Microlending Fails

The Boston Globe published an article in September, subtitled, “Billions of dollars and a Nobel Prize later, it looks like ‘microlending’ doesn’t actually do much to fight poverty.” …

Three important randomised controlled trials were unveiled this year. In one, economists … persuaded a lender in Manila to tweak a credit-scoring computer program so that it randomly awarded or denied loans to marginal borrowers. … Male-owned businesses tended to become more profitable after a loan, and female-owned businesses did not. … The loans produced no improvement in diet or income about 18 months down the line.

[In] a second trial … a leading microfinance operator agreed to randomise … The company chose 104 suitable areas of the city but at first only marketed loans in 52 of them. … Households seemed to use the loans to buy more expensive goods and then cut back on everyday spending to repay the loan, but income did not rise, nor were there improvements in health or women’s empowerment. Business owners did manage to improve profits. The time horizon, again, was less than two years.

A third [non-randomized] trial, of a micro-savings scheme in rural Kenya … found that the savings accounts were popular among women and helped them save, invest in businesses, spend more and cope with bad luck. All this was despite the fact that the accounts paid no interest and charged hefty withdrawal fees. …

The reason for the backlash is obvious: microfinance was supposed … to emancipate women, create millions of entrepreneurs and get rid of stubborn stains on your collar. … “Suppose microfinance is not having much average impact on poverty, but is giving millions of people a modicum of greater control over their lives … is that so bad?”

More from Tim Harford.  Gee, another anti-poverty silver bullet turns to rubber.  Guess we’ll need another decade or two to build up another great white hope, before it too disappoints.  Yet we’ve known for many decades how to help the world’s poor while actually benefiting us in the process: allow more immigration.  Yes even immigration has limits, but that’s no excuse for why we haven’t done all we can.

GD Star Rating
loading...
Tagged as: ,
Trackback URL:
  • Sean C.

    a) it’s not clear from the formatting that most of your post consists of direct quotes from the article. You should clean that up

    b) Why do you quote the first two thirds of the article but not the paragraph where he qualifies his statements?

    Microfinance fans should not feel too defensive about these mildly positive results, especially when microfinance itself has passed a market test by growing very rapidly, often without subsidies. All such trials are context-specific and have other limits: the Manila study targeted marginal borrowers, while in the Hyderabad study, Spandana was not the only microfinance lender in town.

    • http://williambswift.blogspot.com/ billswift

      In the twentieth century, recreational drug use grew very rapidly, not only without subsidies, but while being actively prosecuted, does this make recreational drug use good?

      • http://yudkowsky.net/ Eliezer Yudkowsky

        It certainly puts a large burden of proof on anyone who wants to argue that recreational drug use is bad.

      • Joe

        …a burden of proof never provided, save for anecdotes and focus on a minority of users (e.g., the 5% of heroin users for whom use negatively impacts daily living, ability to hold a job, and support a family).

    • anon

      Exactly. It makes ittle sense to expect miracles from any development interventon. Even if all microfinance does is displace inefficient (and thuggish) local moneylenders, this should be seen as a welcome development. Also, some of these studies have reported positive results, despite their narrow, short-term focus.

      Also, it is misleading to say that most microloans are spent for consumption. These borrowers are not eating at restaurants, buying fancy clothes and cars and renovating their homes; they’re most likely spending the money on basic education and healthcare, or buying cellphones or mopeds. This may not be business entrepreneurship, but it is a form of investment, not consumption. What we’re really seeing is the removal of credt constraints.

  • http://www.nancybuttons.com Nancy Lebovitz

    A couple of things were left out of the experiments– the original Grameen bank was set up to free people from dependence on money-lenders who charged very high rates of interest, and there was an elaborate social structure where small groups were responsible for repaying loans. The loans weren’t exactly made to individuals.

  • Michael Turner

    I still remember the good ol’ days when microlending was considered a modestly successful form of aid, instead of how it’s come to be viewed: as some kind of panacea that was also wildly profitable (somehow, after aid administration costs) to boot.

    t also got kind of feministed — sure, women in very poor countries are oppressed by their men, but so are those men, by yet other men, all the way up to the corrupt oligarchs running their benighted countries. Did anyone actually try controlled studies to see whether the lot of women in those societies improved when microfinance was doled out in a gender-neutral manner, instead of in a way favoring women?

  • http://www.nancybuttons.com Nancy Lebovitz

    Sorry, I didn’t read the article carefully enough– the Spadana experiment did use groups. I wonder why it didn’t work as well as Grameen. Was it structured differently, or were the positive news stories cherry-picking the most optimistic anecdotes?

    More generally, why not study outcomes for people who’ve used Grameen, which is the largest and most successful micro lender?

    You seem very quick to believe that some rather small surrogate studies are definitive.

    • Rob

      I wonder why it didn’t work as well as Grameen.

      Or maybe Grameen doesn’t work well. Nice anecdotes aren’t good evidence.

      I’m sure there are a lot of people who would jump at the chance of doing a good controlled study of Grameen Bank’s work. If there’s no study in the works, it’s probably because Grameen doesn’t want it.

  • Curt Adams

    These are very weak studies for criticizing microcredit:

    Study 1: Marginally creditworthy borrowers have marginal benefits from credit. Ooh, big shocker! Look for new groundbreaking research showing marginal investments in factory capacity have marginal effects on profitability!

    Study 2: Grossly underpowered. Has any randomized economic study, ever, shown that an intervention affecting a small minority of households has community-wide benefits? Microlending would have to be the Second Coming to provide significant benefits on that kind of test.

    Study 3: Well, actually, microcredit *did* help.

    And Hayden himself implies microcredit works if rational expectations is true:

    microfinance itself has passed a market test by growing very rapidly, often without subsidies

    You could make a better case that microcredit is approaching the economically optimal limit in some cities, but inadequate power becomes a really big issue there.

  • Dog of Justice

    Yet we’ve known for many decades how to help the world’s poor while actually benefiting us in the process: allow more immigration. Yes even immigration has limits, but that’s no excuse for why we haven’t done all we can.

    I would suggest that “doing all we can”, on the immigration front, might entail FIRST figuring out how to assimilate and educate existing Mexican immigrants effectively, and THEN scaling up the solution we come across.

    California’s budget and K-12 public school performance have been run into the ground. Its middle class quite obviously has NOT benefited from current levels of immigration. This may just be because we’re doing it wrong, but how do you propose to ensure that we won’t continue doing things wrong if we follow your suggestion of increasing immigration even more?!

    • http://hanson.gmu.edu Robin Hanson

      Have a cite showing that Calif schools and budget were destroyed by immigrants?

      • Dog of Justice

        Have a cite showing that Calif schools and budget were destroyed by immigrants?

        Schools: http://www.cde.ca.gov/ta/ac/ap/, click on the “Descriptive Statistics and Correlation Tables…” link, note the -0.73 correlation between “% Hispanic or Latino” and Academic Performance Index (on page 3 of the pdf). Also note that, 40 years ago, metrics of California secondary school performance placed it in or near the top quartile among US states; now, after a vast increase in immigration, it’s hovering around the tenth percentile.

        Budget: There are other factors involved here (e.g. Proposition 13), but what are the demographics of those providing the most tax receipts? How does that compare with immigrant demographics?

        I don’t actually believe that there’s no way to “do it right” w.r.t. Mexican immigration, and by extension non-cognitive-elite immigration from other poor countries (though it’s worth noting that even today, both China and India are poorer on a per capita basis than Mexico). But it’s irresponsible to imply that those blaming California’s problems on immigration are just ignorantly applying anti-foreign bias.

      • Stuart Armstrong

        Schools: http://www.cde.ca.gov/ta/ac/ap/, click on the “Descriptive Statistics and Correlation Tables…” link, note the -0.73 correlation between “% Hispanic or Latino” and Academic Performance Index (on page 3 of the pdf). Also note that, 40 years ago, metrics of California secondary school performance placed it in or near the top quartile among US states; now, after a vast increase in immigration, it’s hovering around the tenth percentile.

        The relevant statistic is whether the non-immigrants are doing worse because of immigration, not the overall score. If an immigrant arrives, and isn’t very good at school, he probably still ends up with a better schooling than back home. This will drag down the overall statistics for his new school, even if everyone is now better off.

      • Dog of Justice

        The relevant statistic is whether the non-immigrants are doing worse because of immigration, not the overall score.

        In a dictatorship, maybe. Not in a democracy where unassimilated children of immigrants (even illegal ones!) have the same right to vote as you do. The negative consequences of this may take a while to fully materialize, but they are all too real.

  • http://www.gulzar05.blogspot.com gulzar

    sometime back I had posted on the same issues based on much the same studies. microfinance remains stuck in the same framework with which it started off, even after a lapse of two decades

    see here
    http://gulzar05.blogspot.com/2009/07/analtsing-tshg-led-micro-cedit.html

  • 2999

    Certainly economic liberalization is a far more potent poverty-reduction tool than immigration. Most poor countries are poor because they have awful laws and governments, so robust markets can’t develop.

    Immigration enables more labor to be utilized more efficiently, but the natural resources of the benighted lands are still being underutilized, so it seems that the best option of all is economic liberalization.

    • http://hanson.gmu.edu Robin Hanson

      Human resources are far more important that natural resources.

      • Stuart Armstrong

        Yep; you just need to look at (most of) those countries “blessed” with oil to see that…

      • Eric Johnson

        Oil (and natural gas) are really the sole truly profitable raw commodity, from what I understand. There are other precious materials, to be sure, but they dont have a large market (though actually diamonds might semi-qualify in that way).

        So, your asterisk on Robin’s statement does not imply to me that there are many other asterisks (as is often the implicit intention in noting an exception to a rule).

  • quanticle

    California’s budget and K-12 public school performance have been run into the ground. Its middle class quite obviously has NOT benefited from current levels of immigration.

    Well, I don’t know how you can justify that assertion, though. California’s government and budget are in dire straits, but there’s no evidence that a high level of immigration caused, or even contributed to the current situation. In fact, from what I can see, it wasn’t the recent immigrants who drove California over the cliff. It was the existing citizenry who voted for many unfunded mandates via the proposition system, while at the same time hampering the state government’s ability to raise funds to enforce those mandates.

    In fact, in this view, the immigrants are hapless victims. They’re being dragged down by a government that they did not (could not) support.

    • Dog of Justice

      In fact, from what I can see, it wasn’t the recent immigrants who drove California over the cliff. It was the existing citizenry who voted for many unfunded mandates via the proposition system, while at the same time hampering the state government’s ability to raise funds to enforce those mandates.

      This does explain part of California’s budget crisis. But certainly not all of it. And it has next to nothing to do with California’s school system.

      In fact, in this view, the immigrants are hapless victims. They’re being dragged down by a government that they did not (could not) support.

      There is a subset of legal immigrants for which I might agree with your “dragged down” comment. However, most of their kids are not themselves dragging down school performance statistics.

  • John Maxwell IV

    Given that the marginal borrowers were *marginal*, you’d expect that we should be close to indifferent between awarding or not awarding them loans–and this result was borne out. That study says little about those that qualify easily for loans. You’d think an *economist* would realize this. 🙂

    It doesn’t surprise me that the loans were ineffective for households.

    • Dog of Justice

      Since we don’t live in a frictionless Coasean bargaining world, there is a difference between “marginal” from the perspective from the bank and “marginal” in terms of overall utility.

  • David E

    Possibly opening a can of worms but what about something that’s somewhat similar to immigration: annexation.

    I’ve little doubt that the lot of people living in Mexican border towns would greatly improve if they could be made part of the USA.

    Not that I’m proposing this as a practical solution in most cases—the governments of the territories being annexed, even if the residents were all for it (and that’s pretty doubtful), would probably tend to frown on the practice.

    • Dog of Justice

      I’ve long thought that our current de facto immigration policy amounts to a slow-motion annexation of Mexico. It’s long past time that we explicitly discuss the pros and cons of this.

      • David E

        But we’re only annexing people. Not territory.

        Sometimes I think, if it could be accomplished (and, politically, that seems dubious), the US should just invite Mexico to become part of the US—perhaps doing it a few small regions at a time over several decades to make it less disruptive.

        Doubtless, though, there are a thousand objections to this.

      • Doug S.

        Extending Pax Americana the Roman way…

  • http://mariocruz.yolasite.com Mario Cruz

    Maybe the missing link is what Hernando De Soto has mentioned in his research (book The Msytery of Capital). In a nutshell he thinks that the foundations for a strong economy is strong laws of property and commercial transactions. If there is microlending without a good underlying framework to do business maybe the results will not be as expected.

  • Robert Speirs

    I thought microlending was to be judged as a failure or success by whether the investors made a profit or not. Isn’t the important advantage of microcredit the ability to spread risk widely? Have the investors profited or not?

  • Julian Morrison

    Business owners did manage to improve profits.

    Why is this not considered an important good result?