Skip to content

The Counter-Revolution of Development Economics: Hayek vs. Duflo

This post is by Adam Martin, a post-doctoral fellow at DRI.

F.A. Hayek, well known as a critic of central planning, also criticized what he called “scientism,” a blind commitment to the methods of the physical sciences beyond their realm of applicability. In The Counter-Revolution of Science, Hayek opposed to “scientism” the genuine spirit of scientific inquiry.

Esther Duflo’s emphasis on small-scale experimentation has affinity with Hayek’s critique of grand schemes of central planning. As Duflo said in an interview with Philanthropy Action: “I think another untested and potentially wrong idea is that you have to do everything at the same time or else. This is a pretty convenient untested belief because if you live in that world, it is almost impossible to evaluate what you do.”

But Hayek’s concerns about “scientism” might yet apply to Duflo. She continues in the same interview:

Whereas if you say, I am going to press on this button and see whether it provides this result, you might find there are many things that do work surprisingly well with surprising consistency. So it is not that the world is so incredibly complex that every place needs a unique combination of five factors just to produce anything. I don’t know that we would have been able to say the same thing five years ago, but now we are starting to be in the position to say that a number of things, if well designed, just work pretty well in a lot of contexts.

Hayek, in contrast, argues that sheer, context-independent experimentation is not a viable path to development:

An experiment can tell us only whether any innovation does or does not fit into a given framework. But to hope that we can build a coherent order by random experimentation with particular solutions of individual problems and without following guiding principles is an illusion. Experience tells us much about the effectiveness of different social and economic systems as a whole. But an order of the complexity of modern society can be designed neither as a whole, nor by shaping each part separately without regard to the rest, but only by consistently adhering to certain principles throughout a process of evolution. (Law, Legislation, and Liberty Vol. I, p. 60)

Experimentation, for Hayek as well as Duflo, is the chief instrument of social change. Making experimentation work for development requires institutional feedback mechanisms which can fit together newly-discovered ways of doing things in mutually reinforcing ways. What Hayek defends as “liberal” principles are the ways of coordinating individual experiments in a way that enhances human welfare. Ad hoc, “pragmatic” approaches might solve some local problem, but without coordination with other projects the progress will not be reinforcing and self-sustaining.

Promoting progress is like playing leap-frog in the dark. Big leaps into the unknown can easily end in disaster. Experiments are small leaps. Only when we combine those small leaps together according to some rules do we leap-frog in a definite direction and reinforce each other’s progress, rather than ambling about and running into each other. Without systematic feedback mechanisms that are effective at coordinating different projects, randomized trials are like those small leaps. They might be able to solve particular problems–especially in mitigating the ill effects of poverty–but they would not lead to the self-reinforcing process of wealth generation necessary to eliminate poverty.

Be Sociable, Share!
This entry was posted in Aid policies and approaches, Metrics and evaluation and tagged , , , , . Bookmark the permalink. Follow any comments here with the RSS feed for this post. Both comments and trackbacks are currently closed.

19 Comments

  1. Frank wrote:

    That post is awfully dense; you lost me at “affinity.”

    Posted June 2, 2010 at 2:57 am | Permalink
  2. gregorylent wrote:

    some good concepts floating here, but the soup is an example of academicism

    Posted June 2, 2010 at 9:18 am | Permalink
  3. A thoughtful and penetrating post, thanks.

    I don’t see a necessary conflict between Hayek and Duflo. Hayek wants a process of evolution whereby many ideas are independently tried and the bad ones weeded out by mechanisms of accountability. I don’t believe that there is a fundamental reason why randomized controlled trials cannot serve well as one way to try things and generate information to help weed out unsuccessful ideas.

    Scientism would hold that after a few randomized trials “the answer” would be known and then some authority would impose that answer, but to me Duflo is not saying that. In the passage cited above, to me she is saying that we don’t need one randomized trial for every possible context before we have generated useful information about whether an idea should or should not get weeded out.

    All that scientific experiments do is generate information; and in some contexts they do that more reliably than many other methods. Whether that information becomes an input to scientistic central planning, or to a decentralized competition of ideas and results, is not inherent to experiments. I doubt that Duflo would espouse the former; I doubt that Hayek would reject the latter.

    Posted June 2, 2010 at 9:29 am | Permalink
  4. Brian Clendinen wrote:

    See, I do not think Hayek was ever against templates or best practices. He argued for many very broad principles.

    What I think is what Duflo is saying that we now have research that gives us a set of best practices for specific categories. Much like each industry has a set of best practices, and a well run firm in a industry will use a majority of the best practice and modify a few to fit their specific business model. 40 year ago relatively this was a lot less true especially when it came for industry specific practices. Now that is not to say the details of how they implemental the best practices are different in many important ways even today. So Heyka is still right in that one cannot take practices for a specific category and say use these and do not deviate from them if you are in this category. However, I think Duflo is saying 40 years ago we only had broad best practices, now we have better research and know in a higher level of detail the differences in best practices between specific categories (industries). We do not have to spend as much time trying to customize solutions because we now have more details at the lower levels which give us good solutions.

    There is a continuum and the real question is where one is on it for a specific issue. The real question is one can only go so far down this continuum to scientism before Haykas points kicks in and provides failure yet again.

    Posted June 2, 2010 at 9:34 am | Permalink
  5. Greg Ransom wrote:

    If you read closely Hayek implies that the “scientistic” picture of science (e.g. held by Mill or Neurath or Carnap or Samuelson or Friedman) was a false picture even for physics. It was just a botched understanding of what science does, based on a false understanding of the hard sciences.

    An attempt was made to impose this false picture of the physical sciences on other fields, such as economics.

    So, “scientism” as Hayek defined it was a false picture of science derived from a bogus understanding of physical science imposed on the social sciences.

    Hayek’s exemplars of this are Bacon and Neurath — who Hayek directly says didn’t understand physical science.

    Hayek liked Kuhn and Popper and Polanyi because they exposed the bogus picture of physical science held by so many in the 20th century.

    You write:

    “[Hayek] criticized what he called “scientism,” a blind commitment to the methods of the physical sciences beyond their realm of applicability. “

    Posted June 2, 2010 at 1:02 pm | Permalink
  6. Greg Ransom wrote:

    Hayek talked about a false image of physics as a two or three variable science with linear law relations — and the imposition of a crude and false understanding of this image on economics.

    The model is the problem — a failure to look at the nature of the problem to be explained and the causal mechanism that provides a causal explanation for the phenomena. In biology the causal mechanism is open ended and can’t be “limned” in closed terms, i.e. natural selection over rival adaptations is functionally open ended and doesn’t give you simple closed and linearly related set of physically defined causes and effects.

    So to economics. The central causal explanatory elements are entrepreneurial learning in changing conditions and common ways of going on together and rule following, which cannot be defined in a simple closed set of physically defined variables nor somehow related according to a linear equation.

    Posted June 2, 2010 at 1:13 pm | Permalink
  7. Greg Ransom wrote:

    It looks to me like Duflo is looking to uncover “explanations of the principle” of just the sort Hayek recommended in his essays on complex phenomena.

    How else are we to understand what she is doing?

    Posted June 2, 2010 at 1:15 pm | Permalink
  8. Greg Ransom wrote:

    Note well that Hayek attacked Popper’s picture of “science” and forced Popper to acknowledge that his falsifiability criterion of “science” was itself falsified by the examples of Darwinian biology, neuroscience, economics and other sciences of complex phenomena.

    Popper conceded to Hayek in his memoir.

    Hayek explicitly talks about how Popper’s picture of science is itself falsified by the examples of Darwinian biology and economics, etc. — clearly Hayek was not just being ironic in choosing the word “falsified”, but was showing how the Popper “criterion” was false on its own terms.

    Posted June 2, 2010 at 1:42 pm | Permalink
  9. Manuel wrote:

    What’s this? Development Celebrity Deathmatch? What next? Sachs vs. Easterly? (ooops, that would be a re-run)

    Posted June 2, 2010 at 1:57 pm | Permalink
  10. Alvaro Santos wrote:

    Thanks to Brian Clendinen and Greg Ransom for this thoughtful discussion.

    The other three commenters above seem to feel a need to tell the whole world when substantive writing makes their brains hurt.

    Aidwatch: Ignore the goons who are telling you they can only handle fluff. Keep the substantive posts coming. Thanks.

    Posted June 2, 2010 at 2:56 pm | Permalink
  11. April wrote:

    My take is that Hayak’s comments/ insights are not incompatible with RCTs (as a methodology). But they are incompatible with the way RCTs are used in the real world by the randomistas….(yes, now we have the answer…demand curves slope downward, so health promoting goods should be free; or school-delivered worming was highly cost effective in this small group of NGO supported schools in Kenya, so worming is a best buy everywhere – even where public schools are seriously dysfunctional and poorly attended by staff and students).

    Posted June 2, 2010 at 4:54 pm | Permalink
  12. Adam Martin wrote:

    Thanks for the thoughtful comments. Just a couple brief responses:

    I am certainly not accusing Duflo of a belief in central planning. The relevant passages quoted from Hayek are from his critique of ad hoc, pragmatic interventions.

    RCT’s are potentially a great tool for solving particular problems. They give us information of the form: Project A (whether a government intervention or a private initiative) effectively creates outcome X. This is the means-end sense in which they can establish “best practice.” The information it doesn’t give us is how well Project A meshes with Projects B and C, designed to create outcomes Y and Z. For that you need some sort of feedback mechanism, the price system being the prime example.

    Fundamentally what seems to be at stake here is confusing economic (or social) problems with technical problems. RCT’s give us great information on how to solve the latter, not the former. Even though they deal with the effects of incentives, they do in terms of solving a particular problem. Development, I’m arguing, requires more than solving those technical problems.

    All that said, I’m critiquing Duflo’s argument, not her practice. Of course RCT’s have social-institutional feedback in the realm of scientific discourse (“Republic of Science” in Polanyi’s words) and the aid funding world. Showing that the practice of RCT’s won’t cause development would require showing that those sources of feedback are an inadequate substitute for feedback from domestically rooted wealth-enhancing institutions (which may or may not be present).

    Posted June 2, 2010 at 5:03 pm | Permalink
  13. Tim Ogden wrote:

    Adam,

    if Duflo is guilty of scientism, then you are guilty of central planning based on this quotation: “Only when we combine those small leaps together according to some rules do we leap-frog in a definite direction and reinforce each other’s progress, rather than ambling about and running into each other.”

    Ambling about and running into each other has always been the justification for central planning. Yet ambling about and running into each other has been remarkably effective throughout history at generating economic growth.

    If you read further through the interview you quote from, you’ll also see Abhijit Banerjee critique the idea that “big leaps” can be had; specifically, “There is no evidence that big changes were ever wrought by big levers.”

    The scientism Hayek critiqued was the use of limited (and often useless because of poor design) experiments to justify big levers. Duflo and Banerjee are arguing the exact opposite.

    Posted June 2, 2010 at 5:17 pm | Permalink
  14. Greg Ransom wrote:

    Duflo really isn’t looking for explanations — and so isn’t even actually doing science in Hayek’s sense of providing explanations for patterns in our experience.

    Mostly she’s looking to identify patterns, which she doesn’t try to explain. Often the explanations are not a mystery and don’t require the identification of mechanisms unknown to common knowledge already. You don’t need “science” to make sense of the patterned results.

    Some people call the identification of patterns they can’t explain “science” — for Hayek these are just the beginnings, the raw material, for doing science. if patterns in our experience raise problems demanding causal explanation, then, for Hayek, if you can provide those explanations, coming up with causal “explanations of the principle”, then you are doing science.

    E.g. systematic statistical data patterns of the boom / bust cycle raise problems in our experience that call for causal explanation, explanations in terms of systematic causal mechanism operating through actual causal pathways.

    Posted June 2, 2010 at 5:49 pm | Permalink
  15. Greg Ransom wrote:

    If you learn and practice rigorous procedures and formalisms and to care for patients they call it “nursing science”.

    If you learn and use rigorous math and procedures to lay out the parameters of often typical social patterns some people call in “economic science”.

    I think the kind of “economic science” Duflo practices has more in common with “nursing science” than it has to do with, say, the sort of causal explanation that Darwin provided for systematic patterns in biology., or Adam Smith and Friedrich Hayek provide for systematic patterns in the marketplace.

    Posted June 2, 2010 at 6:02 pm | Permalink
  16. Dan Kyba wrote:

    The greatest ongoing experiment is the market economy as it strives towards adaptive efficiency with its constant process of trial, error and imitation of success. There is a constant feedback mechanism between A (investment) and C (the market) which is broken in the NGO world with the insertion of B (the NGO) between A and C. In response to the constant changing environment that is the market economy we create hierarchies to anticipate and plan for change. The overall success of such hierarchies are limited but they are useful enough to earn their keep. If the hierarchy (management) is not particularly useful, the direct feedback mechanism will remove the business.

    In the NGO world, the presence of hierarchies and their tools including RCT also have a limited but necessary use. The problem is that since there is no direct feedback mechanism the not so useful hierarchies and their NGOs will survive far longer than their cousins in the market economy.

    This is especially obvious in business advisory in developing countries where broken down to its simplest, we have businesses, either registered or unregistered, which compete on their own merits, businesses supported by government and businesses supported by NGOs. Anecdotally, the most efficient and effective businesses are those surviving on their own merits, the least effective and efficient tend to be those supported by NGOs. So what to do.

    We are back at the broken ABC feedback loop; if the aid industry can be restructured to tie A and C closer together we will then have a tighter feedback loop within which NGOs might still have a limited use as will RCTs, or a version thereof.

    Posted June 2, 2010 at 8:44 pm | Permalink
  17. Greg Ransom wrote:

    Note that Adam’s considerations apply also to RCT results used in nursing science — RCT proven nursing care practices need to fit within what Hayek called “negative rules of just conduct”, e.g the First, Do No Harm principle. It would be a costly mistake to give up a broad and clear general principle proven by long experience — but by no RCT — for brittle ad hoc measures rigoriously “tested” to work only a handful of unique local cases.

    Adam wrote:

    “Experimentation, for Hayek as well as Duflo, is the chief instrument of social change. Making experimentation work for development requires institutional feedback mechanisms which can fit together newly-discovered ways of doing things in mutually reinforcing ways. What Hayek defends as “liberal” principles are the ways of coordinating individual experiments in a way that enhances human welfare. Ad hoc, “pragmatic” approaches might solve some local problem, but without coordination with other projects the progress will not be reinforcing and self-sustaining.

    Promoting progress is like playing leap-frog in the dark. Big leaps into the unknown can easily end in disaster. Experiments are small leaps. Only when we combine those small leaps together according to some rules do we leap-frog in a definite direction and reinforce each other’s progress, rather than ambling about and running into each other. Without systematic feedback mechanisms that are effective at coordinating different projects, randomized trials are like those small leaps. They might be able to solve particular problems–especially in mitigating the ill effects of poverty–but they would not lead to the self-reinforcing process of wealth generation necessary to eliminate poverty.”

    Posted June 2, 2010 at 11:00 pm | Permalink
  18. Robert Tulip wrote:

    Hayek combined opposition to central planning with the view that setting the rules of the game is the key to economic growth and poverty reduction. Rule of law is about strong government in support of free markets. Duflo says ‘press on this button and see whether it provides this result, you might find there are many things that do work surprisingly well with surprising consistency’. You might also find surprising unintended consequences, including distraction from the focus on reforming the complex institutional context that Hayek argued is foundational for growth. Aid projects can have high opportunity cost. Investment won’t deliver sustainable development without an effective incentive structure embodied in institutions.

    A nice Hayekian line is from the Gospel of the Hebrews quoted by Clement of Alexandria: “He who seeks will not give up until he finds; and having found, he will marvel; and having marveled, he will reign; and having reigned, he will rest.”

    Posted June 3, 2010 at 12:48 am | Permalink
  19. Steve wrote:

    Greg said “Duflo really isn’t looking for explanations . . . [she] isn’t doing science.” A lot of people say that but I don’t think they’ve read all of her papers or esp. her lecture notes.

    I suggest “Giving Credit Where Credit is Due.”

    Also, if the point of this post was that RCTs can’t generate economic growth, that was quite a long-winded way to say it.

    Posted June 5, 2010 at 11:28 am | Permalink

5 Trackbacks

  1. [...] This post was mentioned on Twitter by William Easterly, Ajay Menon, akhmad bayhaqi, AudienceScapes, Nancy Muigei and others. Nancy Muigei said: RT @bill_easterly: Hayek vs. Duflo: our most insightful unreadable post ever http://bit.ly/ds2rTF [...]

  2. By Hayek and Duflo on June 2, 2010 at 10:28 am

    [...] Provocative post speculating on what Hayek would think of the idea of using micro-experimentation as a way of improving aid efforts to help poor people. [...]

  3. [...] Is Hayek incompatible with Duflo? June 2, 2010 | Posted by Greg Ransom You decide. [...]

  4. [...] also The Counter-Revolution of Development Economics: Hayek vs. Duflo (Aid Watch) Tags: Aid, Esther Duflo, TED [...]

  5. [...] at Aid Watch, the blog of the Development Research Institute at NYU, Adam Martin discusses the possible [...]

  • About Aid Watch

    The Aid Watch blog is a project of New York University's Development Research Institute (DRI). This blog is principally written by William Easterly, author of "The Elusive Quest for Growth: Economists' Adventures and Misadventures in the Tropics" and "The White Man's Burden: Why the West's Efforts to Aid the Rest Have Done So Much Ill and So Little Good," and Professor of Economics at NYU. It is co-written by Laura Freschi and by occasional guest bloggers. Our work is based on the idea that more aid will reach the poor the more people are watching aid.

    "Conscience is the inner voice that warns us somebody may be looking." - H.L. Mencken

  • Recent Comments

  • Archives