Just came across a fascinating study by Jonathan Morduch & co-authors, who find that an increased presence of formal banks in a given economy pushes “commercially-oriented” (for-profit) microbanks towards serving a poorer segment of clients. I find this simultaneously intuitive and counter-intuitive: intuitive that competition pushes banks to expand their target market, but rather surprising that the expansion of large banks (who are generally seen as bystanders in discussions of financial access for the poor) may increase access to financial services in the end. Interestingly, the authors believe this to be one of the first studies of competition between microbanks & formal banks, pointing to a fantastic avenue for research as the microfinance sector continues to grow & formalize.
It’s been a matter of some curiosity to me that the debate over for-profit vs. non-profit microfinance has remained quite vociferous over the past few months, even whilst more balanced views of the social role played by microloans have been gaining acceptance. It’s clear that a good deal of the heat generated by this particular debate is attributable to the inclusion of market-based anything – as Amanda recently wrote, “people conflate ‘market’ with ‘unfair and exploitative’ all the time,” which is obviously a sensitive issue when one is talking about banking with the poor. But then, MFIs by their very nature – even non-profit ones – are committed to helping clients participate in business and markets. A purely market-based critique falls a bit short of explaining all the differences between the two sides.
I was thinking about this when I read Matt’s recent post on arbitrary deadlines in development. The most insightful observation was that it’s “highly likely that the types of policies we should implement when thinking about poverty for the next 100 years differ significantly from those we would use when thinking about poverty for the next 20, or 5 years.” And this, I think, goes straight to the heart of the microfinance debate: there is often a temporal mismatch between the effective goals of for-profit MFIs and non-profit MFIs. For-profit MFIs, whose clients appear to be on average less poor than those of non-profits, fit into an economic space where real growth at least seems possible over the long run. Non-profit MFIs, on the other hand, are often better suited to meet the immediate consumption-smoothing needs of poorer clients, rather than long-term investment needs. The goal that one prioritizes probably indicates where one stands on the for- vs. non- debate as well.
This temporal perspective also goes a long way towards explaining certain criticisms that each side has launched at the other. For instance, among the various non-profit complaints about the Compartamos IPO is the valid criticism that poor clients who are left behind by a privatizing MFI may not have access to another financial institution. The subtext is, is short-term harm to poor people worth longer-term growth for their descendants? Conversely, one also finds criticisms of non-profits for funding small businesses that fail to generate jobs or economic growth – an interesting moralizing critique of MFIs that use their capital inefficiently, and perhaps help fewer people than they ultimately could in the long run.
I am, of course, generalizing to some degree about the characteristics of for-profit and non-profit MFIs. This is based on my current understanding of the debate, but I wouldn’t wish anyone to accept this uncritically, as I’m sure there’s a non-negligible number of MFIs out there for whom these descriptions would be wrong. (I’ll even point to the first one myself – FINCA-DRC is a for-profit institution that’s quite committed to small group lending with women, which is a short-term consumption-smoothing tool if ever there were one.) But the larger take-away here is that we as practitioners need to look more closely at the temporality of economic growth and poverty reduction, and discuss our allocation of resources between the short and long terms more explicitly. I’m guessing there’s a place for both types of MFI in most, if possibly not all, markets.
Ah, I have so much to write about after 2.5 months with FINCA in the DRC, and so little free time in which to do it! But one comment that I’ve been turning over in my mind for some time now is something my father said to me when I was explaining the concept of group microlending to him. I pointed out that most clients don’t borrow for discrete events, like buying a car or paying for college, but rather take out multiple sequential small loans to finance everything from business activities to children’s educations. (Granted, many microfinance clients aren’t supposed to be using loans for consumption purposes, but even the social pressure of group lending can’t always prevent this.) My father’s response was that this type of lending sounded much more analogous to owning a credit card than taking out a “traditional” Western loan. And I’ve been increasingly fascinated with this idea as its implications have sunk in.
Consider the long-dominant narrative of microloans as a pathway out of poverty, now coming up against a more nuanced narrative of microloans as consumption smoothing tools whose efficacy in moving borrowers out of poverty is also dependent on the individual borrower. (See the end of this post for more on individual variation.) I don’t actually find these positions contradictory, just varying in their levels of nuance, but for those who do find them mutually exclusive, I think the credit-card-vs.-traditional-loan idea may resolve many of the apparent contradictions here. “Traditional” loans are usually aimed at smoothing consumption around the type of significant investments that provide positive, and lasting, long-term shocks to income levels. Borrowing to finance a university education is a good example of this. Credit card companies, on the other hand, make few claims about their beneficial effects on one’s long-term financial health – and are better served overall when a portion of their customers spend themselves into debt and pay hefty fees for the privilege. The whole point is to smooth consumption of inexpensive-to-midrange products & services, and few people consider them a substitute for a larger “traditional” loan unless they’ve no other choice. (Have you discussed putting a $180,000 undergraduate education on a credit card with your friendly local admissions officer? [Hello, Dartmouth, you expensive old dear.] I didn’t think so.)
In short, the “pathway out of poverty” narrative expects microloans to function similarly to “traditional” Western loans, and play a significant investment and capacity-building function on a personal level. The consumption smoothing narrative, on the other hand, is redefining microloans as something closer to cash-only versions of credit cards. (Without so many options for punitive fees for defaulters, and with a greater risk of an angry group of Congolese women coming to your house at 6 am to repossess all your plastic chairs as punishment for a missed payment.) There are a thousand interesting directions one could go with this thought, but what I think it mostly points to is the fact that the formal economic structures of microfinance are still going through a dramatic period of evolution. There aren’t many Western banks that wake up wondering if their primary product is a mortgage loan or perhaps really a credit card, after all. And that leads into implications for product design, and the unbanked’s perceptions of and interactions with formal banking institutions, and on and on into what I’m sure will be many future blog posts.
There’s been a seeming lot of discussion lately about the non-traditional savings habits of the poor, what with Portfolios of the Poor coming out, and I recently saw another interesting document on product-specific savings plans in India. Women (usually) may sign up with a store to save towards the purchase of a specific product, and may be rewarded with an extra several percentage discount at the end – the original installment plan (although they don’t get to take the product home first!). The shortcomings of this plan are that the money is non-fungible (i.e. can’t be later withdrawn to pay for a different product), and that there’s no guarantee that the price of the product won’t rise over the course of the saving period, forcing the client to save more than originally expected. However, a number of clients also identified the social pressure to complete their savings and make the purchase as a valuable incentive to continue saving, which they wouldn’t have had with a formal savings account. This echoes the findings of other studies of non-monetary rewards for savings, and could be an interesting concept around which to design new savings schemes.
I’ve seen some interestingly conflicting reports lately on the impacts of microfinance upon education. Someone directed me to one which showed that microcredit clients were more likely to pull their children out of school to do the domestic work which parents gave up in order to run their microcredit-supported businesses, but I’ve also seen another which found that the children of microcredit clients were actually more likely to be in school, as their parents could more easily pay their school fees. I can think of several different ways in which these elements of small businesses, children’s labor, and schooling could interact:
- Maybe the parents who pulled their children out of school to cope with successful businesses worked far from their homes and didn’t have the time to handle both business and domestic work, while successful clients whose children stayed in school worked closer to home and were more easily able to balance the two.
- Perhaps there’s some sort of U-shaped curve of parental income (as supported by microcredit) and the likelihood of children’s schooling. Parents who are suddenly busier with work than previously, but are still too poor to afford to hire domestic help, may be the ones more likely to pull their children out of school – the low point of the U. Parents whose businesses are quite successful, on the other hand, might better be able to pay for both domestic help and school fees.
- This is probably also correlated with whether parents were successfully able to repay their loans. There must be some exogenous shocks to parental earnings that affect both their ability to repay their loans and their financial capacity to send their kids to school – drought, for instance. So one might find microcredit clients pulling their kids out of school for reasons unrelated to their loans.
Lending group sizes: Read something recently stating that repayment rates were highest among a studied selection of lending groups when the group had 14 people in it. I find this fascinating – is it something like Dunbar’s number for close friends, instead of general social connections? What I don’t totally understand is why the number would be as large as 14. It seems to me that the social pressure exerted by any one individual on others would decrease linearly with the number of group members – that is, I can put more pressure on 5 other people to repay their loans than on 14, because of my own time constraints. Or perhaps the limits on social pressure are outweighed by the implicit pressure of having 13 other examples of people successfully repaying their loans, instead of just 4? Freakonomics mentioned some research recently about the power of implicit social pressure (or sanction) in guiding behavior, with regard to people seeing examples or merely believing that a majority of others were acting in a certain way. Perhaps one could think of living up to an example (of loan repayment, in this case) as a way of earning social capital, instead of actively spending it by pressuring other group members to repay their loans.
Social determinants of lending group access: Another predictable but still interesting fact is that self-selecting lending groups are sometimes reluctant to take on members who are too poor – lacking social capital as well as financial (and isn’t that a metaphor for poverty in general?). It does make sense to view current poverty as deterministic of future entrepreneurial success, from the POV of another lending group member, but I also wonder if group members would view occasional poverty (i.e. brought on by a recent illness) differently than chronic poverty. How long does someone have to be abjectly poor before a lending group is more likely to reject them? That’s an interesting question. [NB: Can’t find citation for this, although I think it may have come from Understanding Poverty.)
Monthly vs. weekly repayment schedules: This was a wonderful study – an analysis of whether Indian MFI clients were more likely to repay their loans if they made weekly repayments, monthly repayments, or monthly repayments with weekly group meetings regardless. Repayments didn’t actually vary in a statistically significant manner across any of the three repayment schemes, although the monthly group actually had the lowest rate of missed payments according to the raw data. I say “wonderful” because this knowledge is a great step towards designing lending programs that are better suited to the variable and unpredictable incomes of the poor – it’s quite valuable to understand that time-consuming weekly repayments aren’t necessary to pressure clients into repaying, which may give them more flexibility with their repayments (and use of loans).