Preferential Software for the Poor

Millions of families who have never accessed the Internet, used landline telephones or even postal service have now begun using mobile phones. Increasingly governments, charities and tech companies see this new infrastructure as an opportunity to re-imagine the way that we deliver health care. While this field is young and riddled with disappointments, some mobile health, or mHealth initiatives have yielded impressive results. For example, inexpensive text messaging programs have improved quality of care for children with Malaria (study), increased adherence to HIV medication (study), and made community-based support as much as 137 times faster and four times less costly (study).

I wish that I could say such studies prove that communication technologies boost health outcomes, but sadly I cannot. Perhaps this field’s most inconvenient truth, observed again and again by practitioners and scholars, is that technologies that are effective in one setting often have very different effects in other settings. It is not difficult to see what makes mHealth so complex; the daily routines by which people make use of technology as well as the technologies themselves are diverse and continually changing. It is more difficult to determine how we might cope with unpredictability and make the most of mHealth opportunities. Currently popular prescriptions include selecting technology that is scalable, fostering collaboration and planning for financial sustainability from day one.

I would like to suggest that we begin by listening to the poor and marginalized individuals who stand to benefit from our work. Listening is a matter of such practical importance that I often think of it as the core competency of the team at Medic Mobile, the mHealth social enterprise I co-founded. We begin by making time for long conversations with people from each of the groups any new initiative might touch. No project moves forward until we have listened long enough to begin seeing the central strengths and constraints of each group from their point of view. Hearing of their daily concerns goes hand in hand with imagining alternative futures, and we always draw sketches of how our collaboration might change their community. By ‘we’ in this case I am also referring to the beneficiaries because sketching and brainstorming is collaborative wherever possible. When we sketch what a future service might look like, we consistently receive more detailed and often more critical feedback than through conversation alone. Making it easier for people all over the world to give feedback on our intentions forces us to listen not only more patiently but also more pragmatically. The number of times we have thrown away old sketches and drawn new ones, always based on feedback, is a reasonable gauge of how well we have listened.

Listening meaningfully is also a great intellectual challenge because appreciating the circumstances of any poor person involves understanding global poverty. The first time that a community health worker in Malawi told me that he had to walk 15 kilometers and spend a third of what he would earn in a typical day just to charge his phone, I understood that an mHealth initiative designed for him would need to reimburse for electricity and use phones which hold a charge for a week or two. Having now heard similar stories from dozens of health workers, it frustrates me how the mHealth community obsesses with fancy smartphones that need to be charged everyday. I have had similar conversations with facility-based staff who lament that the internet is slow, unreliable or absent from their clinics. These are the clinics most in need of support, and yet nothing is more in fashion than to boast that one’s mHealth application is “cloud-based.” International development agencies often write requests for proposals that explicitly call for cloud-based apps because they are deemed ‘more scalable,’ which in turn pressures charities that rely on such grants to write cloud-biased proposals. After many cycles of such investment, it should not surprise us that cloud and smartphone apps are among the most ‘tried and true’ tools in the global mHealth community. Neither should it surprise us that when the ‘best tools’ excel in more urban and middle-income environments, the tools themselves come to incentivize organizations to focus not on the poorest, but on middle-of-the-road needy communities where the tools are likely to be relevant.

To be sure, smart phones and cloud servers can be tremendously helpful in many settings. It seems almost counter-intuitive to suggest that we could ever go wrong by maximizing ‘value for money’ with technologies that are ‘cost-effective’ enough to ‘go to scale’ nation-wide. No one seems to be intentionally marginalizing the poorest. An alternative view may only emerge when we not only listen to the poor, but spend enough time in their company to appreciate that they are subject to a perpetual train of abuses and unfair disadvantages. The most difficult to perceive are the abuses which are nobody’s fault, the suffering which seems inevitable because it flows not from evil people but from patterns of society that have long been taken for granted. Some use the term ‘structural violence‘ to describe such social arrangements that tend to put specific groups in harm’s way. Revisiting this concept has helped me attend to who gets left behind when we design technologies with a utilitarian focus on the characteristics of the average needy community. Many who call attention to structural violence have also argued that the only means of addressing such systemic unfairness is to build a preferential option for the poor. What would preferential software for the poor look like? Perhaps I can explain this most clearly with an example.

My first mHealth project was with a rural Catholic hospital in Malawi called St Gabriel’s. We gave phones to about one hundred community health workers so that they could exchange logistical information such as “Mary Banda in Msangwa village might have Tuberculosis and I want you to come test them.” The project was successful enough that many other organizations asked us to replicate it, but they wanted to use more quantitative data in more complex ways than our initial approach to simple prose text messaging. At the time, industry experts referred to durable, inexpensive and long-battery-life phones as “dumb” or “SMS-only” phones because they “couldn’t run apps” that would support forms-based data collection or menu-based decision support. Bucking conventional wisdom, I fixated on the idea of putting apps on $20 phones because it seemed the only way to move beyond the limitations of text messaging without leaving behind communities like St Gabriel’s.

During a trip to Kenya I learned of the popular mPesa mobile banking service. It involved simple menus and worked on the cheapest phones–it was possible! How ironic that the experts in my field had neglected to discuss a technology that 70% of adults in Kenya were using. I learned that mPesa was installed not on the phone itself but on the SIM card, so I began talking to mobile network operators about putting health apps on their SIM cards. I was told again and again that it simply would not be possible unless I were prepared to pre-order a custom batch of 10-50 thousand SIM cards (and my funders said that would be impossible). More than a year later I discovered a small Eastern European manufacturer of paper-thin ‘parallel-SIM’ cards that slide underneath a typical SIM card. Using this technology, we could install our apps on the parallel SIM and keep using ordinary SIMs to connect to the mobile network. Medic Mobile soon began developing a parallel SIM card app called Muvuku, which means ‘to listen’ in Chichewa, one of the indigenous languages of Malawi. Since then we have used it in more than a dozen countries and it has won a few awards. Conveniently, it also works well for programs that could afford more expensive phones and that do have access to electricity and local smartphone suppliers. It was harder and took much longer to develop than an Android app though. It was only worth the effort because of a commitment that has more to do with who we are than it has to do with maximizing impact or going to scale. While we’re open to working in any setting that would benefit from our support, we design our core technologies in partnership with, and to suit the circumstances of poor and marginalized communities.

What does it mean to make preferential software for the poor? I do not mean to argue that the solution lies with any one technology or in dealing with any particular constraint, be it electricity, usability, adaptability or local availability and appropriateness. Rather, it means standing with the poor, spending time in their company and appreciating their strengths and constraints—as they see them—before attending to technical or financial concerns. I believe this matters because engineers and economists too often advocate for their expertise in ways that make the aspirations of the poor seem impossible, or even a waste of resources. In recent decades we have heard that HIV prevention is ‘cost effective’ but treatment is not and that patients who do not respond to the cheaper classes of Tuberculosis drugs are ‘untreatable.’ Today many still argue that poor community health workers can volunteer while their bosses receive comfortable salaries and that the ‘dumb phones’ which have proliferated all across the African continent are not ‘smart’ enough to extend the highest standard of medical care. These social fictions stem from a poverty of imagination about what is possible. They stem from a traditional deference to the expertise of foreign technical elites over the lived experience of beneficiaries. To be sure, spending time in the company of the poor will not render all problems easily solvable. But listening patiently and with a deep sense of humility is a decent place to start.

 

Isaac Holeman, Medic Mobile 

Isaac Holeman is a designer and a scholar striving for global health equity. Through ethnography and innovation, his work is about seeing through the eyes of the poor and marginalized and responding pragmatically. He has pursued this work as a cofounder of the mHealth social enterprise Medic Mobile, as an Echoing Green Fellow and as a Gates Cambridge Scholar.

Leave a Reply

Your email address will not be published. Required fields are marked *