suggests that payments either cash or in kind should arise from the community to those undertaking the responsibili ty and work load of management. This is currently being explored by programmes all around the world and “best practice” had yet to be defined. Regarding emergencies and rehabili tation, where the social structures that make up a society have been destroyed e.g. due to war, rehabili tation programmes often concentrate on building up leadership institutions e.g. water management committees. Blench (1998)


argues that in these circumstances, these types of institutions are unlikely to be sustainable. In this instance, greater emphasis should be placed on the role of individuals.

3.1.5 The role of women Smout and Parry-Jones (1999) assert that it is now widely accepted by NGOs that gender issues must be integrated into project development. Some NGOs have positive experiences of involving women, particularly in community mobili zation and hygiene education. However, they point out that inclusion of women in committees does not guarantee that they will be involved in decision making. It is not easy to ensure genuine involvement of women in management structures as this may run counter to traditional roles, and contrary to economic structures. There are many examples of projects where women have been manipulated by men in order to satisfy the requirements of NGOs. For example, when a female Grameen Bank centre was set up in a Bangladeshi vill age, a group of male relatives of bank members took out a loan for a tube well in women’s names and chose a female manager. Once the loan was repaid and the bank had withdrawn, the men installed an exclusively male management committee (Van Koppen and Mahmud (1995)). Manikutty (1998) supports this view with examples from his case studies in India. He found that in some vill ages, women were reluctant to take part in meetings in the presence of men. If this is the case, then women appointed to committees were not making any meaningful contribution to the management process, and he suggests an alternative mechanism should be devised for eliciting the opinions and needs of women into the management structure. Van Koppen and Mahmud looked at a number if irrigation schemes in Bangladesh, which included different management arrangements. One variable was the male / female composition of irrigation groups responsible for managing and operating schemes. They found that even where all group members were female, the initiatives taken tended to be based on the interests of men. This was not surprising, as crop irrigation was traditionally men’s work. However, they also found that women’s non-economic status improved as a result of their involvement in irrigation enterprises. Their contribution was most appreciated where they had provided a better service than the previous pump management. This brings into the equation the wider development goals. It is possible that the inclusion of women in water committees can play an important role in the empowerment of women in rural societies. This can often be more important to overall development objectives than contributions they may make to the work of water committees. This effect tends to be observed in Islamic societies e.g. Pakistan (Aziz and Halvorson (1999)). Regmi and Fawcett (1999) also consider the role of women in the wider context of development and argue that sustainable impact can only be achieved if women are active partners in the development process and strategic gender needs are linked to practical issues. They suggest that as women are the primary collectors of water, they will have a better understanding of water resources than men e.g. seasonal availabili ty, quality. They


can, therefore, make valuable contributions to project planning processes in terms of technical information, and bring an understanding of social issues associated with water collection and use e.g. washing practices. Finally it is worth noting Smout and Parry-Jones comments on the mechanisms by which women are appointed to positions within management structures. They say that these mechanisms are also important in determining how effective women are in those positions. Where women are appointed by individuals or small groups of powerful people, most of whom will be men (e.g. chiefs, NGOs, vill age elders), they will be reluctant to disagree with any opinion expressed by these people, to whom they will feel a sense of obligation. In more extreme situations, male committee members may simply refuse to include token female members imposed by NGOs. By contrast, they will tend to adopt a more active role in defending women’s interests if they feel accountable to other women in the community e.g. if elected through a community wide process.

3.1.6 Training Brett (1996) notes that management of boreholes through community participation in water committees will only be effective where members are given adequate training in technical as well as organisational issues. The key to the above concerns of participation, management and gender lie in appropriate training. In suggesting that the institutional capacity of a community should be capable of sustaining a pump, Arlosoroff et al (1987) recognise the importance of management structures, and the contribution that training can make towards building up the capacity of the community in both technical and management skill s. In addressing a complete lack of managerial and technical skill s in irrigation committees in Bangladesh, an NGO provided both technical training and assistance with organisation (van Koppen and Mahmud (1995)). They stress that the assistance with organisation was critical in achieving effective operational management. Regmi and Fawcett (1999) point out the dangers associated with dividing training between men and women. Whilst women tend to be offered training in “soft” skill s such as health, hygiene and administration, men receive training in practical skill s such as masonry and mechanical maintenance. This can exaggerate differences in status, as technical training provides the trainee with portable income generating skill s that tend to be in demand. They go on to say that lower literacy rates of women and lack of general participation experience in projects tend to reinforce male prejudice against women’s involvement in management structures. Training should, therefore, be designed to address these issues if women are to be equipped as effective players. Women’s involvement is also hindered by their work load, leaving little time to attend to additional duties. Training initiatives aimed at encouraging men to share work loads at home can play an effective part in addressing this (Regmi and Fawcett (1999)). Communities are normally expected to finance and manage the operation and maintenance of a pump, which requires capacity building in both money management and organisation.


Ongoing training provides a number of benefits as it can help to motivate community organisations, to refresh skill s, and to counter migration of skill ed personnel.

3.1.7 Links to local government, private supp liers and community Regarding participation, the agency should interface with the community. The level of participation achieved is likely to be higher if a greater variety of mechanisms for interfacing with the community are in place, they should include direct communication with the community, and they should involve two way communication. These can include existing local representative bodies which, if they are effective, will enhance community participation and sustainabili ty. Government departments generally have a technical focus and are not equipped to deal with socio-economic aspects of development programmes. Research indicates that specialist agencies with these skill s can be effective in promoting community participation, and provide a valuable service in co-ordinating agencies. Although NGOs may be capable of taking this role, arrangements have to be made to ensure sustainabili ty for when NGO activities are complete and they withdraw. Manikutty (1998) makes a similar point. “One of the problems associated with water departments is that they tend to be made up of engineers, with little appreciation of the socio-economic aspects of project management and community development. One way of addressing this is to undertake training of government personnel in these disciplines, but this is not a trivial task and should not be underestimated. Good communication between community and government bodies can achieve effective delineation of responsibili ties, leaving vill age committees responsible for community mobili sation, but this must be corresponding sharing of power and decision making.” Regarding ongoing technical support, there needs to be a balanced view. Lammerink et al (1999) advocate community management systems in order to improve rural water supply systems. But they point out that there is always the danger that advocating increased community accountabili ty and responsibili ty will be seen as a way for governments to cut spending and to wash their hands of community contact. They suggest that there is a need for continued support from government - without this, the approach will simply not be sustained. Smout and Parry Jones take an alternative view, that where a comprehensive local government structure exists, it may seem attractive to make use of these, and indeed it may be required of an NGO to involve the appropriate government departments e.g. water department, geology department, public health etc. However, these should never be allowed to substitute vill age level user groups. Both sets of authors would probably agree that a sustainable system requires both community mobili sation for VLOM, and support from the local government. The levels of responsibili ty for each will depend on how the spares are being supplied, policy within a government (hands on or supervisory) and the level of complexity of the water supply system. Partnerships and co-operation. Smout and Parry Jones point out the need for NGOs to forge partnerships with local government agencies and the private sector in order to


achieve the continuity needed to ensure sustainabili ty of programmes after the NGO withdraws. Although some find working within political and administrative frameworks frustrating and time consuming, most agree that such partnerships are central to sustainabili ty. Manikutty noted in India that learning from communities, and from experience, is important in improving performance, but government departments tend to be bureaucratic and inflexible, so they are not adept at responding to new ideas. Systems tend to be designed to ensure adherence to the rules rather than maximising benefits to the target group. However, health benefits associated with improved water supplies will only be realised if programmes are complimentary to health, sanitation and hygiene programmes. Sustainabili ty can, therefore, be enhance by linking water programmes to government departments that are active in these areas. In their guidelines for water sources in emergency situations, House and Reed (1997) give a list of reasons why an implementing NGO should liase with local government organizations, in which they recognize that they are likely to be responsible for looking after facili ties after the NGO leaves. Other than this, there is no mention of pre-positioning of intervention for sustainable hand over when the NGO/agent leaves.

3.1.8 Techno logy Choices So far we have discussed the social mobili sation required to develop a sustainable water supply. It is widely recognised that the choice of technology can be an important factor in the ultimate sustainabili ty of community water supply projects (Arlosoroff et al (1987), Lammerink et al (1999)). For example, familiar technology can contribute to success (treadle pumps in Cambodia) and unfamiliar technology to failure (hydraulic ram in Mozambique). Arlosoroff et al (1987) propose a guideline on choice of technology for community water supply planners:

"The technology chosen should give the community the highest service level that it is willing to pay for, will benefit from, and has the institutional capacity to sustain".


Definitions of Reliabilit y

The reliability of mechanical equipment is commonly measured in terms of the “mean time before failure” (MTBF) (represented in survey data by “frequency of breakage”). However, in community water supplies, the mean down time is at least as significant as the MTBF i.e. the average period for which the pump is out of service when it does break down (represented in survey data by “average downtime”). If pump A needs a mobile team to carry out repairs, and stands idle for an average of two months each time it breaks down, pump users will not regard it as more reliable than pump B which can be put right in one week by an area mechanic. A better indicator is what the mechanical engineering profession terms “availability” i.e. the probability that equipment will be in operating condition on any one day, calculated as the operating time as a percentage of total time. On this basis, pump B has a higher reliability (97%) than pump A (90%). source: Arlosoroff et al (1987).

There were a series of problems with both the function and durabili ty of the Afridev design; Wood (1994) identified the following from experience in Ethiopia:

� PVC riser pipes, which were sometimes glued together, needed to be removed so that sections where the rod connector had worn holes could be replaced. The caretaker alone could not do this. Also, risers in wide-diameter wells flexed during pumping and eventually cracked because the pipes were not secured.

� Pump rod jointing. The clip-on device, where used, came off after a few months. A special fishing tool was then needed to extricate fallen rods - the ordinary tool was not suitable.

� Plunger seal and O-ring absorbed water, expanded and were difficult to remove. Colin 1999 notes that “Thus many pumps proved far harder to repair at vill age level than had been envisaged; many people were forced back into using traditional, unprotected sources. A very common problem was with the rising main, meaning that even the Afridev could not be fully repaired by a vill age caretaker. Despite design improvements, this problem remains today.”

3.1.9 Emergency Rehabili tation Environments The focus of large scale relief programmes is on preserving life, but Whiteside (1996) ill ustrates how almost any intervention affects local infrastructure which in turn affects the conditions for longer term development. A common example is the procuring and distribution of emergency food aid; both importing and purchase of local produce by aid agencies distorts market prices, hiring of local transport facili ties upsets market infrastructure, and issues of dependency and uncertainty of supply influence farmers’


planning. He argues that injecting cash into local economies may be one of the best ways of helping rehabili tate communities. With regard to emergency water provision, there is scope for cash (or food) for work schemes in the use of local labour, and in payment of mechanics, committee members etc. Payment of management staff also promotes motivation and status, which contributes to long term sustainabili ty. However, a cash based approach can only be appropriate where basic market infrastructure exists. When planning interventions, even in emergency conditions, it is necessary to adopt a participatory approach to understanding priorities within communities; similarly, Chalinder (1994) stresses the need to take a broad view of principal needs. Whiteside (1996) urges caution at scaling down operations too quickly, stressing the need for ongoing support for the poorest in moving towards longer term development activities. Rural people will take account of aid resources when planning, so it is important for agencies to keep communities informed of their intentions, and to plan gradual withdrawals in order to give people sufficient notice to plan accordingly. Chalinder (1994) emphasises that water and sanitation programmes contribute only in part to the overall environmental health of a population. Although this publication acknowledges the link between emergency interventions to meet immediate needs and longer term development activities, it concentrates on effective service provision rather than future sustainabili ty.


3.2 DFID Guidance Manual The Guidance Manual on Water Supply and Sanitation Supplies (DFID 1999) has been strongly promoted in the last year or so “to help DFID staff involved in each stage of the (project) cycle (regarding water and sanitation)” . “The manual is also intended to help DFID’s project partners…..” In order to make our research useful to practitioners this section specifically relates the context and findings of the case studies to statements made in the guidance manual. It is hoped that this will facili tate relating the research to the project cycle. One of the central questions to the research project was the role of the exit strategy, particularly with regard to emergency water supplies, to the final sustainabili ty of the supply. The guidance manual discusses water programmes for poverty alleviation. In particular it encourages entry strategies that are participatory and “demand driven” approaches. We note that the manual states clearly:- “This manual does not cover water supply and sanitation in emergency contexts.” Nevertheless the heart of the research was to answer the question “what factors should be included in an exit strategy to ensure maximum possible sustainabili ty” . Secondary questions include “can a good exit strategy make up for a non-participatory entry strategy?”

3.2.1 Participation Let us first consider what the manual states about participation. In Chapter 2.2 the manual discusses Social development perspectives. It notes that participatory approaches to Water Supply & Sanitation (WS&S) planning and implementation are very different from supply-driven technology-based approaches of the past. They are said to be :- (Page 41)

• Their inclusive nature ensures that voices of the poor and under privileged are heard

• They empower people to take responsibili ty for their own services • They respect the traditions and culture of different societies and use them to

develop appropriate solutions • They recognise that women have a beneficial influence on key decisions, and that

gender considerations matter • They mobili se public and private partnerships to make optimum use of all

stakeholders • They make user demand, demonstrated by Willi ngness To Pay (WTP), a primary

criterion for selecting levels of service Here we see a set of criteria very much at the heart of the research. When assessing the impact of the strategies used in these water programmes, the research has included participation and ownership variables. However the manual states that:- “Implementation is often driven by the engineering components of a project; the social dimensions such as developing local ownership or


capacity to manage water supply and sanitation can get left behind. This can happen either because social development takes longer and is seen to hold the process up, or because new actors who are unfamiliar with the process approach and social development issues become involved at the implementation stage.” This is all the more so for emergency supplies and rapid rehabili tation programmes.

3.2.2 Cost recovery – a sign o f sustainabili ty? “A causal relationship between cost recovery on the one hand and project ownership and sustainabili ty on the other is by no means automatic. Influential study Feachem et al 1978, demonstrated that level of cash contribution collected from vill agers was so modest and the administrative costs so high, that cost sharing had a negligible effect on project sustainabili ty. Also, rather than instilli ng in vill agers a sense of ownership and responsibili ty, contributing cash, labour and local material towards construction convinced them that they had already paid their fair share and that the government should take on the longer term responsibili ty of operation and maintenance. This example serves to underscore the importance of on-going consultative processes and the development of a mutually agreed and shared agenda from the outset. It does not imply that water supply and sanitation services should be free to users. Some contribution from users is certainly desirable.” (emphasis ours) One area that the research project should explore is the relationship between cost recovery and ownership. Is there a correlation between expressions of ownership and willi ngness to contribute. The guidelines note that Willi ngness To Pay (WTP) varies in rural communities. Poor households without good alternative supplies are often willi ng to pay much more for improved water supplies, in both absolute terms and relative terms, than richer families pay for their existing supplies. Time and monetary costs of obtaining water from alternative sources is a key influence on WTP for improved supplies. Family characteristics, such as level of education and family size - thought to be related to the opportunity cost of time - will also influence the perceived attractiveness of improved supplies and affect WTP for different standards of service. Where people believe government should provide free water, WTP is very low.

3.2.3 Private supp liers “Private entrepreneurship currently plays a significant role in service delivery to people in poverty and building capacity in the private sector can improve both affordabili ty (and accessibili ty) and livelihood opportunities, particularly when it includes local level and informal sector enterprises.” The role of private entrepreneurs in the sustainabili ty of the whole system is important. No matter how motivated the communities are – the pumps will have lengthy downtime if spares are not readily available. The difficulty of the private supply of spares seems to be the combination of low turnover and small profit margins. Even in the guidance manual there is a boxed example from Malawi. It discusses the Danida supported Karonga integrated groundwater supply project on the northern shores of Lake Malawi. They supplied shopkeepers with spares such as o-rings and u-seals who received a 10%


commission. Having had to supply the initial stock the system only continues with the goodwill of the shopkeepers. “The vill age shopowners felt that stocking the spare parts is mainly a community service because of the small profit margins, but all expressed willi ngness to continue to stock the fast moving spares.” This situation remains true for the situations within the research project. Shop keepers explained that the profit margins were very small and that the turnover was minimal. From the point of view of the community this is desirable – it is good that the pumps do not break down often and that the o-rings are not wearing out too fast. However from the viewpoint of provision of ongoing supplies, it creates some difficulties. There is also the question of who is the right “private” supplier of spares. Most of the agencies in the project have been working with shopkeepers (retailers) who have established businesses and add the spares to their selection of grocery or hardware stock. However one research partner felt that the local government water department was the best mechanism for resale of spares. This idea was not tested as the donor prevented spares being issued to local government.

3.2.4 Sustaining community based op eration and maintenance The following guidelines are given in the manual as a highlighted box:- Implication for practice

• Participatory methods of working are centrally important, and not simply participatory contributions of labour and money in response to agency instruction, but continual processes of shared decision making between the agency and primary stakeholders at all stages of the project cycle, with the aim of developing community members sense of responsibili ty for and control over the local operation, maintenance and management of the system.

• Increase as far as possible the room for manoeuvre that community members have in relation to aspects of the service design – choice and location of service, and operation, maintenance and management of the system.

• Where community management is required, capacity building should be a project output, especially the development of skill i n management, planning, analysis, decision-making, and problem solving. The time scales for construction work and capacity building are different. Capacity building requires separate resources of time, resources and personnel.

• Provision for the transfer of technology and responsibili ty should be built in from the beginning of the project, with clear recognition that this is a process not an event.

• Recognition that tackling the problem of non-payment of recurrent costs is not simply a matter of adjusting payment levels, but addressing all aspects of effective community management, including institutional and technical

• As far as possible, make change pay, that is by creating paid jobs in service operation and management wherever possible and reducing or eliminating reliance on volunteer labour.

• Keep technology very simple to maintain and repair, where possible, with a reliable supply of spare parts and technical assistance available locally.

Derbyshire and Vickers, 1997 quoted DFID Guidelines.


3.2.5 Agency links The last point above calls for there to be some form of technical support available locally. The Guidelines paint a picture of the critical success factors in water institution performance. They include nine critical success factors or performance categories:-

• Organisational autonomy • Leadership • Effective management and administration • Commercial orientation • Customer orientation • Technical capabili ty • Human resources development • Organisational culture • Interactions with key external institutions

These factors apply to both the external agency that is in process of exiting, the local government, private systems of technical support and to a certain extent any local co-operatives specialising in the technical support of the water systems. We can note that many local government departments fail to meet these standards – not just on the commercial and customer orientations, but also in terms of items such as organisational autonomy. Municipal water departments, for example, which are not able to hire staff or raise tariffs to meet their projected costs, have insufficient autonomy to manage effectively. The research project should take into account the capacity and role of local government and private suppliers.

3.2.6 Techno logy The DFID Guidelines explicitly give the Afridev as an example of a VLOM pump. “The Afridev handpump is the result of a design and development process which started in 1972 and has been evolving ever since. The original pump was designed to the following criteria. It:-

• Used appropriate technology • Used lightweight, non corrosive components • Could be maintained by women • Could be manufactured locally to an exact specification • Needed only one of two simple tools for installation and maintenance • Was relatively cheap • Was designed with preventative maintenance in mind

Over the years some design features have been modified and improved to aid VLOM. The Afridev has been specified for standardisation in many countries including Mozambique and Malawi.” Wood 1993 Skinner 1996 quoted DFID Guidelines


3.2.7 Measuring sustainabili ty? The Guidelines discuss indicators of progress in water and sanitation programmes. While all of these factors do not apply to the research project they nevertheless provide a good basis for designing the research parameters. Sustainabili ty Reliability of systems :- Quality of water source, Number of facili ties in working order, and Maintenance. Human capacity development :- Management abili ties, Knowledge and skill s, and confidence/self-concepts Local institutional capacity:- Autonomy, supportive leadership and systems for learning and problem solving Cost-sharing and unit costs:- community contribution, agency contribution and unit costs Collaboration among organisations:- planning and activities.

Effective use Optimal use:- number and characteristics of users, Quantity of water used (all purposes), time taken to use facili ties, management of water resources Hygienic use:- Water quality at home, water transport and storage practices, home practices to improve water quality, site and home cleanliness, personal hygienic practices Consistent use:- pattern of daily use, pattern of seasonal use Replicabili ty Community ability to expand services:- additional water/latrine facili ties built, upgraded facili ties, new development activities initiated Transferability of agency strategies:- Proportion and role of specialised personnel, established institutional framework, budget size and sheltering, documented administrative/implementation procedures, special/unique conditions Some of the above does not concern exit strategies per se. Knowledge and practice of hygiene behaviour should be part of an integrated water project and requires very long term inputs; it is covered only briefly by the research. Agency factors such as transferabili ty of strategies are not considered part of the research.


4 Methodo logy

4.1 Basis for the investigation

4.1.1 Data gathering A combination of methods was adopted to gather information regarding the effectiveness of the different exit strategies adopted and the factors, which appear to most influence their success. The main process was an open enquiry of the community and key informants followed by semi structured interviews. Open enquiry allows the community to set the agenda, and to mention the factors which according to their perceptions are important to the sustainabili ty of the water supply. The methods involved key informant interviews, site visits and observations, interviews with water-point users in each of the sampled communities and a review of secondary data involving project documents and pertinent literature.

4.1.2 Analysis As stated above, the main approach adopted was to capture the perceptions of community members regarding the development, use, management and performance of their pumps. These open enquiry perceptions were used to form the conceptual construct of the proceeding enquiry. These issues provided the structure for the continuing data collection, i.e. a semi structured interview process. The data from the pump user and key informant interviews, and observations were coded and entered into computer software for analysis. Given the nature of the data, non-parametric tests were applied to identify possible relationships, i.e. Mann-Whitney test for differences, and Spearman’s correlations. Significance. Throughout the text the word significant is used in the statistical sense. There was a greater than 95% confidence that there was a difference between the two samples. Where the data suggests a particular link between two samples but this is not statistically significant the text uses the phrase “the data tends to suggest” . In order to keep the text easy to read, the report avoids putting the p values of the significance in the text as recommended in academic reports.

4.1.3 Verification Initial findings were also discussed with the project holders to further explore and verify that the issues pertinent to the users were also the main concerns of the key players. A workshop was held, initially with the project partners and later opened to other agencies in Mozambique. The workshop discussed the issues surrounding sustainable water supplies and verified the data generated by the local communities. A wider verification of the issues was achieved by presentation of interim findings at the WEDC annual conference in Ethiopia and distributing a questionnaire among practioners at the conference.


4.2 Sample

4.2.1 Sample Size The following calculations were applied to determine the minimum sample size. These were taken from the World Health Organisation’s (WHO) manual for determining sample size (WHO, 1987:2-3).

Minimum sample size for estimating population proportion with absolute precision.

n = z21-αp(1-p)/d2

where: 1-αp = 50% (assumed proportion of population) z = 95% (degree of confidence)

d = 10% (degree of absolute precision) n = ? (Minimum sample size)

n = 96 Minimum sample required in the case of a simple random sample. n (2) = 192 Minimum recommended sample to allow for the design effect of cluster sampling. Actual sample used = 96(3) The design effect used was “3” in order to help the issue of greater representative ness. = 288 (1.1) Additional 10% to allow for sampling problems = 316

______________________________________

4.2.2 Sampling Process A random stratified cluster sampling process was used to identify the sample of pump users. This approach was used given the lack of reliable census data and the recent population movements within the areas surveyed. The clusters were the communities dependent on pumps installed by the programmes under study but now independent of further agency support. The selection of the clusters was divided evenly between the three agencies and further divided evenly between three of the regions in which each programme was or had worked. I.e. 16 pumps were selected from each programme, 5 from each of the 3 randomly selected districts. In the case of the CARE programme only two districts were visited due to the condition of the roads not permitting access to the most isolated region. In this instance two subsets of 7 and 8 clusters each was selected. The clusters were selected at random from the respective agencies records. Given that some of the World Relief sites were more isolated than the Care ones, the lack of isolated Care sites was not considered significant.


The sample unit was an adult member (over 18 years of age) of selected community dependent on a pump installed by one of the three programmes under study. In each selected cluster approximately 7 member were randomly sampled.

4.2.3 Description o f acquired sample In all 305 usable sets of data were collected from 47 communities or clusters. The description of this sample is presented in the following Table 1 :

Table 1 Sample Descriptors

Description % of valid responses

Gender Male 49.3 Female 50.7 Age <30 19.1 30 to 50 49.0 50> 31.9 Position Chief 6.9 Water committee member 10.1 Pump mechanic 7.5 No position 74.7 Proximity to pump < 1 kilometre 74.7 1 kilometre > 25.3 Sett lement Respondents settled due to pump

14.2

Others settled due to pump 35.8

Figure 1 Household size.

Household size.

0123456789

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Number in Household

% o

f sam

ple


The following gives the basic description of the pumps

Table 2 Pump Descriptors

Description % of

sampled pumps

Depth of pump Set <30metres 42.5 Date of installation <1995 59.0 Emergency installations 83.0 Type of pump Volanta 10.6 Afridev 83.0 Other 6.4 Non operational at t ime of visit

21.2

Distance from Agency op’s centre

<50 Km 21.3 50 to 100 55.3 100 to 150 10.6 151 to 200 4.3 200> 8.5

Of the pumps investigated, 83% were described by the agency as being installed in emergency or rehabili tation conditions. The other 18% were installed post emergency and would be described as being installed in the “development” phase of their work. In practice this gave a wide spread of years that the pump had been installed. .


Figure 2 Year of pump installation

0

5

10

15

20

25

30

91 92 93 94 95 96 97 98

Year installed

Per

cent

age

of s

ampl

e

Series3

In approximately one third of the pumps the depth of the water was less than 20m, a third between 20m and 40m, and the remaining third over 40m. 76% were new boreholes by the agency while the remaining were reconditioned. 83% had the Afridev pump installed, while 10% had the Volanta and 6% had the Nadia. 80% were due to the drought or emergency and the remainder were post drought and part of a development process. The valid percentages for each variable are available in Appendix (11)


5 Find ings

5.1 The Conceptual Construct Based on the categories of interest identified by the communities, a number of issues that appeared to have an effect on the sustainabili ty of the pump were identified and the relationship between these explored. These were therefore also seen to be pertinent to both existing and future exit strategies. The main conceptual categories used in the analysis of the data can be divided into the following presented in the Table 3 below. These in turn can be considered as three types:

• Programme context – social and physical factors outside the immediate control of the intervention.

• Programme design – specific issues directly under the control of the programme • Programme outcomes – issues which an intervention can influence

These three groupings reflect the more intuitive discussion held with the partners, who effectively came up with the same basic framework (see below). In order to investigate issues of project design and sustainabili ty, the variables created from the coding of survey data can therefore be regarded as either independent (those issues that affect the outcome of a project) or dependent (indicators of outcomes). The independent variables generally represent the entry and exit strategies, whilst the dependent variables represent community social and water related issues which are affected by water supply interventions and sustainabili ty. Note that several variables can be considered as either. Programme context variables are presented “fait accompli” . For development programmes these are physical items such as depth of pump, but for programmes attempting to exit from an emergency intervention this may include the entry strategy which was undertaken by the former programme management. Programme Outcomes will be both the physical availabili ty of the pump plus the social mobili sation. Borehole installation programmes include support elements such as community mobili sation, training, spares supply and maintenance network. The extent to which these are addressed varies according to the time and resources available at planning and pre-installation stages, which is heavily dependent on whether or not emergency conditions were prevaili ng at the time of installation. As these elements are widely regarded as necessary for sustainabili ty, the degree to which programmes are successful in achieving these features may in themselves be regarded as indicators of success and sustainabili ty. In an attempt to investigate the impact of the projects on these social mobili zation issues, three categories of dependent variables were identified:-

• Organisation • Participation. • Competence

Programme Design includes all independent variables such as training levels, etc. And for this research the context included both the actual physical context and the historical context (which captured entry strategies of emergency programmes).


Table 3 Conceptual Categor ies for Analysis

Categories Description

Programme context Context

Depth Alternative sources of water

Pressure

Contextual variables are considered regarding their impact organisational and pump performance indicators, however, it is recognised that the elements which make up the context category are not open to manipulation via specific exit strategy designs. The main element considered in this category are the depth of the borehole and pump setting, the presence of alternative sources of water. Recording the time spent collecting water per day captures population pressure on the pump.

History Settlement

Emergency installation

The historic background of the pump is taken into account by recording the instance of settlement due to the pump and whether the borehole was drill ed in response to an emergency need. These two elements are important to the questions addressed by this research. Although it is recognised history cannot be changed the abili ty to compare differences in the background of the pumps permits an adjustment of exit strategies to take account of the differing histories.

Programme Design Education

Education represents the training provided, its level and length, when the training took place, i.e. at the time of installation or some time later. Also if follow-up has training has taken place. The other issue which is considered within this category is who was trained, the whole committee or specific local mechanics.

Programme Outcomes Function or performance of the pump

This was taken to be the primary indicator of sustainabili ty and was viewed as an expression of both frequency of breakage and length of time inoperative when broken. Other factors such as the condition of pumphead care were also considered. (This is taken to be the main dependent variable category).


Organisation A group of measures which capture level of

organisation and structured management of the community pump. Issues such as the frequency of meetings, the existence and nature of rules for pump use, the payment of water quotas, the application of sanctions for non-payment of these contributions. Again this category is seen to be an expression or result of other planned or contextual circumstances.

Participation The degree, to which the community and its various social sectors participate in the management structure or have participated in the installation and maintenance of the pump in the past i.e. The instance of prior contribution to the purchase of the pump head is one of the factors taken into account in this instance.

Competence

Competence is seen to be an outcome of educational input. The category is represented by the communities own perception of the competence of their water committee and /or local mechanics

From this construct over 70 variables were identified and coded. The variables are described in detail in an Appendix. The appendix presents the valid percentage of respondents to each variable. Since the process was not based on a formal fixed questionnaire, some variables did not attain a valid sample. However they are found to helpfully inform the discussion. The results of the Mann Whitney test on the relationship between the independent and dependent variable clusters are summarised in an Appendix.


5.2 Verification b y the partner agencies

5.2.1 Key Issues The issues surrounding the research were discussed individually with each agency as the project progressed. Towards the end of the project the issues surrounding the existing water programmes were jointly discussed in a workshop as a prelude to verifying the details of the data. This section combines all the discussions. The ideas presented are the opinions of individuals within the three partner agencies, and do not represent official agency policy.. At the workshop the partners were asked to brainstorm the most important issues surrounding their programmes. The partners mentioned:-

• arranging finance within the community, • bank accounts and how do communities save their money, • training of mechanics (single or groups, area or community) • training of trainers • spares issue – how to supply spares sustainably, • shift from centralised repair to local • local government - three ministries involved – Water, Health, Gender, Youth &

Community services • ongoing presence – small but critical inputs • use of water from cisterns (and other water supplies),

From this brainstorm list we can organise the main questions of partner agencies as follows:-

5.2.2 The question o f community involvement All partners agreed that community involvement was important. The community approach encourages involvement and decision making. In principle it should give room for the community to choose affordable options. It should encourage their own monitoring and feedback helping them to come to their own conclusions on how much things cost and helping them decide how much they can afford on maintenance. Any stakeholder analysis or participatory approach must be gender sensitive noting the different roles for men and women. In both countries the government recommendations are for committees to be 60% women but this is not often followed to the letter and the committees gender mix varies quite considerably. The structure of the committee is important. Should there be non traditional leadership i.e. the agencies or local government encouraging the formation of a different committee from the local leadership.


All partners noted the trend towards decentralisation, but wished there was a coherent set of guidelines and some communal responsibili ty (across agencies and government). They also noted the trend towards democratisation, i.e. local ownership with elected leaders to co-ordinate VLOM. SCF Malawi, who was not a full partner in the research but kindly gave extensive interviews, raised the question of vill age social power structures. They noted that although one could set up a committee, the initial committee would probably reflect the power structures within the vill age. If, over time, these social dynamics changed then a committee might cease to function and the training or capacity building would have been wasted. SCF have considered mapping the social dynamics of a community as a first step in the development process.

5.2.3 The question o f f inance On the question of finance the partners questioned the availabili ty of money for communities – do they have money, what is the poverty level of that community? It was felt that communities varied in their abili ty to find cash. It was noted that the rural per capita for Mozambique is only $50. The partners have experience of credit programmes and noted that credit doesn’t work well in the rural areas. To raise 2 to 3000 per month is enough for an Afridev but not a Volanta handpump. Most rural communities don’t have money but do have capital (goats, chickens, etc). In some cases interviewees reported that money was raised by working for someone in the vill age who needed labour and was rich enough to pay for it. The question of “How do they raise the finance?” provoked considerable discussion. What were the prioritisations within a vill age, how was the collection divided between men and women? Examples from the interviews suggested that families paid on the basis of how many women were in the household (since a household may have more than one wife per male). The agencies experience and the interviews showed that another critical issue for the vill age was the timing of collections. Should money be collected in advance of breakdowns and stored in some form, in order to give a rapid response to the breakdown? Or should the money be collected when the repairs are required? In practice most vill ages collect the money after the breakdown in order to repair it. This can be difficult if the breakdown is during a season when resources are low. On the other hand there was a case of one vill age that had 3,000,000 Mt collected and stored under the bed of the treasurer. All the agencies have tried to encourage the communities to store the collected finance in the form of pump spares (inflation proofed) which are then readily available to attend the breakdown. The difficulty of this is that some spares may be stored for too long (perishing of rubber) and may be the wrong spares for the breakdown (perhaps the problem might be a broken handle when the spares stored are the O-rings).

5.2.4 The question o f t raining Should the communities be trained only in simple handpump maintenance, or in maintenance and repairs? The government is unable to respond quickly and on time. Private mechanics have a key role. Some agencies have created mechanics for the


locality. In some cases the government through UNDP undertook rehabili tation of boreholes, and trained communities in preventative maintenance, simple repairs and complex repairs in a three phase training. A complementary approach proposed is to train the local shops so that the spare parts are sold correctly. Regarding training and forming the committees, it was noted that elected members are there to serve the community but what is their motivation and how can we give them the incentive to continue working for many years?

5.2.5 The question o f techno logy In addition to the difficulties outlined above, there is the added question of availabili ty and accessibili ty of spare parts. They are often not within easy reach of the communities. Agencies are currently addressing this problem through private entrepreneurs. However this strategy has a number of diff iculties:- • The government has in some case been supplying spares hence the high figure for

spares availabili ty. One agency has had spares at the office but has worked towards handing the spares over to government. Another agency has been trying to reduce the number of middlemen. There is approximately a 10% handling charge every handler. So if the spares start in the Capital then move to the provincial centre, then district, and then handed to the local private retailer the handling charge could be as much as 30%. On the other hand, commercial outfits in Kenya charge 100% handling fee for any item costing less than $5, and this is common practice around the world. So the hope expressed by agencies that private retailers will only charge 10 to 30% handling fee seems to reduce the traders profitabili ty to an unrealistic level.

• One difficulty highlighted was the “donor constraint” . Some donors require three

quotations to purchase the initial pumps as a form of accountabili ty. This means that the agencies purchase the pumps from a supplier who is either in the Capital or even from an importer. Having purchased the pumps from out of the locality the agency then attempts to provide a sustainable supply of spares by asking local retailers to stock spares. The link between local retailer who can supply spares to the vill age, and the original supplier of the pump is therefore tenuous.

• Technical questions arose on matching the technology to both the depth of the

borehole and to the capacity of the community. The Volanta handpump is often installed in the poorest and most remote areas. For some of the agencies they have installed Volanta handpumps in areas which World Food Programme have classified as uninhabitable. Supply of spares to these very remote pumps is likely to be unsustainable.

• All the programmes follow Government policy. In Malawi, the Maldev and Afridev

are standard, while in Mozambique the Afridev and Volanta are the preferred pumps. This is good for encouraging local retailers to stock a limited number of spares. However in the past there have been subsidised sales of both whole pump systems and spares – this has been not very helpful to profit making. A number of different approaches were discussed which included the option of paying for some spares and receiving free spares from others.


• In Malawi, UNICEF has been funding revolving spares, except that some partners noted the spares are not revolving. Some of the spares are out of stock too quickly and others are not moving. The policies involved over the recent years are confusing. New policies are introduced without sufficient time to develop participation of all stakeholders. To give an alternative example the donors have stated in Malawi that programmes must work with hygiene and water development together, but the local government hasn’t had time to think about this and adjust accordingly.

The partners were agreed that one of the major constraints to a private supply of spares which reaches down to the vill ages through local retailers is the constraint of profit margin. The private sector needs a minimum turnover to make a it worthwhile handling the stock. It is a credit to the pump design that they do not breakdown very frequently – but that in turn means that there is very low turnover of spares for retailers. They then need to add a substantial handling charge if they are to make a worthwhile profit. In the interviews with agencies and communities there was very limited discussion about what breaks on a handpump. There was some consensus that rods remain a slight problem, it is easy to either crack them that leads to a break, or drop them during the process of maintenance or repair. Once dropped it is a particular skill to fish them out again. The stress on the rods seems very much related to the depth of borehole. However, although this is an example of some necessity for extra repair skill s, the partner agencies all agree that the Afridev is a very appropriate pump and can be maintained at vill age level.

5.2.6 Agency constraints There were a number of questions arising about the resources of agencies to create a sustainable supply. How much time is required to engage and then exit from a community? Partners questioned whether this was actually possible in a 2 year programme. Although each agency has been engaged with water programmes for over 7 years, every step has been on a maximum 3 year programme. Examples have been given of donor pressure on the agencies. Donors were quoted as saying, “other agencies ran a 6 month programmes where the supplies were considered sustainable at the end. If agency X did it in 6 months why cant you?” The agencies also feel that they are often constrained in their other resources too. Another quote from a donor was “If agency Y undertook a similar programme with 1 car, why do you need 3 vehicles?” At the workshop, a Malawian government representative discussed the lack of resources available to local authorities responsible for the water supplies. Most water agents are responsible for about 100 vill ages. It is not practical to do a preventative maintenance circuit with the resources available. There is a history in Malawi of a very hierarchical tiered maintenance system, and the partners acknowledged that in some cases it takes time for people to adjust their understanding of participation and community involvement.


5.2.7 Partner sugg estions for data framework The partner agencies proposed the following framework for gathering data. Three main states can be considered to the supply. 1) = prior condition (context) 2) = input given by the agency 3) = output or end state The partners then went on to brainstorm the essential elements of Sustainable Water Supply from their own experience. Lack of water 1 Without a suitable demand for water, a supply is

unlikely to be sustainable. Clear policy, poli tical commitment

1 This was seen as a pre-requisite to a sustainable supply.

Role of key individuals 1/3

Key individuals such as the chief might be considered a starting condition which cannot be changed by the programme, or as a desirable end condition.

Prior contribution 2 This is part of the entry process to a community Training 2 Part of an ongoing process. Training needs to be

defined more clearly – see other text Follow up (post pump visits) 2 Part of the process of exiting Information (education, dissemination)

2 The processes should include hygiene education and result in an increased desire for clean water.

?Selection of affordable technology?

2 This was presented as part of the process. It relates to the question as to whether the Volanta is an affordable technology.

Community involvement and participation

2/3

Considered here an end state rather than process, because it is desirable for a sustainable supply

Gender, changing committees

3 The abili ty of a committee to change members and to retain some gender balance is an essential element for a sustainable supply

Spares (Availabili ty/ Access) 3 There must be some way for the communities to access spares.

Abili ty of someone to repair pump

3 There needed to be someone capable of repairing the pump who was available at short notice and who could be paid (either in kind or by cash)

Ongoing contributions 3 VLOM must be paid for – there needs to be a mechanism in place for financial contribution

Savings of contributions 3 This was a debateable point since finance could be collected at the time of the breakage – hence this is a desirable end condition rather than an essential

We should note that the in the above the inputs (1) cannot be change by a single agency. The processes (2) are part of the entry and exit strategy, while the outputs (3) were those elements partners felt needed to be in place if the water supply was going to be sustainable. This framework was used as part of the conceptual construct and is explained more fully in later sections.


5.3 Verification from Conference Delegates The opportunity of the 1999 WEDC Conference in Addis Ababa was used to circulated a questionnaire amongst delegates at the 1999 whose programmes were concerned with handpumps. An analysis of responses is given in Appendix 2(12). The responses confirmed that training is considered as the single most important issue in determining programme sustainabili ty, and that there are many aspects of training that need to be thought about (e.g. who, what, when). They also highlighted the importance of technical issues, mentioning the need for VLOM technologies, and reliable mechanisms for supplying spares. Community participation at a number of levels was widely viewed as a key issue; the major aspects were participation at or before installation, participation in management and maintenance, and community ownership of boreholes. Although most programmes had been handed over to community organisations, a recurring issue was that of ongoing relationships with external agents – local government or implementing agency. One aspect concerned ongoing monitoring and evaluation of both community structures and local government, and the other was the need for local government to support communities. Other points raised included entry strategy issues of participation and prior contribution, either financial or in kind, links to other programmes e.g. health and hygiene, and inabili ty of people to pay. Interestingly, corruption did not appear to be regarded as a major barrier to sustainabili ty, ranking alongside other motivation issues such as migration of trained personnel. Regarding the issue of whether emergency conditions during installation are likely to have a detrimental affect on the future management of boreholes, most respondents felt this need not be the case, although most gave conditions for achieving this e.g. training, cohesive community. This exercise confirmed that experts involved with implementation in the field regard the following as the key issues for sustainabili ty:

• training • technology • participation • ongoing relationships with local government • community mobili sation at installation • links to wider development programmes.

5.4 Sustainabili ty Issues for Framework of Research The combination of the open enquiry process plus verification by the practitioners and the review of the literature, has showed that the following general areas are regarded as important or of interest:

• Community participation • Management and water committees • The role of women • Training • Links to external organisations • Technology


The DFID guidance manual specifically states that it does not address emergency projects. However the issues above are addressed in the context of best practice for development projects. These broad categories were then used as the framework for questioning in the interviews, and each is represented in the variables coded from the survey data. Issues regarded as important by the partner agencies were raised throughout the project, and the data gathered was jointly discussed at a workshop towards the end of the project. This raised some specific and practical questions that partners have to address. Again, points raised fell into the principal categories listed above - community involvement, cost recovery, training, and technology. Feedback from delegates to the WEDC conference revealed categories of concern that were remarkably similar to those from the literature review. This confirmed that the framework used for data collection did indeed cover all of the appropriate subject areas. A general observation is that issues raised by project partners and delegates to the conference tend to reflect either broad opinions, or specific problems. They are unlikely to mention issues that have a positive influence on their work, as most of their attention at work is dedicated to dealing with issues that cause problems. Most items raised in the literature are specific points of fact or interpretation that help explain a more general assertion. Table 4 summarises relevant issues raised in the literature, and indicates where this has been mentioned by practitioners (either as a concern or an opinion).

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

47

Tab

le 4

Sum

mar

y of

Iss

ues

rais

ed b

y L

iter

atur

e an

d P

artn

ers

as im

port

ant t

o ex

it st

rate

gies

Lite

ratu

re

Issu

e ra

ised

by

pr

actit

ione

rs

Par

tici

pati

on

D

esig

n

reco

mm

enda

tion

s C

omm

unit

y pa

rtic

ipat

ion

at

or

befo

re in

stal

lati

on

Des

pite

the

time

take

n, e

ffor

t sho

uld

be m

ade

to m

axim

ise

the

room

for

man

oeuv

re th

at

com

mun

ity

mem

bers

hav

e in

rel

atio

n to

asp

ects

of

serv

ice

desi

gn.

x

Pr

omot

es o

wne

rshi

p C

omm

unit

y pa

rtic

ipat

ion

is im

port

ant

to d

evel

opin

g a

sens

e of

res

pons

ibili

ty f

or, a

nd

cont

rol o

ver,

the

loca

l car

e of

sys

tem

s x

R

ole

of w

omen

W

omen

hav

e a

ben

efic

ial i

nflu

ence

on

dec

isio

n m

akin

g x

Des

ign

C

onst

rain

ts

Abi

lity

and

trai

ning

Pe

ople

shou

ld o

nly

be a

llow

ed to

par

tici

pate

in d

ecis

ion

mak

ing

whe

re th

ey h

ave

appr

opri

ate

info

rmat

ion

and

ski

ll

C

onfli

ct in

com

mun

ity

Part

icip

ator

y sy

stem

s w

ill n

ot b

e ap

prop

riat

e w

here

con

flict

s of

int

eres

t as

so s

ever

e th

at

they

can

not b

e re

solv

ed th

roug

h a

gree

men

t

Su

stai

nabi

lity

of

part

icip

atio

n A

ltho

ugh

a p

roje

ct c

an b

e im

plem

ente

d w

ith

a h

igh

deg

ree o

f com

mun

ity

part

icip

atio

n, if

th

e is

sue

of h

ow p

arti

cipa

tion

is to

be

sust

aine

d th

erea

fter

is n

ot a

ddre

ssed

it w

ill n

ot

cont

inue

.

x

G

enui

ne c

ontr

ibut

ion

by

wom

en

Initi

ativ

es ta

ken

can

be

base

d on

the

inte

rest

s of

men

, who

ven

tril

oqui

se th

eir

plan

s th

roug

h w

omen

, esp

ecia

lly

in a

reas

whi

ch h

ave

trad

ition

ally

bee

n th

e re

spon

sibi

lity

of

men

.

C

ultu

ral s

tand

ing

of

wom

en

In s

ome

vill

ages

, wom

en w

ere

relu

ctan

t to

take

par

t in

mee

ting

s in

the

pres

ence

of

men

.

H

eavy

wor

kloa

ds o

f w

omen

W

omen

’s in

volv

emen

t is

hin

dere

d by

thei

r w

ork

load

, lea

ving

litt

le ti

me

to a

tten

d to

ad

diti

onal

dut

ies.

Out

com

es

Acc

ess t

o lo

cal

know

ledg

e an

d sk

ills

The

ben

efit

s of

com

mun

ity

part

icip

atio

n in

clud

e lo

cal k

now

ledg

e of

wat

er s

uppl

ies,

un

ders

tand

ing

of s

ocio

-cul

tura

l str

uctu

re a

nd im

prov

ed c

omm

unic

atio

n.

C

apac

ity

buil

ding

Pa

rtic

ipat

ory

appr

oach

es e

mpo

wer

peo

ple

to ta

ke re

spon

sibi

lity

for

thei

r ow

n s

ervi

ces.

T

he e

mpo

wer

men

t of

wom

en c

an o

ften

be

mor

e im

port

ant

to o

vera

ll de

velo

pmen

t ob

ject

ives

than

con

trib

utio

ns th

ey m

ay m

ake

to th

e w

ork

of w

ater

com

mitt

ees.

x

Pa

rtic

ipat

ion

in

mai

nten

ance

x

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

48

Org

anis

atio

n

Des

ign

re

com

men

dati

ons

Rol

e of

ext

erna

l age

ncie

s It

can

be

attr

acti

ve to

use

exi

stin

g st

ruct

ures

e.g

. loc

al g

over

nmen

t. P

ubli

c an

d pr

ivat

e pa

rtne

rshi

ps c

an m

ake

optim

um u

se o

f al

l st

akeh

olde

rs.

O

n-go

ing

supp

ort

On-

goin

g co

nsul

tati

ve p

roce

sses

are

impo

rtan

t in

inst

illin

g ow

ners

hip

and

mai

ntai

ning

sh

ared

res

pons

ibili

ties

. x

E

xter

nal s

uppo

rt

(gov

ernm

ent)

C

omm

unit

y m

anag

emen

t and

mot

ivat

ion

initi

ativ

es re

quir

e a

low

leve

l but

ong

oing

inpu

t fr

om e

xter

nal a

gent

s an

d as

sum

e th

at a

wat

er a

genc

y ha

s no

t act

uall

y ex

ited

from

the

area

.

x

Pa

ymen

t of

mem

bers

C

reat

e pa

id jo

bs w

here

ver

poss

ible

and

red

uce

reli

ance

on

vol

unte

er la

bour

. T

his

is

criti

cal t

o en

sure

that

mem

bers

are

mot

ivat

ed, a

nd w

omen

’s n

on-e

cono

mic

sta

tus

is

impr

oved

if t

hey

earn

inco

me.

C

ompl

emen

tary

act

iviti

es

Ben

efit

s as

soci

ated

wit

h im

prov

ed w

ater

sup

plie

s w

ill o

nly

be re

alis

ed if

pro

gram

mes

are

co

mpl

imen

tary

to h

ealt

h, s

anit

atio

n a

nd h

ygie

ne p

rogr

amm

es.

x

Des

ign

C

onst

rain

ts

Loc

al c

omm

unit

y ch

arac

teri

stic

s A

bilit

y to

pay

- le

vel o

f ed

ucat

ion

and

fam

ily s

ize

will

inf

luen

ce W

TP;

abi

lity

to p

ay

vari

es b

etw

een

com

mun

ities

. x

M

otiv

atio

n o

f m

embe

rs

Onc

e th

e m

otiv

atio

n o

f a

cap

able

mem

ber

has

been

ero

ded,

it c

omm

only

giv

es w

ay to

co

rrup

tion

and

abu

se o

f po

siti

on.

x

L

ocal

pow

er s

truc

ture

s M

echa

nism

s fo

r ap

poin

ting

wat

er c

omm

ittee

s ar

e of

ten

irre

leva

nt, a

s ch

iefs

and

pow

erfu

l le

ader

s m

anip

ulat

e pr

oced

ures

beh

ind

the

scen

es.

Fr

agm

ente

d co

mm

uniti

es

Whe

re th

e so

cial

str

uctu

res

that

mak

e up

a so

ciet

y ha

ve b

een

des

troy

ed, l

eade

rshi

p ty

pes

of i

nstit

utio

ns a

re u

nlik

ely

to b

e su

stai

nabl

e.

M

etho

d of

ele

ctio

n

(wom

en)

Whe

re w

omen

are

app

oint

ed b

y po

wer

ful p

eopl

e, th

ey w

ill f

eel o

blig

ed to

agr

ee w

ith

th

ese

peop

le.

Mal

e co

mm

ittee

mem

bers

may

sim

ply

refu

se to

incl

ude

toke

n fe

mal

e m

embe

rs im

pose

d by

NG

Os.

x

In

still

ing

owne

rshi

p C

ost r

ecov

ery

does

not

aut

omat

ical

ly le

ad to

pro

ject

ow

ners

hip

and

sust

aina

bilit

y. W

TP

is

low

whe

re p

eopl

e be

liev

e go

vern

men

t sho

uld

prov

ide

free

wat

er.

C

ost r

ecov

ery

Prob

lem

s of

non

-pay

men

t and

aff

orda

bilit

y re

quir

e eff

ecti

ve c

omm

unit

y m

anag

emen

t. x

C

olle

ct m

oney

reg

ular

ly

or a

t bre

akdo

wn?

[s

peci

fic

issu

e ra

ised

by

part

ners

] x

Tra

inin

g (E

duca

tion

)

Des

ign

re

com

men

dati

ons

Top

ics

to in

clud

e te

chni

cal,

org

anis

atio

nal

and

fina

ncia

l are

as.

Com

mitt

ees

will

onl

y be

eff

ecti

ve w

hen

giv

en tr

aini

ng in

tech

nica

l, o

rgan

isat

iona

l and

fi

nanc

ial s

kill

s. A

ppro

pria

te tr

aini

ng is

the

key

to s

olvi

ng c

once

rns

of p

arti

cipa

tion

, m

anag

emen

t and

gen

der.

x

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

49

and

fina

ncia

l are

as.

man

agem

ent a

nd g

ende

r.

C

ontr

ibut

es to

mot

ivat

ion

Ong

oing

trai

ning

pro

vide

s a

num

ber

of b

enef

its

as it

can

hel

p to

mot

ivat

e co

mm

unit

y or

gani

sati

ons,

to r

efre

sh s

kill

s, a

nd to

cou

nter

mig

rati

on o

f sk

illed

per

son

nel

Des

ign

C

onst

rain

ts

Gen

der

issu

es

Wom

en te

nd to

be

offe

red

trai

ning

in “

soft

” sk

ills

such

as

heal

th, h

ygie

ne a

nd

adm

inis

trat

ion,

whi

lst

men

rec

eive

trai

ning

in p

ract

ical

ski

lls

such

as

mas

onry

and

m

echa

nica

l mai

nten

ance

; th

is c

an e

xagg

erat

e di

ffer

ence

s in

sta

tus.

L

evel

of

trai

ning

[s

peci

fic

issu

e ra

ised

by

part

ners

] m

aint

enan

ce an

d/or

rep

airs

x

Com

pete

nce

D

esig

n

Con

stra

ints

C

apac

ity

of li

nk

orga

nisa

tion

G

over

nmen

t dep

artm

ents

gen

eral

ly h

ave

a te

chni

cal f

ocus

and

are

not

equ

ippe

d to

dea

l w

ith

soc

io-e

cono

mic

asp

ects

of

deve

lopm

ent p

rogr

amm

es

Con

text

Des

ign

re

com

men

dati

ons

Tec

hno

logy

mus

t be

appr

opri

ate

Tec

hno

logy

sho

uld

be si

mpl

e to

mai

ntai

n a

nd re

pair

. Fa

mili

ar te

chno

logy

can

con

trib

ute

to s

ucce

ss an

d un

fam

iliar

tech

nolo

gy to

fail

ure.

x

A

vail

abili

ty o

f sp

ares

A

vail

abili

ty o

f sp

ares

is c

riti

cal i

n a

chie

ving

low

dow

ntim

es.

Exi

stin

g go

vern

men

t st

ruct

ures

may

be

able

to s

uppl

y sp

ares

. x

Des

ign

C

onst

rain

ts

Alt

erna

tive

supp

lies

W

TP

is h

ighe

r w

here

no

alte

rnat

ive

supp

ly e

xist

s. W

itho

ut a

dequ

ate

dem

and,

a w

ater

su

pply

is u

nlik

ely

to b

e su

stai

nabl

e.

x

Pr

oble

ms

for

the

priv

ate

sect

or

A d

iffic

ulty

for

the

priv

ate

supp

ly o

f sp

ares

seem

s to

be

the

com

bina

tion

of

low

turn

over

an

d sm

all

prof

it m

argi

ns.

x

The

re w

ere

a fe

w is

sues

whi

ch w

ere

rais

ed in

all

four

sou

rces

: •

Ow

ners

hip

– al

l end

orse

d th

e ne

ed fo

r co

mm

unity

par

ticip

atio

n in

dev

elop

ing

a se

nse

of o

wne

rshi

p;

• T

rain

ing

– al

l rec

ogni

se th

e ne

ed fo

r co

mpr

ehen

sive

trai

ning

, and

the

wid

e ra

ngin

g be

nefit

s th

at t

his

can

brin

g.

• T

echn

olog

y –

the

tech

nolo

gy in

volv

ed s

houl

d m

atch

the

capa

city

of

the

com

mun

ity, p

rinc

ipal

ly t

he is

sue

of s

impl

icity

for

rep

air

and

mai

nten

ance

.


6 Results and Analysis

6.1 General discussion The general framework above highlights the key elements of a sustainable water supply. This section now looks at the three case studies as a whole to discuss some of the details involved with sustainable supplies. We will then use these results and this discussion to comment on issues and processes that should be addressed for a successful exit strategy. We have seen that while some policy and planning issues influence the sustainabili ty of the water supply, the main focus of the research was on the community’s abili ty to sustain the supply. In particular we consider how the different variables1 affect the functioning of the water supply. It has been stated that mechanical functioning of the borehole is taken as the primary indicator of sustainabili ty, primarily the frequency of breakage and average downtime when broken (see Box 3.1.8). The analysis starts with this primary indicator of sustainabili ty, and presents the links within the context of a flow chart representing the decisions and delays encountered in getting a pump repaired. This is followed by the analysis of data relating to the three categories of social factors (Section 5.1).

6.2 Programme Outcome - Pump reliabili ty There are two key parameters that describe the functionality or reliabili ty of the borehole. They are “how frequently the pump breaks down” and the “average downtime”. We can now discuss what factors affect these two elements2. Regarding Frequency of breakdown:- logically this should be affected by the technical factors (depth of borehole, quality of the pump manufacture, suitabili ty of the pump to its working conditions, age, etc). Social factors will only really come in to play in terms of the pressure on the pump (the amount it is used) and the quality of the operation, maintenance and repairs. If rough handling is allowed, or children can play on the pump, one might expect it to break down more often. If no preventative maintenance is undertaken one might expect the pump to breakdown more frequently as it gets older. A poor repair might only temporarily solve the problem, and a new breakdown might occur soon after a repair.

1 The data is treated as a whole and differences between agencies are not presented. All variables were checked for agency and country differences and the results taken into account. However, it was decided that since this is not a project based evaluation of an agency’s work, these results should not be presented. Therefore the data is always treated as a whole. 2 Other pump function variables. It is important to note that although the pump is considered to be non-functional if the pump is out of action, the status of the pump-head in general could also be considered as part of the functioning aspect of the water point. The word function has been used as opposed to performance, as measurements of actual yields over time were not taken into account in the survey. Two other depended variables were taken into account alongside the two key indicators of pump function discussed above. These are:

• The operational status of the pump, i.e. whether it was functional at the time of the visit. This was a snapshot observation only. This is considered to be of only limited importance

• Observed status of the pumphead, i.e. observation by the researcher on the care taken to keep the pumphead and its surrounds in a clean and controlled state. This issue is questionable as a measure of function as many of the pumps that had never experienced problems also had some of the worst cared for pumpheads.

However, since these variables did not highlight any new insights or contradict those documented above, the full analysis is not presented here.


These variables were captured in the interview responses and will be discussed below. However before doing so, let us look at the process of repair. The diagram in Figure 3 attempts to present the decision making flow for repairing a rural pump as indicated by the interviews. Surprisingly, the literature survey did not identify an equivalent diagram. Regarding the average downtime;- the decision flow diagram below has been created to assist the discussion of the time taken to repair the pump. The decisions required each affect the time it takes to repair and it is the cumulative time taken that is captured in the downtime variable. Using this diagram, we may now consider each decision and using the data indicate whether this factor was of significance to the formation of the total downtime. For instance – “ is finance available to fund the mechanic?” will probably be strongly influenced by whether the community collects money regularly or whether it waits for the pump to break and then collects the money (as many communities do).

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

52

Bre

akag

e

Is

mec

hani

c

with

in c

yclin

g

dist

ance

?

Is

tech

nica

l ad

vice

on

han

d?

Is fi

nanc

e

avai

labl

e to

fu

nd

m

echa

nic?

Is p

robl

em

with

in

capa

bilit

y of

m

echa

nic?

A

re sp

ares

av

aila

ble?

Prel

imin

ary

diag

nosi

s

Are

mec

hani

cs

still

on

site

?

Rai

se

mon

ey fo

r bu

s fa

re

Res

pons

e

Rep

air

W

as re

pair

O

K?

Rai

se

fina

nce

Is

fina

nce

avai

labl

e fo

r sp

ares

?

Rai

se

fina

nce

C

an sp

ares

be

pur

chas

ed

loca

lly?

Purc

hase

sp

ares

C

an sp

ares

be

pur

chas

ed in

co

unt

ry?

C

an fi

nanc

e be

ra

ised

for

trav

el to

get

sp

ares

?

Una

ble

to

repa

ir p

ump

N

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

Y

N

N

N

N

N

N

N

N

N

N

Car

ry o

n

pum

ping

Det

aile

d di

agno

sis

F

igur

e 3

Flo

w C

hart

for

Ana

lysi

s of

Dow

ntim

e

Key

A

ctio

n th

at r

equi

res

time

Dec

isio

n re

quir

ed


6.3 Building the picture on average downtime.

6.3.1 Diagno sis The first element on the diagram is the preliminary diagnosis. This will be the chief, the committee or others in the community making a first guess at what might be wrong with the pump. They then have to decide who should repair the pump. Within the sample there are a variety of possible responses. Some of the early emergency pumps have been left with the impression they should call the Government. For some pumps a local area mechanic was created, sometimes a government employee and sometimes a private specialist (e.g. car mechanic). Other pumping communities had various numbers of people trained within their community, and the case studies have differences in the levels of training these groups or individuals received.

6.3.2 Access ing the mechanic The first decision point is whether there is suitable technical competency on hand. Realistically, for those situations where there is an “area mechanic”, the mechanic might be within another community but within a bicycle ride – so no funds are required for the initial consultation. However, where the mechanic is a long distance away, perhaps at a district centre or further, there may be a need to raise transport costs for the messenger, and this may delay the start of the process. Data on “who usually repairs pump” is presented in Figure 4, and shows that the majority of pumps are repaired internally i.e. by committee or local mechanic.

0%

10%

20%

30%

40%

50%

60%

Com

mite

e

Loca

l pum

pm

echa

nic

Gov

ernm

ent

agen

t

Piv

atel

yco

ntra

cted

mec

hani

c

Age

ncy

No

one

Figure 4 Frequency Distr ibution: Who usually repairs pump

There may also be some delay if the community mechanic(s) is away. The early projects trained men, and sometimes only one man. Men often travel to find work and may be away from the vill age. There may, therefore, be a difference in time delay according to whether training was given to one man, or a sub group of two or more. If the above really affects the downtime, one would expect a difference on “who repairs the pump” between external and internal (community) agents with respect to “downtime”. There is such a difference, and the Mann Whitney test shows a statistical significance. Looking at the data in more detail reveals that downtime is significantly shorter where


committees repair pumps rather than local mechanics, which confirms the potential delay in getting hold of a single person. We can also look at the difference between training one person and several persons. This to some extent is captured by the variable “Who was trained” . The average downtimes are lower where the whole committee was trained rather than just mechanics. (The discussion on the words “committee” or “caretaker” has been taken into account).

6.3.3 Fund ing the repair An element of finding the mechanic and asking he/she/them to look at the broken pump, is that the community must have, or have the abili ty to promise, payment to the mechanic. Where communities have internal mechanics, this may be in kind or by preferential treatment at the pump head when collecting water. In the interviews there were a wide range of responses as to how a community mechanic was paid. Where external mechanics are required, there will probably be some element of payment. In some cases within the research interviews, a few respondents claimed that they could approach local or central government to repair the pump. These few cases stated that this was the “government’s responsibili ty and would be undertaken for no payment” . However respondents from the same vill age or community also talked about having to pay. The availabili ty of payment was difficult to measure. It was not possible to get clear answers on how much the committee had in hand. It varied between one treasurer who claimed to have several milli on under his bed at home, to other committee members who described how they gave the money to a wealthy man because he would not run off with it, and that he stored it in his own livestock – to be called upon as and when required. Transparency and accountabili ty varies from vill age to vill age. However if payment is a critical factor in the total downtime, one would expect a significant difference to be found on the variables “Pump maintenance payment” and “Payment of quota”. The former captures whether people make regular payments or not, the implication being that if people pay regularly then the committee might have reserves readily accessible. The latter variable is whether they make a payment or not. This latter variable is weaker in its implication that there are reserves held by the committee. Neither of these variables shows significance with respect to downtime. The implication of this is that the holding of reserves does not shorten downtime significantly; the corollary being that collecting money to fund the repair process does not delay it significantly.

6.3.4 Mechanic capabili ty Having got the mechanic to the pump, they may undertake a more detailed diagnosis of the problem to determine if it is within their capabili ty to repair the system. It is possible that the person summoned does not have the capabili ty to repair the problem. The case studies have trained people to varying levels of capabili ty. A government service, or an area mechanic, will have the tools and skill s to tackle the more difficult problems. The community mechanics were often only trained in maintenance and simple repair. The


agencies in the study are undertaking training to upgrade the technical skill s of the community mechanics. If this inabili ty to repair the pump by the local mechanic is critical to the total downtime, one might expect a difference with downtime (Mann Whitney MW) to be shown by the variables “Length of training” , “Level of training” and possibly “Follow-up training” . MW differences show on the length of training. It might be expected that a significant difference in downtime would result from the length of training received. The longer the training the shorter the downtime. Length of training should reflect the level of skill imparted. In the cases where training was provided for longer than a week there was a significantly shorter downtime when the pump was broken. This is logical, as one would expect the more highly trained to be able to reduce the period of downtime. Neither the variable “Level of training” nor “Follow-up training” show on the average downtime. However, the level of training only has response from the committee members and therefore does not have a very high sample of responses. Similarly Follow-up training was questioned by the agencies as a potentially unreliable variable.

6.3.5 Spares availabili ty If the mechanic has the capabili ty and willi ngness to repair, there is still the question of spares availabili ty. This again depends to some extent on finance. Do the community have the financial resources available to purchase spares (or have they pre-purchased and stocked them)? This part of the process is very susceptible to spares availabili ty generally, and this leads to the discussion on private suppliers. Reported problems with pumps were divided into six categories, and indicate that, for the programmes covered by this study, the provision of spares is not foremost in the communities mind.

0%

5%

10%

15%

20%

25%

30%

35%

tech

nica

l

soci

al

orga

nisa

tiona

l

com

pete

nce

acce

ss t

osp

ares

popu

latio

npr

essu

re

chief

cttee

mech

none

Figure 5 Incidence of repor ted problems divided into catagor ies

However, one must keep in mind that the community may not be aware of problems with spares. This is born out by the breakdown of stated problems by type of respondent. It shows chiefs to be feel technical, social and organisational problems more keenly. Not


surprisingly, committee members see few problems with organisation and access to spare – presumably they are responsible; however, they are more aware of shortcomings in technical competence. Again, it is not surprising that pump mechanics feel their work is hindered by poor access to spares. A small number (49) of comments explaining delays in pump repairs were recorded mainly from those involved in the actual repair. These can be categorised:

1. Access to materials (mainly glue for major repairs involving raising pipes!) 2. Due to individual or organisation 3. Low priority or motivation 4. Organisational problems 5. Technical problems

0

5

10

15

20

25

30

35

40

mat

eria

lsav

aila

bilit

y

orga

nisa

tiona

l

indi

vidu

al /

orga

nisa

tion

low

mot

ivat

ion

tech

nica

l

%

Figure 6 Reasons for delay in repairs

Figure 6 shows that materials availabili ty is quoted as being responsible for most delays. Within organisational issues, comments referred exclusively to problems with funds. In one agency the comments predominantly referred to “finding a mechanic”. The other feature to consider is year of installation. Mechanisms for providing access to spares appear to have been sorted out in later installations. And we should note that later installations appear to have been installed in locations with high population pressure (to be discussed below)


05

101520253035404550

tech

nica

l

soci

al

orga

nisa

tiona

l

com

pete

nce

acce

ss t

osp

ares

popu

latio

npr

essu

re

pre 95

95 -

Figure 7 Repor ted problems according to year of installation

6.3.6 Other factors high lighted by the data While the above follows the logic imposed by the flow diagram, the data also highlighted a few other variables that seemed critical to the downtime. Social Conflict. Figure 5 shows how most of the problems reported can be attributed to social issues and population pressure, although these are related to some extent. Many of the social complaints relate to payment (e.g. people don’t pay; quarrelli ng over where to keep money collected), but other frequently mentioned social issues include conflict (e.g. fights), lack of motivation (e.g. community not active), and preference for alternative water sources. Most population pressure type complaints clearly relate to too many households. Other complaints tend to be consequences of the same problem e.g. long queues, insufficient water, need more pumps etc. This confirms the importance of achieving good social organisation, especially for collecting money. However, it also shows that there are certain social factors which can disrupt sustainable functioning of management structures, irrespective of the amount of planning, participation, and the quality of organisational structures. In particular, the data highlights social conflict, and poverty, both of which are referred to in the literature. Installation. It can be noted that downtime varies significantly with whether people were present at the installation or not. There are significant differences with downtime (MannWhitney) with the variable “present at the installation of the pump”. A slightly higher significance was recorded for the variable “participated in installation” . This may reflect the involvement of the community in the care of the pump and this can be explored in the following sections on organisation and participation. Of those near to pump (proximity) 82% were present at installation, compared with 90% of those far from pump. Of those near, 69% participated in installation, compared with 71% of those far. These results are not statistically conclusive.


Participation in installation can be a relatively simple act. Responses to “Participation in Installation” were divided into the following categories: 1. labour in preparation 2. labour in installation 3. materials 4. organising 5. money matters 6. transport We see from Figure 8 that the contribution of labour and materials (sand and gravel) were the main mechanisms for involvement.

0

5

10

15

20

25

30

35

40

45

mat

eria

ls

labo

ur in

inst

alla

tion

labo

ur in

prep

arat

ion

mon

eym

atte

rs

orga

nisa

tion

tran

spor

t

%

Figure 8 Par ticipation in installation (how)

Emergency installation. The variable “emergency installation” shows a significant difference. Since this is an important part of the research it is discussed in greater detail below. Year of installation. The age of the pump appears to influence only the average downtime but not the frequency of breakage. Shorter downtimes for pumps installed from 1995 onwards probably reflects the shift towards greater community mobili sation in later programmes, in which case repairs carried out by local mechanics tend to be carried out more quickly. Type of pump. Two main types of pump are compared - the Volanta for the deeper pump, and the Afridev for the shallower pumps, ideally those with pumps set at less than 30 meters. Only 5 Volanta pumps were captured in the sample. In Malawi 3 Nadia shallow pumps were also observed, however, in this comparison the Nadia have been excluded. The Volanta pumps breakdown more frequently, and are down for longer. It is interesting to note that the influence on the performance indicators remains the same even when only considering pumps set below 29 meters. This suggests that there is a major problem with the Volanta pump, particularly as it is specifically designed for the deeper setting compared to the Afridev.


The interesting issue is that with the Afridev deeper pumps the downtime when broken was not significantly different. The respondents using deep pumps expressed a greater sense of pressure which corresponds to the significant rise in time waiting to draw water. Alternative unreliable source. Access to alternative sources of surface water, even if unreliable, has a significant impact on the attention to the care and urgency of repair of the pump at the time of breakage. Where no other water source exists, the care and speed of response to breakage is greater (average downtime is shorter). This suggests that in the long term health education focusing on reinforcing the dangers to health of drinking surface water might have an impact on the downtime of the pumps. Quality of water : The salt content of the water appears to have an influence on the downtime. The downtime when broken is also significantly longer when the pump's water is more salty, . A number of interviews showed that people used salty boreholes for other domestic needs and not for drinking. If this is so then the slow response to repairing the borehole is a logical one. Sett lement. Where the respondents claim to have settled due to the presence of the pump (n = 30 or 16%), the average downtime when broken was significantly higher. The instance of settlement due to the presence of a pump does appear to influence the operational status and sustainabili ty of the pump negatively. Interestingly, there is no significant correlation between settlement and felt pressure or the time waiting for water. This indicates that where settlement occurs due to the presence of a functioning pump the social cohesion may not be as strong and this may reflect a weaker commitment to the care of the pump. Community Involvement. The average downtime was lower where committee meetings were held regularly, which confirms the positive effect of organisational capabili ty within the community. However, whether people made payments regularly or irregularly had no significant influence on pump performance, so issues of community organisation have only a weak influence on performance.

6.4 Frequency of Breakdowns Given the above, we may now look at the data surrounding frequency of breakdown. Most respondents seem to distinguish between maintenance activities (preventative) and an actual breakage requiring repair. Depth and Type of pump. There is a very high difference on frequency of breakdown based on depth and type of pump. In this case the comparison was made between pumps set above and below 30 metres. The deeper the pump is set the more frequent the breakage . These findings are logical and are to be expected, especially when setting Afridev pumps below 29 metres. The depth at which it is needed to set the pump was not open to much alteration as it was found that pump and static water depths only differed on average by 1.5 metres. Where water is deeper and its withdrawal more problematic, the community appears to respond positively to the challenge. The change of pump to one more appropriate for the depth


does not appear to have provided a solution, as is discussed above in the context of average downtime under type of pump. The Volanta pumps breakdown more frequently , and are down for longer than the Afridev pumps. The snapshot given from visits also shows a higher proportion of the Volanta pumps were out of action . The general status of pumpheads at Volanta pumps is, however, higher ; this shows that care of the pumphead cannot make good technical issues pertaining to pump design. It does however, demonstrates a positive response to problems by the community. Better care of the pumphead may have been seen as a way of attracting continued external help from the implementing agency or an expression of the importance of the pump to the community. They are doing what is in their power to maintain it. It is worthwhile noting that Pump problems tend to be more severe (hardware) where pumpheads are better kept. This runs contrary to what might be expected and has some explanation in the predominance of good care in the Volanta pumps. It is interesting to note that the influence on the performance indicators remains the same even when only considering pumps set below 29 meters. This suggests that there is a major problem with the Volanta pump, particularly as it is specifically designed for the deeper setting compared to the Afridev.

6.4.1 Reliance on mechanics Who repairs the pump: As discussed above this is a comparison between external and internal (community) agents. This issue has a significant influence on both of the pump performance indicators, i.e. frequency of breakage and downtime. Pumps tend to break down more frequently where they are repaired locally as opposed to by external staff , but downtimes tend to be shorter . As indicated on the decision flow diagram (Figure 3) this seems to indicate that the quality of repairs carried out by local, non-specialised, personnel may be poorer than those carried out by more highly trained external agents. It also shows that where a community has the internal resources to maintain a pump, they can be mobili sed quickly resulting in shorter downtimes. These two important issues work against each other in making up total reliabili ty or availabili ty. However, where it is a local personal who carries out repairs, their abili ty is held in high regard. This apparent anomaly indicates that perceived abili ty is more strongly influenced by the abili ty of mechanics to get a pump repaired quickly; people tend to forget how often it breaks down. It may also reflect the social cohesion amongst communities which have the capabili ty to carry out repairs, and a certain pride in that abili ty. Conversely, it may reflect a lack of interest and absence of opinion in the abili ty of external mechanics. This confusion explains why perceived abili ty has no affect on the frequency of breakdowns. Who was trained: This represents a comparison between cases where one or a few local mechanics were trained in pump maintenance as opposed to the whole committee. A significantly better performance appears to be achieved where training is given to the whole committee rather than just to mechanics - both frequency of breakage and average downtimes are lower.


6.4.2 Use of the pump. One might expect that the way the community use the pump affects how often it breaks down. However the data seems to suggest a counter-intuitive response. Where rules are applied there is a significant difference regarding frequency of breakage and the status of the pumphead. Existence of rules appears to correspond to a higher instance of breakdown . The issue of the status of the pumphead is counter-intuitive as the existence of rules appears to correspond to the less well-kept sites . This may also reflect a retroactive action on the part of the committee given the lack of care. A related variable is Sanctions, which does not appear to influence the performance of the pump, however, there is a tendency (not significant) for the communities who do not apply sanction to have fewer breakdowns. Breakdowns may motivate the committees to apply sanctions given the more constant demand for funds. Payment of quota. Those who say they don’t make payments have lower frequency of breakage than those who do. How can this be possible? One explanation might be that those who are socially organised to make payments also use local community mechanics who, as we have seen above, have a lower quality of work than external mechanics. This link might explain the unusual significant difference of payments with respect to frequency of breakdown. Pressure on the water source. The time spent waiting to draw water, which is an indicator of pressure on the borehole, shows that where people have to wait on average three or more hours, the frequency of breakage is higher . The pressure on the borehole, as expressed by the community, registers a difference with regard to status of the pump at the time of the research visit. Both of these indicate, as might be expected, that the borehole is more likely to break down where it is subject to higher pressure.

6.4.3 Initial involvement Pump requested: if the pump was requested by the community or provided without prior request. There is a significant difference such that more frequent breakage occurs in those instances where the pump has been requested. However, downtime tends to be higher where the pump was not requested (not significant). Siting, who chooses the site for the pump - the community or an external agent. The frequency of breakage is significantly higher where the community claim to have sited the pump as opposed to the drill ers. This suggests that with the issue of siting the pump it is better left to the experts with only some consultation with the community. Prior contribution. It should be noted that this only applies to one agency, and was introduced with the more recently drill ed pumps. Its value as a comparative variable is, therefore, questionable in this instance. In the cases where a prior contribution was made, there is a significantly lower instance of pump breakage . The time waiting for water is also significantly lower with those respondents who reported making a prior contribution . The correlation between prior contribution and proximity to the pump may suggest their closer association with the pump and shorter distance and time to draw water. There is a close correlation between the prior contribution and participation at the installation of the pump.


6.4.4 Involvement and Ownership Ownership. It is also important to note variables that do not show a significant difference on pump functioning. The “ownership” variable, which was a stated response of “who owns this pump?” does not show any significant difference on either downtime or frequency of breakage. One might expect that those who stated that the vill age or vill age authorities owned the pump might respond quicker to pump repairs than those who state that the government owns the pump. However the data does not support this logic. Similarly Community involvement (in the committee), the number of women on the committee and the election process for the committee all do not show as significant to pump function. This will be discussed further below.


6.5 Social factors

6.5.1 Section introdu ction In the above discussion we have noted two key parameters of pump functioning. Whilst reliabili ty (the combination of “frequency of breakdown” and “downtime”) may be the best indicator of whether a pump is effectively serving a community with water, NGOs take a wider view of interventions. The primary purpose of providing a clean water supply is to improve health, so pump installations in a development context are usually included as part of wider integrated development programmes, accompanied by hygiene and sanitation promotion (and possibly other development actions). Therefore when we come to discuss ownership and the social mobili zation to undertake Vill age Level Operation and maintenance (VLOM), we must consider the question from two angles. On the one hand we have literature stating that VLOM is the best practice for long term sustainabili ty, and therefore we would expect to see that VLOM results in greater availabili ty of water. Our analysis above questions whether this is actually so. Vill age or community level maintenance and repairs tend to lead to a greater frequency of breakage than if an area mechanic or government team take action. The downtime is lower for local repairs (as long as spares are available). However the combined effect of higher frequency of breakage and lower downtime for a VLOM response does not actually result in a higher availabili ty of the pump when compared with area mechanic systems or even Government responses (in general). On the other hand we must also consider the wider agenda of social mobili sation as an objective in its own right. If the community have a greater capacity to solve problems, then they may mobili ze this capacity to be applied to other problems in the vill age. For water programmes health and sanitation are an integrated response and the agencies in the case study undertake health and sanitation training as part of their projects. The following analysis considers issues influencing the three categories of social factors, and examines interactions between them. The primary relationships are represented in the figure below. This figure makes a distinction between preset conditions, over which project planners have little or no control, and programme design features, and shows how each makes an impact not only on borehole performance, but also on social factors. The historical issue of settlement is presented as a secondary level programme output.

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t St

rate

gies

” R

7136

P

age

64

F

igur

e 9

Pri

ncip

al R

elat

ions

hips

hig

hlig

hted

by

Cas

e St

udy

Dat

a

KE

Y

Pres

et c

ondi

tions

Pr

ogra

mm

e de

sign

Pr

ogra

mm

e pr

imar

y ou

tcom

e Pr

ogra

mm

e se

cond

ary

outc

omes

U

npl

anne

d o

utco

mes

Com

pete

nce

Org

anis

atio

n P

arti

cipa

tion

of

the

wid

er

com

mu

nity

Ed

uca

tion

and

tr

ain

ing

Pre

ssur

e on

we

ll:

Tim

e qu

euin

g

Te

chn

ical

issu

es:

D

epth

, pum

p ty

pe, q

ual

ity o

f w

ater

Alte

rnat

ive

wa

ter

sour

ces

Em

erg

ency

co

nditi

ons

Pre

set c

ond

itio

ns

Fea

ture

s of

pr

ogr

amm

e

desi

gn

Are

as o

f pr

ogr

amm

e

imp

act

Bor

ehol

e

func

tion

Pe

ople

se

ttlin

g du

e to

wel

l


6.5.2 Competence Competence was considered as a whole by looking at the variables:-

• Who usually repairs pump • Number of trained pump mechanics • Last agency pump maintenance visit • Abili ty of local committee or trained pump maintenance rep's

The last variable considers people’s perceptions as to the competence of their vill age representatives, be they committee and/or pump mechanic. We have already seen how these factors influence pump functioning. We have noted that actually the greater skill that lies in external mechanics results in a lower frequency of breakdowns. However, even though in terms of pump availabili ty the gains of lower downtime offset the loss on frequency of breakage, in this section we are considering local capacity a desirable end state. We consider local capacity to be part of the project’s wider objectives on health, sanitation and general development. However, it is diff icult to extract the relationship between local competence and health and sanitation. A full Knowledge And Practice survey was not undertaken. Some insights are given through the care of the pumphead and these are explored in a section below. As might be expected, training issues have the greatest influence on competence, especially who was trained, and when they were trained. Where committees received training they tend to carry out repairs, and where it was mechanics who were trained, it is they who carry out repairs. Providing training for committees has resulted in a larger number of trained mechanics, which helps reduce dependence on a single mechanic, as pointed out above. Where training was given some time after installation, fewer people were trained (concentrating on mechanics rather than committees), but they tend to be better supported by more frequent agency visits. Where it is a local person who repairs the pump, the abili ty of the mechanics and committee are held in higher regard. Perceived abili ty is also higher where longer training has been given, as might be expected. Training is a component of project design, and reflects changes in emphasis over the years. Later installations (post 1995) have a higher number of trained mechanics, and are repaired locally, and management personnel are regarded as very able, all of which reflect a high degree of internal capacity within the community. This indicates that later initiatives concentrated on sustainabili ty issues such as management and maintenance payments. It is interesting that follow up training appears to have no affect on technical competence, indicating that it addressed other issues (e.g. community mobili sation , payment mechanisms etc.). People tended to settle where there were more pump mechanics, where the last agency visit was more recent, and where the local committee and mechanics are regarded as competent. This indicates that technical competence and external support are important factors in attracting people to a water point. However, there is a strong correlation between people settling and year of installation such that people tended to settle due to pumps in later years, which indicates that insecurity factors may be involved.


It is interesting that certain entry issues (community requesting the borehole, and people being present at installation) have no affect on technical competence, yet others (prior contribution, and participation in installation) do show a link. Where boreholes were requested, agency pump maintenance visits tend to be more recent. Where a borehole is requested, this may be the first stage of a longer term and closer relationship with the community. This is confirmed by the large proportion of boreholes which were requested, that received follow-up training. It is only in cases where the committee repairs pumps that a significant proportion of prior contributions were paid, and people participated in installation. Prior contributions were only introduced in later installations (post 1995), and a greater proportion of these cases are repaired by committees. Therefore, this link more probably reflects changes in emphasis of project design in later years, with a greater degree of community focus – both prior contributions, training more mechanics, and committee activity. On-going community involvement also appears to have little affect on technical competence. Perceived ownership has no link with any indicator, and the only link with community participation in committees is that where the time since last agency visit is shorter, confirming greater agency involvement keeps the communities motivated to take part in water programmes. This supports the idea that local social mobili sation is more about the wider development goals than about pump functioning.

6.5.3 Organisation The organisation performance variables chosen can be divided into two principal subsets - those dealing with payment, and those dealing with organisational function. The payment subset involves: • frequency of payment schedule • actual payment • sanction for non-payment. The function subset: • existence of water committee • number of members • percentage of female membership, • community election • frequency of meetings • existences of rules for pump use. There are two main factors, which appear to have most influence on community organisation performance. The first is the perceived capabili ty of the local mechanics or committee to repair the hardware as well as software problems of the pump. Where this higher abili ty exists the people tend to pay more regularly and sanctions for non-payment are enforced. The implication for future exit strategies is the apparent importance of ensuring that training is given to someone within the community to manage the hardware as well as software issues, particularly in the case of deeper pumps.


The second primary influence on community organisational performance is the level of community involvement in the management. Community involvement is implicit in many of the other influential factors such as the committee election process and siting of pump. Involvement is also implicit in whether the community requested the pump or not. It is assumed that where a request was made the community has already mobili sed itself to address the water problem and delegated the task of identifying support, an indication of involvement and organisation from the outset. Age of installation has a strong influence on organisation indicators; again, this represents a shift in emphasis of later programmes to focus on community mobili sation issues. The data raises some rather controversial issues regarding community participation in the committee. Firstly, the involvement of women on committees appears to have no affect on any aspect of technical competence. The development community has developed the logic that if women are on the committee then responses to water problems will be quicker and more effective. Although the data would seem to dispute this, care must be taken to consider points raised in the literature review. Two main areas of concern are evident - token representation of women, with resistance to genuine power sharing from a range of mostly cultural beliefs, and the need for training to enable women to contribute effectively, and to address the prejudices of men. Further study would be required to investigate detailed reasons why the involvement of women has had little affect. Secondly, the fact that the election process and community involvement in water committees have almost no affect on issues of technical competence confirms that this area is independent of community participation. To a certain extent this is backed up by comments in the literature to the effect that participation issues are complex, and traditional power structures are well established. Training which took place before or at the time of pump installation appears to result in a more positive performance on the functioning of the committee. However, training, received sometime after the installation resulted in an improved performance on the payment subset of indicators (e.g. regular payments). This may be due to the later training placing greater emphasis on the development and management of a maintenance fund by a committee structure. In contrast, the earlier training's focus appears to have been on pump mechanics and pumphead maintenance. Turnover of caretakers. One agency interviewed reported a frequent turnover of caretakers. The most common cause being marriages or death. The type of pump affects organisational indicators. Communities with Volanta pumps make more regular payments and apply sanctions, yet their committees meet less often. This suggests that Volanta pumps are more expensive, and that the technically demanding nature of the pump requires external expertise to carry out repairs, so a lower level of participation exists. Dependency on the borehole as the only source of clean water encourages good organisation e.g. higher incidence of payments. An interesting finding is that factors such as prior contribution to the cost of the pump, participation in the process of pump installation and the percentage women on the committee still appear to have little influence on the organisational performance indicators. This finding seems to contradict the expectations of implementing agencies and the importance placed on these issues within current exit strategies. It indicates that good organisation can be achieved independently of entry conditions.


Long term financial management could not be measured. There will be a need to replace the whole pump and possibly re-drill the borehole should it collapse with age. Complete Willi ngness To Pay (WTP) should include replacement, however none of the systems were old enough to have warranted such an action, and therefore no data was generated on the full WTP of the communities. Participation in maintenance links to a range of organisation indicators e.g. where people take part in maintenance, committee meetings are held more frequently, it engenders payment of quotas and the application of sanctions and rules. People also tended to settle where the maintenance of boreholes was better organised e.g. regular payments, more sanctions and rules. However, as noted above (discussion of competence issues), there is a strong correlation between people settling and year of installation such that people tended to settle due to pumps in later years, which indicates that insecurity factors may be involved. When considering exit strategies. Follow-up training needs to ensure both local fund and hardware maintenance management capabili ty. Each appears to reinforce the other. However, the most important aspect is the encouragement of community participation in the pump management structure and process.

6.5.4 Participation o f the wider community In an attempt to look solely at participation of the wider community the following variables were investigated.

• Ownership of pump • Community involvement in water committee • Participated in installation • Participate in maintenance • Prior contribution (pre drilli ng/pump provision)

Date of installation is a significant issue and appears to highlight the difference between pumps provided with and without prior social mobili sation. It therefore draws attention to one of the initial assumptions of the research project, that where pumps have been installed rapidly during an emergency response there may be a level of dependency with minimal social commitment to maintaining the water point. It is also evident that the entry strategy changed significantly between the pre '95 period and those pumps installed later. These changes in entry strategy do appear to have led to greater community participation in the installation and care of the water-point. The noted influential changes in entry strategies are the insistence on a prior contribution to the pump purchase before drilli ng is initiated (only one agency) and the training of the whole committee. Training, particularly prior training and training which lifts the sense of local pump management competence, is possibly the most important stimulant to future participation. Prior training appears to act as a social mobili ser and correlates with the instances of


contribution to the initial cost of the pump. (It is recognised that only one agency has introduced the concept of prior contribution and that this was combined with initial committee training.) Issues of competence, such as who repairs the pump, and the perceived abili ty of local maintenance people, have an affect on participation indicators. Where demonstrated abili ty to deal with pump maintenance exists within a community (related to training mentioned above), participation will tend to higher e.g. greater community involvement in committees, participation in installation and on-going maintenance. Responding to the request of the community as opposed to pre-empting it, is also an important factor in insuring future local participation. Possible related to this are the issues of social cohesion and the involvement in the siting of the pump. Where members were present at the time of installation there was a far greater propensity for future participation in the maintenance of the water-point. It should be noted that one agency interviewed had difficulty to stimulate interest in Community Based Management before the community had seen the borehole. This project facili tated selection of the committee after the drilli ng and therefore the vill age headman did all the mobili sation during the construction phase. This, it is reported, tends to compromise community participation. Proximity to the pump does have an influence on participation. Either special effort has to be invested to make those living more removed from the water-point part of the process or it should be accepted that their level of involvement will be less. Within the five dependent participation indicators, the ownership and prior contribution variables could equally be considered as independent or causal variables. Note: An issue worthy of note is that prior contribution does help stimulate initial participation but not necessarily the ongoing care of the pump.

6.5.5 Section conclusion So we have seen that while the social factors may be a goal in themselves, and that they have links to localisation and ownership, they do not directly affect the reliabili ty of the pump. Competence is the main link and there seems to be no substitute for good technical training.

6.6 Outcome indicators for Health and Hygiene It has been pointed out that the indicator of care of the pumphead whilst not representative of the technical reliabili ty of pumps does coincide with it. Where the observed status of the pump is good, the boreholes tend to have lower frequency of breakdown and lower downtime. But good care of the pumphead may also represent better understanding of health issues and a propensity to better hygiene. Since all the communities have had hygiene training, and the research did not include a full KAP study of the communities, it


was not possible to test this supposition. However, all water supplies have been installed as part of integrated water supply and sanitation programmes, so data on this indicator may give some insights into these wider objectives. Where pumps have been installed in emergency situations (due to drought) the general condition of the pumphead tends to be poorer. There is no significant difference in the condition of boreholes installed pre 1995 and those installed from 1995 onwards, so it does appear to be the emergency conditions rather than age that influence care of the pumphead. This reflects an increased emphasis on community involvement in maintenance and care in development installations.

The data highlights that good care of pumpheads is evident in communities which have a high degree of internal capabili ty:

• Pumps tend to be better kept where repairs are carried out locally as opposed to external mechanics. However there is no difference in care between boreholes repaired by local mechanics as opposed to committees. This shows a link with general local capabili ty.

• Pumpheads also tend to be better kept where the abili ty of the committee or mechanic is held in high regard.

• Where people regard boreholes as being owned by their community, they tend to take better care of the pumphead.

All of these indicate that community mobili sation and training initiatives which focus on local capabili ty have made some impact on the care of the pump (keeping in mind our supposition that a cared for pump might represent some hygiene awareness).

In one evaluation document, one of the partner agencies states the following:- Story with a gap SG Finding #1 Tool not effective in identifying problems within vill age because it is too “ leading” . Comparison of “ good” and “ bad” picture makes answers obvious and all vill agers in each vill age immediately agreed that the following problems were in their vill age – unkempt borehole, unhygienic kitchens/homes, dirty surroundings and many households without pit latrine. In these statements it is clear that both the partner agencies and the communities see a link between the condition of the borehole and general hygiene awareness. Since the report states of an earlier exercise “exercise identifies knowledge but not practice of hygiene” , it seems a reasonable supposition that the actual state of the borehole at the time of visit is mildly indicative of the practice of hygiene awareness.


The variable “Rules” highlights a counter intuitive relationship with observed status. The application of rules and sanctions correspond to other indicators of community organisation and participation e.g locally elected committee, frequency of meetings and payments, community involvement, participation in maintenance. However, whether repairs are carried out locally / externally has no significant bearing on rules and the data indicates that the care of the pumphead is better where there are no rules. This may indicate that it is the local capabili ty of O&M that encourages care of the pumphead, rather than actual community mobili sation (ie participation factors). A good local mechanic or caretaker creates a well cared for pumphead, rather than peoples understanding and participation in the O&M of the pump.

Figure 10 Categor ies of Rules mentioned

This is supported to a certain extent by the awareness of the rules. Figure 10 summarises the rules mentioned by the respondents. As we see, general hygiene and payments are mentioned much more than caring for the actual wellhead. Another link that appears to contradict the first few points in this section is that the general status of pumpheads at Volanta pumps is higher. For Afridev pumps we find that :-

• more comprehensive training was given – higher level of training, longer training, health and sanitation training, and follow up training.

• more mechanics were trained; • committees were elected locally and met more frequently;

Closer inspection of the data reveals the influence of a number of issues. Volanta pumps:

• tended to be put in earlier and at the request of the community • people tend to live close to the pump, and have settled because of the pump; • quota payments are higher; • agency visits are more recent. • queuing times were shorter.

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

50%

hyg

en

e

pa

yme

nt

so

cia

l ord

er

limite

da

cce

ss

care

of

we

llhe

ad


This raises the importance of issues of dependency on water supply and pressure on the pump. Volanta installations have a high dependency on the borehole expressed by a higher WTP, they were requested and settlement. Also, people living closer to the pump combined with less pressure (shorter queuing times) make it easier to achieve good care of the pumphead. In themselves these do not provide clear indicators as to how hygiene awareness is increasing within these communities and it may be that the more recent agency visits also puts pressure on communities to take care of the pumphead.

6.7 Programme History - The Influence of an Emergency Context The research project is particularly interested in the impact of entry strategies on long term sustainabili ty and whether the shortfalls of a non participatory entry (caused by an emergency) can be made up by an exit strategy. The issue of whether the pump was installed in response to an emergency significantly influences the frequency of breakage, the downtime when broken and the observed status of the pumphead. In all three cases the performance against these three indicators is more negative in the case of emergency installations. The pumps break more frequently, they take longer to mend and the pump head is less adequately cared for. Most emergency installations took place prior to 1995, but the fact that these issues are not reflected to the same degree when considering the date of installation suggests that there are other factors at play that are peculiar to the emergency response process. This is not surprising, as the whole point of a development approach is to ensure greater overall sustainabili ty for the water supply. However, when we investigate the situation in more detail we can see the role of the “entry” and “exit” strategies. By considering the cases of emergency / “development” installations as a whole, the data shows there are significant differences in entry strategy issues in terms of “participated in installation” , and “prior contribution” . Both confirm that there was less community mobili sation in emergency installations. When looking at indicators of social factors, the following trends are evident in emergency installations: Community organisation:

• appointment of water committee members tend to be influenced by external agents, • committee meetings are held less frequently;

Competence: • pumps tend to be repaired by external representatives, • there are fewer pump mechanics, • the committee or local mechanics are perceived as less able;

Participation: • people perceive the pump as being owned externally, • lower community involvement in water committees.

This confirms that social participation, organisation and training, which tend to be more prevalent in the non-emergency pump installation process, do indeed have a positive impact on social factors within a community.


“Development” pumps have the “advantage” of entry strategy (participatory with prior contribution), follow up (many of the development committees had follow up training) and availabili ty of spares (no one at a development pump site complained about lack of spares being available locally). Emergency pumps have a mixture of some training at installation, a mix of entry involvements (people helping with siting, or helping form a road for the drilli ng rig), various levels of follow up and in some cases lack of available local spares. In an attempt to investigate this complex relationship of variables, the data was sub divided. Since the Volanta pumps have particular diff iculties in terms of reliabili ty and spares, these were removed from the sample. MW tests of frequency of breakage, downtime and ownership were then run on the emergency holes only across all the variables that represent the design or strategy of the programme. . The five variables that show as statistically significant are siting of the pump, present at installation, spares availabili ty, where spares are accessed and number of women on the committee. This would confirm that even the simplest of involvement during installation makes a long term difference, that the spares issue is important and influential, and that the involvement of women does affect ownership and reliabili ty. However, the data at this second level analysis is quite weak and statistical significance must be treated with caution. It is more helpful to look at the trends of the data as represented by the Mean Rank. While this must be handled qualitatively rather than statistically, it nevertheless confirms a number of important elements of the exit strategy. Table 5 summarises the results

Table 5 Independent variable (on emergency ho les wi th Afridevs on ly)

Downtime Frequency of breakage

Ownership

Last agency pump maintenance visit - - = Election of members - - + Follow-up training - + + Sanitation and health training = = = Length of training + + = Level of training given + + + Number of members + - + Number of trained pump mechanics + + + Participated in installation + + + Present at drill/installation * + + + Siting of well * + + + Number of women on committee * + + + Spares available * + = = Where are spares acquired * + - = Ability of local committee or trained pump maintenance rep's

+ - +

Water committee appointed = - = Community involvement in water committee + - + Key + More of the var iable, or localisation of the var iable, results in a “better” outcome, eg less downtime, less breakage, or more ownership. - More of the var iable, or locali sation of the var iable, results in a “worse” outcome, eg more downtime, more breakage, or less ownership. = No difference in the samples. * statistical significance (but with borderline sample number – treated with caution)


Table 5 confirms two themes that have emerged from the wider analysis. First, that technical training is important. Competence will be formed by technical knowledge, which in turn is influenced by the length of training, the level of training and the number of trained mechanics. Second, that the more local mobili sation there is the greater the ownership and lower the downtime, but the higher the frequency of repair. This higher frequency seems to be due to local people having a lower level of technical competence and whilst local community mobili sation lowers downtime, it can increase frequency of breakage. The two variables that are counter intuitive are “ last agency visit” and “follow-up training” . The more frequent the last agency visit the longer the downtime and the higher the frequency of breakage. This may indicate that communities have come to rely on the agency, or it may mean that the agency is visiting frequently because this is a troublesome pump (or community). Similarly the more there is follow-up training the longer the downtime. The follow-up training given often emphasised the social mobili sation, the formation of committees and how to raise finance. One would expect that this would result in shorter downtime and more frequent breakages (given the rest of the analysis). In fact it results in longer downtime and less frequency of breakages. This again may be because follow-up training is given where the pump is maintained by area mechanics and the training has yet to take effect. In general the analysis highlights what would be expected – that technical training affects competence which in turn affects pump reliabili ty, and that social mobili sation results in an increasing level of ownership and local wider community participation.

6.8 Programme Design - Can Training make up for a weak entry strategy?

Given that the length of training, level of training and number of trained mechanics have been highlighted as important elements of an exit strategy, the data was used to investigate whether these elements could compensate for a weak entry strategy. The data was compared using the following:-

• cases of emergency boreholes, with the highlighted element from the relevant variable

• cases of emergency boreholes, without the highlighted element from the relevant variable

• cases labelled “development” (i.e. with a participatory entry). I.e. comparing those emergency installations that have longer training with those that have shorter training, and comparing both with “development” pumps. This approach confirms the following:- Length of training – frequency of breakage and downtime are both reduced with longer training. More importantly; “emergency pumps” which have had longer training tend to have less frequent breakdowns and shorter downtimes than “development” pumps. Ownership was increased with the longer training for emergency pumps and the


ownership levels were almost the same for emergency pumps with more than 1 week training and “development” pumps. Level of training – the frequency of breakdown of emergency boreholes tends to be reduced with more technical training, although the downtime tends to be slightly shorter when sanitation and simple maintenance is emphasised. There was a trend to greater ownership of the pump with increased level of training which matched the “development” pump’s expression of ownership. Number of mechanics – emergency pumps with more than 2 trained mechanics tend to have lower frequency of breakage and shorter downtimes. They do not as such match the “development” pumps who have even shorter downtimes and less breakages, but nevertheless this confirms the importance of the number of mechanics. The number of mechanics has no impact on levels of ownership of emergency pumps, which is as expected. To summarise:- there are confirmed benefits to the availabili ty of the pump from making trainings more than 1 week, including both soft (sanitation, health, community management and simple pump maintenance) and hard (complete pump maintenance and second level pump repair) subjects, and training as many mechanics or caretakers as possible. The expressions of ownership are also enhanced by the longer training and increased levels of training. There are clear indications that there is no substitute for early involvement and that even simple involvement during a pump installation makes a long term difference to the sustainabili ty of a borehole. However, while an emergency pump with weak entry strategy can probably never fully match the sustainabili ty of the “development” pump, ensuring competence through technical training (over a suitable period of time and with enough people) can increase emergency pump sustainabili ty.


7 Exit strategies – a discussion In the results and analysis a clear distinction was made between the pump function and social mobili sation. Both are desirable. However, in the literature, the writers often make the connection that good social mobili sation results in good pump function. This view is not supported by the data from this research project. In the Figure 9 Principal Relationships highlighted by Case Study Data” we have attempted to ill ustrate the relationships between the constructs used in this project. We will use this diagram to guide the discussion on exit strategies. The following discussion assumes the following scenario:- An emergency is drawing to an end. A number of new or rehabili tated water supplies have been installed with emergency funding from various donors. If the external agencies withdraw instantly the likelihood is that these water supplies will not be sustained by the local community and the local authorities do not have the capacity (or policy) to manage all the holes from a central location. An agency (either local government, NGO or multilateral) is considering spending some finance on either building government capacity or local community capacity to sustain the water supplies. The external agency will spend a limited time in a “planned exit strategy” . The end goal is that the existing emergency water supplies will be as sustainable as other supplies in that country that might have been put in by a developmental or planned approach.

7.1 Is ex it actually possible? Before discussing the above scenario, where an agency is attempting to exit, we must have a small diversion into the question as to whether exiting is actually possible and desirable. There are four points to be considered here.

1) At the heart of a good water, sanitation and health strategy is the desire to create sustainable water supply. However, the literature survey shows how the current emphasis is on an integrated strategy inclusive of sanitation and health education. Changes in knowledge and Practice of Health and hygiene concerns are unlikely to occur with only the input of a two year programme. An evaluation of one of the partner agencies notes that although the agency completed all the target activities, outputs and intermediate goal (actually exceeding on some points eg number of pumps and latrine slabs.) there was actually no change from baseline to evaluation of the Goal which concerned diarrhoea in both under and over 5 yr old and conjunctivitis. With another agency, there was some improvement in Bilharzias and malaria (due to not travelli ng to stagnant water sources). However the general principle seems to be that changes in health require long term health education and cannot be fit into a two year programme. In Malawi the Ministry of health have agents (HSA) that regularly visit vill ages, and they have taken the water supply under their wing as well.

2) Each of the partners have been working with water supplies in the country for over 7 years. Each of them has done this in 2 or 3 year projects, and each effectively


promised to exit at the end of each project. Consider the impact of designing a 7 year programme which had a more realistic single timetable.

3) There is a strong need, acknowledged by the literature and data, for technical support on the second level repairs. The communities are unlikely to ever be able to undertake the very serious repairs (such as collapsed borehole). Therefore there must be some agency present (even if private supplier or local government authority) that has that extra technical capabili ty.

4) The literature review clearly shows that “VLOM” is a combination of vill age level resources with local area resources (either local government or private suppliers). Complex second level repairs will probably need an agency external to the vill age. The local authority may be present as an implementing agent or as a regulator. In either case it is not true to imply that the vill age is left purely to its own resources to operate and maintain the pump.

5) Motivation within any social grouping will change and more than likely diminish over time. Again literature and data suggest that there must be “ongoing consultation” in order to keep the community motivated to pay. In theory this could be undertaken by the vill age water committee. However there is also the element that participation wanes after a while, and it may require very small but important external support to maintain participation and motivation.

The above seems to imply that the word exit can be misleading. It is unlikely that all external influences on a community will l eave at the end of, say, a two year water supply programme. The local authority will continue with health education within their capacity, and it may be that other external agencies will spend a number of years in the area undertaking other development initiatives. Training of children in schools on health education has a particular impact on the long term response of the community. The above also suggests that some small localised external input is necessary to maintain participation and motivation, and to provide technical support for the more severe water supply problems. One might use the term “handover” rather than exit to cover “handover” to communities and “handover” to local authroties and agencies. However, the word “exit” has been used to communicate the need for international agencies to look to local sustainabili ty rather than keep a long term presence in an area.

7.2 Essential elements of an exit strategy Given the above caveat, let us now consider the possible issues and strategies for an external agency to consider so that it leaves a water supply in such a way as to be (as near as possible) “sustainable”.

7.2.1 Participation Exit issue:- A good entry strategy is helpful. Given our scenario one is presented with a “fait accompli” . Even the simplest of community involvement at installation will enhance long term ownership. Exit strategy:- If at all possible alter the entry strategy to involve the people in such basics as making the road to the borehole site! if this has not been possible then use follow up training and visits to dialogue with people on ownership and social mobili sation


issues. Explain the project process as much as possible so the community understand the timing of any project (when exit will occur). Exit issue:- Participatory dialogue with the wider community is useful for long term development goals of social mobili sation but may have limited impact on pump functionality. Exit strategy:- promote dialogue on ownership, participation and organisation capabili ty for long term goals (within mobili sation/training budget). However, where resources are scarce it is more important to have local technical competence to ensure pump availabili ty. Exit issue:- the vill age social power structures affect the organisation and management of the community response to the pump and will change over time. Exit strategy:- promote dialogue on ownership, participation of the wider community and organisation. In conflict (post conflict) situations this must be undertaken with due care to the sensibili ties of the conflicting parties. Methods of election to the management of a water supply should avoid appointment by powerful groups or individuals or being seen as appointments of the agency. Exit issue:- participation will drop off over time, and motivation of committees will need to be sustained by incentives. Exit strategy:- Water committees will respond to small inputs of ongoing dialogue and interest with external agencies. However it is important that these ideas are generated by the community and not imposed from outside. Exit issue:- involvement of women may be culturally constrained. For wider goals of development and for longer term impact on water supplies it is important to involve women as much as possible. Exit strategy:- when encouraging the formation of the water committee, women should be included. Training should try not to reinforce traditional roles, eg only soft skill s of health and hygiene for the women, technical skill s for men. Exit issue:- local participation of the wider community does not directly shorten downtime or improve maintenance. Exit strategy:- Local participation of the wider community can strengthen local organisation (committees), and this in turn can shorten downtime if there is enough local competence and spares are available. Build in participation as part of a general capacity building strategy.

7.2.2 Organisation Exit issue:- “Government departments generally have a technical focus and are not equipped to deal with socio-economic aspects of development” Exit strategy:- “Agencies need to mobili se public and private partnerships to make optimum use of stakeholders” . Make strategic networks with other agencies enlisting those with a socio-economic understanding. (EG HSA workers in Malawi.) Exit issue:- there is a need for some technical support from a local agency on severe second level technical problems.


Exit strategy:- identify local area resources that can address these more complex technical repairs to broken pumps. This may mean capacity building of local government, or training private suppliers. Exit issue:- there is a need for some ongoing support from a local agency to encourage community management and motivation initiatives. Exit strategy:- Although an NGO may plan its exit from the arena, it should build the socio-economic understanding in local agencies and authorities, so that they may give small but strategic ongoing support to the communities. Exit issue:- Technology must be appropriate to abili ty and willi ngness to pay Exit strategy:- Vill age profili ng will give some indication of both abili ty and willi ngness to pay. If technology not appropriate, consider replacing technology (see below) Dialogue and training can affect WTP.

Exit issue:- the vill age social power structures affect the organisation and management of the community management of the pump and will change over time. Exit strategy:- promote dialogue on ownership, participation and organisation. Methods of election to the management of a water supply should avoid appointment by powerful groups or individuals or being seen as appointments of the agency.

Exit issue:- there was no direct significant relationship between cost recovery and sense of ownership. Exit strategy:- the message that the water point is the communities needs to be continually reinforced both by the community committees and by external agents. However, NGOs should note that strong leadership can collect the necessary finance and that poor wider community participation and ownership does not in itself imply poor VLOM. Exit issue:- raising of funds is necessary for maintenance and repair of the borehole. However so many other factors, such as alternative sources and pressure on pump, affect the flow of finance, that one cannot clearly say that regular collection of a quota actually shortens downtime. Exit strategy:- Training should include discussion on the various options of financing repairs but the mechanisms for collecting money and storing it are best left to the communities to decide. Exit issue:- Raising of funds is necessary for replacement of the system in the longer term future. The data did not yield any insights to this challenge. Exit strategy:- Replacement needs to be discussed with the community. Strategic connections between the community and the financial service sector may need to be made, this will vary based on the country context.

7.2.3 Training (Education)

Exit issue:- Neither participation of the wider community nor organisation guarantees good pump functioning. While it may be desirable for integrated water, health and sanitation programmes and longer term development, it is not critical to good pump function. Competence is more important.


Exit strategy:- If time and resources are limited it is better to focus trainings on the technical skill required to maintain and repair pumps rather than organisation and wider community participation skill s – UNLESS the programme is a part of a longer term programme on health and sanitation. Exit issue:- There is a higher frequency of breakage when local communities repair the borehole than when an external person or group does (area mechanic or local government). i.e. the external person has a greater competence, possibly resulting from longer training and more experience. Exit strategy:- The key element is competence. Local community based mechanics need as much technical training as possible. Follow up trainings should always have technical components alongside any social elements. Longer trainings were more effective – exit trainings should be as long as necessary to cover clearly all the subjects. Exit issues:- It is clear that pumps with women mechanics or caretakers, function as well as equivalent pumps maintained by men Exit strategy:- Train women in both soft and hard skill s. Women can make use of transferable practical skill s such as pump maintenance and repair. Exit issue:- Shorter downtimes are found when many local people are trained. Exit strategy:- It is clear from the research that a good exit strategy trains as many people as possible (minimum more than 2). The mechanics can be men or women. An emphasis on people from within the user community will shorten downtime.

7.2.4 Context Exit issue:- health and hygiene training is best applied when in the context of community mobili sation for pump O&M (and repair). It does not seem possible to legislate hygiene – it is based on relationships and awareness. Exit strategy:- Training of local committees should include technical, organisational, financial and health areas. Trainings should encourage people to problem solve on health rather than legislate (ie rules and sanctions do not produce as good a result as ongoing dialogue between community agents and their own community). Exit issue:- Some of the strongest factors influencing pump function are the depth and technology. If emergency supplies for resettled populations are to be addressed then an agency is generally presented “fait accompli” with depth and technology. Exit strategy: Consider changing the technology. One agency in the study used Afridev pumps beyond their recommended depths because even though they would be subject to a higher frequency of breakdown the net result was more sustainable than the alternative Volanta. Exit issue:- There must be a supply of spares which can be accessed by the communities. Private local suppliers seem to be a reasonable option although there are considerable problems with the profitabili ty of spares. Low turnover and low profit margins combine to make stocking pump spares an unattractive proposition from a business point of view. Exit strategy:-. It is important to either find retailers who are willi ng to do it out of a sense of duty to their community, or to allow retailers nearer a 100% handling charge. If


such a handling charge is allowed, it must not be undermined by agencies or governments supplying spares at a subsidised rate. Exit issue:- Those communities with access to alternative sources, both boreholes and unreliable surface water, will be slower in repairing their pumps than those without. Exit strategy:- The response of the community may be influenced over the long term by a programme of health and hygiene awareness.

7.2.5 Optimal use of water Exit issue:- Those pumps with salty water which is only used for washing and not regularly used for drinking, will have longer downtimes. Exit strategy:- make these holes low priority. Do not spend time and effort on these holes, there are other holes where the same effort will produce a better result regarding sustainabili ty.

7.2.6 Sett lement Exit issue:- Settlement will be affected by the borehole and in particular the confidence the people have in the pump being repaired quickly. Exit strategy:- If the agency wants to encourage settlement in a particular location, or set of vill ages, then it must promote competence, and make sure it is seen to do so.

7.2.7 Environmental Sustainabili ty This section has not commented on environmental sustainabili ty of the pump. Data was not available that could determine the key issues. Environmental sustainabili ty must be considered when planning an exit strategy. This will i nclude issues such as:-

• Lowering of water tables due to high extraction rates from boreholes, can causes local hand dug wells to dry up.

• Contamination of water sources - from human or animal wastes, or run-off of agro-chemicals.

• Spread of disease - from inadequate water purification and inadequate care and maintenance of water supply systems.

• A side effect of the provision of a reliable water source where none existed before is that it will attract pastoralists and farmers, resulting in over use of land and soil erosion.

Readers are referred to The Manual of Environmental Appraisal (ODA 1996b)

7.3 Fund ing o f exit strategies

7.3.1 Who fund s the exit? Exit Issue:- One of the issues surrounding this project concerns – who funds the post emergency or resettlement exit. Emergencies seem to be generally handled by a specific department within an agency. Funding is often very time bound and post emergency resettlement and rehabili tation is from another tranche of funds or is handled by another department within the agency. Agencies (or department within agencies) that specialise in development may then be reluctant to pick up on “emergency” assets, citing that the


entry strategy (participatory planning) is the most critical feature of the development cycle, and that it would be better to fund new projects rather than “rescue” water supplies that were put in with “poor entry strategies” . One of the conclusions of this research is that a good exit strategy can make up for a poor (non participatory) entry strategy. The challenge remains though – who will fund these exit actions in order to leave “emergency water supplies for resettlement populations” in a sustainable condition. Exit Strategy:- The case studies suggest that both funding and implementing agencies working with water supplies and/or having integrated programmes on water, sanitation and health, should not abandon water points that were the outcome of an emergency or resettlement programme. They should work with the local authorities and local communities on a planned exit strategy (see guidelines section above) to turn the water points into sustainable supplies owned and managed by the local population.

7.3.2 How much to spend on an exit strategy? Exit Issue:- Having suggested that agencies seriously consider investing extra resources to turn “emergency” water supplies into sustainable supplies, we must ask the question regarding cost effectiveness. The direct costs per borehole for the agencies in this research project vary considerably, from a few thousand to approximately ten thousand dollars. The support costs are more consistent approximately the same amount again. Where one agency is implementing a development strategy of participatory entry including the communities participating in the purchase of the initial equipment, the cost per supply including all post installation trainings is approximately $12,000 (Evaluation 1997). Exit Strategy:- Exit strategies need not cost a lot of money although the case studies seem to indicate that to stimulate effective ownership of a water supply and to put in place the technical competency required to VLOM and repair that borehole, costs about the same amount as the original hardware of the water supply. However a planned exit strategy that utili ses the minimum guidelines would cost about half the original cost of the borehole (assuming normal borehole costs, not low cost techniques). Exit Strategy:- Emergency water supplies are often put in by agencies with considerable sums of money at their immediate disposal (funds raised to save human life in response to the emergency). Development agencies often have a lot less money for disposal. Emergency agencies could put aside an amount equal to the original cost of putting in the hardware for a longer term planned exit strategy. Exit Issue:- All the agencies commented on how they have been working for many years on water supplies, but have always had short contracts of 1, 2 or rarely 3 years. With longer term contracts a more strategic plan could be made with phased exit from communities and handover to local authorities or private suppliers. Exit strategy:- Donors could consider the longer term goals of an emergency programme, and fund external agencies accordingly. They also need to be flexible to achieve stated goals, to allow for change in circumstances and time scales during implementation.


7.3.3 Supp ly of Spares Exit issue:- We have said above that the supply chain is important and that there must be a supply of spares. The case studies indicate that spares through private local suppliers can be effective. The research project has generated only qualitative data regarding the supply chain pre local supplier. There are issues of initial purchase of the pump. If the pump can be purchased by an in-country manufacturer then the supply of spares will be more likely to be ongoing and sustainable. However in many emergency situations, local manufacturers may not have enough stock to supply the programme, the local manufacturers may not be able to manufacture new pumps quickly enough (or if they can it is at the cost of quality), and emergency programmes (like all large programmes) may have rules about tendering which can favour lower cost external suppliers. Since the cost of materials in Asia is considerably cheaper than Africa, pumps imported from Asia often have a price advantage despite the transport costs. The original supplier of pumps ideally should have an arrangement with a network of local retailers to supply spares. It should be noted that this is considered a major issue in Mozambique. Work with the private sector is crucial as Government policy is not to provide spares. The implementing agency should also not provide spares as this is not sustainable in the longer term. The private sector therefore seems to be the way forward for Mozambique. Exit Strategy:- Given our scenario that an emergency programme has been undertaken and the agency is now considering how it may exit leaving sustainable supplies, we must assume that a contractual arrangement with the supplier to have the local network of retailers is not in place. What then can an agency do? The agency should work with the government to source a supply of spares and capacity build to establish the relevant network. This action may be being undertaken by other agencies, and these macro issues must be tackled by a co-ordinated effort, not a single agency response. We have said above that private suppliers should be given the opportunity to make a profit by placing a reasonable handling charge on small i tems, and not being undermined by free gifts from local government or agencies. This supply of spares should not need to be subsidised at any stage of the long term process either as an incentive to the suppliers or to the distributors. The research indicates that communities can raise finance for spares if they are committed to the water supply. We have also noted that private suppliers could also undertake the more complex repairs and become “area mechanics” with good technical competence providing ongoing technical support to the community. Training by the agency can be offered during the exit process perhaps for a nominal training fee to ensure commitment by the new businesses. While the above encourages a purely private enterprise response to the supply of spares based on economics, it must be noted that the existing supply chain is strongly influenced by “good intention” by the local retailers. They see their action of supplying spares not so much as a profitable enterprise but as a service to the community. This goodwill i s commendable. However it may or may not be sustainable. It should also be noted that some communities may prove to be very technically competent. They may then provide their skill s for payment to other nearby communities, effectively becoming private enterprise. Examples of this were found in the research case studies.

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

84

Tab

le 6

Sum

mar

y of

sug

gest

ed e

xit

stra

tegy

com

pone

nts.

Par

tici

pati

on

D

esig

n

reco

mm

enda

tion

s C

omm

unit

y pa

rtic

ipat

ion

at o

r be

fore

inst

alla

tion

Exi

t is

sue:

- A

goo

d en

try

stra

tegy

is

help

ful.

Giv

en o

ur s

cena

rio

one

is

pres

ente

d w

ith

a “

fait

acco

mpl

i”.

Eve

n th

e si

mpl

est o

f co

mm

unit

y in

volv

emen

t at

inst

alla

tion

will

en

hanc

e lo

ng te

rm

owne

rshi

p.

If a

t all

poss

ible

alt

er th

e en

try

stra

tegy

to in

volv

e th

e pe

ople

in

such

bas

ics

as m

akin

g th

e ro

ad to

the

bore

hole

site

!

Exi

t st

rate

gy:-

if t

his

has

not b

een

pos

sible

then

use

foll

ow u

p tr

aini

ng a

nd v

isit

s to

dia

logu

e w

ith

peo

ple

on o

wne

rshi

p an

d so

cial

mob

ilisa

tion

issu

es.

Pr

omot

es o

wne

rshi

p E

xit

issu

e:-

Part

icip

ator

y di

alog

ue w

ith

the

wid

er c

omm

unit

y is

use

ful f

or lo

ng te

rm

deve

lopm

ent g

oals

of

soci

al m

obili

sati

on

but

may

hav

e lim

ited

impa

ct o

n p

ump

func

tion

alit

y.

Exi

t st

rate

gy:-

pro

mot

e di

alog

ue o

n o

wne

rshi

p, p

arti

cipa

tion

an

d or

gani

sati

on c

apab

ility

for

long

term

goa

ls (w

ithi

n

mob

ilisa

tion

/tra

inin

g bu

dget

). H

owev

er, w

here

reso

urce

s ar

e sc

arce

it i

s m

ore

impo

rtan

t to

hav

e lo

cal t

ech

nica

l com

pete

nce

to e

nsur

e pu

mp

avai

labi

lity.

R

ole

of w

omen

Se

e be

low

Des

ign

C

onst

rain

ts

Abi

lity

and

trai

ning

C

onfli

ct in

co

mm

unit

y E

xit

issu

e:-

the

vill

age

soci

al p

ower

st

ruct

ures

aff

ect

the

orga

nisa

tion

and

m

anag

emen

t of

the

com

mun

ity

resp

onse

to

the

pum

p an

d w

ill c

hang

e ov

er ti

me.

Exi

t st

rate

gy:-

pro

mot

e di

alog

ue o

n o

wne

rshi

p, p

arti

cipa

tion

of

the

wid

er c

omm

unit

y an

d or

gani

sati

on.

In c

onfli

ct (

post

co

nfli

ct)

situ

atio

ns th

is m

ust b

e un

dert

aken

wit

h d

ue c

are

to th

e se

nsib

iliti

es o

f th

e co

nfli

ctin

g pa

rtie

s. M

etho

ds o

f el

ecti

on to

th

e m

anag

emen

t of

a w

ater

sup

ply

shou

ld a

void

app

oint

men

t by

pow

erfu

l gro

ups

or in

divi

dual

s or

bei

ng s

een

as

appo

intm

ents

of

the

agen

cy.

Sust

aina

bilit

y of

pa

rtic

ipat

ion

Exi

t is

sue:

- pa

rtic

ipat

ion

will

dro

p of

f ov

er

time,

and

mot

ivat

ion

of

com

mitt

ees

will

ne

ed to

be

sust

aine

d by

ince

ntiv

es.

Exi

t st

rate

gy:-

Wat

er c

omm

ittee

s w

ill r

espo

nd to

sm

all i

nput

s of

ong

oing

dia

logu

e an

d in

tere

st w

ith

ext

erna

l age

ncie

s.

How

ever

it i

s im

port

ant

that

thes

e id

eas

are

gene

rate

d by

the

com

mun

ity

and

not

impo

sed

from

out

side

.

Gen

uine

con

trib

utio

n

by w

omen

, Cul

tura

l st

andi

ng o

f w

omen

, H

eavy

wor

kloa

ds o

f w

omen

Exi

t is

sue:

- in

volv

emen

t of

wom

en m

ay b

e cu

ltur

ally

con

stra

ined

. F

or w

ider

goa

ls o

f de

velo

pmen

t and

for

long

er te

rm im

pact

on

w

ater

sup

plie

s it

is im

port

ant

to in

volv

e w

omen

as

muc

h a

s po

ssib

le.

Exi

t st

rate

gy:-

whe

n e

ncou

ragi

ng th

e fo

rmat

ion

of

the

wat

er

com

mitt

ee, w

omen

sho

uld

be in

clud

ed.

Tra

inin

g sh

ould

try

not

to r

einf

orce

trad

ition

al r

oles

, eg

NO

T o

nly

soft

ski

lls

of h

ealt

h

and

hygi

ene

for

the

wom

en a

nd te

chni

cal s

kill

s fo

r m

en.

W

omen

sho

w s

imil

ar c

ompe

tenc

e on

tech

nica

l ski

lls

if tr

aine

d.

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

85

Pa

rtic

ipat

ion

in

mai

nten

ance

E

xit

issu

e:-

loca

l par

tici

pati

on o

f th

e w

ider

co

mm

unit

y do

es n

ot d

irec

tly

shor

ten

do

wnt

ime

or im

prov

e m

aint

enan

ce.

Exi

t st

rate

gy:-

Loc

al p

arti

cipa

tion

of

the

wid

er c

omm

unit

y ca

n

stre

ngth

en lo

cal o

rgan

isat

ion

(com

mitt

ees)

, and

this

in tu

rn c

an

shor

ten

dow

ntim

e if

ther

e is

eno

ugh

loca

l com

pete

nce

and

spar

es a

re a

vail

able

. B

uild

in p

arti

cipa

tion

as

part

of

a g

ener

al

capa

city

bui

ldin

g st

rate

gy.

O

utco

mes

A

cces

s to

loca

l kn

owle

dge

and

skill

s

C

apac

ity

buil

ding

O

rgan

isat

ion

D

esig

n

reco

mm

enda

tion

s R

ole

of e

xter

nal

agen

cies

E

xit

issu

e:-

“Gov

ernm

ent d

epar

tmen

ts

gene

rall

y ha

ve a

tech

nica

l foc

us a

nd a

re n

ot

equi

pped

to d

eal w

ith

soc

io-e

cono

mic

as

pect

s of

dev

elop

men

t”

Exi

t st

rate

gy:-

“A

genc

ies

need

to m

obili

se p

ubli

c an

d pr

ivat

e pa

rtne

rshi

ps to

mak

e op

timum

use

of

stak

ehol

ders

”. M

ake

stra

tegi

c ne

twor

ks w

ith

oth

er a

genc

ies

enli

stin

g th

ose

wit

h a

so

cio-

econ

omic

und

erst

andi

ng.

(EG

hea

lth

wor

kers

in M

alaw

i.)

O

n-go

ing

supp

ort

Exi

t is

sue:

- th

ere

is a

nee

d fo

r so

me

tech

nica

l sup

port

fro

m a

loca

l age

ncy

on

seco

nd le

vel t

ech

nica

l pro

blem

s.

Exi

t st

rate

gy:-

iden

tify

loca

l are

a re

sour

ces

that

can

add

ress

th

ese

mor

e co

mpl

ex te

chni

cal r

epai

rs to

bro

ken

pum

ps.

Thi

s m

ay m

ean

cap

acit

y bu

ildi

ng o

f lo

cal g

over

nmen

t, o

r tr

aini

ng

priv

ate

supp

lier

s.

E

xter

nal s

uppo

rt

(gov

ernm

ent)

E

xit

issu

e:-

ther

e is

a n

eed

for

som

e on

goin

g su

ppor

t fro

m a

loca

l age

ncy

to

enco

urag

e co

mm

unit

y m

anag

emen

t and

m

otiv

atio

n in

itiat

ives

.

Exi

t st

rate

gy:-

Alt

houg

h a

n N

GO

may

pla

n it

s ex

it fr

om th

e ar

ena,

it s

houl

d bu

ild

the

soci

o-ec

onom

ic u

nder

stan

ding

in lo

cal

agen

cies

and

aut

hori

ties

, so

that

they

may

giv

e sm

all

but

stra

tegi

c on

goin

g su

ppor

t to

the

com

mun

ity.

Paym

ent o

f m

embe

rs

No

find

ings

from

this

stud

y

Des

ign

C

onst

rain

ts

Loc

al c

omm

unit

y ch

arac

teri

stic

s E

xit

issu

e:-

Tec

hno

logy

mus

t be

appr

opri

ate

to a

bilit

y an

d w

illin

gnes

s to

pay

Exi

t st

rate

gy:-

Vill

age

prof

iling

will

giv

e so

me

indi

cati

on o

f bo

th a

bilit

y an

d w

illin

gnes

s to

pay.

If

tech

nolo

gy n

ot

appr

opri

ate,

con

side

r re

plac

ing

tech

nolo

gy (

see b

elow

) D

ialo

gue

and

trai

ning

can

aff

ect W

TP.

Mot

ivat

ion

of

mem

bers

N

o fi

ndin

gs fr

om th

is st

udy

L

ocal

pow

er

stru

ctur

es

Exi

t is

sue:

- th

e vi

llag

e so

cial

pow

er

stru

ctur

es a

ffec

t th

e or

gani

sati

on a

nd

man

agem

ent o

f th

e co

mm

unit

y re

spon

se to

th

e pu

mp

and

will

cha

nge

over

tim

e.

Exi

t st

rate

gy:-

pro

mot

e di

alog

ue o

n o

wne

rshi

p, p

arti

cipa

tion

an

d or

gani

sati

on.

In c

onfli

ct (

post

con

flict

) si

tuat

ions

this

mus

t be

und

erta

ken

wit

h d

ue ca

re to

the

sens

ibili

ties

of t

he co

nfli

ctin

g pa

rtie

s. M

etho

ds o

f el

ecti

on to

the

man

agem

ent o

f a

wat

er

supp

ly s

houl

d av

oid

appo

intm

ent b

y po

wer

ful g

roup

s or

in

divi

dual

s.

Fr

agm

ente

d co

mm

uniti

es

See

abov

e “L

ocal

pow

er s

truc

ture

s”

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

86

M

etho

d of

ele

ctio

n

(wom

en)

See

abov

e “C

onfli

ct in

com

mun

ity”

In

still

ing

owne

rshi

p E

xit

issu

e:-

ther

e w

as n

o di

rect

sig

nifi

cant

re

lati

onsh

ip b

etw

een

cos

t rec

over

y an

d se

nse

of o

wne

rshi

p.

Exi

t st

rate

gy:-

the

mes

sage

that

the

wat

er p

oint

is th

e co

mm

uniti

es n

eeds

to b

e co

ntin

uall

y re

info

rced

bot

h b

y th

e co

mm

unit

y co

mm

ittee

s an

d by

ext

erna

l age

nts.

How

ever

, N

GO

s sh

ould

not

e th

at s

tron

g le

ader

ship

can

col

lect

the

nece

ssar

y fi

nanc

e an

d th

at p

oor

wid

er c

omm

unit

y pa

rtic

ipat

ion

an

d ow

ners

hip

does

not

in it

self

impl

y po

or V

LO

M.

Cos

t rec

over

y Se

e be

low

“C

olle

ct m

oney

reg

ular

ly o

r at

br

eakd

own”

C

olle

ct m

oney

re

gula

rly

or a

t br

eakd

own?

Exi

t is

sue:

- r

aisi

ng o

f fun

ds is

nec

essa

ry

for

mai

nten

ance

and

repa

ir o

f th

e bo

reho

le.

H

owev

er s

o m

any

othe

r fa

ctor

s, s

uch

as

alte

rnat

ive

sour

ces

and

pres

sure

on

pum

p,

affe

ct th

e fl

ow o

f fin

ance

, tha

t one

can

not

clea

rly

say

that

reg

ular

col

lect

ion

of

a q

uota

ac

tual

ly s

hort

ens

dow

ntim

e.

Exi

t st

rate

gy:-

Tra

inin

g sh

ould

incl

ude

disc

ussio

n o

n th

e va

riou

s op

tion

s of

fin

anci

ng r

epai

rs b

ut th

e m

echa

nism

s fo

r co

llec

ting

mon

ey a

nd s

tori

ng it

are

bes

t le

ft to

the

com

mun

ities

to

dec

ide.

Exi

t is

sue:

- ra

isin

g of

fun

ds is

nec

essa

ry f

or

repl

acem

ent o

f th

e sy

stem

in th

e lo

nger

te

rm fu

ture

. T

he d

ata

did

not

yie

ld a

ny

insi

ghts

to th

is c

hall

enge

.

Exi

t st

rate

gy:-

Rep

lace

men

t nee

ds to

be

disc

usse

d w

ith

the

com

mun

ity.

Str

ateg

ic c

onne

ctio

ns b

etw

een

the

com

mun

ity

and

the

fina

ncia

l ser

vice

sect

or m

ay n

eed

to b

e m

ade,

this

will

var

y ba

sed

on th

e co

untr

y co

ntex

t.

T

rain

ing

(Edu

cati

on)

Des

ign

re

com

men

dati

ons

Top

ics

to in

clud

e te

chni

cal,

or

gani

sati

onal

and

fi

nanc

ial a

reas

.

Exi

t is

sue:

- N

eith

er p

arti

cipa

tion

of

the

wid

er c

omm

unit

y no

r or

gani

sati

on

guar

ante

es g

ood

pum

p fu

ncti

onin

g. W

hile

it

may

be

desi

rabl

e fo

r in

tegr

ated

wat

er,

heal

th a

nd s

anit

atio

n p

rogr

amm

es a

nd

long

er te

rm d

evel

opm

ent,

it i

s no

t cri

tica

l to

good

pum

p fu

ncti

on.

Com

pete

nce

is m

ore

impo

rtan

t.

Exi

t st

rate

gy:-

If

time

and

reso

urce

s ar

e lim

ited

it i

s be

tter

to

focu

s tr

aini

ngs

on th

e te

chni

cal s

kill

requ

ired

to m

aint

ain

and

re

pair

pum

ps ra

ther

than

org

anis

atio

n a

nd p

arti

cipa

tion

ski

lls

– U

NL

ES

S th

e pr

ogra

mm

e is

a p

art o

f a

long

er te

rm p

rogr

amm

e on

hea

lth

and

san

itat

ion.

C

ontr

ibut

es to

m

otiv

atio

n E

xit

issu

e:-

The

re is

a h

ighe

r fr

eque

ncy

of

brea

kage

whe

n lo

cal c

omm

uniti

es re

pair

the

bore

hole

than

whe

n a

n e

xter

nal p

erso

n o

r gr

oup

does

(are

a m

echa

nic

or lo

cal

gove

rnm

ent)

. i.e

. the

ext

erna

l per

son

has

a

grea

ter

com

pete

nce,

pos

sibly

res

ulti

ng fr

om

long

er tr

aini

ng a

nd m

ore

expe

rien

ce.

Exi

t st

rate

gy:-

The

key

ele

men

t is

com

pete

nce.

Loc

al

com

mun

ity

base

d m

echa

nics

nee

d as

muc

h te

chni

cal t

rain

ing

as

poss

ible

. F

ollo

w u

p tr

aini

ngs

shou

ld a

lway

s ha

ve te

chni

cal

com

pone

nts

alon

gsid

e an

y so

cial

ele

men

ts.

Lon

ger

trai

ning

s w

ere

mor

e eff

ecti

ve –

exi

t tr

aini

ngs

shou

ld b

e as

long

as

nece

ssar

y to

cov

er c

lear

ly a

ll th

e su

bjec

ts.

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

87

grou

p do

es (a

rea

mec

hani

c or

loca

l go

vern

men

t). i

.e. t

he e

xter

nal p

erso

n h

as a

gr

eate

r co

mpe

tenc

e, p

ossib

ly r

esul

ting

from

lo

nger

trai

ning

and

mor

e ex

peri

ence

.

com

pone

nts

alon

gsid

e an

y so

cial

ele

men

ts.

Lon

ger

trai

ning

s w

ere

mor

e eff

ecti

ve –

exi

t tr

aini

ngs

shou

ld b

e as

long

as

nece

ssar

y to

cov

er c

lear

ly a

ll th

e su

bjec

ts.

Des

ign

C

onst

rain

ts

Gen

der

issu

es

Exi

t is

sues

:- I

t is

cle

ar th

at p

umps

wit

h

wom

en m

echa

nics

or

care

take

rs, f

unct

ion

as

pum

p as

equ

ival

ent p

umps

mai

ntai

ned

by

men

Exi

t st

rate

gy:-

Tra

in w

omen

in b

oth

sof

t and

har

d sk

ills.

W

omen

can

mak

e us

e of

tra

nsfe

rabl

e pr

acti

cal s

kill

s su

ch a

s pu

mp

mai

nten

ance

and

repa

ir.

L

evel

of

trai

ning

E

xit

issu

e:-

Shor

ter

dow

ntim

es a

re fo

und

whe

n m

any

loca

l peo

ple

are

trai

ned.

E

xit

stra

tegy

:- I

t is

cle

ar f

rom

the

rese

arch

that

a g

ood

exit

stra

tegy

trai

ns a

s m

any

peop

le a

s po

ssib

le (m

inim

um m

ore

than

2)

. T

he m

echa

nics

can

be

men

or

wom

en.

An

em

phas

is o

n

peop

le fr

om w

ithi

n th

e us

er c

omm

unit

y w

ill s

hort

en d

ownt

ime.

Co

mp

eten

ce

D

esig

n

Con

stra

ints

C

apac

ity

of li

nk

orga

nisa

tion

Se

e al

l ab

ove

Con

text

Des

ign

re

com

men

dati

ons

Com

plem

enta

ry

acti

viti

es

Exi

t is

sue:

- he

alth

and

hyg

iene

trai

ning

is

best

app

lied

whe

n in

the

cont

ext o

f co

mm

unit

y m

obili

sati

on fo

r pu

mp

O&

M

(and

rep

air)

. It

doe

s no

t see

m p

ossib

le to

le

gisl

ate

hygi

ene

– it

is b

ased

on

re

lati

onsh

ips

and

awar

enes

s.

Exi

t st

rate

gy:-

Tra

inin

g of

loc

al c

omm

ittee

s sh

ould

incl

ude

tech

nica

l, o

rgan

isat

iona

l, fi

nanc

ial a

nd h

ealt

h a

reas

. T

rain

ings

sh

ould

enc

oura

ge p

eopl

e to

pro

blem

sol

ve o

n h

ealt

h r

athe

r th

an

legi

slat

e (i

e ru

les

and

sanc

tion

s do

not

pro

duce

as g

ood

a re

sult

as o

ngoi

ng d

ialo

gue

betw

een

com

mun

ity

agen

ts a

nd th

eir

own

co

mm

unit

y).

T

ech

nolo

gy m

ust b

e ap

prop

riat

e E

xit

issu

e:-

Som

e of

the

stro

nges

t fac

tors

in

flue

ncin

g pu

mp

func

tion

are

the

dept

h a

nd

tech

nolo

gy.

If e

mer

genc

y su

ppli

es fo

r re

sett

led

popu

lati

ons

are

to b

e ad

dres

sed

then

an

age

ncy

is g

ener

ally

pre

sent

ed “

fait

acco

mpl

i” w

ith

dep

th a

nd te

chno

logy

.

Exi

t st

rate

gy: C

onsi

der

chan

ging

the

tech

nolo

gy.

One

age

ncy

in th

e st

udy

used

Afr

idev

pum

ps b

eyon

d th

eir

reco

mm

ende

d de

pths

bec

ause

eve

n th

ough

they

wou

ld b

e su

bjec

t to

a h

ighe

r fr

eque

ncy

of b

reak

dow

n th

e ne

t res

ult

was

mor

e su

stai

nabl

e th

an th

e al

tern

ativ

e V

olan

ta.

A

vail

abili

ty o

f sp

ares

E

xit

issu

e:-

The

re m

ust b

e a

supp

ly o

f sp

ares

whi

ch c

an b

e ac

cess

ed b

y th

e co

mm

uniti

es.

Pri

vate

loca

l sup

plie

rs se

em

to b

e a

reas

onab

le o

ptio

n a

ltho

ugh

ther

e ar

e co

nsid

erab

le p

robl

ems

wit

h th

e pr

ofit

abili

ty

of s

pare

s. L

ow tu

rnov

er a

nd lo

w p

rofit

m

argi

ns c

ombi

ne to

mak

e st

ocki

ng p

ump

spar

es a

n u

natt

ract

ive

prop

ositi

on fr

om a

bu

sine

ss po

int o

f vi

ew.

Exi

t st

rate

gy:-

. It

is im

port

ant

to e

ithe

r fi

nd r

etai

lers

who

are

w

illin

g to

do

it ou

t of

a se

nse

of d

uty

to th

eir

com

mun

ity,

or

to

allo

w r

etai

lers

nea

rer

a 1

00%

han

dlin

g ch

arge

. If

suc

h a

ha

ndli

ng c

harg

e is

all

owed

, it

mus

t not

be

unde

rmin

ed b

y ag

enci

es o

r go

vern

men

ts su

pply

ing

spar

es a

t a su

bsid

ised

rat

e.

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t Str

ateg

ies”

R71

36

Page

88

spar

es a

n u

natt

ract

ive

prop

ositi

on fr

om a

bu

sine

ss po

int o

f vi

ew.

D

esig

n

Con

stra

ints

A

lter

nati

ve w

ater

re

sour

ces

Exi

t is

sue:

- T

hose

com

mun

ities

wit

h a

cces

s to

alt

erna

tive

sour

ces,

bot

h b

oreh

oles

and

un

reli

able

surf

ace

wat

er, w

ill b

e sl

ower

in

repa

irin

g th

eir

pum

ps th

an th

ose

wit

hout

.

Exi

t st

rate

gy:-

The

resp

onse

of

the

com

mun

ity

may

be

infl

uenc

ed o

ver

the

long

term

by

a p

rogr

amm

e of

hea

lth

and

hy

gien

e aw

aren

ess.

Pr

oble

ms

for

the

priv

ate

sect

or

See

abov

e A

vail

abili

ty o

f sp

ares

Add

itio

nal I

ssue

s

O

ptim

al u

se o

f w

ater

E

xit

issu

e:-

Tho

se p

umps

wit

h s

alty

wat

er

whi

ch is

onl

y us

ed fo

r w

ashi

ng a

nd n

ot

regu

larl

y us

ed fo

r dr

inki

ng, w

ill h

ave

long

er

dow

ntim

es.

Exi

t st

rate

gy:-

mak

e th

ese

hole

s lo

w p

rior

ity.

Do

not s

pend

tim

e an

d ef

fort

on

thes

e ho

les,

ther

e ar

e ot

her

hole

s w

here

the

sam

e eff

ort w

ill p

rodu

ce a

bet

ter

resu

lt re

gard

ing

sust

aina

bilit

y.

Se

ttle

men

t E

xit

issu

e:-

Sett

lem

ent w

ill b

e af

fect

ed b

y th

e bo

reho

le a

nd in

par

ticu

lar

the

conf

iden

ce th

e pe

ople

hav

e in

the

pum

p be

ing

repa

ired

qui

ckly

.

Exi

t st

rate

gy:-

If

the

agen

cy w

ants

to e

ncou

rage

sett

lem

ent

in a

pa

rtic

ular

loca

tion

, or

set o

f vi

llag

es, t

hen

it m

ust p

rom

ote

com

pete

nce,

and

mak

e su

re it

is

seen

to d

o so

.


8 Conclusions and recommendations The previous section has grounded the research results into practical advice for those considering their exit strategies on programmes that have water supplies and resettlement (emergency or post emergency). Key Findings:- There is a need to balance the training given to include suff icient technical focus. Building the capacity of a committee on subjects such as community mobilisation and mechanisms of community management enhances long term development. However, there is no substitute for technical training. Technical competence influences the sustainability of the pump. There is also no substitute for gett ing the community involved early on. Even the simplest of involvement, such as clearing roads or providing materials, enhances the future sense of ownership. To be inclusive in the installation process does not necessarily takes a lot of time and it should be possible for even emergency programmes to be mindful of exit during the entry process. External agencies as part of their exit strategies should set in place mechanisms for ongoing support. There should be a recognition that communities will need assistance for complex (albeit infrequent) second level repairs. Building the capacity of the private sector or local water institutions to be able to provide this technical support is part of a realistic exit strategy. Community mobilisation for problem solving, including Village Level Operation & Maintenance of the water supply, is a long term process. Communities need long term encouragement to maintain the motivation and develop the skills required for development. Social mobilisation of the communities may (will l ikely) come from local institutions other than water departments staffed by engineers (eg health and Community services). Local institutions able to provide ongoing social mobilisation should be identified and enabled to include the water system in their programmes. The research has provided data for assessing what the important elements of a sustainable water supply are. For pump function, the important links are with technical capabili ty (competence) and unavoidable starting conditions such as pump depth. The local organisation of communities and particularly the participation of the majority of the community in their own pump management actually has little impact on pump function. However it is an important goal in itself in terms of long term development and problem solving, and in health education, hygiene and the impact of clean water on the community. The literature states that integrated programmes which include hygiene education have a greater impact on health than a programme concentrates on the provision of clean water alone. However, for pump reliabili ty, local technical competence is the key factor. Ideally, a programme should start with participation of all stakeholders. However the reality of emergency and resettlement water programmes is that necessity can marginalize participation. The research has shown that this need not be critical to the sustainabili ty of a water point. The entry strategy certainly strongly influences the ownership and commitment of the community to VLOM. Even simple actions such as asking the


community to make a contribution to installation by materials or labour can enhance ownership and sustainabili ty. However, a strong exit strategy can enhance sustainabili ty of the system. Post installation training on technical, organisation, finance and health aspects of the water supply can promote a certain level of community ownership and responsibili ty. After initial trainings, small ongoing inputs can ensure longer term participation and motivation. The need for small ongoing support is not unique to exit strategies on emergency supplies. Even with a strong participatory entry strategy, the participation and motivation of the community and committees will decay over time, and small ongoing inputs are important for long term sustainabili ty. Therefore an important element of any exit strategy of an NGO, is the need to build support capacity within a local agency, possibly local authorities. In this one should not confine the support to the Water department, as other government departments (health or community development) may have more capacity to be proactive in their support and may have a greater sensitivity to the socio-economic aspects of pump management. Although participation is desirable for longer term development, the pump availabili ty is strongly influenced by the technical competence of the mechanics. Exit strategies should train as many caretaker mechanics as possible within a community. They should train women as well as men, since this provides women with transferable skill s and encourages problem solving which breaks free of cultural constraints. The technical level should be as “deep” as possible. If the communities have had some experience of maintaining the pump it may be possible to identify members of the community who have the capabili ty to learn how to make second level repairs to the pump. Even if communities are trained in repairs, it is necessary to have some capabili ty within the area for the complex repairs. This may be the local water authority or it may be private suppliers. Maintenance and repair needs a supply of spares. Before an agency exits from an area, it should ensure that there is some mechanism for supplying spares to the local communities. Experience indicates that private suppliers are a favoured option. However to encourage private suppliers or retailers to stock the low turnover, low cost components of an appropriate pump like the Afridev, agencies will have to face the realities of private enterprise. Retailers may need to mark up the components with a handling charge by as much as 100% in order to make them profitable. The alternative is to appeal to the goodwill of merchants and ask them to stocks the spares as a service to the community – with payment in the form of prestige rather than cash. Any private supply of spares should not be undermined by subsidies from other agencies. It seems that it is important that the exit strategy includes introducing the local retailers to a national supply system whether that be government controlled or preferably a sustainable private supply. In terms of accessing and paying for spares, it seems best to leave communities to raise finance according to their traditional mechanisms and not impose a regular collection from outside. Communities will find their own mechanisms to save the money. While keeping a supply of spares on hand in the community can be encouraged it does not ensure short


downtimes. For long term development it seems better to let communities do a problem solving exercise on raising finance and this may include using their traditional forms of saving and risk alleviation. Training course should discuss the problem of raising finance but leave the details to the community. The research has shown how an exit strategy addressing these points above, can leave a sustainable supply. Investment in such a strategy might be as much as the original direct cost of the water point or more. However, when faced with a choice between investing in a good exit strategy on resettlement water points or investing in completely new water points, agencies should invest in good exit strategies. The total cost for an exit strategy on an existing water point will be less than that for a strong entry strategy, direct costs and exit strategy on a new water point. Government policies for post emergencies should reflect this where possible and encourage their NGO partners to undertake planned exit programmes.

8.1 Recommendations:-

• Agencies that have been involved with water supplies during emergency or resettlement programmes should have planned exit strategies regardless of their entry strategies. A planned exit strategy can enhance sustainabili ty.

• Donor agencies should be prepared to fund follow up work after an emergency or

resettlement programme. Strategic inputs to a planned exit strategy can be a cost effective mechanism for leaving sustainable water assets in an area.

• Donor agencies should create policies which link departments concerned with

emergency work with those for longer term development so that it is clear which department should fund the necessary exit strategies.

• Agencies should note that a key factor in creating a sustainable water supply is to

have adequate technical and management competence within the locality. Competence has a greater influence on the total availabili ty of water than the details of community organisation and participation. Limited budget or resources for an exit strategies should be used to enhance technical competence, especially if traditional forms of community organisation are effective.

• Agencies wishing to exit an area must acknowledge that some ongoing support to

the community is required. Ongoing technical support is required for diff icult technical repairs and ongoing institutional support is required to encourage ongoing social mobili sation. Those responsible for water are not necessarily the most effective local authority or agency for the social mobili sation and other departments such as health or community development should be considered. Agencies should make strategic links and build capacity to leave some form of ongoing support. The private sector is a strong possibili ty in present day Africa for both the supply chain and some form of area mechanic.

• Exit strategies for sustainabili ty should include the following practical elements:-


�Social profili ng – to know in advance of potential conflicts of within the community. �Working with existing structures if they exist. Do not attempt to introduce new democracy – let the vill age community suggest the most appropriate way of selecting people for training, committees and care-taking �Training of the community selected committee in technical, management, financial and health skill s with sufficient emphasis on technical skill s to ensure a competent repair process. �Training on the financial mechanisms should leave the community to decide if regular payments is the best mechanism for their community. (Collecting quotas at the time of breakdown does not tend to delay repairs). �There is a need to save towards replacement of the system – strategic connections between the community and the financial service sector may need to be made, this will vary based on the country context. �Training on management should not be dogmatic on the need for rules and sanctions. Rules can discourage ownership by the general community. Training should include discussion of the most appropriate mechanisms to manage the pump. �Training of as many people as possible (within a community and within budget) in those skill s. Training of both men and women in all the skill s. (Do not reinforce stereotypes) �Working with the government and other agencies to ensure a supply chain of spares. Preferably the supply chain should result in spares being available to the community from a local retailer. �Capacity building of local government or private enterprise, to form a group that can undertake more difficult repairs on an ongoing basis after the agency has exited. �Working with local government other than just the Water Department. Health and Community Development (Youth, Gender) workers often have more resources (time and capacity) to visit communities regularly. Agencies should encourage local government workers to encourage communities to sustain their motivation.

8.1.1 Recommended immediate actions • The current Guidance notes circulating in DFID do not seem to address the overlap

between an emergency WS&S programmes, resettlement or rehabili tation, and longer term development WS&S programmes. Guidance notes based on this research should be disseminated to complement and support existing literature.

• If possible, review DFID policies and guidelines surrounding the transition from emergency WS&S programmes to longer term development. This may require a workshop/consultation with relevant stakeholders to discuss the findings and suggest practical administrative changes.

• Guidance notes targeted at different institutions should be generated based on the information in this report, eg a guidance note for national governments experiencing an emergency and for NGOs working with WS for resettlement and development


The research has identified the important factors that make up a good exit strategy that leaves the water point as sustainable as possible within the country context. While there is no substitute to a good participatory entry strategy, this gives some hope for retaining the assets of an emergency or resettlement programme beyond the life of the programme and presence of the external agency.


9 References Afonso, Nhassango & Sutton 1998, Community Water and sanitation project, Inhambane Province, Final project Evaluation, Care International 1998 Appiah J.O. 1999 Sustainable water and sanitation committees. Integrated Development for Water Supply and Sanitation. 25th WEDC Conference. Ethiopia. 1999. Arlosoroff, Tschannerl, Grey, Journey, Karp, Langenegger and Roche. 1987 Community Water Supply; the Handpump Option. World Bank 1987. Aziz.N and Halvorson. S 1999 Women's involvement: a switch in thinking, Hoto, Pakistan PLA Notes 35. 1999. Blench. R 1998 Fragments and Sentiments: Why is “The Community” the Focus of Development? ODI Agricultural Research Extension Network (AgREN) Network Paper No. 81. 1998. Brett. E.A. 1996 The Participatory Principle in Development Projects: the Costs and Benefits of Co-operation. Public Administration and Development. Vol.16. 1996. Buttle M and Smith. M 1999 Out in the Cold: Emergency water supply and sanitation for cold regions WEDC 1999 Chalinder. A 1994. Water and Sanitation in Emergencies. ODI Relief and Rehabili tation Network Good Practice Review 1. 1994. Concern Univeral, Emergency safe Water Supply and Environmental sanitation Project, 1996 Direccao nacional de Aguas, Fiscalizacao de Construcao de furo nos distritos de Vilankulo, Inhassoro, Govuro e Mabote, 1997 Colin J 1999 VLOM for Rural Water Supply: Lessons from Experience. Task No 162, WELL, London & Loughborough Direccao nacional de Aguas, Politica nacional de Aguas, 1997 DFID 1998 Guidance Manual On Water Supply and Sanitation Programmes. Water and Environmental Health at London and Loughborough 1998 FAO 1995 Reforming Water Resources Policy. A Guide to Methods, Processes and Practices. FAO Irrigation and Drainage Paper 52. 1995. Hill yard. D and Mthunzi. M. 1998 Concern Universal, BOND 1998 “Introduction of Low Cost Rotary Drilli ng to Malawi: Specifically the Eureka Port-a-Rig” http://info.lut.ac.uk/well/bond98/contents.htm.


House. S and Reed. B 1997 WATER SOURCES EMERGENCY - GUIDELINES FOR SELECTION AND TREATMENT 1997. House. S and Reed 1997 Emergency Water Sources: Guidelines for selection and treatment WEDC 1997 Lammerink . M.P, Bolt. E, De Jong. D and Schouten. T, June 1999 Strengthening community water management PLA Notes, June 1999. Manikutty .S. 1998 Community Participation: Lessons from Experiences in Five Water and Sanitation Projects in India. . Development Policy Review. Vol.16. 1998 McKemey & Nghatsane. 1997 End of Term Evaluation, Agricultural and Well maintenance Programme Gaza Province World Relief 1997. Noppen D (Editor) 1996 VLOM of handpumps: Experiences from Karonga, Malawi. Project and programme Papers. IRC International Water and sanitation Centre, The Hague Obiols, A. L. and Baumann, E. (1998): Performance of Afridev Pumps in the CARE Community Water and Sanitation Project, Inhambane, Mozambique, SKAT News No. 33 pp. ODA (1996b) Manual of Environmental Appraisal (revised edition) Overseas Development Administration (Now DFID), London. Ray. S 1993 Emergency water supply programme, Limpopo Corridor Gaza Province Final Report World Relief 1993 Regmi. S. C and Fawcett. B 1999 Integrating gender needs into drinking water projects in Nepal. Gender and Development Vol.7 No. 3 1999. SCF 1998 (Mill enium Consulting Group) Salima Water and sanitation Promotion Project – Lessons Learned Evaluation, 1998 SCF Mchinji Community water Project 1998 – Lessons Learned Evaluation, 1998 SCF, 1998 Dowa District Safe Water and sanitation, Proposal, 1998 Simpson et al 1996. An evaluation of the HESP/VLOM Programme, submitted to CU, UNICEF, Water Department, Ministry of Health, Ministry of Women Children Community Services, Concern Universal 1996 Skinner, B. (1996): Handpump Standardisation, Proceedings of the 22 nd WEDC Conference, WEDC, pp.102-105 Smout.I and Parry-Jones. S 1999 Lessons Learned from NGO Experiences in the Water and Sanitation Sector 1999


Van Koppen. B, Mahmud. S 1995. Women and waterpumps in Bangladesh: the impact of participation in irrigation groups on women’s status. Bangladesh Institute of Development Studies. Various Monitoring documents. All agencies gave a selection of monitoring reports and examples of agreements. Eg Agreement between CU and Mangola Community for mangola Safe Water Supply Project submitted to Malawi Social Action Fund (MASAF). White, J. (1996): Evaluation Report EV: Evaluation Synthesis of Rural Water and Sanitation Projects, ODA. Whiteside. M 1996 Realistic Rehabili tation. Linking Relief and Development in Mozambique. Development in Practice. Vol.6 No.2. 1996. World Relief. 1993 Rural water supply proposal, Limpopo Corridor Gaza Province 1994-95 Phase II , 1993. World Relief. 1993 Perfil Socio-economica da populacao no corredor do limpopo, Chokwe Gaza, World Relief 1993


10 Appendix MAPS OF MALAWI AND MOZAMBIQUE


11 Appendix - LOGICAL FRAMEWORK OF THE RESEARCH

Narr ative summary Measurable indicators

Means of verification Impor tant assumptions

• Improved Assess ment and Management of Water Resources for Resett lement Popu lations

that at least two agencies incorporate the identified elements in their next water programme in southern Africa

programme proposals submitted to funding agencies (DFID & others)

• That the recommendations are accepted and used by agencies involved in water actions with resettlement populations.

• To identify elements of a successful exit strategy for water programme with resettlement rural populations.

at least four definable factors which influence successful exit.

Report summary available for distribution

• That the recommendations are accepted and used by agencies involved in water actions with resettlement populations.

• thematic evaluation of three water programmes in Southern Africa

• analysis of common factors which have led to effective handover

• workshop attendents endorsement of common factors

• project management system installed and operational

detailed data for three programmes is collated and available

data has undergone analysis by both project team and management of 3 agencies working on resettlement water programmes

workshop has reached conclusion and published report.

Visit reports

Interim reports

Thematic evaluation report

Workshop report

Final report

• That there are identifiable common factors

Activities : • Initial lit erature review

• three project visits including focus group discussions

• workshop of major players in case studies

• production of report

Inputs: • project team • 6 airfares for project

team • expenses for support

staff (translators) • travel expenses for

agency management to workshop

• communication • report production

• that agreed verbal co-operation of agencies becomes actual co-operation

• that there is working agreement of governments for the project

• that communities are willi ng to co-operate

• that there are no major disturbances in the resettled populations.

• availabilit y of support staff (translators, etc)


12 Appendix Respon ses from WEDC Conference A questionnaire was distributed to delegates at the WEC conference in Addis Ababa (1999). The following results are based on 14 returned questionnaires.

12.1 Boreholes and Programmes All respondents worked with handpumps in development situations (one in both emergency and development), and all but one in the rural sector. The figure below shows that most boreholes cater for 200 – 500 people.

0

1

2

3

4

5

6

100

- 15

0

150

- 30

0

300

- 50

0

500

-10

00

1000

-20

00

2000

-40

00

Figure 12-1 Frequency distribution – number of people served per borehole There was a wide range in the length of training given to community members on pump maintenance, as shown in Figure 12-2. The data sample was insufficient to state with any confidence that there was a link between length of training and proportion of operating boreholes. 80% said they provided follow up training, mostly within a year of training, although some offered this as required.

0

1

2

3

4

5

6

7

< 1 wk 1 week 2 weeks 3 weeks

Figure 12-2 Frequency distribution – duration of training on pump maintenance Initial participation and payments from the community were a strong feature of these programmes (only one did not encourage this). After installations, most agencies provided support to communities over a period of two years, with one supporting for over 10 years. Most respondents handed responsibili ty for maintenance to community organisations, although some specified that they were to be supported by government agents.


Nearly one third of respondents reported a shift in settlement around water points provided during times of partial water shortage. Common themes between these were that they handed water points over to community based organisations, and they all stressed the importance of technical issues and spares.

12.2 Key Issues Respondents were invited to list activities and policies leading to sustainabili ty, and barriers inhibiting sustainabili ty; the following tables summarise comments received.


12.2.1 Key activities leading to local sustainabili ty Category 1 Community responsible for management 111 1 Community ownership 11 2 Self initiation 1 2 Participation at/before installation 1111 2 Compulsory preparation by users (legal) 1 2 Community contribution (capital cost) 1111 3 Training (local gov’ t & community structures; watsan

committees) 111111

3 Follow up training and monitoring 11 4 Participation in management / maintenance 111111 4 Appointing community mobili ser 1 4 Legal status to committees and associations 1 4 Involvement of women 1 4 Vill age pump mechanics 1 4 Vill age water committees, formation and training 11 5 Use VLOM pumps 111 6 Health and hygiene link to water 11 6 Health education 1 7 Community contribution (maintenance) 1 7 Close co-operation with agency for follow up support 1 7 Reliable pump and spares programme 11 7 Proximity and convenience of alternative source 1 7 Regular monitoring and evaluation (local gov’ t &

community) 1

7 Communication between gov’ t and community 1 7 Support from local gov’ t in O&M 1 7 Water fund 1 7 Minimise distance to source 1 7 Family wells rather than community 1 7 Availabili ty of water for other uses 1


12.2.2 Principal blockages Category 3 Inefficient training 1 3 Government office lack of capacity 11 3 Lack of institutional capacity (all levels) 1111 3 Lack of training 11 3 Lack of maintenance abili ty 1 5 VLOM pumps require technical support not available in

vill age 1

5 Use of traditional sources (free) 1 7 Poor perception of effectiveness of committees 1 7 Subsidising water sales 1 7 Conflict between WSCs and local councils 1 7 Policies not encouraging private company involvement 11 7 Spares availabili ty; Government spares distribution not

effective 111

7 Lack of technical support 1 7 distance 1 8 Community value attached to water systems 1 8 Hand over to corrupt / disinterested local government 1 8 Trained technical staff move to better paid jobs 1 8 People reluctant to take collective responsibili ty 1 8 Local disagreements 1 8 Squandering of maintenance income 1 9 Economy of the community, willi ngness to pay 111111 9 cost 1 9 High capital cost 1


12.3 Summary of Key Issues Category Description Total

requirements Total Barriers

1 Community involvement and ownership 5 2 Entry strategy e.g. prior contributions 10 3 Training 8 10 4 Management structure 12 5 Technical 3 2 6 Links to other issues e.g. health &

sanitation 3

7 Programme design 12 10 8 Personnel motivation / integrity 6 9 Costs / abili ty and willi ngness to pay 8 The most commonly occurring themes for sustainabili ty were training, and participation of wider sections of the community (e.g. all users) in management of water points. Comments on training covered who (for committees, caretakers, mechanics, local government), what (management), and when (follow-up training); the lack of capacity and need for training was also the highest rating barrier to sustainabili ty. Entry strategy issues such as community participation in planning, and prior contributions (financial or in kind) were perceived as important. The significance of the technology used was recognised by a number of comments on the need for VLOM technology. It was further commented that even VLOM pumps require technical support not available at vill age level, so ongoing technical support is required of an external agent. The availabili ty of spares is quoted as both a barrier and a prerequisite for sustainabili ty. It is interesting that a couple of respondents identified lack of private sector involvement as a barrier to sustainabili ty, referring to the provision of spares. Links to other development issues (health and hygiene) was raised as key policies for sustainabili ty. It is not clear whether this is regarded as essential to achieve sustainabili ty of water points, or whether links are required to achieve sustainabili ty of overall development projects. Ongoing monitoring and evaluation was raised by a couple of respondents as a key activity for local sustainabili ty. This raises interesting questions over ownership of water points, as some external agent would need to assume responsibili ty for M&E activities. Although some comments refer to relationships between community and agency (pre-handover), others refer to support and communication between community and local government, implying that local government appears to be the most likely organisation to provide ongoing support needed for sustainabili ty.


The most commonly identified single barrier to sustainabili ty is lack of payment – variously attributed to economic condition of the community, abili ty of individuals, and willi ngness of individuals. A major problem is presented in motivation of communities and staff. Although corruption, over which programmes may have little influence, is mentioned, it does not dominate, and issues arising from management structures are equally important e.g. migration of trained mechanics.

12.4 Emergency Installations Respondents were asked whether boreholes installed in times of water shortage would be more susceptible to poor management than those installed during times of adequate water supply.

12.5 Reasons given Yes Poor siting 1 They will not be interested to look after it 1 No People will return to original sources if not educated 1 Depends on value attached to system 1 As long as there are health clubs 1 People have stronger sense of felt need 11 In Ethiopia most of population in emergency situation 1 Depends on cohesion of community 1 Needs effective training for maintenance 1 75% of respondents felt that drought conditions should make no difference to the future management of boreholes, although most gave conditions for achieving this e.g. training, cohesive community. Stronger felt need was given as a reason why the management of boreholes installed during drought could be equally good.


13 Appendix Valid percentages of variables

VALID PERCENTAGES OF VARIABLES This appendix presents the variables considered. The following is a simple frequency analysis for those interested in the profile of the responses. Data for certain variables has been integrated on a community basis, whilst most is based on individual responses.

Implementing agency

Valid Percent

Cumulative Percent

World Relief 29.2 29.2

Care 34.1 63.3

Concern Universal 36.7 100.0

Valid

Total 100.0

Coun try

Valid Percent Cumulative

Percent

Mozambique 63.3 63.3

Malawi 36.7 100.0

Total 100.0

Region/Province


Percent

Gaza 29.2 29.2

Inhambane 34.1 63.3

Central Region 22.3 85.6

Southern Region 14.4 100.0

Total 100.0

District/ Municipali ty

Valid Percent

Cumulative Percent

Guija 6.9 6.9

Mabalane 7.5 14.4

Chicualacuala 14.8 29.2

Vilankulo North 20.3 49.5

Vilankulo South 13.8 63.3

Ntcheu 12.1 75.4

Dedza 10.2 85.6

Chikwawa 14.4 100.0

Valid

Total 100.0

Distance of community from Agency op's centre

Valid Percent

Cumulative Percent

<50 klm 19.3 19.3

50 to 100 klm 55.7 75.1

100 to 150 klm 12.1 87.2

151 to 200 Klm 4.6 91.8

200> klm 8.2 100.0

Valid

Total 100.0

Type of Bore (Communitiy)

Valid Percent

Cumulative Percent

new 76.6 76.6

Reconditioned 23.4 100.0 Valid

Total 100.0


Year of pump installation (Community)

Valid Percent Cumulative Percent

91 2.2 2.2

92 8.7 10.9

93 19.6 30.4

94 28.3 58.7

95 15.2 73.9

96 8.7 82.6

97 13.0 95.7

98 4.3 100.0

Valid

Total 100.0

Type of Pump (Community)


Percent

Afridev 83.0 83.0

Volanta 10.6 93.6

Nadia 4.3 97.9

Other 2.1 100.0

Valid

Total 100.0

Gender of interviewee


male 49.3 49.3

female 50.7 100.0 Valid

Total 100.0

Age of interviewee


Percent

<30 19.1 19.1

30 to 40 49.0 68.1

50> 31.9 100.0 Valid

Total 100.0

Position in community

Valid Percent

Cumulative Percent

chief/headman 6.9 6.9

commii ttee member 11.1 18.0

mechanic/maintenance 7.5 25.6

no po sition 74.4 100.0

Valid

Total 100.0

There was debate at the workshop over the translation of chief and headman, and over committees. By consolidating the variable to include both chief and headman as one item, we hope we have overcome the translation problems. The committees are of complex makeup and there may have been translation errors over whether someone was a committee member or a member of the maintenance sub committee.

Proximity to well


close 74.1 74.1

distant 25.3 99.3

Valid

Total 100.0

Quali ty of water (Community)

Valid Percent

Cumulative Percent

sweet 70.2 70.2

slightly salty 23.4 93.6

very salty (diff icult to drink)

6.4 100.0 Valid

Total 100.0


Operational Status of pump (Community)

Valid Percent

Cumulative Percent

Operating 78.7 78.7

Broken 19.1 97.9

Missing 2.1 100.0 Valid

Total 100.0

Frequency of breakage (Community)

Valid Percent

Cumulative Percent

less than on ce a year

34.8 34.8

once a year 19.6 54.3

twice a year 17.4 71.7

more than twice a year

28.3 100.0

Valid

Total 100.0

Average down-time when broke (Community)

Valid Percent

Cumulative Percent

less than on e week

53.3 53.3

one to two weeks

22.2 75.6

two to four weeks

13.3 88.9

more than on e month

11.1 100.0

Valid

Total 100.0

Main cause of breakage (Community)

Valid Percent

Cumulative Percent

softwear 44.4 44.4

hardwear 55.6 100.0

Valid

Total 100.0

Who u sually repairs pump

Valid Percent

Cumulative Percent

Committee 47.8 47.8

Local pump mechanic

28.7 76.5

Privately contracted mechanic

4.3 80.9

Government agent 16.1 97.0

Agency 1.7 98.7

No on e .4 99.1

Don't know .9 100.0

Valid

Total 100.0

Who repairs the pump may have been subject to a few translation errors.

Number of t rained pu mp mechanics


1 24.6 24.6

2 21.3 45.9

3 13.1 59.0

4 21.3 80.3

5 9.8 90.2

6 9.8 100.0

Valid

Total 100.0

Last agency pump maintenance visit

Valid Percent

Cumulative Percent

<3months 19.8 19.8

Six months

14.4 34.2

One year 5.4 39.6

One year> 14.4 54.1

Not known 45.9 100.0

Valid

Total 100.0


Abili ty of local committee or trained pu mp maintenance rep's

Valid Percent

Cumulative Percent

Very able (soft and h ard elements)

58.2 58.2

Able (soft element only) 26.7 84.9

Not able (not trained) 7.5 92.5

Not known 6.8 99.3

Valid

Total 100.0

Who was trained in pu mp maintenance

Valid Percent

Cumulative Percent

local maintenance mechanics 31.8 31.8

All the committee 50.6 82.5

No on e 9.1 91.6

Not known 8.4 100.0

Valid

Total 100.0

When pu mp maintenance training g iven

Valid Percent

Cumulative Percent

Before or at t ime of installation 38.9 38.9

Sometime after installation 52.8 91.7

Not known 8.3 100.0

Valid

Total 100.0

Length of t raining

Valid Percent

Cumulative Percent

less than a week 64.6 64.6

more than on e week

35.4 100.0 Valid

Total 100.0

A variable “ followup training” was deleted following workshop recommendations

Level of t raining g iven

Valid Percent

Cumulative Percent

Sanitation and environment only 5.6 5.6

Softwear maintenance 46.3 51.9

Soft and h ardwear maintenance 37.0 88.9

Not known 11.1 100.0

Valid

Total 100.0

Sanitation and h ealth training


yes 97.9 97.9

no 2.1 100.0 Valid

Total 100.0

This variable has a significant number of “missing system” . Agencies involved are surprised at this response since training has been given in most communities. In the light of this, this variable has not been considered reliable.

Water committee appo inted

Valid Percent

Cumulative Percent

yes 91.3 91.3

No 6.4 97.7

Not known 2.3 100.0

Valid

Total 100.0

Clarification – the committee is not appointed by the agency. It is elected within the community by their own processes (sometimes prompted by the agency).


Number of members


2 2.8 2.8

3 10.4 13.2

4 4.7 17.9

5 10.4 28.3

6 17.9 46.2

7 6.6 52.8

8 3.8 56.6

9 2.8 59.4

10 35.8 95.3

11 1.9 97.2

12 .9 98.1

14 .9 99.1

30 .9 100.0

Valid

Total 100.0

Number of women on committee


1 13.4 13.4

2 20.9 34.3

3 19.4 53.7

4 20.9 74.6

5 11.9 86.6

6 7.5 94.0

7 3.0 97.0

10 3.0 100.0

Valid

Total 100.0

Election o f members

Valid Percent

Cumulative Percent

community 64.1 64.1

Chief or headman 24.2 88.2

agency 1.3 89.5

government agent 4.6 94.1

Not known 5.9 100.0

Valid

Total 100.0

Again it should be noted that in fact agencies do not elect members to a committee. However since the above is the perception of individuals within the community the variable stands.

Frequency of committee meetings

Valid Percent

Cumulative Percent

twice a month 25.4 25.4

once a month 27.5 52.9

occas ionally (irregular)

13.8 66.7

Only when pu mp breaks

10.1 76.8

Not functioning 13.8 90.6

Not Known 9.4 100.0

Valid

Total 100.0

There are differences between the two countries on payments. In Mozambique, agencies encourage feedback of finance regularly within the community, whereas in Malawi, the feedback tends to be as a small part of more general meetings.

Community involvement in water committee

Valid Percent

Cumulative Percent

always 61.0 61.0

occas ionally 19.2 80.2

never 18.7 98.9

Valid

Total 100.0


Pump Maintenance payment

Valid Percent

Cumulative Percent

once a month 39.2 39.2

once a year 1.4 40.6

only at t imes of breakage 47.2 87.8

no quo ta payment 12.2 100.0

Valid

Total 100.0

Local current quo ta payment

Valid Percent

Cumulative Percent

1.00 1.3 1.3

2.00 2.2 3.6

3.00 3.1 6.7

5.00 11.6 18.3

10.00 8.0 26.3

20.00 .4 26.8

200.00 .4 27.2

1000.00 13.8 41.1

1500.00 .4 41.5

2000.00 20.1 61.6

2500.00 2.2 63.8

3000.00 3.6 67.4

4500.00 .4 67.9

5000.00 13.8 81.7

10000.00 15.6 97.3

12000.00 .9 98.2

20000.00 1.3 99.6

30000.00 .4 100.0

Valid

Total 100.0

Payment of quo ta


yes 81.3 81.3

no 18.7 100.0 Valid

Total 100.0

The financial variables are of course subject to people saying what they think an external person might want to hear.

Sanction for non e payment


yes 36.3 36.3

no 63.7 100.0 Valid

Total 100.0

In Malawi there are no sanctions, some cases have a person by the well , but it differs from Mozambique where they encourage daily control and a lot of sanctions.

Rules re Pump use


yes 74.2 74.2

no 25.8 100.0 Valid

Total 100.0

Spares available

Valid Percent

Cumulative Percent

yes 88.9 88.9

no 3.7 92.6

not known 7.4 100.0

Valid

Total 100.0

We note the large number of missings but this is consistent with the number of interviewees who are part of the VLOM mechanism.


Where are/have been spares acquired

Valid Percent

Cumulative Percent

Local stores 45.3 45.3

Agency 37.7 83.0

government agent

5.7 88.7

not known 11.3 100.0

Valid

Total 100.0

There was some confusion over whether spares have been supplied by agencies. In practice some initial spares have been supplied but each agency has a policy of promoting a sustainable supply of spares (where it is out of the loop).

Ownership of well

Valid Percent

Cumulative Percent

community 71.5 71.5

agency 2.6 74.1

government 21.2 95.3

Don't know 4.7 100.0

Valid

Total 100.0

These perceptions are interesting but must be treated in context. Groundwater does indeed belong to the government, and therefore some people may have interpreted the question in a different way from others.

Well /pump requested by community

Valid Percent

Cumulative Percent

yes 68.5 68.5

no 26.8 95.2

not known

4.8 100.0 Valid

Total 100.0

Present at drill/i nstallation


yes 83.5 83.5

no 16.5 100.0 Valid

Total 100.0

Sett led because of well


yes 14.2 14.2

no 85.8 100.0 Valid

Total 100.0

In Mozambique it should be noted that people re-settled in their home area because it had water but did not settle solely because there was water.

Others sett led because of well


many 35.8 35.8

few 17.9 53.6

non e 46.4 100.0

Valid

Total 100.0

Other sources of water

Valid Percent

Cumulative Percent

river 15.2 15.2

lagoon s 8.6 23.8

other open wells 24.6 48.4

neighbou ring boreholes 45.3 93.8

no o ther source 6.3 100.0

Valid

Total 100.0


Participated in installation


yes 68.6 68.6

no 31.4 100.0 Valid

Total 100.0

This variable is of course linked to the year of installation and the transition from emergency supplies to a more developmental approach. Some food for work activities may have been seen by vill agers as a contribution or particpation by them in the installation.

Participate in maintenance


yes 85.4 85.4

no 14.6 100.0 Valid

Total 100.0

This variable encompasses people perceptions. They may only clean the head of boreholes, or sweep a few leaves and perceive this as participating in maintenance.

Siting o f well (Community)


Percent

community 25.0 25.0

Chief /elders 10.0 35.0

drill ers 27.5 62.5

agency 7.5 70.0

government 2.5 72.5

existing site 27.5 100.0

Valid

Total 100.0

Even in emergency situation there was a consultative process of as many stakeholders as possible.

Emergency installation (Community)

Valid Percent

Cumulative Percent

due to drough t

83.0 83.0

post drough t 17.0 100.0 Valid

Total 100.0

Prior contribution (pre drilli ng/pump provision)


yes 12.7 12.7

no 87.3 100.0 Valid

Total 100.0

Since Care is the only agency implementing a prior contribution (monetary) strategy one must treat this variable with caution.

Number in hou sehold


1 1.2 1.2

2 6.7 7.9

3 10.3 18.2

4 9.1 27.3

5 10.9 38.2

6 12.7 50.9

7 9.7 60.6

8 15.8 76.4

9 5.5 81.8

10 4.8 86.7

11 4.2 90.9

12 3.0 93.9

13 1.2 95.2

14 1.8 97.0

15 2.4 99.4

16 .6 100.0

Valid

Total 100.0


Number of buckets per day (20 li tres)


Percent

1 2.5 2.5

2 16.7 19.1

3 16.7 35.8

4 22.2 58.0

5 13.6 71.6

6 11.7 83.3

7 2.5 85.8

8 6.2 92.0

10 3.1 95.1

12 2.5 97.5

14 1.2 98.8

15 .6 99.4

20 .6 100.0

Valid

Total 100.0

Felt pressure on well


high 72.8 72.8

low 27.2 100.0 Valid

Total 100.0

We note that pressure is subject to the households planning – most people collect water during a few hours of the day creating a “rush hour” . Many interviewees acknowledged that out of hours time for waiting was considerably less.

Time waiting for water (hou rs)


.10 6.2 6.2

.20 2.1 8.2

.30 3.1 11.3

.50 9.3 20.6

1.00 12.4 33.0

1.30 1.0 34.0

2.00 20.6 54.6

3.00 20.6 75.3

4.00 14.4 89.7

5.00 3.1 92.8

6.00 7.2 100.0

Valid

Total 100.0


Number of hou seholds using well


30 2.6 2.6

44 2.6 5.1

50 5.1 10.3

60 7.7 17.9

62 2.6 20.5

72 2.6 23.1

78 2.6 25.6

81 2.6 28.2

84 2.6 30.8

89 2.6 33.3

90 5.1 38.5

100 2.6 41.0

110 2.6 43.6

130 2.6 46.2

140 2.6 48.7

145 2.6 51.3

150 2.6 53.8

175 2.6 56.4

200 15.4 71.8

250 5.1 76.9

300 2.6 79.5

312 2.6 82.1

375 2.6 84.6

500 2.6 87.2

600 2.6 89.7

1000 2.6 92.3

1100 2.6 94.9

1440 2.6 97.4

1446 2.6 100.0

Valid

Total 100.0

This variable has not been used because of the many missing cases. However it has been left in this appendix for interest.

Observed status of well head (Community)

Valid Percent

Cumulative Percent

well kept 21.3 21.3

average 36.2 57.4

poo r (li tt le evidence of care)

42.6 100.0 Valid

Total 100.0

Awareness of interviewee

Valid Percent

Cumulative Percent

low 39.4 39.4

average 41.1 80.5

high 19.5 100.0 Valid

Total 100.0


14 Appendix - DATA ANALYSIS – WELL PERFORMANCE INDICATORS

The following tables present results from Mann-Whitney U tests (non-parametric) performed on coded data. The test identifies where there are differences between two populations. The figures presented represent the statistical significance of the difference i.e. p = 0.05 means 95% confidence.

14.1 Summary Table To help pick out those variables which make a difference, the summary table presents only those results where the date shows a difference with a high degree of confidence.

*** = 100 % confidence ** > = 99 % confidence and above * >= 95 % confidence and above.

Supplementary Tables All of the significance figures are presented in the following supplementary tables:

• Well performance indicators • Competence indicators • Organizational indicators • Community participation indicators.

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t St

rate

gies

” R

7136

Pa

ge 1

17

Tab

le 1

4-1.

Sum

mar

y of

sig

nifi

cant

link

s (b

ased

on

Man

n W

hitn

ey t

ests

)

(C

onfi

denc

e li

mit

s:

**

* =

100

%;

**

>=

99%

; *

<=

95%

)

W

ell P

erfo

rman

ce

Com

pete

nce

Org

anis

atio

n C

omm

unity

Par

ticip

atio

n

Operational Status of pump

Frequency of breakage

Average down-time when broke

Observed status of well head.

Who usually repairs pump

Number of trained pump mechanics

Last agency pump maintenance visit

Abil ity of local c’ ttee or trained rep's

Water committee appointed

Number of members

percentage of women

Election of members

Frequency of committee meetings

Pump Maintenance payment

Payment of quota

Sanction for none payment

Rules re Pump use

Awareness of interviewee

Ownership of well

Community involvement in c’ ttee

Participated in installation

Participate in maintenance

Prior contribution (pre drill ing/pump)

Pre

ssur

e (h

igh;

low

) **

**

Alt

erna

tive

sour

ce (b

ore;

unr

elia

ble)

**

**

**

**

**

*

*

**

*

*

year

of i

nsta

llati

on (p

re ’9

5; ’

95 o

n)

***

**

* **

**

*

**

* **

*

**

**

*

* **

* **

**

**

* Q

ualit

y of

wat

er (

swee

t; s

alty

) **

*

**

**

*

**

*

* **

* O

ther

bor

ehol

e ac

cess

(bor

e; u

nrel

iabl

e)

**

**

**

**

***

*

***

*

E

mer

genc

y in

stal

latio

n (d

roug

ht; p

ost d

roug

ht)

*

* **

* **

* **

*

* **

* **

*

**

* S

ettle

d (y

es; n

o)

*

*

*

*

**

Oth

ers

sett

led

1 (y

es; n

o)

*

***

*

*

***

**

**

* **

***

G

ende

r (m

ale;

fem

ale)

*

***

*

A

ge 2

(<30

; 30

& o

ver)

*

**

P

ositi

on 2

(pos

ition

; non

e)

*

**

**

* *

* **

* **

**

*

*

Pro

xim

ity to

wel

l (cl

ose;

dis

tant

)

*

*

*

**

W

ho re

pair

s 1(

loca

l; ex

tern

al)

**

*

**

***

***

**

**

**

* **

*

**

* *

***

* A

bilit

y(co

ded)

(ve

ry a

ble;

not

ver

y ab

le)

**

**

**

***

**

**

* *

*

*

**

Who

was

trai

ned

(mec

hani

cs; c

omm

ittee

) **

**

*

**

* **

**

**

***

**

* *

**

**

Whe

n tr

aine

d (b

efor

e/at

inst

alla

tion

; aft

er)

**

*

* *

*

*

*

Len

gth

of tr

aini

ng (<

1wk;

>1w

k)

**

*

**

**

**

**

*

Fol

low

up

trai

ning

(yes

; no)

**

*

L

evel

of

trai

ning

(sof

t; so

ft &

har

d)

**

**

**

*

***

*

O

wne

rshi

p1 (c

oded

) (c

omm

unity

; oth

er)

*

*

*

**

* **

*

* R

eque

sted

(yes

; no)

**

**

*

**

**

**

**

*

***

**

**

P

rese

nt (

yes;

no)

**

*

*

***

*

Siti

ng (c

omm

unity

; dri

ller

s)

**

*

*

**

**

*

**

**

***

**

Siti

ng 1

(cod

ed)

(int

erna

l; ex

tern

al)

**

*

*

*

**

*

* **

*

**

P

erce

ntag

e of

wom

en (<

50%

; 50

**

*

**

Ele

ctio

n 2

(cod

ed)

(com

mun

ity;

oth

ers)

*

**

**

* **

* **

***

**

***

**

San

ctio

n (y

es; n

o)

***

*

***

**

***

**

**

*

**

*

* **

R

ules

(yes

; no)

**

*

* *

**

***

***

***

**

***

**

Com

mun

ity in

volv

emen

t (al

way

s; n

ever

)

**

**

*

***

***

***

**

***

***

**

* **

**

P

rior

Con

trib

utio

n (y

es; n

o)

**

**

**

**

**

*

***

**

Par

ticip

ated

in in

stal

lati

on (y

es; n

o)

**

*

* **

*

**

***

***

**

P

ump

dept

h pd

1 (2

9m; 2

9m &

ove

r)

***

***

*

*

P

ump

type

(Afr

idev

; Vol

anta

) **

**

* **

* **

*

* **

*

**

***

**

*

Pum

p pr

oble

m (s

oftw

are;

har

dwar

e)

**

**

*

**

**

* *

**

peop

le se

rved

per

buc

ket (

>1.

5ppb

; 1.5

ppb

& l

ess)

**

**

queu

ing

tim

e (>

2 hr

s; 2

hrs

& u

nder

)

*

**

* P

artic

ipat

e in

mai

nten

ance

(yes

; no)

**

*

* **

**

* *

**

**

**

*


Table 14-2 Well Performance Indicators

Operational Status of

pump

Frequency of breakage

Average down-time when broke

Observed status of well

head. Pressure (high; low) 0.005 0.102 0.284 0.911 Alternative source (bore; unreliable) 0.003 0.230 0.001 0.106 year of install ation (pre ’95; ’95 on) 0.161 0.957 0.000 0.256 Quality of water (sweet; salty) 0.001 0.192 0.041 0.852 Other borehole access (bore; unreliable) 0.003 0.230 0.001 0.106 Emergency install ation (drought; post drought) 0.225 0.012 0.044 0.000 Settled (yes; no) 0.957 0.090 0.024 0.739 Others settled 1 (yes; no) 0.032 0.119 0.518 0.745 Gender (male; female) 0.187 0.060 0.435 0.916 Age 2 (<30; 30 & over) 0.851 0.335 0.484 0.416 Position 2 (position; none) 0.805 0.719 0.277 0.032 Proximity to well (close; distant) 0.462 0.828 0.087 0.184 Who repairs 1(local; external) 0.123 0.001 0.022 0.008 Abili ty(coded) (very able; not very able) 0.010 0.739 0.157 0.010 Who was trained (mechanics; committee) 0.008 0.009 0.028 0.370 When trained (before/at installation; after) 0.896 0.660 0.158 0.089 Length of training (<1wk; >1wk) 0.062 0.164 0.007 0.251 Follow up training (yes; no) 0.892 0.097 0.895 0.466 Level of training (soft; soft & hard) 0.245 0.173 0.850 0.458 Ownership1 (coded) (community; other) 0.158 0.202 0.799 0.016 Requested (yes; no) 0.908 0.001 0.075 0.162 Present (yes; no) 0.532 0.200 0.001 0.702 Siting (community; drillers) 0.805 0.000 0.946 0.314 Siting 1 (coded) (internal; external) 0.505 0.001 0.676 0.044 Percentage of women (<50%; 50 0.798 0.859 0.747 0.367 Election 2 (coded) (community; others) 0.044 0.816 0.650 0.830 Sanction (yes; no) 0.277 0.404 0.751 0.775 Rules (yes; no) 0.440 0.003 0.213 0.029 Community involvement (always; never) 0.119 0.114 0.406 0.701 Prior Contribution (yes; no) 0.779 0.007 0.288 0.474 Participated in installation (yes; no) 0.058 0.055 0.001 0.176 Pump depth pd1 (29m; 29m & over) 0.000 0.000 0.861 0.540 Pump type (Afridev; Volanta) 0.003 0.000 0.000 0.001 Pump problem (software; hardware) 0.776 0.384 0.323 0.003 people served per bucket (>1.5ppb; 1.5 ppb & less) 0.633 0.837 0.924 0.435 queuing time (>2 hrs; 2 hrs & under) 0.502 0.015 0.952 0.279 Participate in maintenance (yes; no) 0.220 0.654 0.932 0.898


Table 14-3. Competence Indicators

Who usually repairs pump

Number of trained pump

mechanics

Last agency pump

maintenance visit

Abili ty of local committee or trained pump maintenance

rep's Pressure (high; low) 0.848 0.066 0.589 0.431 Alternative source (bore; unreliable) 0.001 0.637 0.889 0.002 year of install ation (pre ’95; ’95 on) 0.000 0.001 0.288 0.008 Quality of water (sweet; salty) 0.486 0.008 0.920 0.264 Other borehole access (bore; unreliable) 0.001 0.637 0.889 0.002 Emergency install ation (drought; post drought) 0.000 0.001 0.153 0.035 Settled (yes; no) 0.779 0.904 0.029 0.375 Others settled 1 (yes; no) 0.932 0.137 0.000 0.038 Gender (male; female) 0.170 0.407 0.595 0.015 Age 2 (<30; 30 & over) 0.617 0.980 0.111 0.101 Position 2 (position; none) 0.545 0.280 0.002 0.514 Proximity to well (close; distant) 0.614 0.034 0.086 0.024 Who repairs 1(local; external) 0.000 0.881 0.079 0.000 Abili ty(coded) (very able; not very able) 0.001 0.830 0.059 0.000 Who was trained (mechanics; committee) 0.000 0.003 0.001 0.561 When trained (before/at installation; after) 0.194 0.001 0.047 0.741 Length of training (<1wk; >1wk) 0.302 0.171 0.833 0.039 Follow up training (yes; no) 0.130 0.136 0.242 0.540 Level of training (soft; soft & hard) 0.058 0.296 0.001 0.092 Ownership1 (coded) (community; other) 0.273 0.508 0.893 0.189 Requested (yes; no) 0.663 0.985 0.008 0.398 Present (yes; no) 0.476 0.084 0.640 0.888 Siting (community; drillers) 0.864 0.714 0.029 0.057 Siting 1 (coded) (internal; external) 0.080 0.539 0.128 0.113 Percentage of women (<50%; 50 0.075 0.202 0.322 0.701 Election 2 (coded) (community; others) 0.284 0.133 0.257 0.655 Sanction (yes; no) 0.533 0.737 0.000 0.033 Rules (yes; no) 0.595 1.000 0.041 0.011 Community involvement (always; never) 0.198 0.952 0.003 0.119 Prior Contribution (yes; no) 0.004 0.001 0.183 0.178 Participated in installation (yes; no) 0.043 0.320 0.103 0.176 Pump depth pd1 (29m; 29m & over) 0.838 0.775 0.763 0.751 Pump type (Afridev; Volanta) 0.015 0.028 0.000 0.342 Pump problem (software; hardware) 0.001 0.774 0.057 0.566 people served per bucket (>1.5ppb; 1.5 ppb & less) 0.073 0.172 0.698 0.079 queuing time (>2 hrs; 2 hrs & under) 0.217 0.469 0.675 0.220 Participate in maintenance (yes; no) 0.000 0.627 0.027 0.003

Gam

os L

td

Fina

l Tec

hnic

al R

epor

t- “

Exi

t St

rate

gies

” R

7136

P

age

121

TA

BLE

14-

4. O

rgan

isat

iona

l Ind

icat

ors

W

ater

co

mm

ittee

ap

poin

ted

Num

ber

of

mem

bers

pe

rcen

tage

of

wom

en

Ele

ctio

n of

m

embe

rs

Fre

quen

cy o

f c’

ttee

mee

tings

Pum

p M

aint

enan

ce

paym

ent

Pay

men

t of

quot

a Sa

nctio

n fo

r no

ne

paym

ent

Rul

es re

P

ump

use

Aw

aren

ess

of

inte

rvie

wee

Pr

essu

re (h

igh;

low

) 0.

675

0.52

9 0.

884

0.44

8 0.

256

0.31

7 0.

647

0.18

0 0.

700

0.00

2 A

ltern

ativ

e so

urce

(bor

e; u

nrel

iabl

e)

0.00

0 0.

173

0.75

1 0.

042

0.06

9 0.

000

0.30

0 0.

385

0.54

0 0.

532

year

of

inst

alla

tion

(pre

’95;

’95

on)

0.

034

0.22

3 0.

269

0.00

0 0.

000

0.70

3 0.

006

0.00

7 0.

012

0.37

5 Q

ualit

y of

wat

er (

swee

t; sa

lty)

0.08

3 0.

773

0.91

1 0.

002

0.66

7 0.

222

0.84

8 0.

037

0.11

7 0.

940

Oth

er b

oreh

ole

acce

ss (b

ore;

unr

elia

ble)

0.

000

0.17

3 0.

751

0.04

2 0.

069

0.00

0 0.

300

0.38

5 0.

540

0.53

2 E

mer

genc

y in

stal

latio

n (d

roug

ht; p

ost d

roug

ht)

0.29

6 0.

568

0.06

5 0.

046

0.01

0 0.

353

0.06

2 0.

406

0.20

0 0.

640

Settl

ed (y

es; n

o)

0.65

5 0.

650

0.75

9 0.

065

0.94

0 0.

059

0.01

4 0.

097

0.26

9 0.

037

Oth

ers

settl

ed 1

(yes

; no)

0.

174

0.76

0 0.

612

0.04

6 0.

079

0.00

0 0.

004

0.00

1 0.

027

0.00

4 G

ende

r (m

ale;

fem

ale)

0.

352

0.09

4 0.

588

0.08

6 0.

145

0.12

0 0.

740

0.07

4 0.

935

0.00

0 A

ge 2

(<30

; 30

& o

ver)

0.

326

0.39

6 0.

029

0.38

0 0.

334

0.57

2 0.

260

0.20

2 0.

082

0.01

0 Po

sitio

n 2

(pos

ition

; non

e)

0.11

8 0.

144

0.33

7 0.

064

0.00

0 0.

011

0.04

3 0.

000

0.01

0 0.

000

Prox

imity

to w

ell (

clos

e; d

ista

nt)

0.43

6 0.

797

0.34

6 0.

845

0.62

4 0.

217

0.21

5 0.

634

0.98

9 0.

245

Who

repa

irs

1(lo

cal;

exte

rnal

) 0.

001

0.36

9 0.

678

1.00

0 0.

010

0.00

0 0.

000

0.11

5 0.

498

0.69

7 A

bilit

y(co

ded)

(ve

ry a

ble;

not

ver

y ab

le)

0.22

8 0.

059

0.26

2 0.

568

0.11

7 0.

002

0.00

0 0.

044

0.02

3 0.

962

Who

was

trai

ned

(mec

hani

cs; c

omm

ittee

) 0.

238

0.00

3 0.

956

0.09

1 0.

610

0.00

0 0.

283

0.00

0 0.

017

0.89

2 W

hen

trai

ned

(bef

ore/

at in

stal

latio

n; a

fter

) 0.

212

0.44

3 0.

617

0.02

0 0.

031

0.02

3 0.

867

0.44

2 0.

371

0.15

1 L

engt

h of

trai

ning

(<1w

k; >

1wk)

0.

444

0.01

0 0.

230

0.00

3 0.

295

0.33

0 0.

632

0.00

9 0.

054

0.42

2 Fo

llow

up

trai

ning

(yes

; no)

0.

481

0.08

3 0.

542

0.21

8 0.

159

0.00

9 0.

430

0.18

6 0.

049

0.30

0 L

evel

of

trai

ning

(sof

t; so

ft &

har

d)

0.82

8 0.

004

0.05

7 0.

158

0.82

1 0.

000

0.18

6 0.

000

0.07

0 0.

285

Ow

ners

hip1

(cod

ed)

(com

mun

ity; o

ther

) 0.

388

0.88

9 0.

728

0.14

5 0.

011

0.84

5 0.

436

0.03

9 0.

003

0.02

2 R

eque

sted

(yes

; no)

0.

024

0.55

4 0.

424

0.00

1 0.

395

0.00

7 0.

006

0.00

9 0.

020

0.33

0 Pr

esen

t (ye

s; n

o)

0.05

0 0.

679

0.09

3 0.

286

0.09

7 0.

055

0.13

5 0.

248

0.73

2 0.

020

Siti

ng (c

omm

unity

; dril

lers

) 0.

254

0.62

1 0.

212

0.05

7 0.

006

0.00

3 0.

020

0.00

1 0.

006

0.29

8 Si

ting

1 (c

oded

) (i

nter

nal;

exte

rnal

) 0.

229

0.37

4 0.

034

0.18

0 0.

061

0.03

7 0.

386

0.00

5 0.

011

0.97

3 Pe

rcen

tage

of

wom

en (<

50%

; 50

1.

000

0.39

9 0.

000

0.47

3 0.

496

0.52

7 0.

010

0.66

9 0.

415

0.88

8 E

lect

ion

2 (c

oded

) (c

omm

unity

; oth

ers)

0.

010

0.46

8 0.

429

0.00

0 0.

000

0.00

8 0.

753

0.00

0 0.

003

0.21

6 Sa

nctio

n (y

es; n

o)

0.87

3 0.

936

0.56

0 0.

000

0.00

4 0.

000

0.00

4 1.

000

0.00

0 0.

121

Rul

es (y

es; n

o)

0.09

0 0.

111

0.36

6 0.

004

0.14

8 0.

000

0.00

0 0.

000

1.00

0 0.

889

Com

mun

ity in

volv

emen

t (al

way

s; n

ever

) 0.

006

0.04

4 0.

147

0.00

0 0.

000

0.00

0 0.

001

0.00

0 0.

000

0.12

3 Pr

ior

Con

trib

utio

n (y

es; n

o)

0.64

7 0.

400

0.50

8 0.

002

0.06

6 0.

146

0.23

7 0.

007

0.07

4 0.

739

Part

icip

ated

in in

stal

latio

n (y

es; n

o)

0.77

7 0.

915

0.12

3 0.

024

0.00

0 0.

875

0.39

6 0.

054

0.32

7 0.

001

Pum

p de

pth

pd1

(29m

; 29m

& o

ver)

0.

268

0.02

1 0.

571

0.20

2 0.

285

0.60

1 0.

155

0.22

6 0.

126

0.94

6 Pu

mp

type

(Afr

idev

; Vol

anta

) 0.

148

0.06

0 0.

079

0.05

1 0.

009

0.00

0 0.

094

0.00

6 0.

204

0.59

4 Pu

mp

prob

lem

(sof

twar

e; h

ardw

are)

0.

894

0.04

2 0.

363

0.51

2 0.

009

0.00

3 0.

201

0.35

1 0.

905

0.15

7 pe

ople

serv

ed p

er b

ucke

t (>1

.5pp

b; 1

.5 p

pb &

les

s)

0.00

8 0.

472

0.37

4 0.

788

0.45

4 0.

002

0.51

1 0.

106

0.39

6 0.

949

queu

ing

time

(>2

hrs;

2 h

rs &

und

er)

0.09

6 0.

962

0.29

1 0.

790

0.25

9 0.

158

0.33

0 0.

716

0.42

7 0.

806

Part

icip

ate

in m

aint

enan

ce (y

es; n

o)

0.24

0 0.

464

0.45

1 0.

054

0.00

6 0.

721

0.04

9 0.

014

0.00

1 0.

159


Table 14-5. Community Participation Indicators

Ownership of well

Community involvement

in water committee

Participated in

installation

Participate in

maintenance

Prior contribution (pre drilling

/ pump) Pressure (high; low) 0.821 0.521 0.068 0.700 0.349 Alternative source (bore; unreliable) 0.658 0.835 0.011 0.122 0.068 year of install ation (pre ’95; ’95 on) 0.014 0.000 0.006 0.002 0.000 Quality of water (sweet; salty) 0.298 0.002 0.038 0.030 0.000 Other borehole access (bore; unreliable) 0.658 0.835 0.011 0.122 0.068 Emergency installation (drought; post drought) 0.019 0.001 0.016 0.053 0.000 Settled (yes; no) 0.482 0.002 0.145 0.863 0.823 Others settled 1 (yes; no) 0.640 0.000 0.534 0.329 0.837 Gender (male; female) 0.080 0.836 0.027 0.317 0.269 Age 2 (<30; 30 & over) 0.324 0.136 0.758 0.645 0.244 Position 2 (position; none) 0.458 0.291 0.015 0.163 0.621 Proximity to well (close; distant) 0.050 0.277 0.831 0.982 0.003 Who repairs 1(local; external) 0.061 0.000 0.026 0.000 0.038 Abili ty(coded) (very able; not very able) 0.085 0.041 0.130 0.007 0.126 Who was trained (mechanics; committee) 0.186 0.005 0.990 0.345 0.005 When trained (before/at installation; after) 0.094 0.017 0.058 0.861 0.012 Length of training (<1wk; >1wk) 0.944 0.008 0.188 0.375 0.029 Follow up training (yes; no) 0.474 0.531 0.771 0.670 1.000 Level of training (soft; soft & hard) 0.362 0.025 1.000 0.884 1.000 Ownership1 (coded) (community; other) 0.000 0.074 0.089 0.051 0.016 Requested (yes; no) 0.078 0.000 0.001 0.110 0.002 Present (yes; no) 0.183 0.942 0.000 0.034 0.127 Siting (community; drillers) 0.233 0.000 0.513 0.002 1.000 Siting 1 (coded) (internal; external) 0.045 0.000 0.373 0.004 0.277 Percentage of women (<50%; 50 0.951 0.066 0.404 0.180 0.523 Election 2 (coded) (community; others) 0.124 0.000 0.302 0.243 0.001 Sanction (yes; no) 0.063 0.000 0.054 0.014 0.007 Rules (yes; no) 0.002 0.000 0.327 0.001 0.074 Community involvement (always; never) 0.899 1.000 0.000 0.001 0.002 Prior Contribution (yes; no) 0.031 0.000 0.001 0.123 1.000 Participated in installation (yes; no) 0.082 0.000 0.000 0.001 1.000 Pump depth pd1 (29m; 29m & over) 0.344 0.758 0.041 0.155 0.065 Pump type (Afridev; Volanta) 0.438 0.753 0.135 0.621 0.015 Pump problem (software; hardware) 0.809 0.105 0.011 0.018 0.004 people served per bucket (>1.5ppb; 1.5 ppb & less) 0.177 0.146 0.792 0.963 0.966 queuing time (>2 hrs; 2 hrs & under) 0.007 0.663 0.913 0.185 0.037 Participate in maintenance (yes; no) 0.058 0.002 0.000 0.123

Exit Strategies for the Resettlement of Drought Prone Populations: Technical Report

Documents

Transcript of Exit Strategies for the Resettlement of Drought Prone Populations: Technical Report