Portfolios of the Poor tells a story you won’t hear from a typical microfinance charity:
Sankar was a landless, illiterate rickshaw driver, whose wife had Grameen membership. They had borrowed from Grameen Bank a few times – in fact one loan had helped him buy his rickshaw. Suddenly his wife told him they would have to open a GPS [savings account] in order to get the next loan. He was suspicious, he told us. “And now?” we asked. He chuckled. “Now, we try to avoid loans and just use the GPS.” Pressed to explain, he said that his income was small but sufficient for their daily needs and they had nothing to invest an expensive loan in. Their priorities now were for their children, and the GPS seemed, compared to borrowing, a cheaper, more relaxed, longer-term way of providing for their future (marriage for the girl, a business for the boy). Like Jharimon, Sankar borrowed sometimes and saved always. “Grameen should have done this years ago,” he said, echoing what many others had told us. (Page 170; emphasis mine)
What do the poor need more, savings or loans? There isn’t much information on this question – something I find pretty shocking given that microfinance is decades old. But the limited evidence seems to point to savings, at least for the poorest clients:
- The general impression given by Portfolios of the Poor (quoted above), the only systematic attempt I’m aware of to see how the poor manage their financial lives.
- A 2001 survey of SEWA Bank clients, which found that the bank was “far more important as a depository for savings in early 1998 than it was as a source of credit” (page 79) and that “39 per cent of borrower households were below the $1 a day line, compared to 53 per cent of saver households and 67 per cent of [non-client] households.” (page 81) This is the only study we’ve located directly assessing savings vs. loans clients at an MFI – please let us know if you know of more – but it is consistent with charities’ and scholars’ off-the-record comments to us that the poorest clients tend to be more interested in savings than in loans.
- The strongest studies of microfinance, which show weak/no effects for two traditional microcredit program and positive effects for a savings program.
- The purely logical argument that savings provides the same “risk management” benefits as loans, without the same risk of harm that comes with putting poor people into debt at high interest rates. (David Roodman among others makes this argument.)
But loans are far more prevalent
Looking at the 537 MFIs listed on MixMarket under US-based networks (source data), we note that
- 517 of 537 appear to have introduced loans before they introduced savings; 0 of 537 appear to have introduced savings first. (This may be an artifact of changing data reporting standards, but if Mix collected loans data before savings data, that fact is also indicative).
- In their most recent year, 503 of 537 MFIs had more outstanding loans than savings; 34 of 537 had more outstanding savings than loans.
Why?
Why are loans so overwhelmingly common compared to savings, when the little available evidence points to a greater need for savings?
Is it for purely logistical reasons? Savings and loans each present different challenges, but the necessary “innovations” seem to have come first and been expanded more for loans.
Or is this another case where the best program loses to the best donor story?
Comments
Holden, I think it is more than the best donor/best story thing? Would you trust a start-up NGO with your savings? Probably not. Would you particularly mind taking a loan from a start-up NGO? Probably not so much. The regulatory barriers to taking savings (and selling insurance) ought to be and generally are higher than for making loans. Mind you, I still want to see those barriers overcome more often.
I have to agree with David. I think most MFIs would love to take savings but are not allowed (especially if they are running as a ngo and not as a regulated bank). Just compare how hard it is in the USA to start a ngo with starting a bank. Same in most countries.
I would have to disagree with David & Basti.
MFIs do not want to provide savings products for many different reasons, but most of all because it is less profitable. The ROI on savings is far lower than that for MFIs, and so MFIs avoid providing savings alternatives. (Compartamos, for example, one of the financially most successful MFIs – think $467M IPO – has had savings pilots for years they have not launched)
The real challenge behind this, as Holden correctly points out, is that the priorities of the ‘financial services for the poor’ sector is skewed too far toward monetary return on investment and not enough toward social returns on investment. This creates entities that look good financially but are not delivering as much social return.
I am also in the process of reading Portfolios of the Poor and studies like these clearly articulate the needs of the poor, and savings is a need that is not receiving enough attention. This is partially because of regulatory hurdles, which are quickly being changed here in Africa, but it is mostly because of skewed priorities: financial over social returns.
Cell phone technology promises to reduce the costs of servicing even the poorest and least-densely populated regions to the point where it will become a more viable social enterprise with better financial ROI. In the meantime, however, I think we all need to re-adjust our priorities and focus more on Social returns. With additional funding and focus, savings can and will be realized for the bottom billions.
I think both David R/Basti and David V raise good points. The factors you’re citing could by themselves explain the MixMarket figures I cited.
It’s still worth keeping in mind that if savings institutions are both rare and more needed, the focus of microfinance charities ought to be on expanding savings. The focus of their marketing clearly isn’t.
Here is a good blog entry on this topic: http://www.seattlemicrofinance.org/microfinances-circular-firing-squad/2009/11/10
David, thanks for the link.
We are a little wary of swinging too far in this direction and promoting microsavings as a panacea, as this article seems to do to some extent. The evidence for savings’s impact is also limited, and there are still critical questions that should be asked of a given savings program.
I’ve worked for two microfinance networks, including one particularly known for microsavings.
The biggest barriers are definitely regulatory – and for good reason. It requires a much more sophisticated institution to manage savings, since they involve semi-random cash flows, as opposed to credit, which can be forecast more easily. Add that to the MIS, internal controls and staff issues, and your typical NGO couldn’t realistically offer savings even if it wanted to.
Credit got much of the initial focus b/c the assumption was that the poor are too poor to save. Which is BS, because the biggest problem most poor households have in the developing world is that their income is highly irregular. Savings helps these households smooth income over the course of the year. Credit does this as well, but they are paying effective interest rates of 50-90% for the privilege of doing so. Unfortunately, subsidized loans to MFIs have crowded out savings. The biggest benefit to MFIs for savings is the reduced cost of funds – interest rates on savings may be 10-20% lower than what they would pay for commercial debt.
The existence of informal savings networks (ROSCAs, tontines) around much of the wrold indicates a huge pent up demand for savings. The rule of thumb we used was always that there were 6-8 savers for every borrower. The key is to identify institutions (cooperatives, progressive banks, more mature MFIs) and technologies (mobile money, ATMs) that facilitate convenient savings.
John, thanks for the very interesting comment. It’s particularly interesting to me that the need for savings was “assumed away” for many years, though now people seem to agree that the need for it is clear. To me it points to a fairly basic failure to understand the setting.
Comments are closed.