Portfolios of the Poor: the power of qualitative data

The power of the portfolios of the poor research is that it yields high frequency, detailed data. While some might argue that this comes at the expense of learning broad lessons of the type "This is just a few people in a few locations", I don’t think this would be a valid criticism. There is a continuum of data sources from anthropological studies and market surveys to high frequency rich data such as financial diaries to nationally representative household surveys; and each of these data sources complements, rather than supplements the other. The criticism would be valid if the authors claimed that their results were representative of the poor globally or of Bangladesh, India or South Africa as a whole, but they are very straightforward about the limitations of their data. One way to illustrate this complementarity is by comparing portfolios to “The Economic Lives of the Poor” by Banerjee and Duflo. Through survey data, Banerjee and Duflo are able to paint broad strokes of what it means to be poor on a daily basis. Portfolios paints a more nuanced picture of the lives of the poor that can help to shed light on some of Duflo’s findings and thus pose further questions that could be included in future survey rounds. For example, Duflo and Banerjee seem puzzled by the proportion of income spent by the poor on festivals, while portfolios might allow us to hypothesize that participation in festivals allows households to be full members in the community, which in turn gives them access to financial instruments such as lending and borrowing among neighbors. Another example of complementarity between sources of data is the use of ethnographic studies to explain the results of impact evaluations. An impact evaluation of a nutritional supplement distribution program might find that the supplement is not reducing the micro-nutrient deficiency at the expected levels, while an in-depth observation study might provide the hypothesis that since the supplement is viewed as food, it is shared among all the members of the family and therefore its impact is diluted. This hypothesis might in turn lead to new product designs that make the supplement be seen more as a medicine and in turn administered only to the undernourished child in the household.

While Portfolios offers great insight into the lives of the poor, It would have been interesting to go a step further and make the connection between financial management and other outcome variables such as health and nutrition status. The book makes a very compelling case for the centrality of financial instruments to the survival of the poor and for the unexpected level of complexity of their portfolios. However, it would also be interesting to know what instruments are more effective in allowing poor households not only to eat, but to be well nourished, not only to pay the pharmacy bills, but to achieve a healthier lifestyle, etc. I think this data was not collected because the aim of the authors was to understand how the poor manage cash. Furthermore, linking financial management to these outcome variables would have increased the cost of the study and required a bigger sample and a more complex study design, both of which would probably only be possible by sacrificing frequency and detail.