Microfinance charities make small loans to very poor people. The Unit of Caring has a post up answering a reader’s question on microfinance:
intomeans asked: So based on your post about microloans, do you think it's better to give $1000 to one person one time, or to lend it out through microloans and then, as the money's repaid, keep relending it to other people indefinitely? That's the main argument that pushed me to lend through microloans (in addition to giving to charities like AMF), and I don't think Givewell's analysis addresses that.
I think it’s better to give $1000 to one person one time.
The business model of micro loan organization is to loan $1000, take back $1200 if the recipient is able to pay it back, hope the additional $200 covers the money they are spending on identifying recipients and ensuring repayment, and loan $1000 again.
That this constitutes ‘the money doing good indefinitely’ is listed on GiveWell ‘six myths about microfinance’, which also links this really useful article. Basically: there is a lot of overhead involved in selecting and monitoring recipients, such that every time the loan is re-loaned out a significant fraction is lost. Overhead isn’t inherently bad but even if all the loanees repay the loans, it’s misleading to suggest that with some fixed amount of money to start, a microloan charity could make loans indefinitely. And not all the loans are repaid. (And sometimes, charities that report really high repayment rates, higher than American banks achieve, are a sign they have a lot of coercive power to get their money back, not a sign that the program is going brilliantly.)
So, a thousand dollars enables more than one thousand-dollar loan. But almost certainly less than ten, and some of those people repaid at great personal cost and ended up in a worse position than they started in (because they didn’t understand the terms of the loan or similar.)
This seems true as far as it goes - but even if the empirical premise were true, this case for microlending seems pretty weird. This argument for microlending is that each time you make a loan, you help the borrower - and because they typically pay back the loan, you can keep relending the principal, thus continuing to help people.
Let’s think about it disjunctively. For any microloan recipient, either they have a good way to invest the money, or they need the loan for short-run consumption. If they’re financing consumption, then either having to pay back the loan puts them even worse in the hole, or they’re using it for consumption smoothing and what they really need is savings.
If, on the other hand, they have a good way to invest the money, then they might pocket a profit even after paying back the microloan, which can then be lent out to someone else with an investment opportunity, a clear instance of “the money doing good indefinitely.” But what happens if you keep not making them pay you back? If they reinvest that money, then that’s also an instance of the money doing good indefinitely! Reinvestment of earnings is a thing, even in poor places, and so is helping one's neighbors.
When deciding between microloans and cash transfers, you’re not deciding between doing good one time and doing good indefinitely. The only thing that makes microloans feel like the impact is longer-lasting is because you can feel like you’re holding onto control for longer. The charity doesn’t just give the money and go away - at the end of the loan’s term, it gets to decide who gets the money next - and again, and again, and again. [UPDATE: Per Paul's comment below, there are some reasons to think that this kind of control control can be a good thing. My problem is with the assumption that it is.]
You the donor don’t even have the control here. You aren't lending to people you know or have otherwise personally verified can use the money. The only question you get to decide is: should your donation be administered by a big official charity? Or should it be administered by some random person in a poor village who knows the people and situation there? If they end up with a lot of money - and microlending would be a good idea - then wouldn’t the recipient of your cash transfer be motivated to do their own microlending?
The first option, picking a charity to administer your donation, might do better at weeding out obviously irresponsible recipients, but on the other hand, it comes with massive overhead costs that likely outweigh this benefit.
(As usual, the form itself is not the problem. I expect there are cases where microfinance is in fact helping. I expect that most of these are for-profit. The problem is the automatic deference to the form.)
I’m embarrassed not to have noticed this obvious flaw in the argument for microloans earlier. This seems like the sort of pathological thinking Sarah Constantin was trying to describe in Ra. Long-run wealth accumulation due to cash transfers doesn’t count because it’s in the hands of some specific individual as real concrete things. Repeatedly re-loaned microcredit keeps counting because it stays under the control of a large respectable institution, as the abstraction of money.
This is bonkers. It has little to do with doing the most good, and a lot to do with the worship of smooth, respectable official-seeming vagueness. Where else am I still making this mistake?
It's not at all obvious to me that a microloan recipient who becomes prosperous would choose to become a microlender himself. And that seems like the crux of your argument, right? Are you assuming that microlending is the most profitable choice for a small entrepreneur, and that small entrepreneurs necessarily will perceive that fact and act on it? I think I disagree with all of those premises.
More generally, this post seems like an example of what I feared would happen with Sarah's "Ra" concept: whenever people don't immediately understand the reason for our society doing X, they assume the reason must simply be prestige. It reminds me of the way many rationalists quickly jump to "civilizational inadequacy!" as an explanation for some feature of our society that seems suboptimal to them, but they're actually just missing the nuances that make that feature sensible.
I think you're reading me as making a bolder claim than I meant to make.
I'm not saying here that microfinance per se is bad, though there are often reasons to believe that this is the case. I'm saying that I was originally persuaded it was good for reason X, other people also report being persuaded by reason X, and reason X is nonsense.
To clarify about on cruxes, I'm not assuming a small entrepreneur will do microlending per se. More that they'll do something sensible, and that it's not obvious that this is worse in expectation than microlending.
The case for wealth transfers to the developing world (whether by subsidizing loans* or giving money) is that money goes farther there than it does here. Putting aside local macroeconomic effects, the difference in paper wealth between a poor person before and after a small cash transfer is much, much smaller than the difference between either of them and middle-class people in rich countries. Even if the original recipient holds onto the wealth and reinvests it for personal gain, it still takes a long time before they become wealthy enough that we should no longer be excited about helping them. Moreover, increasing the real wealth of one person in a poor area likely has spillover benefits. Lending to their neighbors is just one among many potential investments they can make with substantial spillover benefits to their neighbors, and one that seems comparatively promising if they run out of ways to make real investments in their own home or business. Higher-wage employment is another potential benefit they could provide.
I'm not sure how to turn this into a crisp "crux," but I think the "keep doing good forever" argument for microcredit over cash transfers depends on the assumption that the efficiency loss in implementing your values from ceding control of the money to some random person in the 3rd world is greater than the efficiency loss from ceding control to a respectable microcredit charity. Seems to me like the reverse could easily be true, so by the same logic that you "keep doing good forever" via microcredit, you also "keep doing good forever" via cash transfers! I meant to bring up the prospect of the recipient lending the money back out as one particularly obvious example, not the only way this could happen.
*Howie Lempel (via private message, named with permission, representing only himself and not his employer) suggested the "subsidized loans" framing, which makes the issue a bit clearer. Instead of thinking of a $1,000 donation to microcredit as $1,000 loaned out over and over, we can think of it as $1,000 distributed in the form of lower interest on many $1,000 loans, at the price of a complicated infrastructure to handle repayment. It's not obvious that this is better than just giving the same number of people the per-loan subsidy now as a cash transfer.
In general, it seems to me that we should expect that the loans that create the most value will be the ones that can also be made on a for-profit basis, so we're at best subsidizing the marginal loans - the loans that are just barely worth it.
I'm not bullish on microfinance, but reason X doesn't seem like obviously nonsense to me. If we think we can help the world unusually much with our dollars relative to dollars in society at large, we should be much more interested in cases where we -- or others who are effectively value-aligned --capture (a fraction of) the returns, compared to the returns being held by another part of society.
This is related to what Julia was getting at: if it *were* the case that everyone reinvested in microfinance, we'd have the effective value-alignment, so reason X *would* be nonsense.
I think that your argument favors for-profit microlending, but not charitable microlending. The prospect of a combined consumer and producer surplus is a pretty strong argument that an investment produces social value on net.
As I mentioned in my other reply to Julia, it seems to me that we should expect that the loans that create the most value will be the ones that can also be made on a for-profit basis. A nonprofit either subsidizes loans uniformly (so it's just a straightforward wealth transfer) or makes loans that couldn't be made on a for-profit basis (so they're loans that are only just barely worth it). Neither scenario makes me very excited about the charitable (i.e. subsidized) component of microfinance per se.
Let's say that (as a rich Westerner) my discount rate is 5% per year. By lending out $1,000 for a year, I've effectively given a $50 subsidy to people in poor countries, in the form of loans that still have very high interest rates. Why should I assume this will work better than giving $50 via GiveDirectly, leading to their receiving a bit over $45 (after overhead) free and clear?* If I keep relending the $1,000, then the net present cost of the subsidy has to be $1,000. If all you wanted to do was make your wealth transfer extend out indefinitely over time, you could instead invest the $1,000, and give away the return on investment as a cash transfer every year.
To make it more like microcredit, you'd have to check to see if you approve of what the recipient did with the money - whether they used it in some way they can't convert back to cash, or kept it in a form they could convert back to cash on the check-in date - and if you don't approve, someone else gets next year's subsidy. You could do a "credit check" at the beginning to see if their past behavior seems "worthy" to you, or you could ask them to join a group of people and pledge to behave the way you want in order to get next year's subsidy, holding the group accountable for each member's behavior. This is, as you may have noticed, a thinly veiled description of what microfinance charities actually do, and what the "keep doing good forever" promise of microfinance means in practice: the donor (or, in reality, the respectable microfinance charity they give to) holds onto the debt as leverage over beneficiaries' behavior. This has a lot in common with Scientific Charity's focus on determining whether recipients' needs were legitimate.
*Plus buying a bunch of high-quality experimentation on what giving $ actually does.
> When deciding between microloans and cash transfers, you’re not deciding between doing good one time and doing good indefinitely. The only thing that makes microloans feel like the impact is longer-lasting is because you can feel like you’re holding onto control for longer. The charity doesn’t just give the money and go away - at the end of the loan’s term, it gets to decide who gets the money next - and again, and again, and again.
I don't think this is the only or main reason. Sorry, this is a sort of rambly comment.
In the past, GiveWell appeared to take seriously studies suggesting that recipients of modest cash transfers make high average returns, I don't remember the exact numbers but I think in excess of of 25%/year, from their transfer. At the very least that's an interesting hypothesis and there continue to be studies suggesting that certain kinds of investment-oriented grants produce returns that high or higher. So we might consider what you would do if it's true.
It might be the case that cash transfer recipients use the money this well and continue reusing it this well indefinitely. If so, then giving $10M to a poor person today will result in $10B in poor-person-income within 30 years, and $10T of poor-person-income in 60 years. More realistically, it implies that even modest donations to the poor today would be sufficient to completely eliminate the kind of poverty that gives rise to these opportunities. Unless you think conditions have changed a lot recently, it also suggests that modest donations to the poor in the past would have been sufficient to eliminate these opportunities.
Frankly this seems implausible to me. It seems much more likely that initial recipients eventually stop being able to put this money to such good use (or even half as good use), such that you need to do additional work to find recipients with these returns, i.e. participants who could pay back a loan. If that extra work was significantly less than the excess returns, then it's very plausible you are coming out ahead. And it seems pretty fair to describe that situation as "the money doing good indefinitely" in a way that wouldn't be true for the cash transfers---you have no evidence that the cash transfers continue to earn interest and do good indefinitely, and there is a strong prima facie case against, whereas if you are able to keep lending in this way and making a profit then you have relatively good evidence that the money is doing good indefinitely.
The value of microfinance is coming from the fact that you retain control, but if you believe in the high returns achieved by transfer recipients then that isn't just because you prefer have control, it's because you are somehow identifying people with outsized returns. That is, you are giving your money in a way that is doing much more good than the average resources in poor nations, and so you having control is a good thing.
Apart from the picking people problem, if you think that people are getting obscene returns then it seems super reasonable to actually try to test that hypothesis by actually collecting the returns back from them. Ultimately you might decide to switch to cash transfers, but knowing something about people's ability and willingness to take loans seems extremely valuable. In this case their ability to pay back is evidence for the claim that the money is doing good indefinitely, but it still seems like a reasonable reason to prefer microfinance.
More fundamentally, the situation where people can earn 40% expected returns over a year with say 20% default risk is really messed up. I think that expanding access to credit seems like the obvious way to fix the problem, and setting up self-financing institutions does seem like an example of the initial setup work doing good indefinitely. Could also be true for just giving people money, but as mentioned (a) that's basically implausible on its face, and (b) you wouldn't have any evidence that it was happening.
I understand that this argues for philanthropists setting up self-financing microfinance organizations, which does seem like the thing you'd do as a responsible philanthropists. And as a large donor you probably would do this with investment rather than donation, since dubious investments are typically more tax efficient than donations.
I don't think GiveWell's case for cash transfers is really based on these high estimates of returns. Also I think the actual situation is that the apparent evidence for high average rates of return is bogus, and that the case for microfinance is problematic largely because it is based on a hope that doesn't materialize. But if you did take that hypothesis seriously, I do think that trying to build financial infrastructure is a much more reasonable response than giving cash transfers, and that on a standard leftist view of the world the missing ingredient is likely to be capital, so that it is reasonable to talk about the provision of capital doing good indefinitely. It's kind of a crude approximation, reducing a complicated set of issues to slogan form, but it seems accurate as slogans go. I don't see how to describe this as automatic deference to the form, it seems like the slogan is pointing to an important and real reason why setting up microfinance would be much better than cash transfers if in fact microfinance could be self-sufficient (and hence could "do good indefinitely").
I agree that holding onto control can plausibly be a good thing if the microfinance org is good enough at figuring out who has a good investment to make and who doesn't that it can do better than the locals (who may have higher time preference than donors, or other incentive incompatibility problems). This seems like the sort of paternalism that ought to have a high bar to clear, vs letting locals decide for themselves whether they'd rather invest extra wealth in cash-generating enterprises or consume it or do something else with it. It's not impossible that a microfinance org can do better, but it's also far from certain. The pitch in this case is "keep doing good in the future by only funding investment, because the beneficiaries either have high time preference and would just squander the money on consumption relative to your preferred allocation, or have too little financial savvy to relend the money if that's the most efficient use for it."
Also, note that these investments are only ones that generate cash in the short term, and exclude things like e.g. better childhood nutrition, which may have a quite high ROI (I don't actually know) but not one that can be cashed out for decades.
Your view seems plausible prima facie, but anyone holding that view ought to be surprised and confused about why these things are set up as charities if the high rates of expected return are real.
On reflection, "only" is overstated and I'll update the wording.
"If, on the other hand, they have a good way to invest the money, then they might pocket a profit even after paying back the microloan, which can then be lent out to someone else with an investment opportunity, a clear instance of “the money doing good indefinitely.” But what happens if you keep not making them pay you back? If they reinvest that money, then that’s also an instance of the money doing good indefinitely! Reinvestment of earnings is a thing, even in poor places, and so is helping one's neighbors."
The usual idea of credit constraints is that it's *not* especially profitable to lend to people because of various transaction costs or barriers. Thus the rate of return per dollar loaned can rise to a high percentage rate (fruitful investments are left untouched for lack of funds), but the return in absolute terms may be dominated by transaction costs and default rates. Thus payday loans in rich countries are not particularly profitable despite their high nominal interest rates (even though the money is often spent on high-value things that avert greater costs).
Microlending in the developing world is not particularly profitable for similar reasons, although there have been cases where changing conditions and technology have created new entrepreneurial opportunities along these lines.
Paul says many of the things I would say about doubting that the claimed high rates of return from cash transfers or microloans are sustainable for individuals over long periods. If each person has a few highish-financial-return investments to make in themselves or their families, but not that many, then their returns can't compound for long. And they won't be able to make high returns lending to others, b/c the usual costs of moneylending that created credit constraints in the first place bite them too.
And it's not a shocking claim that most people don't give away their income to others as soon as their marginal rate of return on invested funds fall below those others. A poor person who gets a cash transfer and makes some money from investments (at least somewhat more than the transfer+market returns) probably won't then give away all of the proceeds to some other poor person who hasn't made the investments. Indeed, it would be shocking if it were otherwise.
If it's the case that there are high-return investments available to poor people that they take when given cash transfers, and also that there hasn't already been explosive economic growth, then you would get some picture like this (limited room for any high return investments, recipients can't make super-high profits indefinitely as moneylenders anymore than the existing moneylenders, and people won't altruistically give away the bulk of their proceeds to another family).
I agree with Julia about this:
"More generally, this post seems like an example of what I feared would happen with Sarah's "Ra" concept: whenever people don't immediately understand the reason for our society doing X, they assume the reason must simply be prestige. It reminds me of the way many rationalists quickly jump to "civilizational inadequacy!" as an explanation for some feature of our society that seems suboptimal to them, but they're actually just missing the nuances that make that feature sensible."
I expect that this is going to be a difficult disagreement to resolve because we're arguing about what someone else believes. I don't think that all the arguments for microfinance are based on this kind of cognitive error, but I find it hard to believe that the underlying enthusiasm for it isn't due in large part to an unexamined preference for big official-seeming institutions controlling things.
Why wouldn't the same things that make microfinance unprofitable on a for-profit basis make it unprofitable on an utilitarian basis once you've taken into account the opportunity cost of the inputs? Does it seem like the sort of business proposition with unusually high net positive externalities? Enough to justify the overhead relative to cash transfers?
The intuition favoring microfinance on priors (i.e. before looking into the details of what borrowers actually do) is one that assumes that large institutions run by sophisticated Westerners are going to make better decisions than locals - and sufficiently better lending decisions to justify the very high opportunity cost. People who live in the same town or area as potential borrowers have a huge information advantage. So it could easily go the other way: if the richer locals don't lend to poor entrepreneurial locals, perhaps we shouldn't either.
A case where I would expect wealthy foreigners to have an advantage would be in places with bubuti-like norms that destroy profit incentives, such that a debt obligation enforceable by an outside party might help someone hold onto and reinvest capital (because otherwise they'll eventually have to default on the debt). However, it seems tricky to structure something like that in a way that doesn't extract all or nearly all of the surplus thereby produced.
" but I find it hard to believe that the underlying enthusiasm for it isn't due in large part to an unexamined preference for big official-seeming institutions controlling things."
This seems backward in light of Kiva's behavior. It tries to present (and used to aggressively present) an image of you as an individual making a loan to a poor individual in the developing world, without institutional intermediation.
The scandal that early GiveWell drew attention to was the degree to which Kiva gave donors the illusion (despite nonobvious disclaimers) that it was a direct interaction with an individual and not a faceless institution. That's the same sort of appeal as GiveDirectly relative to the broader charitbale world
Further, there's an element of empowering individuals, by giving a loan instead of a handout (a debtor who repays the loan with profits is higher status than a recipient of handouts).
Another element which I have seen is just confusion about what happens to the money. Because one will very probably receive back the amount one provided, the rare defaults and the lack of investment returns aren't salient to many donors, so it feels like one is getting charitable impact 'for free' or at least in a way that exploits mental accounting and loss aversion better. I recently had a facebook discussion with a donor who gives charitable contributions to GiveWell recommended charities and separately loans money on Kiva. They described the loans to Kiva as not costing money, and not being a charitable donation (since the charitable donation is disguised in the terms of the loan).
There is a general market for confusing structures that hide costs and manipulate mental accounting.
When I think about microfinance and people's enthusiasm for it, I tend to think more of things like Grameen Bank, than things like Kiva. Looks like Kiva's only moved about 5% of the amount of money Grameen has ($1 billion vs $18 billion), but it's plausible to me that Kiva's a bigger deal for retail donors.
I agree that blind trust of institutions isn't plausibly the reason for enthusiasm about things like Kiva that are heavily story-loaded. I still think something like the control fallacy is going on here, but not in the way my original post describes.