What Proportion of the Poor are Capital-constrained Entrepreneurs?

Finally it is time to move on from Theme Two, having decided from currently available evidence that when microfinance providers enter the scene, the poor benefit—at minimum—from gaining access to a greater variety of options for financing their consumption needs, their investment opportunities and their response to major costs, both predictable and unpredictable.

  • Does the microfinance option enable the poor to earn more income?
  • To what extent do they use microfinance loans to invest in their own businesses to earn income?
  • How much return do they actually get on their investments in business?

These are key questions for Theme Three of Freedom from Hunger’s Evidence Project.

The founding narrative for microfinance was that the poor, even those poor enough to be chronically hungry, are entrepreneurial enough to take a loan to start or expand a microenterprise. If the poor are entrepreneurial, they will start businesses when given start-up capital, and they will expand these businesses if given working capital. The principal, most common constraint to their entrepreneurial ambitions was supposed to be lack of access to affordable credit. This set of suppositions created a logic that has driven and justified microfinance for decades now.

This image of the poor entrepreneur suggests that she (or he—but you know why I’ll use “she”) would earn more income for the household if she starts or expands her business activities. And she would be enabled to do this by taking a loan priced so that the interest paid is less than the marginal return on her investment of the loan in her business. As Abhijit Banerjee and Esther Duflos point out so well in Poor Economics (chapter 9), it turns out that marginal return on investment of borrowed capital in a micro-business is often quite high. The ROI is more than enough to cover interest rates that are a lot higher than commercial banks would charge her (if they could be persuaded to lend to her—not going to happen), but still a lot lower than she could typically get from a moneylender.

Making loans to poor entrepreneurs makes sense in this scenario—the poor earn more income and they can repay loans with considerable interest because they invest these loans in profitable businesses. Microfinance is a win-win proposition for both borrower and lender, if the loans are invested in successful businesses.

It seems fairly straightforward to verify the underlying assumptions of this microfinance narrative with real data about loan use, business success, and repayment. But we know better. Nothing in anti-poverty work is straightforward! Let’s see what real data there may be.

Let’s start with this question: “What proportion of the poor are capital-constrained entrepreneurs?”

“Entrepreneurial” is often taken to mean they want to be self-employed business owners, as though they yearn to fulfill a dream to start their own businesses. Certainly some do have such dreams and the will and skill to be successful, if only someone would lend them the necessary start-up capital or the working capital to expand on what they’ve already started. However, most would prefer the stability of a steady job with an established business or the government, as Banerjee and Duflos reasonably assert. They start their own businesses in order to earn income when steady jobs are unavailable.

When referring to most of the poor, the adjective “entrepreneurial” means they are enterprising enough to do what they need to do, and often with considerable determination, to earn income in the absence of or in addition to wage employment or agricultural production. The term “income-generating activities” (IGAs) may be more suitable for “microenterprises” that are not true businesses that will grow to employ non-members of the “entrepreneur’s” own household.

Okay then, let’s ask what proportion of the poor engage in IGAs in which they can usefully invest a loan. From their eighteen-country data set, Banerjee and Duflos (p. 210) provide an average for urban dwellers living on less than 99 cents per capita per day (adjusted for PPP, of course): 50 percent of households operate a non-agricultural business. For the equally poor rural inhabitants, the average is 20 percent (ranging from 7 percent in Udaipur, India to 50 percent in Ecuador).  I don’t know how they define a “business,” but let’s assume they mean what I’ve just called an IGA, because they go on to say that (p. 213) “the vast majority of businesses run by the poor never grow to the point where they start having any employees or much in the way of assets.”

It seems safe to conclude that a lot of households are engaged in IGAs. Regarding the eighteen-country data set of Banerjee and Duflos, I can’t tell what proportion of households was accessing credit from microfinance institutions. On p. 159, they reveal that two-thirds of the poor in rural Udaipur have loans, but only 6.4 percent of these loans were from “a formal source” (the rest were from shopkeepers, relatives and moneylenders, in that order). Note above that 7 percent were operating IGAs. The microfinance narrative might suggest that the frequency of IGAs is related to the frequency of borrowing from formal sources. If there was more microfinance lending, would there be more IGAs? I doubt the connection is so causal in Udaipur—probably a coincidence.

What is far clearer is that a lot of borrowing was taking place for reasons other than starting or expanding an IGA. Given what we learned in Theme Two about financial lives of the poor, this comes as no surprise. A lot of people in Udaipur are borrowing, but few of them are using the loans for IGAs.

What does this mean for the validity of the microfinance narrative?

While we don’t seem to have the data to truly answer the question about the proportion of capital-constrained entrepreneurs among the poor, the Udaipur data usefully leads us to ask further about the proportion of borrowers worldwide who are investing their loans in IGAs.


  • Jan Maes

    I think one of the biggest challenges in trying to answer this question is that linking a particular loan (or savings or grant) to a given IGA is quite impossible. Money is fungilble and IGAs are part of a very diversified household economic portfolio. And there is another issue as well: I have a hunch that the poor often use loans to maintain and repair their most important production capital, i.e. the human one (where maintaining = feeding and repairing = visiting a health clinic or bying medicine). In this case, the loan might have “merely” helped people to keep working as laborers or engaged in already existing IGAs. 

    The best way to research this is to look at the entire household economic portfolio (similar but more involved than financial diaries) and see how the overall leverage (debt/equity) of households relates to total household income. If you consider the household asset portfolio as an investment portflio, we might be able to use the same type of analytics as investors use based on portfolio theory.

    Anyone ever attempted something like this? If not, anyone interested in working on this?  

    • http://freedomfromhunger.org Chris Dunford

      Excellent point about the key use of loans to feed, protect and enhance the human productive assets of the household. The kind of household-level research you recommend has not been done to my knowledge, but my knowledge is still very limited. We would welcome others weighing in on Jan’s question.