Content uploaded by Sara J. Marks
Author content
All content in this area was uploaded by Sara J. Marks on Aug 10, 2017
Content may be subject to copyright.
7th RWSN Forum “Water for Everyone”
7ème Forum RWSN « L’eau pour tous »
29 Nov - 02 Dec 2016, Abidjan, Côte d’Ivoire
PEER REVIEWED
Page 1
Measuring the impact of multiple-use
water services in Tanzania and Burkina
Faso: water service quality, nutrition, and
health
Type: Short Paper
Authors: Sara Marks1*, Ariane Schertenleib1, Honoré Biaou2, Kees Vogt2, Boukari Salifou2, Muganyizi
Ndyamukama2, Ilyasse Kabore2, Emily van Houweling3, Vincent G. Vyamana4, Vivienne Abbott5, Mary
Renwick5
1Eawag/Sandec, Switzerland
2Winrock International, USA
3Virginia Tech, USA
4iWASH/GLOWS Tanzania, USA
5Consultant, USA
*Senior Scientist, Eawag / Sandec, Switzerland, sara.marks@eawag.ch, +41 58 765 5631
Abstract/Summary
Multiple-use water services (MUS) is an integrated service delivery approach that takes into account
households' full range of water needs. Past studies have shown the benefits of MUS in terms of enhanc-
ing income and livelihoods diversification. However, little is known about whether MUS is associated
with improved health, nutrition, or water service quality. We used a matched control design dietary diver-
sity among rural households receiving MUS through two large-scale water supply programs in Burkina
Faso and Tanzania. Data was collected from 2,704 households representing five MUS typologies and a
control group. Key informant interviews, focus group discussions, and water quality testing were also
conducted. Comparisons across different MUS household typologies and the control group reveal a con-
sistent positive trend regarding the benefits of MUS one to four years after project implementation.
Households receiving MUS have experienced fewer injuries, enhanced food security, and use more relia-
ble and safe water sources. These results contribute to a growing global evidence base regarding the varie-
ty of benefits associated with higher levels of water services in rural communities.
Introduction
Multiple-use water services (MUS) is an integrated water service delivery approach that takes into account
a range of household’ needs when planning, financing and managing water services. The MUS model
explicitly acknowledges households’ tendancy to re-allocate water intended for one type of activity to
another, such as domestic water supplies being used for livestock watering or irrigation channels being
accessed for drinking needs. In this way, MUS recognizes a more holistic approach that protects water
systems from overuse, while simultaneously supporting synergistic uses of water at the household- and
community-level. In practice, MUS typically delivers a higher level of water service, with at least 50 liter
per capita per day (LPCD) available on household premises.
Past studies have shown the benefits of MUS in terms of enhancing water-based income generation
(Crow, Swallow, & Asamba, 2011; Noel, Hoang, Soussan, & Lovett, 2010), especially in the presence of
enabling factors, such as markets and electricity (Davis, Hope, & Marks, 2011). In rural Senegal,
7th RWSN Forum “Water for Everyone”
7ème Forum RWSN « L’eau pour tous »
29 Nov - 02 Dec 2016, Abidjan, Côte d’Ivoire
PEER REVIEWED
Page 2
productive uses of water were linked to livelihood diversification among women (van Houweling, Hall,
Diop, & Davis, 2012) and improved technical operation of water systems (Hall, Vance, & van Houweling,
2014).
Other potential benefits of MUS beyond income and livelihoods are less understood. Fully documenting
such benefits is necessary to justify the often higher upfront investment required to establish the higher
level of service offered by MUS.
Context, aims and activities undertaken
Study Objective: The study’s objective was to rigorously assess a range of impacts expected to arise from
MUS, including improvements in child health, safety during water collection, food security and nutritional
status [2]. Sandec’s Water Supply and Treatment group collaborated with United States Agency for
International Development (USAID) and the Global Water for Sustainability (GLOWS) program,
including partner organizations Winrock International, Virginia Tech, and Florida International
University, to systematically evaluate the MUS component of two rural water supply programs in Africa.
Background: The first program, called the West Africa Water, Sanitation and Hygiene (WA-WASH)
program in Burkina Faso (2011 to 2015), offered households the option to invest in subsidized self-
supply (upgraded private wells equipped with rope pumps), along with other program activities. The
second program, called the Integrated Water, Sanitation and Hygiene (iWASH) in Tanzania (2010 –
2015), used a demand-led approach to engage community members during the installation of new or
upgraded communal water supply systems (reticulated networks, upgraded community wells with rope
pumps, and/or livestock troughs). Both programs featured “impact boosting activities” that were tailored
to local conditions and designed to mazimize the systems’ potential for productive use. These activities
included seed distribution networks, market garden demonstrations, support for improved poultry
housing (kinengunengu) and livestock husbandry (Figure 1).
Figure 1. Impact boosting activities within the WA-WASH and iWASH programs
7th RWSN Forum “Water for Everyone”
7ème Forum RWSN « L’eau pour tous »
29 Nov - 02 Dec 2016, Abidjan, Côte d’Ivoire
PEER REVIEWED
Page 3
Study Design: Baseline (pre-intervention) data on outcomes of interest for this study were not collected
prior to the launch of the programs. Thus, it was not possible to directly measure the before-after status
of households receiving MUS. We instead relies on a two-step strategy to estimate impacts: (1)
randomized sampling of various household typologies in WA-WASH and iWASH intervention
communities, as well as a control group, and (2) statistical matching techniques. This article reports results
for the first step only.
Communities that had participated in the iWASH and WA-WASH programs at least one year prior to the
study were eligible for enrolment into the treatment group. Communities that were located within the
program service regions and were pre-qualified for participation (but had not yet done so) were eligible
for enrolment in the control group. From this eligible pool, communities were purposively selected to
optimise variation in the water supply and impact boosting intervention(s) received.
Based on community visits and discussions held with field staff, the study team pre-defined and randomly
sampled several household typologies. Within WA-WASH communities, household typologies include:
(a) investors, (b) neighbours of investors (those who did not invest but who are accessing an investor’s
upgraded well), and (c) non-neighbours (those who did not invest in and do not use an investor’s
upgraded well, i.e., a pseudo-control within treatment communities). Within iWASH communities,
household typologies include: (d) members of MUS interest groups and (e) non-members of a MUS
interest group. Finally, households randomly selected from control communities (as described above)
were defined as: (f) control households.
Data Collection: Between May and October 2015, field teams conducted 2,704 interviews with heads of
households (men and women). Surveys probed on the water sources used throughout the year for any
purpose, health status of adults and children, self-reported food security, dietary diversity, and other
measures of well-being. In addition, semi-structured interviews were held with a key informant in each
community (typically the village chief) to estimate population size, proximity to markets, and other
community-level measures. Focus group discussions were held with men and women (both separately and
mixed) of all household typologies in Burkina Faso to better understand the changes experienced since
community members had participated in the WA-WASH program. Finally, fecal contamination of
households’ main drinking water sources was assessed using compartment bag test (CBT) kits in Burkina
Faso (n = 181) and hydrogen sulfide (H2S) vials in Tanzania (n = 35). Table 1 summarizes household
typologies and sample sizes.
7th RWSN Forum “Water for Everyone”
7ème Forum RWSN « L’eau pour tous »
29 Nov - 02 Dec 2016, Abidjan, Côte d’Ivoire
PEER REVIEWED
Page 4
Category
Household
typology
Household
interviews
Focus
group
discussions
Key
informant
interviews
Microbial
water
quality
WA-WASH
(19
communities)
a. MUS investors
146
8
19
59
b. Neighbours
292
0
c. Non-
neighbours
451
59
Control –
Burkina Faso
(9
communities)
f. Control
438
0
9
63
iWASH
(7
communities)
d. MUS interest
group members
322
0
7
35
e. Non-members
410
0
Control -
Tanzania
(3
communities)
f. Control
645
0
3
0
Total sample size
2,704
8
38
216
Table 1. Household typologies and sample sizes for the WA-WASH and iWASH programs.
Main results and lessons learnt
Illness and Injuries. As compared to control households, MUS households reported experiencing fewer
instances of children experiencing diarrheal episodes in the past week, as well as fewer injuries
experienced by women while fetching water. In bivariate tests, only the difference in the rate of injuries
among households in the iWASH program (3%) and control communities (12%) was statistically
significant (p < 0.05). Other health measures were not found to be significantly different across MUS and
control groups.
Food Security. The survey asked respondents to rate their household’s food security in the past year as
being very secure, somewhat secure, or insecure. Interviewers explained the concept of food security in
the local language. Results show that whereas food insecurity existed to some extent in all communities,
the share of households identifying as insecure was significantly lower within communities receiving
MUS. For example, within iWASH communities, 84% of MUS interest group members identified as
“very secure”, as compared to 65% of non-members and 53% of control households reporting the
same.
Nutritional Status. The household survey included a standardized set of questions designed to assess the
overall nutritional status of women of reproductive age (FAO 2014). Three measures were analysed: (1)
the total number of food types consumed in the past week, (2) consumption of animal products (meat,
7th RWSN Forum “Water for Everyone”
7ème Forum RWSN « L’eau pour tous »
29 Nov - 02 Dec 2016, Abidjan, Côte d’Ivoire
PEER REVIEWED
Page 5
milk and eggs), and (3) consumption of leafy green vegetables. Statistical comparisons revealed that
overall dietary diversity was slightly but significantly improved only among households participating in the
iWASH program (6.4 food types), as compared to control households (5.4 food types). In both programs,
households receiving MUS were more likely to have consumed animal products in the past week, as
compared to control households. For example, 92% of WA-WASH investors and 91% of iWASH
interest group members had consumed meat, milk or eggs in the past week, as compared to only 82% and
77% of control households, respectively (Figure 2).
Figure 2. Share of households consuming leafy green vegetables and
animal products within the past week.
Water Service Quality: We report preliminary results here for the WA-WASH program only. A greater share
of MUS households (34%) reported waiting at some point during the year for their main drinking water
source (traditional or improved wells) to recharge, as compared to control households (19%). However,
MUS households typically waited for the water to return in the well for half the amount time as compared
to control households (median of 60 minutes and 120 minutes, respectively). In terms of technical
breakdowns, MUS households experienced fewer (16%) full day interruptions in water service, as
compared to non-MUS households (23%). Water quality testing revealed that the majority (69%) of
upgraded wells with rope pumps provided water categorized by the WHO as low risk (< 10 CFU/ 100
ml). By contrast, nearly all traditional wells (93%) were categorized as very high risk. A higher share of
improved community wells (84%) had no detectable E.coli as compared to improved wells on premises
(54%). Overall, we find most (75%) MUS investors are accessing a source categorized as safe or probably
safe, whereas only 47% are doing the same in the control group.
Conclusions and Recommendations
Our analysis of the impact of two large-scale MUS programs in Africa is limited due to the lack of
statistical controls for potential confounding factors. Nonetheless, through strategic enrolment and
randomized sampling across various household typologies, direct comparisons reveal a consistent positive
trend demonstrating the benefits of MUS one to four years after implementation of the WA-WASH and
iWASH programs. Households receiving MUS had experienced fewer injuries while fetching water, were
7th RWSN Forum “Water for Everyone”
7ème Forum RWSN « L’eau pour tous »
29 Nov - 02 Dec 2016, Abidjan, Côte d’Ivoire
PEER REVIEWED
Page 6
more food secure, and were more likely to be consuming animal products. MUS systems were more
reliable overall and delivered higher quality water. However achieving WHO standards for microbial
safety for all remained elusive. These results expand the growing evidence base regarding the benefits of
higher levels of water services in rural communities globally. Planned future analyses include using
multivariate modeling to further control for sources of bias, investigating potential spillover effects of
MUS among neighbouring households, and estimating the return on investment for MUS projects.
Acknowledgements
Data for this study were sourced from USAID and the GLOWS program, with funding support from
Winrock International, Florida International University, and Eawag. Data are publicly available and
subject to USAID regulations.
References
Crow, B., Swallow, B., & Asamba, I. (2011): Community organized household water increases not only
rural incomes, but also men’s work. World Development, 40(3), 528–541.
Davis, J., Hope, R. A., & Marks, S. J. (2011): Assessing the link between productive use of domestic
water, poverty reduction and sustainability: Kenya country report. Report to the World Bank Water
and Sanitation Program. Stanford University, Oxford University.
Hall, R. P., Vance, E. A., & van Houweling, E. (2014): The Productive Use of Rural Piped Water in
Senegal. Water Alternatives, 7(3), 480–498.
Noel, S., Hoang, T., Soussan, J., & Lovett, J. C. (2010). The impact of domestic water on household
enterprises: evidence from Vietnam. Water Policy, 12(2), 237–247.
van Houweling, E., Hall, R. P., Diop, A. S., & Davis, J. (2012): The Role of Productive Water Use in
Women’s Livelihoods: Evidence from Rural Senegal. Water Alternatives, 5(3), 658–677.
Contact Details
Name of Lead Author: Sara Marks
Email: sara.marks@eawag.ch
Name of Second Author: Ariane Schertenleib
Email: ariane.schertenleib@eawag.ch
Name of Last Author: Mary Renwick
Email: maryerenwick@gmail.com