My Brain Made Me Do It (Again)

A month ago I blogged some reflections on transparency. Transparency is the motherhood and apple pie of microfinance, I wrote, if not in those words. Microfinance institutions (MFIs) should make it easy for clients to understand the full costs of services, not hide fees in fine print or no print at all. And they shouldn’t introduce charges through the back door such as by overcharging for life insurance bundled with loans. But accepting the necessity for transparency, I wondered about the means. Are annualized percentage rates (APRs) the best way to summarize expenses? There, my reflections stopped.Writing about this for the book just now forced me to push the analysis farther. I realized that transparency is in a way a distorting metaphor for disclosure. After all, pages of fine print are “transparent”: all the information is right there. Accepting that MFIs should give clients a clear window onto costs, there remains the question of how to describe those costs. The window must be framed. One of the most important ongoing developments in economics is the subfield called behavioral economics. The famous duo of Daniel Kahneman and Amos Tversky demonstrated through experiments that how information is presented (framed) often matters at least as much as the information itself for how human beings act. I’ll show you an example below.It was then that I saw the light about the importance of behavioral economics for microfinance, a connection I had only glancingly blogged before, in My Brain Made Me Do It. The starting point of my chapter 7 is Amartya Sen’s definition of development as freedom---“freedom” meaning control over one’s circumstances. Whether microcredit in particular gives people more or less control over their lives depends in part on the quality of their judgment about when to borrow. And that judgment depends in part on their understanding of the costs of credit. Put those statements together and we see that how costs are disclosed can affect the development impact of microcredit. Certainly it did in the United States for those subprime mortgages with teaser rates.So if scientific experimentation points to certain ways of stating microcredit costs as better informing people’s judgment then it is incumbent upon MFIs that are serious about a social mission---or even just aiming to cultivate long-term relationships with clients---to provide information in those ways. Call it “beyond transparency.”Looking for concrete ideas, I was thrilled to find a few good sources. One is a finely written U.K.-focused report by Mike Dixon of the Institute for Public Policy Research in London. Read chapters 5 and 6 for an introduction to behavioral economics and its implications for financial services for regular people. I also found Psychology and Economics: What It Means for Microfinance, by Sendhil Mullainathan, a leader in this field, and Sudha Krishnan. By way of commenting on a study of how people in rich countries invest for retirement, they write:

…people are making mistakes because they are not choosing according to their own preferences. This is a general point: behavioral economics is not about external imposition of preferences. Instead, it studies how people may fail to maximize their own stated or revealed preferences. This discipline is crucial because without it, simply arguing that others make mistakes is too easy. Second, such mistakes occur even when decisions are extremely important and people are motivated to choose wisely…[T]he behaviors we study are “universal”, and are not specific to the poor or some other population…[T]he general psychology that affects all income categories is already powerful enough to shed light onto the behavior of the poor, without invoking a specific psychology of the poor.

(An aside: a thread running through behavioral economics is that our brains were well-adapted to the environments in which they evolved and maladapted for today’s world. Yet I wonder whether the lives of the poor more closely resemble those of our ancestors in the uncertainty of income, financial vulnerability to illness, and reliance for survival on kin and social networks. If so, then human brains might be better suited to the financial problems of the poor than of the rich. Perhaps this helps explain the richness of the financial lives found in Portfolios of the Poor.)I recommend the paper, but was disappointed with how little it did to concretely explain how MFIs should disclose--what I happened to be looking for. Mostly it inventoried familiar human foibles. On average, people think they’re above average in skill and luck, and so overestimate how much debt they can handle. We all need discipline against procrastination, which is why I am not allowed to pay off my whole mortgage with a single check at the end of 30 years. And so on. The paper is conceptual and I suppose its job is to, well, frame questions for direct research.Happily, through a thoughtful review by Slate's Ray Fisman I learned about a new, randomized study of disclosure techniques for payday lenders in the United States. These lenders give people envelopes of cash in exchange for the right to electronically withdraw payment from their bank accounts on the next payday. A typical fee is $15 per $100 lent for two weeks. Yes, 15% for two weeks, for an APR exceeding 400%. Low-income people paying such exorbitant prices naturally raises concerns. Fearful of a political backlash at a time when everyone loves to hate creditors, one payday lender agreed to work with University of Chicago's Marianne Bertrand and Adair Morse to randomly expose some borrowers to displays that might help them understand the full costs of their borrowing. So imagine you go to the window at the payday lender's shop, you do the paperwork to get a new loan or renew an old one, and the clerk behind the window passes you an envelop with cash inside and one of these three images on the front:The first image discloses the APR and shows that it dwarfs those for other kinds of credit that poor Americans might take. The second instead adds up the total cost of repeat borrowing: if you borrow $300 repeatedly for 3 months, you'll pay $270 in fees. The third shows how common repeat borrowing is, as if to say, "If you think you'll climb out of this debt in two weeks, think again."How much would you guess that each message perturbed borrowing trends in the months that followed? Well, I can't cite exact figures---though public, the paper is preliminary and marked "not for citation"---but the general pattern is: image 1 (APRs)---no clear effect; image 2 (total cost)---biggest reduction; image 3 (frequency of reborrowing)---middling. Evidently, total costs meant more to people than APRs. It's possible that the reduction in borrowing with image 2 was harmful---maybe people were inappropriately scared away from taking credit they needed. But the opposite interpretation seems more credible: information that people could relate to about costs led some to be more careful. It's hard not to favor this interpretation if you believe in transparency.Bertrand and Morse appear to have vindicated the forgotten Ralph Pitman, who, as I blogged before, argued in 1931 that disclosing total costs was not only transparent but more helpful than disclosing APRs.Translating the lessons of this study to microcredit, which has different terms, requires some thought. What is the best way to convey total costs honestly and intuitively? Most microcredit loans are for longer periods and, as a result, are gradually paid off like mortgages. This introduces a distinction between the initial and average balance of a loan. A $100 one-year loan with weekly payments has an average balance of $50, so that calling it a $100 loan is arguably misleading, the opposite of transparency. Perhaps MFIs ought to disclose two numbers: the average amount that clients will owe over the course of the loan, and the total fees and interest they’ll pay during that time. The first number would be net of any fees and mandatory savings taken up-front from the loan amount. The second would be net of any interest earned on mandatory savings....or maybe “average balance" is too complicated a concept to work, and would fail to enlighten in the way image 2 did. That, as they say, is an empirical question. So I hope research is underway to do for microcredit what Bertrand and Morse have done for payday loans.