Philanthropy Action

News & Commentary

Archive

We’re live-blogging the Innovations for Poverty Action/Financial Access Initiative Microfinance Conference.

Ever since micro-credit entered the popular imagination as a “silver bullet” poverty intervention, there have been lots of assumptions about what it accomplishes for the poor: Micro-loans alleviate poverty; They allow poor people (especially women) to start businesses and become financially self-sufficient; They correlate with increased spending on education for children and health care, etc.

The reality is that there has not yet been a controlled trial on the impact of the standard microfinance product – a small loan given to a group of women who meet together weekly for repayment. Put another way, all of the above-mentioned “received facts” about what is accomplished by giving people access to small lines of credit have been based on anecdote. The reality is that no one knows what impact, if any, micro-credit has on the lives of the poor.

For this reason, the presentation made here by Esther Duflo from MIT on results from the first randomized controlled impact evaluation on a micro-credit program is so important. The project involved 2400 households and took place in Hyderabad

The early results are literally just in. The first round of analysis was completed in the last 48 hours and there is a lot of data left to gather and analyze. With that caveat, the early results are:

—Take-up of micro-loans was relatively low at 17.5 percent among the sample – this raises a question about whether demand is in fact as high as policy makers and lenders claim, or if there are other countervailing dynamics in play, such as risk aversion
—The loans, unlike many, did not require the borrowers to start a business – they were free to use the loans however they wanted. Thirty percent of women state their purpose for the loan is to start a new business; 22 percent expect to invest in goods for an existing business; 15 percent buy a durable household item; 15 percent use it for household consumption; 6-8 percent pay for a health shock; 6-8 percent pay for a household crisis; 6-8 percent pay for education and 6-8 percent expect to pay for a ceremony of some kind (totals not 100 percent, because loans can be split among expense categories)
—In practice, one-tenth of loans are used for a new business
—Loan-takers have increased expenditures on business and household consumption compared to the control group
—Loan-takers also spend more on home-based durable goods
—All loan takers spend less on “luxury” items such as tea, ceremonies, etc., suggesting that people make more prudent decisions with the money they do have
—The study shows no impact of access to credit on outcomes for education, health, or women’s empowerment

The full study results are not yet out, and not enough time has passed perhaps to register full impact, but the analysis that has taken place to date calls into question a number of assumptions about microfinance as an overarching tool with impact on a wide range of development issues like education, health status and women’s empowerment. Even the question of micro-enterprise establishment and whether loans lead to recipients gaining sustainable sources of incomes is, at this stage, inconclusive.

Let me reiterate again that this is the very early data and analysis. But even these early results are evidence that we really know almost nothing about the real impact of micro-credit and how or whether it works.

Tim and I also had a chance to interview Duflo and Abhijit Bannerjee while here at the conference. We’ll be posting that interview in the next week.

Comments

Thanks for the report.  If you have more updates I’d love to read them. I came here from Tactical Philanthropy which headlines your post like this “The study suggests a lot of what we believe about microcredit is a myth.“  In light of that, a couple of comments:

“The reality is that no one knows what impact, if any, micro-credit has on the lives of the poor.“ 

This is overstated.  There is plenty of evidence of impact though you’re right that there are a limited number of large scale comparative assessments. Your statement seems to suggest that only such large scale assessments are legitimate and I would question that assumption right from the start.

“The loans, unlike many, did not require the borrowers to start a business – they were free to use the loans however they wanted. Thirty percent of women state their purpose for the loan is to start a new business; 22 percent expect to invest in goods for an existing business; 15 percent buy a durable household item; 15 percent use it for household consumption; 6-8 percent pay for a health shock; 6-8 percent pay for a household crisis; 6-8 percent pay for education and 6-8 percent expect to pay for a ceremony of some kind (totals not 100 percent, because loans can be split among expense categories) “

This is not a revelation.  Those of us who work with microcredit have always known that loans are put to a wide range of uses.  People need money for all kinds of things in their lives, not just income generation.  Well functioning programs don’t loan money on a whim though.  Education, health, ceremonies, those are all important contributions to a family’s ability to make a living.  If some have been thinking that microcredit is solely used for direct business investments, they haven’t been paying attention.

“Loan-takers have increased expenditures on business and household consumption compared to the control group”

Presumably this is a good thing, in fact it’s the point isn’t it?  Poor people have greater access to resources and are maybe living better lives?

“All loan takers spend less on “luxury” items such as tea, ceremonies, etc., suggesting that people make more prudent decisions with the money they do have” 

While I question whether ceremonies should be written off as luxuries (ceremonies are social glue, you build relations through them; relations are the basis of everything), the point about prudence again seems to make the case in favor of microcredit: not only are they better able to access resources, they are less “profligate” with what they do have.  A died in the wool modernization theorist would be crowing about seeing Weber in action

“The study shows no impact of access to credit on outcomes for education, health, or women’s empowerment”

No impact because it didn’t address it or the measures didn’t show it?  If the latter, I’d love to see more details.  Measuring “empowerment” for example is very difficult because it’s not easy to see.  For example, I often tell an anecdote about a women’s group we work with that couldn’t get a local government committee to listen to their request for support in repairing a bridge.  So they literally made themselves a uniform (matching saris) and the committee couldn’t help but pay attention, and subsequently allocated a substantial amount of money to the group.  My argument is that empowerment is seen not in the fact that they got the money, but that they devised a strategy to demonstrate collectivity, which is itself an exercise in power.  Anyway, the basic point is that it is difficult for broad scale quantitative assessments to get at the we might call the microphysics of empowerment.  So I’d want to see how the researchers did it.

So the bottom line is, one thing that needs to be questioned, in addition to what microcredit actually does, is the assumptions that others have been making of it.  These results don’t surprise me: so what is the “myth” that the headline was talking about?

October 30, 2008

The premise behind this post is misleading.  There have been many robust studies done on microfinance showing the benefits and the shortfalls of microfinance as a development strategy. For examples check out http://www.cgap.org.  Also, the preliminary results of Ms. Duflo’s study seem to at least imply that approaches such as the one taken by Freedom from Hunger’s “Credit with Education” may be just what is needed to capitalize on the opportunities that microfinance presents.  The United Nations Populations Fund published a very good paper on this approach (backed by empirical research) showing the impact of combining health education with microcredit.  Let’s not throw the baby out with the bath water.  While it is important to have good information on the impacts of microcredit, it is equally important to not form too strong of opinions based on one small study carried out in one small geographic area.

October 30, 2008
Editor

Thanks to both Tim and Chris for the thoughtful comments.

Allow me a quick response:
1) The misleading premise both commenters note is from a site linking to us, not in this post.

2) That being said, we are strong believers that there is a pervasive “myth” about microfinance. The myth is not held by practitioners as both of these commenters seem to be, but are routinely spread by many practitioners in their marketing materials. Ask the general public what they know about microfinance and they will routinely tell you that small loans are given to women, who are the only reliable borrowers, who start their own businesses, work their way out of poverty, and send their children to school. What little we do know about microfinance suggest this marketing narrative is simply not true.

3) Additionally, we strongly believe that the continuation of this marketing narrative is what truly threatens to “throw the baby out with the bathwater.“ If the industry continues to tell donors that microfinance accomplishes something that it does not, eventually those donors will revolt—and the microfinance industry and its clients will suffer greatly as a result. And that is an outcome we want to avoid by telling the truth now rather than later.

4) Both commenters suggest that there is a lot of impact data already available. This is simply not the case. This conference was attended by the leading microfinance researchers around the world and all agreed on two points made by two universally respected researchers: a) Jonathan Morduch’s note that there are only “one or two” methodologically sound studies of impact; b) Esther Duflo’s note that her study was the only randomized controlled trial (which means the only reliable study) of the standard microfinance product.

5) I would encourage both the commenters and others to read the other posts on studies presented at the conference (available on this page: http://www.philanthropyaction.com/archive/category/poverty_alleviation/) which serve to provide context to this post and the debate.

November 03, 2008

Thanks for the re-joinder.  A couple of quick points.

“Ask the general public what they know about microfinance and they will routinely tell you that small loans are given to women, who are the only reliable borrowers, who start their own businesses, work their way out of poverty, and send their children to school. What little we do know about microfinance suggest this marketing narrative is simply not true.“

Overly simplistic is not the same as not true.  That narrative describes what is attempted, and can be used to describe some actual cases (though what “climibing out of poverty” actually means in practice is an open question).  The report you cite itself suggests that some aspects of the narrative are true. 

That said, I wholly agree that hyperbolic marketing does a disservice to everyone.  But, is there any other kind?  Is it possible to market with nuanced information?  In my experience the simple answer is no.  We’re forced to come up with simple messages because few “consumers” of charity really take the time to learn about what they’re buying.  Unhinge the funding of social change work from consumerist marketing and you’ll start to get more truth in advertising. 

“Both commenters suggest that there is a lot of impact data already available. This is simply not the case…. the only randomized controlled trial (which means the only reliable study)...“

That parenthetical clause is so debatable it’s a non-starter for any kind of conversation regarding evaluation.  Just because you don’t like the methods used in existing studies does not mean those studies don’t exist. 

That said, it is very hard to get good evaluation work.  There are various reasons for this It costs money (lots of money) and no one wants to pay for it.  A lot of what is done is intended to puff up beneficial impact; a bad evaluation can kill an organization.  There are plenty that should die rapid deaths.  But there are also very good ones that make mistakes and need to have the luxury to learn how to do things better.  They often don’t have that luxury though, because the punishment for doing so (defunding and fundamental weakening of the organization) creates the conditions where it is better to fudge than to come clean.  In other words, it gets back to the narrow story telling imposed by consumerist marketing, which itself is driven by a shallow, market driven philosophy of social change. 

Let me expand on this last point.  Markets can be wonderful tools for mobilizing people and moving resources around.  But too often market growth is seen as the end of development work rather than a means to an end.  The end, in my opinion, is creating the conditions for human beings to meet their basic needs and have the capacity to be full participants in their society (culturally, economically and politically).  Markets are just one way of doing this.  But much of the enthusiasm for microcredit, I think, derives more from the former assumption, that markets are ends in themselves, rather than from a sober consideration of what markets can do for human dignity and well being.  So, yes, there has been a lot of mythmaking, and no doubt organizations, in their struggle to get funded, feed into it.  But donors/funders too are part of the problem.  Their expectations, in particular their desire for quick, easy market-based magic bullets, forces people who know better to feed them simple, comforting stories.

In the end, I basically agree with you on the general point: that microcredit isn’t all it’s pumped up to be.  But I disagree with the cause (that it’s just the practitioners who are guilty for creating the hype) and think that you are overstating things (imperfect knowledge is not the same as no knowledge).

One final point: microfinance is being treated here as a single, clearly defined ‘thing’.  But there are many activities that fall under the rubric (eg, mf banks vs. cooperative savings and loan funds).  Part of ‘truth in advertising’ needs to start with a more nuanced understanding of what microfinance means (but, good luck trying to advertise that in a workplace giving newspaper supplement!)

November 04, 2008
Editor

Chris,

1) The point is not that I don’t like the methods. The point is that there are, according to the world’s leading experts, practically no studies that use a reliable methodology. While there are lots of studies, those that use methodologies that are unreliable are not materially different from just making things up. “The plural of anecdote is not data”.

2) We agree one hundred percent that nonprofits are not solely to blame for overhyping microfinance or any other poverty intervention. In fact, we spend most of our time pointing out that donors are the root cause and that donors play a significant role in making nonprofits less effective. That being said, your comments come dangerously close to saying, “the ends justify the means.“ It will take a great deal of courage for nonprofits to stand up to donors with unrealistic demands and expectations. But the alternative is simply continuing the downward slide of trust in nonprofits.

3) I think you’ll find if you read through all the posts, that far from microfinance being treated as a single thing, the studies (and our site) cover the vast range of activities in the sector.

November 10, 2008

Actually there have been numerous studies carried out on the impacts on women and families’ lives after receiving microcredit loans…Women at the Center [a book] is a fantastic look at the Grameen system in Bangladesh. The Jameel Poverty Action Lab also has been acquiring hard statistics on real benefits.

Ed. Note: the writer is associated with a documentary about Mohammed Yunus

February 19, 2009
Editor

Once again, a commenter serves to highlight the point that while people think there are lots of reliable studies on the impact of microfinance, there really aren’t. And as the original post notes, the Jameel Poverty Action Lab is a leader in conducting reliable impact studies—and the results, rather than affirming the popular perception, show us that microfinance serves a very different role than the marketers would have us believe.

March 02, 2009

Perharps it will be interesting to ask the question what is a “reliable meaurement” as the word itself is subjected to many different definitions and perceptions. To this date, there is no one research study that can claim to be perfect. All studies do have their weakness and should be treated as such. We seem to be arguing from a purely positivist school of thought and believe that only when control groups are present then it is to be believed. From a research point of view, qualitative studies do exist and should not be looked down at. I agree with Vivian that there are actually many, many studies that have been done which are credible and do support the fact that microfinance in some context do help. On the other hand there are many researchers e.g. Aneel karnani who claim that there is evedence that microfinance is detrimental.

March 11, 2011

Remember my personal information

Notify me of follow-up comments?

Comments may be edited for length. Inappropriate comments will not be published.