High-Tech meets Low-Resource: Applying Innovation in a Global Health Context (part 1)
- 1. m-maji (mmaji.wordpress.com) Sunny Jeon, PhD Candidate, Political Science, Stanford Joshua Cohen, Professor of Law, Philosophy, and Political Science, Stanford Eran Bendavid, Assistant Professor of Medicine, Stanford Katherine Hoffmann, MA, International Policy Studies, Stanford (now at IPA) Terry Winograd, Professor of Computer Science, Stanford
- 2. Water in Kibera is scarce, costly, and contaminated. Kibera, Nairobi, Kenya
- 3. Water in Kibera is scarce, costly, and contaminated. Kibera, Nairobi, Kenya Population estimates range from 170,070 to 1 million. No formal water or sanitation provided—85% rely on vendors/kiosks. On good days, it takes just under an hour to find water. On bad days, the search can take all day. Water shortages hit approximately 4x a month, during which time price of water increases tenfold. Infant mortality in Kibera more than three times the average of Nairobi as a whole.
- 5. Kibera lacks consistent access to water, but not to mobile phones …
- 6. Kibera lacks consistent access to water, but not to mobile phones … Source: Jack and Suri (2009) Source: CCK, Operators’ returns Phone Use in Kenya Average Mobile Tariffs per Minute
- 7. Kibera lacks consistent access to water, but not to mobile phones … Source: M-Maji survey of 1,000 households in Kibera (February 2011)
- 8. M- Maji Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back Use ubiquitous mobile technology to improve access to clean water Buyers Vendors
- 9. Vendors Advertise M- Maji Use USSD to notify M-Maji of water availability, location, price, and quality . Listings updated to reflect current information. Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back Buyers Vendors
- 10. Buyers Query M- Maji Use USSD to find the closest , cheapest , and cleanest water. Water information verified by M-Maji staff on the ground. Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back Buyers Vendors
- 11. Buyers Rate M- Maji Buyers can provide feedback and file complaints about vendors who misrepresent water availability, price, or quality. M-Maji maintains “vendor ratings” and alerts future buyers of any negative histories. Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back Buyers Vendors
- 12. M- Maji Better information will: 1. Reduce water search costs (time, money) 2. Improve clean water uptake 3. Encourage vendor accountability Buyers Vendors
- 13. suggestive evidence Mobile phones can reduce search costs and improve market efficiency (Aker 2010; Jensen 2007). Informing households of water contamination can improve clean water uptake (Jalan and Somanathan 2008; Luoto 2009; Madajewicz et al. 2007). The poor are willing to pay slightly more “with their feet” for clean water (Kremer et al. 2009).
- 14. M-Maji Pilot (2011-2012) Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back pilot area 6 month pilot in 5 villages (Gatwekera, Kambi Muru, Kisumu Ndogo, Makina, and Mashimoni). Led by Umande on the ground. Responsibilities include marketing M-Maji services and manually enroll vendors. Obtain user feedback and finalize design choices.
- 16. M-Maji Pilot (2011-2012) Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back pilot area 6 month pilot in 5 villages (Gatwekera, Kambi Muru, Kisumu Ndogo, Makina, and Mashimoni). Led by Umande on the ground. Responsibilities include marketing M-Maji services and manually enroll vendors. Obtain user feedback and finalize design choices.
- 19. M-Maji Pilot (2011-2012) Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back 6 month pilot in 5 villages (Gatwekera, Kambi Muru, Kisumu Ndogo, Makina, and Mashimoni). Led by Umande on the ground. Responsibilities include marketing M-Maji services and manually enroll vendors. Obtain user feedback and finalize design choices. pilot area
- 20. M-Maji Pilot (2011-2012) Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back Conduct randomized impact evaluation. pilot area
- 21. Randomized Impact Evaluation 500 households recruited Visited 3-5 times over 6 month period Week 1: Baseline survey (household survey, water quality test, GPS/price survey) Weeks 3-4: All households receive quality test results, water ed. course, and GPS/Price survey. Random selection encouraged to use M-Maji. Weeks 12, 24: Return to examine changes in access to clean water and household water quality. 1. Lower Water Prices? 2. Lower Search Costs? 3. Household Water Quality?
- 22. M-Maji Pilot (2011-2012) Welcome to M-Maji! 1. Request Water 2. Sell Water 3. Complain OK < > Back Conduct randomized impact evaluation. M-Maji evaluated on: pilot area - Water prices - Search costs - Clean water uptake
- 23. Negotiate better contracts with local telecoms. Obtain permission from water company to conduct quality tests. Develop model for sustainability. 3 key challenges
- 24. 3 Villages, 20,000 households 10% of households adopt M-Maji 1/week, 150 vendors use 3/week 9,800 USSD sessions/month and 8,000 SMS/month On YU: KSH 13,800 = $172/month On SAFARICOM: KSH 70,800 = $885/month Both Networks: $363/month service charge
- 25. Negotiate better contracts with local telecoms. Obtain permission from water company to conduct quality tests. Develop model for sustainability. 3 key challenges
- 26. Special Thanks Leann Bankoski Prof. Michele Barry Aidah Binale Prof. Joshua Cohen Prof. Ben Crow Prof. Rob Daines Prof. Larry Diamond Erica Estrada Wendy Ju Sam Kitonyi Kelvin Bai Lugaka Prof. Edward Miguel Hilda Moraa Sally Madsen Caleb Ogara Josiah Omotto Prof. Dan Orwa Jackie Sargent Prof. Peter Waiganjo Prof. Terry Winograd and our amazing collaborators at Umande Trust and the University of Nairobi More details available at: http://mmaji.wordpress.com
Editor's Notes
- #25: Safaricom 4 hops = Ksh 6 per USSD, Ksh 1.5 per SMS; means 58,800 + 12,000 =