Wall Street Journal on Microcredit-Moneylender Links

Ketaki Gokhale, who won a Daniel Pearl Memorial Journalism Internship to work at the Wall Street Journal this year, and who ignited some controversy this summer with a story asserting widespread microcredit overlending in India, has continued to cover microfinance in India. (For more on Pearls, see yesterday's post.) A new article continues on the same theme and may also receive a lot of attention...but maybe not as much as before, since this time, I think, Gokhale has done better at staying within the evidence.The headline (whoever wrote it) nicely emphasizes correlation, not causation: "As Microfinance Grows in India, So Do Its Rivals." True, the statistics at the beginning seem a bit misleading:

Even as the government and nonprofit organizations came together to create the Indian microfinance market in the 1990s, traditional moneylenders' share of total rural Indian household debt grew to 29.6% from 17.5%, according to a government survey. Another recent survey by the Reserve Bank of India found that between 1995 and 2006, the number of registered traditional moneylenders increased 56% to 19,627 from 12,601.

I say "misleading" because most growth in Indian microcredit is extremely recent, as Gokhale documented, largely occurring after the 1990s and even after 2006. Through 2006, self-help groups won far more clients than Grameen-style microcredit. Meanwhile India has changed profoundly in the last 15--20 years, so one can imagine lots of reasons for the growth of moneylending apart from microcredit.That said, the rest of the article is fairly careful to distinguish anecdotes, statistics, theories, and opinions. It raises the possibility that one cause of the correlation between microcredit and moneylending is that people go to moneylenders when they need help making microcredit payments---and there is no question that often happens. Whether this is good or bad is also left reasonably ambiguous, as I think it should be. Read it and watch the slide show too.Commenting on my blog post about Gokhale's bubble article, Hugh Allen pointed me to two excellent studies commissioned by CARE on how families manage debt---one on northwest Bangladesh, one on southeast. For richer insight and stronger evidence on the themes in the WSJ read the executive summary of the first and the conclusion of the second and the gray-boxed anecdotes in both. I just read them now and think I need to add them to chapter 7. [Update: done.]Heck, I'll copy the main points here for you. In northwest Bangladesh:

• Indebtedness is a routine livelihood strategy for the majority of poor households in Northwest Bangladesh. The study documents an expansion of loan activity between 1998 and 2003. Thus, in 1998, less than one-quarter of households had received loans, but by 2003, more than three-quarters of households had done so. This pattern applies across all wealth categories and for loan-taking from both formal and informal sources.
• While the average number of loan episodes per household was slightly higher for always poor households as compared with cyclically poor households, the average volume of borrowing (i.e. total value of loans taken each year) for cyclically poor households was almost twice that of the rest of the sample. There was however relatively little variation between wealth categories relative to the sample mean of 0.47 in the debt to income ratio.
• Nearly half of all loans are incurred by women, accounting for over 60% of the total value of credit. While women contracted nearly 94% of NGO loans and half of all other formal loans, men received over 80% of loans from mohazani [moneylenders] and other informal loan sources. Women therefore play a key role in the acquisition and management of household debt, although the qualitative data suggest they do not always control the use of the loans they take.
• In Northwest Bangladesh, NGOs and the Grameen Bank are the most frequently reported source of loans but the mohazans and friends and relatives are also important, and often complementary, sources of credit. Cyclically poor households reported greater use of NGO and formal bank credit, while increasing percentages of usually and always poor households reported mohazan loans. The qualitative data suggest that households increasingly rely on multiple sources of credit by means of cross-borrowing from different sources.
• Across the sample, 60% of all loans taken were used for productive (i.e. income-generating) investments, with the remainder being directed towards consumption needs. While the share of loans for agricultural and non-agricultural activities (rickshaw or boat purchases, small business) has grown at the same rate as the expansion of credit, the amount of borrowed money spent on consumption, dowry, and re-lending increased disproportionately in recent years. Poorer households reported higher proportions of loans used for consumption needs.
• In the context of limited asset-holdings and low incomes and the reported debt income ratios, this pattern of use suggests that household debt burden is heavy for many poorer households, and that there is a linkage between debt urgency and debt burden.
• As loan-taking is often the first response to crisis, sickness, dowry and wedding costs can lead to high levels of debt. Moreover the livelihoods of many small/marginal farmers depend on access to seasonal credit in the form of high interest mohazan loans. In short, debt burden is often a reflection of household vulnerability in terms of the structural features of rural livelihoods in Northwest Bangladesh as well as vulnerability to unexpected shocks and crises,
• Despite this vulnerability, the great majority of households however do not default on their loans. Rather, they manage their debt burden in different ways, including cross-borrowing and re-borrowing, restricting consumption, re-allocation of household labor, the sale of household assets, and migration.

In southeast Bangladesh:

• There has been a marked increase in the frequency of loan-taking over the period covered by the survey (1998--2003), particularly for usually poor and to a lesser extent for always poor households. This pattern applies for loan-taking from both formal and informal sources, with a pronounced increase in mohazan lending activity in 2003.
• The observed incidence of loan-taking is highest for usually poor households followed by the cyclically poor and then the always poor, and the mean amount of outstanding debt is lower for the always poor as compared with the usually and cyclically poor.
• Nearly two thirds of all loans are incurred by women (92% of all NGO loans but only 32% of mohazan loans), and women play a key role in the management of household debt. The qualitative data however suggest that they do not always control the loan funds.
• In Southeast Bangladesh, NGOs and the Grameen Bank are the most frequently reported source of loans, but friends and relatives and the mohazans are also important sources of credit. NGOs and the Grameen Bank are the most frequent source of credit for usually and always poor households, whereas the cyclically poor are more likely to take loans from the mohazan or from the bank.
• Across the sample some 40% of loans taken were reported to have been used for “non-productive” purposes, i.e. purposes that did not generate a cash return, with many loans being taken in response to crisis. The proportion of loans taken for non-productive uses was higher for always and usually poor households than for cyclically poor households. Furthermore, while the proportion of loans used for productive activities in both the agricultural and non-agricultural sector has grown at the same rate as the expansion of credit, the amount of borrowed money spent on consumption, dowry and re-lending increased disproportionately over the last two years covered in the survey.
• In the context of limited asset-holdings and low incomes, the ratio of outstanding debt to income suggests that household debt burden is heavy for most of the usually poor and always poor households.
• The debt/income ratios do not however capture the linkage between debt urgency and debt burden whereby debt burden is related to household vulnerability. As loan-taking is often the first response to crisis, sickness, dowry and wedding costs can lead to high levels of debt. Moreover the livelihoods of many small/marginal farmers depend on access to seasonal credit in the form of high interest mohazan loans.
• Despite this vulnerability, the great majority of households however do not default on their loans. Rather, they manage their debt burden in different ways, including cross-borrowing and re-borrowing, restricting consumption, re-allocation of household labor, the sale of household assets, and migration.