Hoffmann and Jeon on using ICT for clean water in Kibera

The February 10 Liberation Technology seminar titled, Can ICT Improve Clean Water Delivery Systems in Slums? Lessons from Kibera was led by two Stanford students, Katherine Hoffman, M.A. Candidate in International Policy Studies and Global Health together with Sunny Jeon, PhD candidate in Political Science. Hoffman and Jeon presented on the topic of the M-Maji system, a start-up non-profit project that uses mobile phones to empower communities with better information about water availability, price, and quality. M-Maji emerged from the Designing Liberation Technologies course taught at the Stanford d.school, which is dedicated to using mobile phone technology for health improvement in Kibera.  

The M-Maji project was conceived to confront a specific need in one of Africa's largest informal settlements: Nairobi's Kibera slum. Kibera holds about one quarter of Nairobi's population, but lacks a formal water and sanitation system. In short, water in Kibera is scarce, costly, uncertain, and often contaminated. The cost of water can rise up to over $3 per square meter during a shortage, and often takes up 20% of residents' income. Water is uncertain because about 40% of vendors in Kibera do not have access to a constant supply of water, especially since many of the connections that they do use are illegal. Additionally, water is often contaminated because water pipes run through sewer areas and are often plastic, since metal pipes would be taken as scrap. The result of these conditions is that the average person spends about 55 minutes per day looking for water, and about 68% of residents use informal water kiosks to access water.

M-Maji is a mobile platform designed to address problems facing both the seller and buyer of water in Kibera. Sellers report information about water availability and price at their water kiosk, enabling them to attract customers. Meanwhile, buyers access the system to find information about where water is available, at what price and at what level of quality. A third feature of the platform enables complaints and feedback regarding water sources. By coordinating and centralizing water information from multiple sources, M-Maji is designed to empower residents with better information about water availability, price, and quality, which ultimately helps to improve access to clean water.

There are several key reasons why a mobile platform that tackles the information side of the problem may be effective in solving the water problem in Kibera:

1. The water market is distorted, but not coordinated. Also, there is significant variation between neighborhoods.
2. Information asymmetry exists that is not solved by other sources or tools.
3. M-maji is needs-driven.
4. Relative to other interventions, M-Maji is cost-sensitive and low tech.

Although fieldwork conducted by the M-Maji team has indicated great potential for the platform's success, the team has also faced many challenges in implementation. One example stems from the platform's use of USSD sessions. Like SMS, these sessions allow short messages (up to 182 characters) to be exchanged. Users must enter a short code to begin a session, which then lasts about 2 minutes. However, the team has faced significant difficulty in getting mobile operators to allow access to USSD (in Swahili), as well as to negotiate rate agreements with mobile operators. The involvement of gangs and government in the water market introduces another challenges. In some cases, gangs bribe utility people to restrict the supply to push up prices, which may hinder the eventual effectiveness of the M-Maji tool.

Following the formation of a partnership with an important local NGO in Kibera called Umande Trust, M-Maji is moving rapidly into the pilot stage. The team is currently working to carry out a randomized impact evaluation to obtain convincing estimates of M-Maji's cost-effectiveness and impact on water outcomes. With the help of the community partner, the team is running a six-month long pilot on the West side of Kibera. In addition to surveying 1000 households for the baseline survey, the team is also carrying out data collection on water prices and water quality. M-Maji will officially launch in May or June of 2011, and the team will continue to reassess the project to better meet user needs as the results of the pilot become available.