Controlled Experiments in Microfinance

Most proponents of microfinance claim that that the benefits of microfinance are “well proven”, “well accepted” or at least “well-known”.  Being a skeptic, I have been reading up various studies about microfinance and its impact.  Much of what is out there is anecdotal and many of the formal studies suffers from sample bias.

So I was intrigued when I stumbled upon a paper by Abhijit Banerjee et al (MIT Dept. of Economics) about the impact of microfinance on microenterprises and households in general.  On reading the paper I realized two things.  One, I have forgotten much of the statistics I learnt for my own PhD.  And two, this is one of the first “control” experiments I have read about, when it comes to the impact of microfinance.

After a couple of hours of deep reading, I offer a summary.

The Context

The paper studies the impact of microfinance (MF for short) loans disbursed by Spandana in Andhra Pradesh.  The “treatment group” is the set of households that received MF loans.  The “control group” is the population that didn’t avail of the loans.

I won’t go through the details of what the authors describe in terms of their methodology, bias-correction, etc., but rather jump to the conclusions (literally!).

The Findings

Impact of Microfinance is different depending on situation of household.

Those that already run a business, tend to buy more durable goods (investment) for the business.  However, a year later, their revenues, profits or number of employees (though positive) are not statistically higher than control group.

Of those that don’t own a business, the authors categorize them into two groups: a) Those with high propensity to start a business (wife of household head is not a wage-earner, etc.), and b) Those with low propensity to start a business.

About 20% of household that received MF funds, started a new business, against “stated purpose” of 30%.  Among those receiving MF funds, there is a higher propensity to start food/agri businesses (low capital) than rickshaw pulling (greater capital).  Examples of food businesses: tea-coffee stands, food vendors, small kirana shops, and other agri produce.

This seems counterintuitive, until you consider that perhaps the availability of capital is drawing those less business savvy into businesses.  The revenues and profits of new businesses started by MF recipients are indeed lower on average than for the control group, but the difference is statistically insignificant.  However, one thing is more statistically clear: the MF funded new businesses have significantly fewer employees than control group.

Within the 15-18 months of disbursement of loans, the authors found no statistically significant changes in expenditures on education, health, etc.

The MF funds may also provide an additional means of dealing with financial shocks. However, the data does not show reduction in reliance on moneylenders, relatives and friends, though there is some evidence of reduction in non-payment of bills and buying items on credit.

The Implications

While admittedly it is hard to imagine the impact of loans over only 15-18 months, it is important to note that most microfinance loans are also only for similar “short” durations.  So the findings of this research do apply to MF loans.

So, to state things in more general terms:

a)      MF can help households with existing businesses invest in durable business assets.  But it cannot ensure greater income or profits.

b)      MF can enable households to start new businesses in low capital industries such as food stalls.  It is unknown whether it can increase net income after subtracting loan repayment.  (the study doesn’t shed light on this)

c)      MF does not appear to impact social indicators (such as dependence on moneylenders or spending on health, education, status of women) over the duration of the loan.

So does this mean microfinance is useless?

Or does it mean that in addition to MF, there are other ingredients that are important… Without which, MF is ineffective?

This entry was posted in Financial lives of the poor, Micro-enterprises & rural businesses, Microfinance, Research studies and surveys and tagged , , , . Bookmark the permalink.

2 Responses to Controlled Experiments in Microfinance

  1. Babu Rahman says:

    Remind me to not read your blogs when I’m depressed about BOP progress!

    OK, I see how you can say that MF could be useless. But, what if we dig into the MF data only, instead of comparing MF to non-MF and trying to assess statistical significance? Within the MF data, is there any correlation between the variation and the family situations? In other words, what could be those enabling factors that you talk about?

    Have others in the field looked at the data the same way you have? Don’t the MFIs look at this data? Yeah, I know, they have their own agenda to lend and make money, but…

    Looking forward to your thoughts/comments.

    Take care,

    • Richa says:

      First, two clarifications: 1. I did not look at the data — I only wrote about someone else’s research, and 2. I didn’t say that MF is useless, only that it doesn’t do what it claims to do. Providing tiny loans to help the poor manage their finances could be a useful thing in itself (unless it is exploitative or misleading). But providing loans may not alleviate poverty.

      As for digging into MF data only (without comparison), I don’t know how we would use this information. Did you have something in mind?

      The enabling factors — I think we assume that the only input a business needs is money. As anyone who has run or started a successful business knows, it takes more than that. It requires availability and access to raw materials/inputs. It requires an infrastructure that enables profitable sourcing and distribution (at least power and transportation). It requires availability and access to information about markets. Some knowledge about markets, competition, attracting and retaining customers, etc. Ability to understand own business model and operations, and how to squeeze more profits from it. etc. etc.

      And above all, it requires a viable product/service to begin with. I don’t know how many MF providers ask the question: Are paan shops generally profitable? What about rearing 2 sheep versus 6 sheep?

      In the absence of all these inputs, enablers and knowledge, just providing capital is not enough. If that was the case, no VC funded enterprises in Silicon Valley would ever go bust.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s