Turkana women in the Northern counties of Kenya have smart cards that use biometric data to identify their rights to direct cash transfers from the Hunger Safety Net Programme (HSNP) funded by DfID and AusAid.
I had two contrasting reactions when I learnt of this project covering the four poorest counties of Kenya (Wajir, Turkana, Marsabit and Mandera). On the one hand it seems entirely in keeping in the rise of e-Kenya, clever use of technology allied to contemporary methods of distributing aid. On the other hand I find it remarkable: mobile technologies, electronic money transfer, implementing a huge project across vast areas in some of the most remote parts of the world with people who have, until recently, had very little contact with digital technology. It’s worth remembering that mobile technology hardly figured in the high-profile (though limited output) G8 Digital Task Force in 2002, a mere 10 years ago.
Phase One of the programme involved an enormous data gathering exercise, targeting people with a high poverty ranking, those over 65 and others according to a dependency ratio. The second phase is concentrating on those who are ‘unemployed’, in the sense of not having a formal job with regular salary. That means, explained Izzy Birch, an advisor in the Ministry for Northern Kenya and other Arid Lands, data on livelihoods activity will be gathered, identifying those who herd camels or raise cattle, for example. Managing this data will be on the agenda of the new Drought Management Authority which is being set up to continue part of the Ministry’s work.
This particular development, and others like it, raise fascinating questions in relation to the issues of Open and Linked Data that we explored in the Nairobi Research to Impact Hackathon. As explained in earlier blogs, we worked with Development specialists to develop personas and scenarios that provided more detail on the context for the developer’s work. For example, we can imagine an extension worker in Northern Kenya, working with populations in Wajir county. If she has good, affordable – or paid for – Internet access and is confident in using online resources, she may know that it is possible to get alerts on a mobile phone of new research findings for families whose livelihoods depend on camel-herding. And she may have been motivated enough to find out how to set up her profile on a online research platforms such as the DfID funded research stored in the R4D database and the ELDIS site managed by IDS which triggers those alerts (using RSS feeds).
We know that this is something of an ideal scenario: many extension workers are constrained by expensive or unreliable Internet access, particularly if they wish to download the large documents typical of many research reports; many are not confident and skilled in identifying and accessing online resources; and the research abstracts and reports are often dense and written first for an academic context which makes it hard for many on-the-ground intermediaries to access and use the content. But for now, putting aside those constraints, the Hunger Safety Net data illustrates the kind of novel developments which have the potential to provide significant benefits for both consumers of Development research such as knowledge intermediaries and the research community.
Clearly, the data collected for the HSNP could not immediately be used for other purposes, except in the aggregate. To do so would be to transgress the most basic of Data Protection principles. But if the HSNP participants signed up to allow data about their livelihoods, for example, to be shared with other agencies, it could open the door to the kind of creative new mobile applications that could link development intermediaries together and with those they are trying to reach. This was well illustrated in the prize winning entries to the Open Data Hackathon.
MobyDev was the winner. This team focused on the case of Extension Workers, as did almost all the entrants. Their impressive prototype started with a small SMS app which enables basic information to be entered by farmers (name, type of crop, location, for example). This user profile can then be used to access particular categories of research data, which the extension worker could then make available to farmers, in an appropriate form, or use the data to inform the strategies developed with her colleagues.
LD-Connect was a close second. This team focused on adding a ‘social layer’ to a standard newsreader-like app. In this case the profile of the user – the extentions workers – is built up, partly by engaging directly with the user and partly from their search patterns. This profile could then be the basis for assisting in data searches or in feeds which push data based on what your friends are reading, liking or recommending, to complement the standard pull pattern of such applications.
Know2Ext won the third prize. This prototype used a system of feeds – the pull referred to above – to get information to extension workers. Interestingly, the scheme proposed to use the location feature of the phone to add further information to the farmer as s/he received information from the extension worker, such as information about agricultural input suppliers in the locality. Again, this would depend on farmer’s signing up to the service. It has a clear advantage that the link to commercial agents opens up revenue-generating possibilities.