The GiveWell Blog

Cash transfers vs. microloans

We’ve written that people in the developing world can get very high returns – in excess of 20% annually (and sometimes much more) – on cash transfers. We’ve previously argued that this is both plausible and empirically supported. However, it raises the question: “If people in the developing world can get such good returns on capital, why not support microlending rather than cash transfers?”

Our answer is twofold. First, the history of, and studies of, microlending are entirely consistent with – and in my view, provide some support for – the notion that people in the developing world can earn high rates of return on capital. The second is that despite this, making loans has a couple of disadvantages that cash transfers don’t have: it may sometimes cause net harm by causing indebtedness (even as it sometimes does great good when people make returns above the interest they owe), and it requires far more overhead to run a microlending operation than to run a cash transfer operation.

Microfinance has advantages too; the conceptual points above do not, by themselves, make the case for cash transfers. However, given the landscape and evidence that we see today, we think there is a substantially stronger case for cash transfers as meeting our criteria.

What does the evidence say about microfinance?
We have de-emphasized microfinance as a charitable cause, because we don’t think that funding microfinance charities is among the most promising ways to achieve humanitarian good in the current funding landscape. But that doesn’t mean we think microfinance is a failure. In fact, we think the following two points are both fairly well-established and quite impressive when taken together:

1. Microlending institutions have repeatedly become self-sustaining and even profitable institutions. Compartamos in Mexico and SKS in India are the most vivid examples of this, as both are now public for-profit companies. Due Diligence, by David Roodman, further discusses the microfinance industry’s success in building sustainable institutions.

The basic model these institutions have followed is that of making small loans to very low-income people, and recouping the loans with substantial interest rates.

2. High-quality studies of microlending have not found systematic reductions in poverty, but neither have they found systematic increases, i.e., evidence of exploitation. Over the last few years, there have been a series of high-quality studies of microfinance, and most have found mixed results, with no robust impacts on consumption in a positive or negative direction. (We have not done an up-to-date review of all such studies, and are currently getting our rough picture of the findings from summaries by David Roodman, available here and here. Note, in addition, that it’s possible that microfinance does have some negative or positive impacts on income and consumption that are simply too small and/or long-term for these studies to find.)

These studies are disappointing for one who expects that microfinance is systematically improving lives, but should be reassuring and encouraging for one who expects that microfinance is systematically damaging them, by exploiting recipients. One possible interpretation of the findings is that microloans help some and hurt others, much like other sources of credit we’re more familiar with.

Taken together, #1 and #2 seem to me to provide suggestive (though not conclusive) evidence that very low-income people in the developing world – despite limits to their education, rationality, etc. – are regularly (though not universally) able to invest at high rates of return.This is consistent with relevant studies of cash transfers.

Microlending vs. cash transfers
As we see it, microlending and cash transfers have the following advantages relative to each other:

Advantages of microlending

  • Microlending implicitly targets people who expect to be able to pay back loans. It may therefore do a better job of selecting for recipients who can best invest capital.
  • Microlending holds out the potential of creating self-sustaining institutions that don’t need donations to survive (the goal we believe most microfinance charities are focused on) and of reaching more people with the same capital.

Advantages of cash transfers

  • Cash transfers don’t run the same risk microloans do of hurting recipients by encouraging indebtedness at high interest rates.
  • Cash transfers faciltate higher-risk investments than microloans. They thus give recipients more options and potentially open opportunities to earn greater returns, without risking having to default on loans.
  • Cash transfers don’t require nearly the same level of overhead that microloans do. A microlending operation must track and seek repayment from each recipient, something that can be costly both to the institution (hence the need for substantial interest rates despite high repayment rates) and to the borrowers (microloan recipients often have to attend frequent, time-consuming meetings).

Listing these advantages, by itself, does not answer the question of which is better to support, and indeed we think it is something of an open question.

When it comes to our recommendations, we strongly prefer cash transfers because

  • The best available evidence seems to suggest that cash transfers are beneficial – increasing consumption over the short and long run – while the available evidence on microlending does not show clear positive (or negative) impact.
  • We are wary of ventures that combine for-profit and nonprofit motives and have not developed a good process for assessing them.
  • We also put some weight on the argument of Due Diligence (a book on microfinance by David Roodman) that there is already a great deal of money – including charitable money – in microlending, and more is unlikely to have high returns in terms of enabling the building of institutions that would be difficult to build otherwise.


  • I could not agree with you more. Cash transfers are certainly most easy and preferable.

  • Bromberg on January 9, 2013 at 3:30 pm said:

    It seems that microlending provides benefits for those who would otherwise be unable to acquire this type of assistance. Paying higher interest rates is just part of the game.

  • I brought this up on another thread, but I believe Zidisha does a good job of combining the benefits of both methods. Their loans are not guaranteed, so interest rates are about 5%, considerably lower than most microlenders. Loans can be forgiven if the borrower encounters financial hardship.

  • Holden on January 14, 2013 at 2:19 pm said:

    Brian, I’ve taken a look at Zidisha, and I’d note that
    Zidisha borrowers are responsible for accessing the Internet and entering their own profiles and updates (in English), and must be able to document positive borrowing histories and sufficient income sources (see “Who is eligible to apply for loans?”) Given these circumstances, I’d guess that Zidisha borrowers are substantially better off to begin with than typical microfinance clients, and even more well off relative to GiveDirectly recipients. Of course, this wouldn’t in itself make lending to such people a bad idea, but I think it weakens the case for a comparison.
    Zidisha itself takes donations to cover part of its overhead (see “How does Zidisha fund its operating costs?”) So the extent to which Zidisha is creating or facilitating a self-sustaining dynamic – and its efficiency in terms of total donations to total funds made available to clients – is not clear.

  • Holden, thanks for responding. I agree that Zidisha is geared toward slightly more well-off borrowers than many microlenders/charities. I wasn’t attempting to make a direct comparison of the recipients, but rather the financing. I believe that taking donations (they ask for 15% of deposits) to cover overhead is better than charging the borrower additional interest. Also, volunteers translate entries into English.

    I understand that there are certain drawbacks with Zidisha that many charitable givers will not like. I just would like to see it mentioned in microlending articles on this site, as I feel it is a unique alternative to the larger microfinance methods.

  • Ian Turner on January 21, 2013 at 9:08 pm said:

    Brian: From a perspective of opportunity cost, the extra 15% that you give as a donation to Zidisha in order to enable the very low loan interest rates is effectively a cash transfer to loan recipients (as compared to money flows in a traditional microfinance operation). Since, as Holden noted, Zidisha recipients are likely much better off than the average for recipient countries, it’s not clear why additional cash transfers to them should be preferred as compared to the cash transfers to the very poor that GiveDirectly conducts.

    Also, to be a bit pedantic, “unique” does not necessarily mean “better”.

  • Ian: You are correct, which is why I pointed out that Zidisha includes elements of both cash transfers and microfinance. Essentially the cash transfer is hidden in the operational costs, and then the loan can be offered at rates closer to those of more developed nations. I am in no way saying this method is better than GiveDirectly, or that GiveDirectly is a poor choice for a charity. It’s a bit of a blend that offers the first bullet-ed advantage of microlending from the article, as well as the first and second advantages of cash transfers.

    Just my input on the “open question”.

  • Randall on February 8, 2013 at 4:57 am said:

    This doesn’t really conflict with anything you all said, but it’s noteworthy that preventing disasters has a high rate of return.

    That is, if you’re borrowing to get your car fixed so you can keep the job you drive to, or you get medicine so your pneumonia doesn’t get worse, paying a high interest rate can be rational, because the interest is much less than the cost you’re avoiding (lost wages, or the financial and nonfinancial costs of letting pneumonia develop).

    The popular story is that microloans are useful because they boost people out of a stagnant scenarios into better ones. But they might sometimes be useful because they merely boost people from declining scenarios into stable ones. And that’s a big deal too.

    (Portfolios of the Poor mostly left me feeling like even perfect credit and savings would leave the poor without a lot of what folks in the rich world take for granted–insurance is an obvious thing, but there are lots of others. Interested in seeing what microfinance groups do next.)

  • kwame oduro on March 27, 2013 at 3:25 pm said:

    The right meaning of charity is helping people to improve their lives with the only obligation on them being that they do their best to achieve a sustainable possitive change in their circumstance.
    I am all for monitored and down stream support providing cash transfer.

  • Colin Rust on May 26, 2013 at 1:18 pm said:

    What are your thoughts on insurance as a way to address high risk-aversion? So e.g. as I understand it, Third World farmers often under-invest despite high expected returns (e.g. not using fertilizer or even not planting all their land) even when they have some financial resources, due to (understandably) high risk aversion. Micro-finance doesn’t address the issue here (risk-aversion) but cash grants do, which of course is part of the point of this post.

    But free crop insurance might be a more efficient way to address the risk aversion issue. I guess the hard part may be how do you determine when the insurance needs to pay out? (You can have an objective, say weather-based, standard, but then the farmer is exposed to basis risk: what happens if s/he has poor yields despite what would in some model appear to be good weather? Or you can base on the actual yields or sale proceeds, but then how do you guard against fraud?) It seems like another area that might be ripe for innovation, similar to how GiveDirectly found an efficient way to distribute money using the phone system.

  • Holden on May 29, 2013 at 1:05 pm said:

    Colin, I agree that insurance is a potentially interesting area. On the other hand, there is a lot of complexity here, which you allude to. In addition to the issue you raise, there’s also the fact that the consequences of risk aversion play out differently for different people; not everyone is in the same business and faces the same risks. While we have seen some references to studies on microinsurance, we aren’t aware of a strong evidence base for any particular product at this time.

  • williamjacobs on December 20, 2013 at 2:34 pm said:

    Zidisha charges 5% and donations are voluntary.

    5% is NOT a first world loan rate. Lord I wish it were!

    After all the fees and junk, you can’t even get a collateralized mortgage at a rate that low.

    Some borrowers sweeten their deal with additional interest but plenty of Zidisha folk are getting loans at very near the 5% rate zidisha charges to pay overhead.

    Zidisha has another terrific feature. You can get your money BACK. If you don’t have a whole lot of cash, you can loan out temporary surpluses if you’re too nervous to give it away.

    Zidisha also allows tiny amounts to give out where Kiva charges much higher interest, you can’t get your money back, and they require a minimum of $25.

    I’m listening for bad news about Zidisha but they seem to have the best deal in town for altruistically minded people.

  • Ian Turner on December 24, 2013 at 10:50 pm said:

    williamjacobs: It’s not clear to me that your remarks about Zidisha address the topics brought up in the article, or in our earlier discussion. In light of those points, can you elaborate on why you feel Zidisha is “the best deal in town”, especially compared to Givewell’s top charities?

Comments are closed.