Waylaid Dialectic

December 11, 2012

Someone’s Making a Mistake Here

Filed under: Random Musings — terence @ 11:53 am

Ed Carr thinks Esther Duflo is mistaken to think that the poor make mistakes, offering as counter evidence work showing male farmers preventing their wives from becoming the primary bread winners. The farmers do this as, were their wives seen to be earning more than them, their (the men’s) status in the community would be challenged and this would likely lead to reduced material welfare. The men’s choices might appear mistaken but once we understand the context they are rational enough.

It’s an interesting example but Carr is the one who is mistaken if he thinks he’s refuting Duflo et al. Mistaken because to Carr’s one example Duflo’s work provides plenty of cases (not chlorinating water for example) where deeper underlying rationality of the sort evident in Carr’s work is very unlikely.

More than that though Carr’s own example has a big ol’ contradiction running through it, which renders it a very poor example for those who might wish to claim we are rational after all.

The problem is as such:

Carr shows that the farmers he researched are choosing something that seems economically irrational (suppressing the earning potential of their wives) but which is actually rational once one considers the incentives they are offered by the informal institutions/norms associated with their communities. Institutions which see farmers loosing status if their wives are earn more than them. Lost of status has a potential land cost and so allowing their wives to earn more makes bad economic sense in the long run. Which is an excellent and interesting insight but one which *does not* afford evidence sufficient for us to conclude that reason has won the day. Does not because the norms themselves (male status being tied to men earning more than women) are profoundly irrational. There’s no good, rational reason, for status to be tied up in him earning more than her, and yet it is.

Sometimes irrational collective outcomes can stem from collective action dilemmas turning individually rational choices into collectively irrational screw ups (i.e. the tragedy of the commons), but it is very hard to see how this could explain the norms driving the actions of Carr’s farmers.

To be clear, Carr’s farmers are hardly unique in this: sexism is quite possibly the most common human cognitive error. And you could find numerous similar examples where I grew up.

And I’m not disputing that their decisions make sense on their own terms. Rather, I’m simply pointing out that the norms (aka informal institutions) shaping their choices themselves appear to be profoundly irrational. Mistaken in other words.



  1. I think you’ve missed the fundamental point of my post – I am not attempting to radically cultural relativist, and in so doing apologize for what I see as extraordinarily problematic behavior on the part of men in the Ghana case. Instead, I am arguing that labeling behavior irrational creates problems of understanding that in turn impact programs and projects such that we get inferior/inappropriate/unintended results. As I said:

    “We can agree or disagree with the premises of these choices, and their outcomes, but labeling them as mistakes creates a false sense of simplicity in addressing problematic outcomes – because people only require “correction” to get to the outcomes we all want and need. This, in turn, rests on/reproduces a sense of superiority on the part of the researcher – because s/he knows what is best (see a previous post on this point here). That attitude, applied to the case above, would not result in a productive project design aimed at addressing income or other challenges in these villages.”

    I believe your last sentence falls more or less completely into this trap – you are making yourself the arbiter of what “mistaken” means. You have that right, but I would argue that this will lead you into complete trainwrecks on the program front. We’ve got about 60 years of development history (and a similar length of time under colonialism) that demonstrates this point…

    Comment by Ed Carr — December 11, 2012 @ 8:07 pm

  2. Hi Ed,

    Thanks for your comment. I’m not accusing you of cultural relativism. My reading of your post was that you were arguing that:
    Duflo et al are wrong to label the choices of the poor mistaken (by which they mean irrational and at odds with their own welfare) because you have an example showing that apparently mistaken choices are justified when set amongst the incentives associated with community norms.

    I think your post was interesting but not the debunking of Duflo that you think it is. Not, because (a) your example is different from many of the examples that Duflo et al have offered, which do not seem to be explicable as a by-product of informal institutions and (b) because there is still irrationality present in your example — the norms themselves are irrational/mistaken.

    As for me making myself the arbiter of what mistaken means, that is true. The moment any of us venture into normative/prescriptive statements we’re doing that. I’m a utilitarian and so in these circumstances would believe that decisions / institutions that are at odds with people’s welfare are wrong.

    Would all this turn my train to wreckage in practice. I hope not. I’m cautious, and believe that aid work often pays far too little attention to context, and overestimates its power to change things. And I think the sort of analysis you’ve done in your post is excellent — exactly what we’d want to know when scoping a project. My only disagreement with you is whether you’re debunking Duflo and whether the decisions you describe are free from mistakes. At one level, individual choices, they are. At another, institutional, they aren’t.

    Comment by terence — December 12, 2012 @ 6:11 am

  3. Dear Terence
    I’ll let Ed answer for himself but I think you may better understand where he’s coming from if you insert ‘reasonable’ for ‘rational’ in what you’ve written above. Consider this: Is dignity rational? Behaviour is complex.



    Comment by Søren — December 12, 2012 @ 9:12 am

  4. Hi Soren,

    Thanks for your comment. Funnily enough, in my PhD work, I much prefer to use reasonable to rational. And I certainly agree that behaviour is complex — indeed that is Duflo’s point.



    Comment by terence — December 12, 2012 @ 9:54 am

  5. Terence and Søren:

    I’m not sure that I was debunking Duflo, et al as much as trying to offer a reframing. I have no doubt that their findings with regard to outcome are rigorous and defendable. Their interpretation of their findings, however, does trouble me. I think Søren’s intervention is useful – it is important to remember that we all can use the same terms in different ways, and therefore misunderstand each other.

    That said, I think we have to be careful about defining other people’s welfare for them. I am not suggesting that the gender inequality we see in the case I presented is OK or should be valorized, but that it needs to be recognized for what it is – a reasonable response to a complex context that cannot be altered through simple interventions…and indeed, I doubt can be altered significantly via external interventions. When we say that this behavior and outcome is not in their interest, we fail to grasp the ways in which the participants in what I described DO see these behaviors and outcomes as in their interest…and this is what we must wrestle with, fundamentally, if we are to address the challenges people face around the world. I think this failure to really understand people’s actions and decisions on their terms is a significant barrier to development effectiveness…and is the source of a lot of the train wreckage that marks the history of development.

    I will say that I appreciate your transparency in terms of your assumptions – whether or not we agree, the very fact that we can talk about the fundamental assumptions behind our analyses will eventually lead to more rigorous and productive outcomes.



    Comment by Ed Carr — December 12, 2012 @ 7:19 pm

  6. Hi Ed,

    Thanks for your comment. While our tastes in development economists might differ, in practice we probably agree on quite a lot. One of the take away points for aid agencies that I offer to donors from my own PhD work on electoral politics in Solomon Islands is that voters aren’t making poor choices that they can be educated out of, rather they are voting in a clientelist manner because that makes sense given their experiences to date and the collective action dilemma they find themselves in in a country without national political movements. So I agree – bad outcomes can come from good choices and as outside actors we need to pay attention to context and not just assume that people are making mistakes.

    However, this isn’t the only cause of train wrecks in development: the overly parsimonious tales of neo-classical economics where people’s revealed preferences map to their real preferences which map to their utility function are near hegemonic in some areas, and very wrong, and Duflo’s work debunking this is very valuable I think.

    Limited internet access for me now — of into the wilds of rural New Zealand.

    Thanks for engaging.


    Comment by terence — December 13, 2012 @ 2:57 pm

RSS feed for comments on this post.

Create a free website or blog at WordPress.com.

%d bloggers like this: