The Global Campaign for Education – a model of international activism

August 19, 2009

What can you do if teachers don't show up?

August 19, 2009

The backlash against microfinance

August 19, 2009
empty image
empty image

The intellectual battlefield of development is littered with magic bullets. New ideas or technologies such as the internet or mobile phones are picked up, microfinancepromoted as panaceas that will end poverty and transform societies, and then rapidly cut down to size by scrutiny and research. That process seems to be well under way on microfinance.

As it happens, I’m in Bangladesh at the moment, where microfinance, spearheaded by the Grameen Bank , began its meteoric rise. It has been picked up and supported by the World Bank, the UN and increasingly, the private sector and reaching over 100m borrowers by 2004.  Microfinance’s rise culminated with the award of the Nobel Peace Prize to Grameen’s founder, Muhammad Yunus, in 2006.

Cue backlash. The critique has come from several angles. Firstly the evidence. The ‘success’ of MF has largely been judged on the success of the business model (i.e. do people repay their loans?), but does it actually reduce poverty? Two recent papers raise serious doubts, using ‘randomized control trials’ (RCTs) (themselves rapidly becoming a methodological magic bullet). These try and get round some thorny problems neatly summarized by the Economist:

‘Measuring the impact of microcredit is complicated by the fact that the counterfactual—what would have happened to a person who borrowed from a microlender if he had not done so—cannot easily be tested. Many early studies compared borrowers with non-borrowers. But if borrowers are in any case more entrepreneurial than those who do not borrow, such comparisons are likely to overstate hugely the effect of microcredit.’

RCTs, as the name implies, get over this by comparing a randomly selected sample of communities that do receive MF with another group that do not – if the randomisation is robust, any differences between them should be down to the presence (or absence of microfinance).

The results are striking. At least over the initial 18 months, two large RCTs in India and the Philippines found that MF did not reduce poverty.

The India study, conducted by MIT’s Poverty Action Lab (one of the main advocates for RCTs) summarized its findings thus:

‘Half of 104 slums in Hyderabad, India were randomly selected for opening of an MFI branch while the remainder were not. We show that the intervention increased total MFI borrowing, and study the effects on the creation and the profitability of small businesses, investment, and consumption. 15 to 18 months after the program, there was no effect of access to microcredit on average monthly expenditure per capita, but durable expenditure did increase…. We find no impact on health, education, or women’s decision-making.’

MF did have some impact though: pre-existing businesses used it to invest; would be entrepreneurs invested it and then cut back on ‘temptation goods’ like alcohol in order to keep up repayments. Others simply spent their windfalls on food and other consumption goods.

The study concludes: ‘at least in the short-term (within 15-18 months), microcredit does not appear to be a recipe for changing education, health, or women’s decision-making. Microcredit therefore may not be the ‘miracle’ that is sometimes claimed on its behalf, but it does allow households to borrow, invest, and create and expand businesses.’

All this confirms the arguments in a much more comprehensive attack in ‘The Microfinance Illusion’, a draft paper by Milford Bateman and Ha-Joon Chang.  The paper starts off by pointing out that no successful national development stories to date have included a role for microfinance – I think it’s a pretty good rule of thumb that if what you are proposing has never worked at a national scale, you should step up your scepticism levels.

Bateman and Chang also make the point that MF is not ideologically neutral. In fact, MF is rather like the promotion of private individual property rights by Hernando de Soto and others –  ‘by emphasising individual entrepreneurship over all other forms (state, cooperative, etc), the microfinance concept has strong political/ideological serviceability to the prevailing neoliberal/globalisation model.’

So much for the broad brush. Among Bateman and Chang’s more detailed criticisms are:

1. Microfinance ignores the crucial role of scale economies. It is ‘crucially important to invest in small enterprise units (including in agriculture) that can rapidly achieve a minimum efficient scale of operations.’ Instead, MF produces ‘an over-supply of inefficient microenterprises that undermines the development of more efficient small and medium enterprises (SMEs)’. This is a key point – economic dynamism often springs from supporting SME that can grow – MF diverts credit and support from these to much smaller businesses, often in retail rather than manufacturing, which merely flood the cities with tailors and streetsellers.

2. Microfinance ignores the ‘fallacy of composition’. Supporting one streetseller to buy larger quantities of stock at lower prices may make sense for that individual, but if everyone does it, the market becomes saturated and retail prices (and incomes) are driven down. La Paz, Bolivia, where I once calculated that there was one street seller for every 3 families in the city, does not need more vendors.

3. The entry of more commercial MF providers has driven up interest rates – at some point, as in George Orwell’s Animal Farm, the MF providers and the loan sharks start to look indistinguishable. Moreover, commercial providers are often keener to encourage borrowing than saving, even though poor people often want a safe and accessible place to save small amounts, rather than get into debt through microcredit. MF risks encouraging a debt rather than a savings culture.

What do Bateman and Chang suggest as alternatives? Well, how about financial and credit cooperatives, community and state development banks credit unionsor subsidised SME loan and credit programmes. They’ve all worked in the past, but have fallen into disfavour in recent years. Hopefully they can be rehabilitated, now that the market fundamentalism of recent decades is looking so threadbare.

It’s all pretty healthy, really. Development is a messy business. On closer inspection simple solutions are often neither simple, nor a complete solution. Time to work out what MF can and can’t deliver and add it to the repertoire.

And the scepticism has even spread to Bangladesh. Yesterday I met the President of the Bangladesh Economic Association, QK Ahmad, whose own survey of 2,500 borrowers (99% of them women) concluded that microfinance had been grossly oversold. ‘They take out credit, and give it back. That’s all they do’ he said. His book, snappily titled ‘Socio-Economic and Indebtedness-Related Impact of Micro-Credit in Bangladesh’ criticised the ‘severe limitations of such a simplistic approach to poverty alleviation’ which ignored crucial issues such as ‘existing power relations in society.’ Couldn’t agree more.


  1. I guess the problem with magic is that, unless you are a very impressed with magicians, there’s always a feeling of scepticism: what’s really going on?

    I wonder whether the disappointment with micro credit stems from the expectation that it will do more than supporting entrepeneurs. The premise that poor people can get themselves out of poverty by being better in business is flawed in that it ignores (as DG says) that poverty is a structural issue embued with power in its every dimension.

    Wouldn’t it be lovely to find a magic bullet to address structural poverty?

    In the meantime though, is there something in the micro credit idea that sits at odds with the cash transfer element of transformative social protection measures? If people are really poor and vulnerable, let’s see the state supporting citizens to transform their lives without them needing to ‘pay back’.

  2. this is an excellent post thanks
    i work in corproate responsibility where ‘enterprise development’ is a must have for companies with operations in the developing world. As mentioned in the Chang and Bateman paper, developing enterprise (through things like microfinance) is so loaded towards private sector interests it is very blatantly not ideologically neutral. These programmes often form part of CR communications and are NEVER scrutinsed but can have little or negative impacts. Can Oxfam advise all their corporte donors about the ‘mythology of microfinance’?

    cheers Jimmy

  3. Lee, I guess that shows the limitations of economics 101.Reasons include market failure (not enough real competition); the need for commercial providers to make a margin, compared to not for profits, and OK, in some cases a degree of subsidy either via the state or aid donors which keeps rates for not for profit and state MF providers lower.

    Chris, must be tough to watch these things coming round for the second and third time! I recommend self-preserving amnesia so you get just as excited each time.

  4. In this era of the corporate globalization enhanced by structural adjustment programme (since 80s) in poor countries, Micro Credit ‘mechanism’ is a kind of primitive form of accumulation. Micro Credit Institutions (MCIs) like Grameen Bank and BRAC accumulated capital from the poor people, never invested on sectors that generate employment in rural areas, rather collected money from the poor and established business institutions either with big corporations (eg grameen phone, pushti yogurt) or to serve rich people in the cities (BRAC Bank for marriage loan, car loan, appartment loan, BRAC University and may many) etc.

    To my understanding, only two categories of people think that micro credit is pro-poor: criminals and stupids.

    Most people ignore the (violation of) rights perspective of micro credit. The poorest people of the world pay the highest amount of interest rate; while the richest people pay the lowest rate. Find below:

    Effective interest rate actually collected by Grameen Bank is 26.6%
    Rate of effective cost of credit of Grameen Bank 30.5%

    Effective interest rate actually collected by BRAC is 40.8%
    Rate of effective cost of credit of BRAC 44.8%

    Effective interest rate actually collected by ASA is 40.8%
    Rate of effective cost of credit of ASA 44.8%

    Effective interest rate actually collected by Proshika is 38.6%
    Rate of effective cost of credit of Proshika 42.3%

    Once upon a time I thought that these MCIs are (at least) promoting poor peoples’ ‘Right to Credit’. I was wrong. MCIs collect savings from poor, for which they never pay profit to the poor, and invest for its own. In many cases, during the era of ‘Martial Law’ (for enhanced structural adjustment programme) MCIs also received money for investment from government, as well as from donors.

    Never forget that Dr Yunus has received the Nobel Prize for ‘peace’ not for his contribution in ‘economics’.

  5. It strikes me that one could substitute “Fair Trade” for “microfinance” and have a legitimate posting on concerns about that development trend.

    I liked your conclusion, similar to what I try to share with consumers interested in Fair Trade, that “Development is a messy business. On closer inspection simple solutions are often neither simple, nor a complete solution.”

    I’m for keeping microfinance in the tool kit but not letting one development strategy be seen as the “end all be all.”

    Thanks for your good work.

  6. I spoke with Bateman and his comments focus primarily on Eastern Europe. Within that context he has a good point. Money was provided for microfinance – I designed a couple of those programs myself – while the real need in Bosnia and elsewhere in the region was to build middle sized businesses. Bosnia was more like Europe after WWII. The infrastructure was destroyed but the knowledge of how to build effective businesses was not. Effective support to these midsized businesses would have developed local economies more quickly than funding a plethora of micro-businesses that were competing with each other. At Oxfam America through Saving for Change we have initiated a very large scale initiative in rural Mali, Senegal, Cambodia and El Salvador – a totally different setting than Bateman is familiar with – that focuses on efficiently mobilizing the under the mattress savings of the members of self managed group of twenty or so. Members lend this money to each other at interest during the year largely for petty trading and less so for agriculture and also for emergencies, school fees and medicine. All the capital is mobilized internally; there is no loan to the group through an external source.

    With funding from Gates we have built Saving for Change to 300,000 group members organized into 15,000 groups at a cost of about $20 per member in just over four years, a real bargain compared to starting new MFIs that cost $200 or much more per borrower before reaching breakeven. Interest income returns to the group members and the fund plus interest is divvied up to the members who use their payout to invest largely in agriculture. Saving for Change is a response to the credit unions that have fallen out of favor; these groups can best be visualized as “nano-credit unions.”

    The question is if all this helps alleviate poverty. Gates is funding a random control trial study of 500 villages half with Saving for Change groups and half without. In two years we will have a reading on the impact of this methodology where most are illiterate women with five children and where 95% live on less than $1 per day. We will have a fix on changes in saving and lending, business development, food security, use of health and educational services, social networks and women’s empowerment. If this methodology proves effective for this group that almost entirely lacks access to financial services, it will represent a major breakthrough in providing financial services to the poorest.

  7. One explanation for the mismatch between the positive results for microlending on new enterprise formation and the weak effect on poverty reduction is that for every new microenterprise there is a good chance that another one will fail, ending someone else’s livelihood. it is the net creation of new businesses that matter, and beyond that, the proportion of those businesses that are able, and want, to grow beyond mere survival.

  8. Mukta. With the savings led programs such as Saving for Change all the interst income returns to the group. The group often charges high interest on its loans to members because the savings and the accumulated interest is divided among the members typically just before the planing so it can be invested in seeds and food. In Mali the women in the groups receive a 49% retrun on their savings. In Cambodia many of the smaller loans that were once provided by the MFIs are now provided by the group with the interst income rebated to the members often every three months.

  9. Why did you (or anyone) expect one cycle of a small loan to improve all these indicators? Wouldn’t one want to study the impact over a longer period to see how things turn out? Even then, is it possible that we do not expect some indicators to show improvement because of some systemic issues? These questions are important because they need to be answered before one looks at data or an impact evaluation study of the type cited in the posting. I think to use any data or even impact evaluation study, we need to have some kind of well-argued prior thought process in place and state it upfront, and then look at data to test the hypotheses coming from the thought process; otherwise the same data can throw multiple interpretations. In my opinion, in the context of the cited evaluations three important factors would need to be nuanced through a prior thought process: length of time for the intervention, range of services offered, and context of the evaluation.

    If you carefully think about the pathways of impact on health, education, or even women’s empowerment, you may come to a conclusion that the impact on these important areas may take longer, because of the nature of underlying reasons. Frankly, I would have been surprised to see an impact on these variables in one cycle of a small loan, particularly in the context where the India study was situated in.

    Secondly, there is a need to nuance that the evaluations were just for micro credit, and not a range of micro financial services, that could have actually fulfilled all the functions that finance for people. For instance, the impact on health indicators may have been more had a health insurance been offered in the context of the Hyderabad RCT, because some health events are insurable. If savings had been offered, perhaps some more impact would have been there, because people may have been able to save up some money in a liquid instrument (saving account) that could have helped them access services. This is important because we are passing judgments on the importance of access to micro finance as such.

    Thirdly, these assumptions would work in a context where (for example) health services are available immediately, and the only barrier is ability to pay. In many remote rural locations, these assumptions wouldn’t work, and the prior would have to be revised. In sectors like health and education, where public goods and merit goods are involved, the situation is complex and the impact would depend on various factors. So, based on the context, the prior thought process would change.

    This thought process helps us formulate hypotheses that can then be tested using data of the type provided by the evaluation studies. Otherwise, one can do an infinite number of evaluation studies, but still wouldn’t really understand the impact of these interventions. Data doesn’t speak for itself, does it?

    You seem to arrive too quickly at the conclusion that micro finance has no impact, immediately after citing a paragraph from the paper that clearly states that this was a short-term study and that micro credit did have impact in terms of more businesses being opened, existing business improving profitability, and reallocation of consumption expenditure in some cases. Even in the Manila study, the results suggest that micro credit works through risk management and investment at the household level. One cannot say that there was no impact over short term. It would be good to hear what your thoughts on “expected change from one cycle of a small loan” is. And wouldn’t this expectation be different if a range of services were offered, for longer term, in a context where systemic issues are not impeding?

    Would be keen to hear back.

  10. The preference for credit over savings (or hybrid savings and insurance products) is the probably the weakest link in the development of microfinance. This coupled with governance structures that are distanced from local communities has meant the generation of finance solutions for poor communities rather than with or by poor communities. This has led to a model that was ripe for the commercialisation that has strengthened the tendencies towards mission drift.

    Equally, the microfinance communities focus on technical advocacy (what improves the business climate for the industry) rather than public policy (how to build genuinely inclusive financial services) has furthered weakened the potential impact on poverty.

    We need to critically assess the benefits of MF and secure a more reflective and politically engaged industry that has deeper linakages with the communities they aim to serve – placing specific and regulatory obligations on interest , credit recovery, community reinvestment and community participation in governance would be a good placwe to start.

  11. I made similar observations as an intern for BRAC in Bangladesh a few years ago. Microfinance was supposedly the magic pill for the nation’s poverty. Yet it’s now over 30 years since Yunus established Grameen Bank and can we honestly say it’s medicinal effects have kicked in?

    What struck me about the MF schemes is the requirement to initiate repayments two weeks after taking the loan out, hardly enough time to reap any returns, not to mention the high interest rates charged (at least 15%). How could such stringent repayment terms could be conducive to long-term investment?

    There is also great deal of competition between the big three MF agencies- BRAC, Grameen and ASA, which means there’s a scramble for customers- a hard sell. Not everyone is an entrepreneur, yet these agencies are very willing to hand over credit, often enabling individuals to take out a loan to repay another (anyone see the parallels with our very own credit crisis?)

    These observations are only based on what I saw in Bangladesh, where microfinance is big business (it bankrolls many of BRAC’s other programmes). Perhaps smaller MF institutions are more effective at opening up opportunities to people who actually have the entreprenurial skills and imagination to put the money to good use.

  12. Duncan,

    Great post.

    So banking is not helping the poor…

    I never believed it would, anyway, according to me MF is totally overrated.



  13. The great joke is: ‘Dr. Yunus and Company’ always says that the MC is a group-led process. Actually not. Neither the members of the group receive any interest of their savings nor the group own any money (i.e. RLF: ‘Revolving Loan Fund’ consists of profit and savings); all goes to the account of MCIs. MCIs put conditions on the group members that they must not miss any ‘installment’ to get ‘bonus’ on their ‘savings’.

  14. All small initiatives will be destroyed by few large MCIs like BRAC, Grameen Bank and ASA as the recently passed (by the previous government) MC regulation system will help big MCIs to grab the small ones. This new regulation is the greatest success of Fazle Hasan Abed (Head of BRAC) who successfully influenced the four party alliance government (a coalition of religious fundamentalist parties in Bangladesh) against Proshika (that politicized the MGO sector in Bangladesh) and captured the benifits in favour of big MCIs using smaller MCIs.

    Here I should mention that Dr. Yunus was rejected by people while he wanted to form a political party in 2006/7. Fazle Hasan Abed never moved for political power but achieved many things (not for people) using others. It is very risky for a person in Bangladesh to speack about BRAC and Company because most of the national (big) NGOs are linked with him (unity of convenience: either family relation or business relation).

  15. Yes, I spoke with Jeff Ashe, but he mischaracterizes my work. If anything, the problems with microfinance described in the paper with Ha-Joon and in my other research and consulting work are much worse in non-Balkan settings. In Colombia (where I ran into Jeff) some preliminary work we did in Medellin found very high rates of client exit and displacement, meaning that most microfinance programs have had very limited impact indeed. Yes, they are often financially sustainable: but how many times do we have to repeat that being sustainable is NOT development. I also found this in my work in China a couple years back, where microfinance is being pushed by international donors largely for ideological reasons. The Chinese had great Urban Credit Coops and Rural Credit Coops which were the foundation of the initial success with Township and Village Enterprises in southern China. Getting rid of them to bring in more market-driven MFIs makes no sense whatsoever.

    Working in Cambodia in 2006 I saw how MFIs such as ACLEDA were a complete disaster for the poor, though well-managed, highly efficient and profitable. The problem was they were ignoring the agricultural sector which needs scaled-up loans to really establish productive agriculture, adn focusing instead on petty trade. Cambodia is one huge bazzaar and nothign else – a country simply will not develop on such a basis. Microloans mean they can’t really invest to escape poverty through some serious agricultural development program, but just buy a few tools and some seeds. Its all subsistence farming stuff, which creates little development value.

    In this and other senses, then, Jeff’s Saving for Change repeats almost all the old mistakes that we highlighted in our paper. The poor remain in a ‘poverty trap’ no less than under other competing forms of microfinance. I don’t see any change being facilitated here by his revised project.

    Finally, the control trials mentioned by Jeff will show nothing, another point argued in our paper. Simply dropping cash into one village and not into another will inevitably result in a boost (temporary) in the one community and not the other. So would dropping cash from a helicopter. So this tells you nothing about whether the project is any good, just that more money in Village One produes more wealth than in Village Two where you did not invest any new funds. Might be better to compare to a Temporary Employment Program (TEP) using the same resources, rather than simply compared to ‘doing nothing’


  16. Jeff:

    The joke is wrong (I guess that is why it is a joke). Dr. Yunus practiced what he preached. Grameen Borrowers used to receive 8.5% on their group savings. They could not transfer their savings accumulation in the group fund unless they were member for 10 years. Since 2002, Grameen Bank has gotten rid of group savings. Now the borrowers maintain individual voluntary savings with the bank and they are paid 8.5%, which is higher than the rate paid by the commercial banks.

  17. The flipside of credit is debt. It’s important to remember that when microfinance is “pushed” at the poor, it’s also debt that’s being pushed. While credit is a useful tool if used appropriately, debt is always challenging to manage and can lead families into trouble – no matter where you live or under what terms the debt is incurred.

    One of the things that has always frightened me about the MF industry is, indeed, the business plan that emphasizes institutional sustainability. What should have primary focus, if we are to talk about development, is client sustainability. MF clients actually need to stay poor to continue accessing micro-credit. When/if they do start becoming bankable in a mainstream sense, there is little to no transfer of their credit history into the mainstream banking sector.

    I’ve known several cases where a client’s (forced) savings deposits grew over time to represent an amount that would be meaningful to invest – to withdraw and use that money, however, would mean that the client would have to start over at the lowest credit amount if they ever wanted to borrow again. So instead of investing their own free money, MF clients who’ve been in a system for a while have to leave that money in the system and continue to pay interest on borrowed funds, or lose their credit history. This strikes me as a substantial hidden cost.

    I have been extremely thrilled to learn about Equity Bank (Kenya), who have recently merged with Uganda Microfinance Ltd. Their vision is to grow with their clients by offering a full range of services – from micro-credit to corporate banking. Their emphasis is on helping their clients become sustainable and move up through a system that has no ceilings. That’s where the MF industry needs to go.

Leave a comment

Translate »