The AP Crisis One Year On: The NCAER Study

Indian microfinance institutions (MFIs) have been criticized in some quarters for their high interest rates, indebteding clients, and coercive recovery practices. The recent study by the National Council for Applied Economic Research (NCAER), sponsored by the Microfinance Network (MFIN), brings much-needed data into the discussion. Covering over ten thousand households across multiple Indian regions, this study merits attention.

How do some of its assertions hold up?

• The study reaffirms that the banking system has low penetration of credit and that MFIs and SHGs fill an important gap.

Informal lending still constitutes almost half of all loans. However, in regions where either MFI or SHG (or both) types of lending are deeper the level of informal borrowings is lower. The study also confirms the widely accepted notion that the sources of lending vary widely by region.

• The data confirms that MFIs and SHGs reach markedly lower income borrowers; typically with income levels half as high as households who borrow from formal sources.

Bank credit typically goes to better off households. MFI and SHG households have lower incomes even than those relying on informal credit.

• The study is less persuasive that total indebtedness is higher in households using informal credit is less persuasive or that multiple borrowing is driven by small loan sizes alone.

To neutralize loan sizes indebtedness is measured as a ratio of debt-to-income. By this measure households using informal sources are the most indebted. But this does not control for the loan term. Nor does it reflect well weekly or monthly debt service burdens. As CGAP’s recent survey of over-indebtedness illustrates, indebtedness is a difficult concept to capture in a single ratio.

The analysis of multiple borrowing adds some data but leaves the most important questions unanswered. The NCAER data shows that households tend to borrow exclusively from either formal, MFI, SHG, or informal sources. But these do not reflect multiple borrowing within each of these four sources, i.e. it does not measure when one borrower receives loans from two MFIs simultaneously. The authors appear to extend their assertion that multiple borrowing is caused by credit constraints (loans that are too small) beyond what the limited data in the study would allow. This falls well short of a full explanation. It would be important to also weigh supply side factors such as “overselling” as a potential contributor.

On interest rates and costs the NCAER report asserts:

• The transaction costs of borrowing from MFIs are half those of banks and also dramatically lower than borrowing from SHGs
• MFI interest rates are typically 4-6% higher than SHGs
• MFI interest rates are less transparent than credit from any other source

NCAER is to be commended for looking at the total costs to borrowers, beyond just interest rates alone. The cost data shows important variations at the borrower level for the transaction costs of different types of credit. And this speaks more accurately to the real experience of borrowers.

One difficult part of the study to understand is how interest rates were calculated. The annex explaining the interest rate calculation methodology is inadequate when compared to the more detailed supply side analysis done by Microfinance Transparency.

One figure which ought to concern MFI investors, boards and management is the high percentage of MFI borrowers who do not know the interest rate they pay. This is even more stark when compared with formal, informal and SHG loans which tend to be much more transparent.

• The use of agents 3- 10 times more likely in Andhra Pradesh than other parts of India

It is troubling that the use of agents to push MFI loans is substantially higher in Andhra Pradesh than in other parts of India. It is notable that the high use of agents correlates with the crisis in AP, a similar correlation CGAP identified in a Focus Note on Growth and Vulnerabilities in Microfinance in other markets.

• NCAER assertion that MFI clients report to use loans for “productive” purposes at a higher rate still leaves questions.

This assertion appears to go beyond what the survey respondents data supports. The fungibility of capital is extremely difficult to capture through survey questions alone and more detailed diaries-style would be needed to more confidently answer questions about loan use.