Access and appropriate use of financial services help people better manage their lives, capture opportunities, and reduce vulnerabilities. This is particularly true in developing countries where the large majority of poor households lives and works in the informal economy. Not by choice but by necessity, they are self-employed and need to generate their own livelihoods.
But when things go wrong, financial services can also make life more difficult: Too much debt, too little savings, wrong insurance. While policymakers and regulators across the globe have embraced financial inclusion as an important development objective, they are also concerned about financial literacy. In developed countries, some have gone so far as to suggest that we would have avoided the financial crisis triggered by excessive lending in the U.S. mortgage market had consumers been more financially literate.
This notion has always made me uncomfortable, in particular in the context of developing countries where years of experience and the financial diary literature have taught me that poor people are not only very active, but also pretty astute managers of their financial lives. That doesn't mean that poorer people don't make mistakes -- we all do because of increasingly better understood decision biases. One strong empirical observation, for example, is that people are more likely to make sub-optimal decisions when they are stressed. And poor people in developing countries, by definition, are much more often stressed as they typically lack regular incomes and social safety net protection. But the financial literacy push often has a paternalistic ring to it, and in the extreme, seems to put the blame on the victims of predatory provider behaviors.
Recent human-centered research and product design efforts in Pakistan have given me a better-founded way to articulate my unease. Rachel Lehrer from the Boston-based design firm Continuum spent on behalf of CGAP six weeks in Pakistan to better understand how poor women in rural areas use (or don't use) the accounts linked to the transfer benefits that they are paid by a large-scale Pakistani social protection program. She found that the women recipients of the program, while illiterate, had a keen sense for commerce and fairness. They were numerate in their context, even though many could not identify Roman numerals. Or when they could, they might type the mobile pin they were given from the right to left, the way they know the Urdu script is used. The research also showed that words can actually complicate and confuse the women beneficiaries instead of simply being ignored by them. ATM withdrawal receipts that had numbers next to lots of words conveyed to the women that the service was not for them. Much easier to understand was the withdrawn amount in large numbers without any distracting words before or after it.
Being illiterate does not just mean that people can't read and write. It also means that they haven't learned to think abstractly and, as a result, it's harder for them to translate a generic instruction into a different type of action. For example, the research showed that very literal photos were the best way to instruct poor women how to use ATMs. This even proved more effective than verbal instruction which often confused the women by its abstraction outside of a literal context that they understood.
Researchers in other areas have tried to tackle the issue of how to best make services accessible to people who are innately smart, but not literate. For example, Indrani Medhi and her colleagues at the Microsoft Research Center in India have looked at text-free computer interfaces for illiterate and semi-literate users. They built and tested two applications -- one for job search for domestic laborers and another for a map to navigate the city -- in collaboration with a community of illiterate domestic workers living in three Bangalore slums. Their work showed that simple screen-shot pictures or drawings, not abstract diagrams or icons, combined with a complementary voice function and continuous, graphical help functionality dramatically outperformed traditional user interfaces. They essentially allowed illiterate people to become self-guided, independent computer users.
Poor people are smart; they have to be to survive. The onus is on us to design products and communicate their use and functionality in a context that works for them. We may take for granted the short-hand for concepts like interest paid on loans or received on savings, but these need to be translated into language consistent with their everyday realities of poor people. Otherwise, the functionality of formal financial tools may remain too abstract to be practical.