At the GSMA NFC and Mobile Money Summit in Milan on 22-25 October, I presented some new data from one the research projects we are sponsoring at the Financial Services for the Poor team of the Bill and Melinda Gates Foundation. The research project is called the Financial Inclusion Tracking Survey (FITS) and consists of three in depth, nationally representative surveys tracking a panel of 3000 households each in Tanzania and Uganda and 5000 households in Pakistan over the next 3 years.
Here is the Prezi, which you can view online. (The presentation also includes data from a payments module we included on the Gallup World Poll in 11 African countries. Blog posts on that data can be found here, here, and here.)
The primary goal of the FITS surveys is to create a general purpose data set which would provide a “window” in to the market’s development for the field of financial inclusion as well as market participants, regulators, and other local stakeholders. The data also has the potential to capture the impact of mobile money on respondent households by using the roll out of mobile money agents as a “natural experiment.” The surveys will eventually have three waves per country, approximately one per year, where each wave can be sped up or slowed down in response to market events.
This project has as its predecessor and prototype the survey done by Tavneet Suri and Billy Jack in Kenya. Here is the paper describing the data from the project and the paper documenting the impact of M-PESA on risk sharing and related slides. They document significant welfare benefits to households who have mobile money in that they are better able to absorb serious economic shocks without reducing their consumption in critical areas like food, health care, and education. Tavneet Suri is collaborating with me the FITS project as well.
The FITS data has already generated insights into the workings of the market in Tanzania and Uganda (Pakistan data is still preliminary and pending.) Here is a sampling of results:
- Market penetration is significant and continues apace in Tanzania (35% of households have a mobile money user) and Uganda (21% of households), Pakistan is still getting started (~9%).
- FITS data shows Tanzania and Uganda to be two very different markets. Uganda has many more unregistered users and transactions are often mediated through friend’s accounts or conducted by agents on behalf of clients. In Tazania, where there are competing mobile money deployments, clients are almost always registered and often have accounts with multiple brands.
- Women can be a major segment, if properly engaged. Women are active in Tanzania (47% of users) and Uganda (38%) but much less so in Pakistan (we estimate only 5% of users are women.)
- The data helps map adoption pathways. Most people (71% in TZ, 68% in UG) learn about MM through media such as radio, billboards, or TV but most adopt due to recommendations from friends or family (often those who they would send or receive from.)
- Agent issues are still rife, despite usage growth. 80% of users in TZ and UG report agents were not open or had no liquidity during a recent visit.
- Mobile money has made in-roads but many households still use informal options. Tanzania shows the most progress where 62% of money transfers go through mobile money as opposed to 20% hand delivery and 10% sent with bus drivers.) In Uganda, only 31% of transfers go through mobile money. The below figure summarizes the landscape of formal and informal money transfer options in Uganda.
- The data shows major potential to help families with economic shocks. 46% of households in TZ, 45% in UG, 28% in PK suffered one or more major shocks in preceding 6 months. Death and illness as well as crop and livestock loss are major concerns. Remittances are a common coping strategy.
This data collection project is the first of its kind to collect rigorous, nationally representative data focused on understanding the workings of the mobile money markets in several countries. It will produce useful data that will push the field’s thinking going forward.
FITS data is updated regularly with new results from the in main survey waves as well as quarterly follow up phone based surveys. For reports, briefs, blog posts, and access to the data itself, check out the project webpage here: http://audiencescapes.com/fits