Author: Daniel Rozas

UPDATE 13/02:  I just came across a 2009 journal article by Susan Johnson by the same title. Had not been aware of it previously.

The verdict is out. Final publication of six randomly-controlled studies (RCTs) has drawn a pretty thick line under the words of David Roodman: the average impact of microcredit on poverty is about zero. The notion that microfinance lifts the poor out of poverty is officially dead.

Now, the caveats. The studies evaluated microcredit only – not savings or payments or insurance. Nor did they cover so-called microfinance-plus programs, which provide training, health care or other interventions, along with credit. It’s quite possible that these or other specialized branches of microfinance practice do raise the living standards of the poor. But, if I may be so bold, even the best of these initiatives are probably less effective than we might have supposed.

This is good news. We in the microfinance community could use some humility. We’re financiers, not doctors, scientists, or teachers. To think that we can alter the lives of millions is hubris.

That doesn’t mean that the value or even the existence of microfinance should now be called into question. Not everything that is valuable is life-changing. Marginal improvements are just as important. And microfinance – not just credit – can marginally improve the difficult lives of the poor. I am 100% confident that you the reader rely on a multitude of financial services for any number of needs. Life would be far more difficult without them.

That the majority of the 2+ billion unbanked will eventually join the financial system is inevitable. The question is how soon and under what circumstances. We could just wait for economic development to take its course, and as incomes rise, banks will find their customers. What would be the cost of that wait?

First, it would mean forgoing useful financial services to hundreds of millions for decades to come. That’s a real cost to them. Sure, they can continue to rely on the informal financial services – the ROSCAs, moneylenders, pawnbrokers, susu collectors, and the finances that flow between friends and family. Some of these are better than others, and all are better than nothing at all.

Certainly, we should not ignore these informal systems, least of all because they give us a better understanding of how to serve the poor. But to think that we cannot improve on informal finance is to romanticize the lives of the poor. It sustains strengthens families and sustains communities! Perhaps. But people don’t seem to think so – as soon as people have the money to go to a bank, an insurance company, or a pension fund, they nearly always do. It’s nice to have a social network to help out when hard times come to pass, but mostly we seem to prefer to keep our friendships and finances separate.

Second, over the past few years, the microfinance sector has made real strides in developing mechanisms to protect clients from harm. On its own, the banking system would never undertake such initiatives. However, client protection is increasingly becoming adopted by regulators as a core operating principle and this is a development for which the microfinance sector can legitimately take credit. This work is ongoing, and much work remains to be done to make sure these regulations apply to clients in all microfinance markets, and not just a select few.

What’s left is the question of outreach, of reducing that 2 billion unbanked without waiting for them to become rich enough to do it on their own. This shouldn’t be just a game of numbers. The way microfinance currently operates, closing that gap would mean 2 billion loans. Or maybe 2 billion savings accounts and a smaller number of loans. We can measure the numbers, but what’s missing is the quality of the services we provide. Protecting clients isn’t nearly enough if the only choices given are cookie-cutter loans tailored for high-turnover businesses. And it’s here that we’ve barely begun.

We have very little currently that enables MFIs to measure both the quantity and diversity of services provided. If financial inclusion is a stated goal, it’s not enough for an MFI to simply count the number of clients. For that matter, the foundation of all microfinance data – the MIX – doesn’t even allow tracking clients. You could check the number of active borrowers, or number of savings accounts, but without any idea whether these represent the same people or not. And unlike loans, many savings accounts are mere shadows, sitting empty, dormant or both.

This is only scratching the surface. Who can answer how many insurance policies have been issued by MFIs worldwide, and how many of those are currently active? How many money transfers (and by how many people) have been done in the past year? And these are just the most top-level metrics. Remember those cookie-cutter loans? If we’re to measure diverse offers, we should be able to answer some more basic questions:  what share of loans are issued with a term of less than 6 months?  More than 2 years? How many savings accounts are in fact dormant? What is the average turnover (deposits + withdrawals) for savings accounts?  What is the average payout ratio on insurance contracts?

Answering any of these questions currently requires sitting down with an MFI’s data management staff and asking them to generate a special report – and that’s assuming the IT system even supports this type of analysis. We’re still far removed from the day where these questions could be answered at the country-level, let alone globally. Yet that is ultimately what’s needed.

Today, the microfinance sector is embarking on an important journey. We leave behind earlier shibboleths of eliminating poverty, as we adopt new goals of better and broader financial inclusion. But this important goal requires a parallel change in the standards and metrics that will tell us how well we are doing. We can’t expand financial inclusion without being able to measure it.

Comments (3)

Getaneh Gobezie

Dear Daniel, Thank you!! I am just sharing my notes on the Ethiopian study (one of the six studies). ... I already shared this note with the Team which did the study (i.e Alessandro Tarozzi (email withheld), Kristin Johnson (email withheld), and Jaikishan Desai (email with held). My key point is that one of the reasons for low/no impact IS the 'TARGETING’ problem -- i.e areas targeted for 'microcredit' programme could more appropriately be targeted for a 'safety-net' (or BRAC-type ''Graduation'' Model... I have also provided an earlier study in Ethiopia (which applies the AIMS/USAID approach -- rather than the Randomized Control Trial method), presented at a UN-FAO workshop, Rome, Italy .... Please read below: ////////////// Dear colleagues, I read (with interest) your recent study:- ‘’The Impact of Microcredit: Evidence from Ethiopia’’. This is one of the six studies using the Randomized Controlled Trials (RCT). I believe this is a big input to the microfinance sector, where there are a lot of controversy and confusion on impact. I am particularly interested in this because I have worked for ACSI (2000 – 2009) at management level, coordinate some studies, including coordinating the ‘’integrated’’ approach (microfinance + family planning) which was supported by the donor Packard Foundation. I am highlighting some issues – which I believe might have affected the outcome of your impact study: 1) Study areas are not necessarily ‘representatives’ of all ‘’microfinance’’ operational areas. When the areas for implementing the integrated programmes (microfinance + family planning) were selected, those selected areas were among the poorest – especially because the organizations implementing ‘’Family Planning’’ (i.e ODA + ADA) were working prioritizing those poorest areas. For example, North Wollo (in Amhara) is one of the areas in Ethiopia traditionally affected by food in-security, and also famine in 1974, 1984/5, etc, etc. So we can expect that microcredit impact in those areas can not be so high – as most of the target people may be the right targets for ‘safety-net’ programmes, or BRAC-type ‘’graduation model’’, than for mainstream ‘micro-credit’. 2) As your study clearly highlights, I also know that partner MFIs (may be, especially ACSI) did not always comply with the experimental protocol. So, may be, we can NOT say that villages in the ‘control’group did not have access to loan from ‘formal’ sources (including from ACSI & OCSSCo) or others, like cooperatives. 3) Attached is an earlier impact study (for ACSI) presented at a workshop organized by UN-FAO in Rome, Italy. … This study follows a different (but practical!!) approach (than the RTCs) to studying microfinance impact. As your study also highlighted, one key problem on microfinance impact assessment is ‘selection bias’. Following the AIMS/USAID methodology (AIMS = Assessing Impact of Microenterprise Services), this can best be managed by using ‘incoming’ (first cycle) microfinance clients as comparison (or control) group. That is, to measure the difference in outcome between an individual who received the services from the programme and what the outcome would have been for the same individual if he or she had not received the service, we take as control group incoming clients -- those who chose to receive the service and therefore are not inherently different from the ‘mature’ programme participants (those who have been with the programme for over five years) in terms of entrepreneurial abilities, willingness to take risk, ability to find and use information, etc; and yet, since they are just joining they do not possibly observe any impact from the programme. Our study reports some positive changes on participating clients on different livelihoods outcomes. This study can also be viewed from UN-FAO web-site: http://www.fao.org/ag/rurfinconference/docs/papers_theme_5/impact_assess... I hope this helps. And I would be happy to hear from you. [FYI -- This comment has also been posted at NEXTBILLION: ….. ] http://nextbillion.net/blogpost.aspx?blogid=5308 Thanks Getaneh Gobezie (email with held)

Venkateswarlu prerepa

Opportunity at the bottom of the pyramid. If the micro finance is taken as a business opportunity how does poverty alleviation take place? Its again an exploitation of the poor. Thats what happened in Andhra Pradesh of India. How a Micro finance company could get a good rating to go for public issue and its share soarining more than rs1000(Rs 10 Fv). Unless we go with service oriented people for this social causes the situation will not alter. Yes the micro finance is dead.



Microfinance is for access to finance and if there is any transformation of lives, credit should go to the poor client.
If we take example of only two poor who received microcredit from an MFI and have two different outcome or impact on their lives. We can not conclude this result is because of microcredit. There are multiple factors including hard work and entrepreneurship of microfinance clients which is main cause of transformation of their lives. If any microfinance client becomes poorer, there could be many other reasons which contributed this.

Leave a comment

This question is for testing whether or not you are a human visitor and to prevent automated spam submissions.