Second Step in the Microfinance Theory of Change: Invest in a Viable Business – Do They?

The classic microfinance theory of change involves three steps the poor person must take to make this theory true:

1. Take a loan from (or save with) a microfinance institution (or similar entity)

2. Invest the money in a viable business

3. Manage the business to yield major return on the investment.

My last post (# 20) found that many, perhaps most, of the poor do not take Step 1. Still, a very large percentage of the poor do take Step 1, so I move on to Step 2 to ask if they . . .

Invest the money in a viable business

Referring to the potential of loans to the poor, Banerjee and Duflos in their book Poor Economics (p. 213) assert that “… the vast majority of the businesses run by the poor never grow to the point where they start having any employees or much in the way of assets.” I don’t know yet on what data they base this assertion, but Freedom from Hunger’s experience says this is very likely true.

Should we limit the notion of viability to businesses that have “employees and much in the way of assets?”

From the perspective of the very poor “entrepreneur,” a tiny business may generate enough income to make it a viable proposition for the operator—even if it could never grow to a level of commercial viability that would satisfy microenterprise development advocates (more on that in the upcoming discussion of Step 3). Moreover, we cannot reliably distinguish between businesses with potential to grow and those without potential. I’m sure there are studies that offer a proxy measure of commercial viability by sorting a large sample of microenterprises by the number of employees and whether or not these employees are members of the business owner’s own household, but I haven’t found them yet. So, lacking a way to distinguish the viability of tiny businesses, I continue to use the term “income-generating activities” (IGAs) to include all types of businesses operated by the poor, including those without potential to grow and employ non-members of the “entrepreneur’s” own household.

Here, then, is the specific question: What percentage of microfinance borrowers/savers invest their “usefully large sum” of money in an IGA?

You might think studies of loan use (and even savings use) would be a dime a dozen. It seems so basic to ask how clients use your product. The complication is that microfinance providers often prescribe the use of the loan (and sometimes the savings through commitment accounts), either by insisting the client propose a specific business investment or (less frequently) by tying the loan to a particular purchase, such as fertilizer for agriculture, a machine for business or an appliance for the household. Once a cash loan is disbursed, the client has the option of doing something else with the money (fungibility), in part or the whole amount. When loan-use studies are conducted by providers, they are often limited to verifying the use of the loan for the client’s originally proposed purpose. Microfinance providers generally assume that the loan cannot be repaid unless it is used as intended. However, we now know that households can and often do repay part or all of the loan from revenue unrelated to the publicly proposed use. This allows the borrower to divert borrowed money to non-IGAs. So our question is more broad than loan-use verification: How does the borrower or saver actually use the money?

Such studies are surprisingly rare. Larger studies may report loan use as part of their larger research agendas, but often they do not. Here are a couple of directly relevant studies recommended to me by David Roodman.

Helen Todd’s in-depth interviewing of 40 women in Bangladesh found that all 40 committed to use their loans for IGAs (rice-paddy husking, cow fattening, mustard-oil grinding, grocery trade, sale of mustard oil/sweets) but only six (15 percent) actually used the loan for the stated purpose (cow fattening or mustard-oil grinding)(see Roodman’s table 2-2 summary of Todd’s data). This is not to say their actual use was unproductive (land transactions, loan repayment, rice stocking, lending to others, rickshaw purchase, dowry/wedding payment and housing materials), but it is difficult to see in most cases how the investment would generate income, at least to make the weekly loan repayments with interest.

Stuart Rutherford’s similar careful study of 43 MFI borrowers and their 239 loans from Grameen Bank (version II) found: 31% of the loans were used to buy stock for retail or trading businesses or crafts; 15% for asset acquisition and/or maintenance; 11% for on-lending to others outside the household; 10% for paying down other debt; 8% for consumption; and 23% for mixed uses (just under half of these included some kind of investment in assets or business stock). Grameen may not classify “asset acquisition or maintenance” or “on-lending to others outside the household” or “paying down debt” as IGAs, but assets can be used for IGAs, on-lending may generate interest income, and replacing expensive debt with less expensive debt can increase disposable income.

You can see how hard it can be to distinguish an IGA from a non-IGA. But it should be clear from these two examples (both focused on Grameen Bank clients, one before, the other after introduction of the Grameen II policies), that the majority of borrowed cash is used for activities that are not the typical business envisioned in the classic microfinance theory of change.

In their landmark study of Spandana, a very big MFI operating in ill-fated Andra Pradesh (India), Banerjee, Duflos, Glennerster and Kinnan provide valuable data about loan take-up and use. Twenty-seven percent of eligible households took up loans from Spandana or another MFI by the time of the endline survey. Spandana does not insist that loans be used for business purposes; however, 30 percent of Spandana borrowers reported they used their loans for starting a new business and 22 percent to buy stock for existing businesses. Additionally, 30 percent of loans were reportedly used to repay an existing loan, 15 percent to buy a durable good for household use and 15 percent to smooth household consumption.

What is not clear is the degree of overlap among clients investing in new or existing businesses; we know that clients often use loans for more than one purpose. But these India data seem to corroborate the Bangladesh studies: no more than half the clients are investing in business, and these are the clients who experience measureable impacts from borrowing (but more on impacts in the discussion of Step 3).

My next post (Part II of Step 2 of the Microfinance Theory of Change) will offer more study results that may clarify or muddle our understanding of actual loan use.