The Rise, Fall, and Recovery of Microcredit
The rapid expansion of non-state action in the 1990s was largely driven by the success of credit schemes that enjoyed wide popular uptake, appeared to be tackling poverty, and promised financial sustainability; the increasing focus on credit moved non-state action away from its more radical, organizing agenda and onto a market-oriented pathway marked by the delivery of services. In Bangladesh, a race for expansion among the four big MFIs between 2002 and 2007 meant that 'growth targets over-rode good judgment' (Chen and Rutherford 2013, 11).
As we have seen, the events surrounding liberation drew elite attention to the severe and widespread problem of rural landlessness and poverty, in particular that of women. Muhammad Yunus was among those who focused his efforts on doing something for his country, starting with tiny experiments from which he learned, theorized, and gained the confidence to seek financing. From the USD 27 Professor Yunus lent to families in the 1970s, the sector eventually grew to reach an estimated 36 million borrowers, almost a quarter of the entire population, at its peak in 2009.   More than half of all Bangladeshi households have been involved with microfinance at some point, and almost half still are at present (Raihan et al. 2015). In 2000, 90 per cent of villages had at least one NGO branch present (Gauri and Galef 2005), at least 90 per cent of which were providing credit. The availability of aid, a dense population, good communications, and a 'simple' regulatory framework helped the rapid spread of the anti-poverty technology, in particular encouraging the development of the 'franchise model' of local cost-centre branches (Zaman 2004; see also Jain and Moore 2003). Most organizations also have other education, health, or social protection schemes that depend to some extent on the viability of their microfinance programmes, in particular for the bulky costs of establishing offices and staffing infrastructure. 'Pure' MFIs, such as the Association for Social Advancement (ASA), are some of the biggest actors in Bangladesh and the world.
This vast scale grew out of a process of experimentation with different modes of loan delivery, client, personnel and fund management. As Yunus put it, '[a]n empiricist, I was willing to learn by my mistakes and those of others' (Yunus and Jolis 1999, 65). Sir F. H. Abed similarly describes his many experiments and what was learned from the failures in the early days of BRAC (see Korten 1980). The sector is nothing if not dynamic, and it responds closely to signals from the ground. We can see this in how certain of the more celebrated elements of the original microcredit models have been formally dropped from the theory after having being dropped in practice many years earlier. As anyone who has spent significant amounts of time with rural women in Bangladesh will know, the peer collateral, group lending, and 'conscientizing' activities that supposedly went with the distribution of loans and gathering of savings were quietly shelved many years before, as loan officials and borrowers got down to brass tacks, focusing on the giving and taking of cash. Such ideas as peer collateral were arguably always mere frippery. As Lamia Karim has pointed out, the technology worked chiefly because of the disciplines it applied to poor rural women: it was less out of empowerment than out of lack of power, and in particular due to the culturally specific manipulation of women's shame and family honour, that loans were repaid so promptly (2011).
The story of ASA (which reads as 'Hope' in Bangla) exemplifies the transformation of the NGOs into MFIs, and is its most extreme and its most complete instance of the change. The founder of ASA, Shafiqual Haque Chowdhury, was born into a rich peasant family, and early on became a follower of the 'Red Maulana' Bhashani, alert to the exploitation and precariousness of peasant life (see Chapter 2). From his first job in what became the Bangladesh Academy for Rural Development (which hosted Akhtar Hameed Khan's Comilla experiments), Shafique learned all about the business and politics of rural credit (Rutherford 1995). When Shafique and a group of his comrades set up ASA in 1978 it was to be a radical alternative to the staid developmentalism into which BRAC was already settling, 'something in between an NGO and a people's movement' (Rutherford 2009, 53). But ASA found the struggle to mobilize against landlords and moneylenders fruitless, particularly when the people they sought to organize faced the most basic subsistence crises and frequently had none other than the local patron to turn to for protection. By the mid-1980s, ASA was accommodating itself to working as a conventional service delivery NGO, recognizing 'there was no alternative to credit'; by 1991 it had become a microfinance provider, even jettisoning the 'development education' with which it initially dressed up its loan schemes (Rutherford 2009). By the 2000s, ASA was exporting its methodology abroad, and setting up offices in China, Viet Nam, and Nigeria. The NGO's biographer concludes, in what could neatly summarize the state of play in the sector as a whole:
Just as its core product, the small general-purpose loan, is running up against its limits in its home territory, ASA is gearing up to get it going in the vast space of China, India, Nigeria, and elsewhere. If the assessment of the core product presented in this book is right—a useful service that may not always transform poor people's lives but rarely fails to help them—ASA's internationalization is something to celebrate. (Rutherford 2009, 213-14)
By the 1990s, a micro-industry in the study of the impacts of microcredit had sprung up. Classic studies by Mark Pitt and Shahidur Khandker of the World Bank showed that credit raised household welfare, particularly if given to women, and had some impacts on women's power in the households (Pitt and Khandker 1998). Along a range of dimensions, impact evaluations of
BRAC, Grameen, and the government's Bangladesh Rural Development Board (BRDB) programme found that participation benefited the poor, particularly women and children, in particular if women were the main participants (Pitt and Khandker 1996). However, their results proved sensitive to assumptions and setting, and a careful inspection of their study concluded that 'benefits from risk reduction may be as important (or more important) than direct impacts on average levels of consumption' (Morduch 1999, 1606).
On the issue of women's empowerment, too, there was early contention. Studies challenged the idea that women were being empowered by these loans on the grounds that most relinquished control to their husbands or other family men (most notably, Goetz and Gupta 1996), stimulating a rich seam of debate about the meanings and indicators of women's empowerment (see in particular Kabeer 1999). Naila Kabeer argued persuasively that studies that showed 'negative' impacts on women's empowerment typically did so from an external standpoint that was not always relevant to women's lived realities: 'empowerment contains an irreducibly subjective element' which, when taken into account, indicated a more positive impact on women's power within households (as perceived by themselves) (Kabeer 2001a, 79).
The much-replicated anti-poverty technology hit its high point internationally in 2006, when Professor Yunus and the Grameen Bank he founded won the Nobel Peace Prize. Since then, microcredit's star has fallen far and fast, and its reputation for both financial regularity and a poverty mission has been irreparably damaged, going 'from hero to zero' over a mere decade (Ghosh 2013, 1203). The devastating Andhra Pradesh microfinance crisis of 2010 was particularly damning (Taylor 2011; Mader 2013). Jayati Ghosh describes the spiral of indebtedness in a context of agrarian crisis, absent state intervention, as a kind of Ponzi-scheme in that setting (2013). The 2000s had seen growing interest in 'impact investing' by private investors (Reille and Forster 2008), but it tailed off immediately after the global financial crisis period, since when public funding for microfinance has grown faster than private (Dashi etal. 2013). While the big four MFIs (Grameen Bank, BRAC, ASA, and Buro Bangladesh, providing between half and three-quarters of all lending) relied on commercial funding to expand from the 1990s onwards, they have not sold equity to outsiders, and so 'retained a nimbleness that helped them to adapt quickly and avoid a crisis' (Chen and Rutherford 2013, 7).
Although this matter has as yet received little attention, it is clear that growing reliance on credit for everyday livelihood activities and consumption exposed people in new ways to global volatilities (Constantinou and Ashta 2011; Daher and Le Saout 2015; Duflos and Gahwiler n.d.), leaving those worst equipped to cope with food price spikes and credit crunches at the sharp end of the shock. Arguably, microcredit was most useful to people facing spikes in the prices of food, on which the poor could easily spend 60 per cent of their income, but it helped create a spiral of indebtedness. It has even been argued (although never shown) that microfinance may actually retard rural economic development by starving more productive small and medium enterprises of finance to feed 'unproductive' micro-schemes (Bateman and Chang 2012).
'No evidence of impact' is of course not the same as evidence of no impact. But given that this is a multi-billion dollar sector, it is surprising that the evidence does not stand up to contemporary standards of impact assessment rigour, and significant that those same standards have been applied to challenge some of the more cherished myths about the positive-sum possibilities of global poverty reduction through credit. And the main champions of the view that microcredit contributes to poverty reduction, particularly for women, have by no means given up the argument. A panel study post-dating the critical systematic reviews and replication studies examined the impacts of microcredit over two decades and argued that the results 'unequivocally' find significant positive effects on household welfare, including higher per capita consumption, income, non-land assets, net worth, children's education (boys' and girls'), and reduced poverty, the latter in particular if the loans go to women (Khandker and Samad 2014). They also find, however, that the implications of lending have changed over time, that credit had a greater impact in the past, and that many communities show signs of declining returns, partly because microcredit chiefly financed activities in low-skill and low-technology trading activities, with which local markets are now saturated (Khandker and Samad 2014).
Other recent analysis has concluded that microcredit probably accounts for between 9 and 12 per cent of national GDP and (by a low estimate) between 13 and 17 per cent of rural GDP in 2012 (Raihan et al. 2015). The impacts on rural labour markets are also estimated to be significant, at an additional fifty-three employment days per year—an increase of 19 per cent compared to people who do not take credit (Osmani 2015). Microcredit and rural credit in general has been causally connected to the relatively rapid reduction in extreme rural poverty in Bangladesh in the 2000s, in particular through the rise in leased land for cultivation by landless peasants (Sen and Ali 2015).
In its homeland of Bangladesh itself, all has not been quiet on the microcredit front. After an ill-judged foray into national non-party politics during the army-backed caretaker regime of 2007-8, and criticism over his fund management tactics, Yunus became embroiled in national politics. Proshika, another of the big MFIs, has been greatly weakened, having run similarly afoul of the financial and political rules. Problems of indebtedness started to become obvious to anyone who troubled to look (Karim 2011). And yet, despite the loss of its reputation as a fundamentally sound tool for tackling poverty, it is by no means the end of the story for microfinance in Bangladesh.
The market may be 'saturated', yet it weathered the global microfinancial crisis (Chen and Rutherford 2013). MFIs have diversified and innovated further, now addressing both concerns about the 'missing middle' of small and medium enterprise (SME) finance and the exclusion of the extreme poor from poverty reduction strategies that depend on people's capacity to service debt. As of 2016, the Institute for Microfinance (InM) has been renamed the Institute for Inclusive Finance Development, billing itself as a 'centre of excellence in research, training, education and knowledge management', and boasting an impressive array of evidence on the dimensions and impacts of the technology in Bangladesh.
The shift towards discussing the technology as a tool for 'financial inclusion' has been a worldwide phenomenon. The Microcredit Summit Campaign has now set the target of helping to lift 100 million families out of extreme poverty as its contribution to meeting the Sustainable Development Goals (SDGs), a task it approaches as a matter of global 'financial inclusion' of the more than half of adults in the poorest 40 per cent of the world's population who have no access to financial services.
Experimentalism and learning-by-doing continue to characterize the sector, which has evolved further (at least in Bangladesh) after the global financial crisis of the 2000s (Chen and Rutherford 2013). The sector grew particularly aggressively between 2002 and 2007, but the reverberations of the 2008 global financial (and food) crisis on microfinance saw borrower numbers plateau and loan sizes increase. Big players like BRAC and Grameen have diversified, moving up the value chain, lending more to small and medium business, entering banking proper, creating new financial services such as remittance payment systems and insurance and mobile money. The poverty mission is being tackled with more generous programmes of support to the 'extreme poor', including those affected by climate change. These 'ultra-poor' approaches have now themselves been exported to other countries, but this time are being tested with the apparent rigour of randomized control trials (RCTs) (Hashemi and De Montesquiou 2011), the experimental approach to the measurement of impact at present considered the gold standard for international development policy research.
-  Microcredit statistics across the industry from before the 2000s are hard to come by. The 36
-  million figure is from the Credit and Development Forum (CDF), the microfinance industryassociation: http://www.cdfbd.org/new/ (accessed 11 April 2016). Mixmarket.org figures record
-  million borrowers in Bangladesh as of 2014, from their smaller sample MFIs.
-  http://inm.org.bd/about-inm/ (accessed 11 April 2016).