Perplexed about Overindebtedness, Part 2
The previous post in this series ticked off some reasons why we ought to be looking more closely at overindebtedness in microfinance. But what do we mean when we use that term? This question, like most others about overindebtedness, gets annoyingly complicated.
There is no uniform definition of overindebtedness. The term is sometimes associated with loans that don’t get repaid, or with loans whose repayment ties up more than X percent of household income, or with loans whose repayment becomes stressful for the borrower, or with clients taking loans from too many sources at once. Can we find a unifying concept at some deeper level?
“Indebtedness” refers to loan obligations—that’s easy enough. And the “over” part means too much. But too much from whose perspective? If we’re analyzing from the lender’s point of view, then things are still pretty straightforward. Too much debt = debt that doesn’t (or isn’t likely to) get repaid. Looking backwards at least, determining whether there’s been an overindebtedness problem is a simple matter of tracking default indicators.
But for many of us in microfinance, the ultimate objective is the welfare not of the bank but of the borrower. From this perspective, it seems natural to say that overindebtedness occurs when borrowers owe so much that their loans (or at least their last loan) have made them worse off than they would have been without the loan(s).
[Hmmm…is the overindebtedness question any different from the impact question? Aren’t they both about figuring out what would have happened to the borrower’s welfare in the counterfactual event of not receiving—or having access to—a microloan? More on this later. In the meantime….]
In a borrower-welfare frame of analysis, we can no longer equate overindebtedness with nonrepayment. A borrower may repay a loan but still have been hurt by taking it. Conversely, default leaves more, not less, money in the borrower’s pocket, at least for the short term.
Once more: when we say that we don’t want overindebtedness, we mean that by and large we don’t want clients to borrow so much that they’re making themselves worse off. Why do we need to stick “by and large” in that formulation? To illustrate, let’s assume that a woman wants to buy a sewing machine to expand her little tailoring business. There’s no absolute guarantee that she’ll be able to grow sales enough to pay for the machine, but the woman and her loan officer think the risk is a prudent one, based on her talks with potential customers. Let’s also assume that she doesn’t have enough cash or another source of credit, so she is able to buy the machine only because the local MFI gives her a loan for it. No microloan, no machine (and no risk). Now let’s assume that the woman buys the machine but the expected demand doesn’t materialize. And while we’re at it, let’s further assume that someone steals her machine, leaving her with an extra debt to repay but no extra income or asset to help repay it. Without the loan, none of this would have happened, so the loan obviously made her worse off.
I’m hesitant to regard this loan as a case of “overindebtedness,” because I associate that term with a kind of lending that MFIs shouldn’t be doing. But there was nothing wrong with this loan: we certainly want MFIs and their borrowers to continue taking prudent risks. In this hypothetical case there has been no irresponsible lending, just some bad luck. And there will always be some borrowers hit by bad luck. The only way an MFI can make sure that it is never leaving borrowers worse off is to do no lending at all.
Maybe it’s more useful to think in terms of larger groups of borrowers. Provisionally, let’s end for now with this formulation: “MFIs are contributing to over-indebtedness when an unacceptably high percentage of their borrowers are made worse off because of the loans the MFI gives them.”
What’s an unacceptably high percentage? I think most readers would find 2% acceptable and 25% unacceptable. Narrowing the band more than that would be complicated, and there are enough complications to deal with for the moment.
Next time: various proxy indicators for overindebtedness and what’s problematic about each of them.
PS: I’m far from sure that I’ve got the definitional issues sorted out properly (or usefully) here. Further suggestions from readers much appreciated.