The Other Half of the Benefit-Cost Debate

When it comes to costs and benefits, we at FAI tend to focus on benefits. The recent release of the Compartamos microfinance impact evaluation was thus a big event in our office. With our heads in the academic literature, we tend to write a lot about RCTs and other ways to measure benefits of interventions.

We’re contributing to a problem, though. There’s a big danger in conflating impact and value. We can’t say much about  the value of microfinance (or any other intervention) based on benefits alone. The most realistic proposition in favor of microfinance is that relatively small benefits are paired with relatively small costs, leading to a favorable cost-benefit ratio. That’s a hypothesis, of course, and it hinges on a careful reckoning of the cost data.

Oddly, the attention to rigor in cost analysis is far less than the attention heaped on getting impact evaluations right. Studies of cost data are rarely published in top economics journals, and top researchers pay cost data little mind–and when they do, the data are often sketchy. But if your cost estimates are way off, your cost-benefit estimates are way off too–even with the most pristinely measured impact coefficient. There’s a clear attention deficit on the cost side.

I’m in the middle of writing a new paper on subsidies using Mix Market data. It reminds me why economists don’t spend more time on costs–it involves lots of accounting detail. For some folks, the choices are fascinating (and they’re certainly important), but if you’re not the kind of person who gets pleasure from filling out income tax forms, you’re probably not the kind of person who enjoys calculating microfinance subsidies.

The mess that is FSS

The starting point is not promising. The most common measure of microfinance profitability (and, by implication, microfinance subsidy) is the Financial Self-Sufficiency ratio (FSS).  It’s the key indicator in Mix Market analysis, and it gets cited prominently in industry analyses. One problem is that it’s much easier to describe how to calculate the FSS than to describe what it really means.

Here’s how to calculate it. The FSS begins with a microfinance institution’s revenues, then makes a few small adjustments. The larger step is to adjust the institution’s expenses. The most important adjustment is to account for subsidies that arrive via cheap capital from donors and social investors. The adjusted revenue is then divided by adjusted expenses. If the ratio exceeds 1, the institution is deemed “financially self-sufficient”–revenues  are at least as large as expenses.

So what does it mean? What is “financial self-sufficiency”?

The most intuitive interpretation is: The ability to proceed from this point forward (a) without subsidy of any kind and (b) without the need for major operational changes. In other words:  Take away the grants and cheap capital, and an institution that’s financially self-sufficient should be able to continue along its given path.

But if that’s the intended interpretation (and what else would it be?), the FSS ratio fails. The first problem is an inadequate adjustment for what the cost of borrowed capital would be if social investors took their cheap capital away. The FSS calculation uses the benchmark interest rate on deposits in the given country (taken from the IMF database). The logic is that if cheap capital was no longer available for borrowing, then microfinance institutions would take deposits. Maybe some would, if they’re regulated to do so. But, even then, collecting deposits isn’t costless, and the FSS makes no adjustment for transaction costs. The bigger problem is that it’s unclear that deposits would (or could be) an alternative source of capital. Instead, microfinance institutions would have to borrow. The right measure for the alternative cost of capital is thus the country’s lending rate (plus a few percentage points to cover the risk premium). That switch matters a lot to the numbers

The second problem creates much bigger changes to the numbers. The FSS ratio makes no adjustment for returns due to equity-holders. The FSS assumes that holders of equity get zero return, but that can’t be. For the concept to match the measure, equity also needs to get a financial return. Some equity-holders are social investors who, from a financial perspective, are providing an altruistic gift more than making an investment. But not all are purely altruistic, and equity investors certainly would not be philanthropists under a commercial scenario.

(These two issues, and more, are also noted by Manos and Yaron, 2007.)

Why it matters

Put those concerns together, and it turns out that FSS greatly under-estimates actual financial self-sufficiency. That matters, because the current way of measuring FSS yields the finding that microfinance subsidies are mostly small or non-existent. And that’s the key to the claim that: small benefit + small cost = appealing cost-benefit ratio. If subsidies turn out to be sizeable, the claim looks a lot less compelling. More on that–and some numbers–when the study is finished (I hope within the month!).