Esther's baby

Libertarian academic Charles Murray did not emerge glorious and triumphant from a lunch with the FT earlier this month. The Browser described the interview thus: “Manages to convey that the esteemed conservative thinker is a lush, a braggart and a bore without saying as much.”

Public image is delicate thing. For once I found myself agreeing with a commenter at Marginal Revolution by the name of Oderus Urungus: “If he [Murray] hasn’t got the sense to present the right public image in light of the public philosophy he advocates, it’s very hard to take him seriously (even if, as I believe, he is substantially correct in his positions).”

MIT economics professor Esther Duflo may have learnt a thing or two from Murray’s encounter as she prepared for her own lunch with the FT last weekend. Prof Duflo appears in any case to be an unassuming person, not dramatically changed by her rock stardom in the field of development economics (one can’t tell, but this is the received impression).

Over the past year/s it has been difficult to avoid *getting to know* Prof Duflo. After the prizes and books there have been interviews and profiles. I did not expect to learn much new from the FT interview. Didn’t she also “thrust her hands into opposing sleeves” and tuck her short brown hair behind her ears during the popular TED presentation? Yes.

So what did the lunch interview method reveal? Well, Duflo had somehow to explain why she would not be following Murray’s error with expensive wine. She will not drink because she is pregnant. After a “pregnant pause”, Duflo let slip she is having the baby with her co-researcher and co-author Abhijit Banerjee. Well, this really was titillating news. The lunch table was on the brink of hysteria when FT’s interviewer assumed an educated pregnant white woman from the upper tribe should be married. Duflo is not. Still, the interviewer’s assumption does suggest that Murray’s message  about social values had got through at the previous week’s FT lunch despite the faux pas.

But had I joined the lunch for all that? I wanted to know what happened outside the tent on the field trips. Duflo and Banerjee are famous for their use Randomized Control Trials (RCTs). Their book documents this work. Charles Kenny  provides a good summary review (worth reading in its entirety, sections quoted below):

Subscribe now

Exclusive explainers, thematic deep dives, interviews with world leaders, and our Year Ahead magazine. Choose an On Point experience that’s right for you.

Learn More

A central theme of Banerjee and Duflo’s work is an effort to take people as they are and the rules as they might be (to misquote Rousseau). And that means much of their work is about better understanding people—the poor themselves, as well as the government officials and non-government actors who provide them services…

What emerges is a rich understanding of the poor not just as agentless, passive recipients, but as people who make complex and demanding choices every day. Poor people are people, too. They want sweets along with spinach, they make mistakes, and they suffer from the usual range of cognitive biases...

Banerjee and Duflo point out that in rich countries parents aren’t allowed to send their kid to school unless they can show that junior has completed his vaccine regimen. No such guide to correct behavior is available to the vast majority of the world’s poor people…

The good news, according to the authors, is that a small incentive can easily change their minds. An RCT they conducted found that a parental reward of a two-pound bag of dried beans, worth less than $2, was enough to dramatically increase immunization rates in the villages…

In a modest tone, Banerjee and Duflo acknowledge that RCTs may not answer the big questions of development, but assert that getting answers to some of the “smaller” questions can make a real and sustained impact on the quality of life of many people…

Similar points come across in the Duflo's FT lunch interview.

The RCT method in development is distinctively and deliberately and perhaps passionately *micro*:

Duflo and Banerjee are students of detail. Is it better to give people mosquito nets or make them pay? What is the best method of getting children into schools, and ensuring that they learn? Should you encourage immunisation by dispatching clinics to villages or reward parents with bags of rice? Or both? Or neither?…

For those who long for simple solutions to poverty, their results are frustratingly intricate. Some ideas work better than others but nothing amounts to a magic bullet...

We conclude by discussing whether her work is frustrating in its lack of simple conclusions. One experiment, for example, raised vaccination rates in Indian villages from single figures to nearly 40 per cent – but still a minority…

“On balance, it is encouraging,” Duflo insists. “The fact that policies often fail for no good reason is annoying, but less depressing than the view that it is a big conspiracy against the poor. Name your favourite enemy – capitalism, corruption ... Our view is easier. You think hard about the problems and you can solve them…

Here was another thing I learned from the interview, Esther genuinely is modest.

Social science in all its forms is fascinated by the micro-macro link. Economics as a discipline divides itself into micro and macro. Periodically debate heats up again. Some days ago Tyler Cowen noted that one or another macro perspective is possible only because it’s built on microfoundations. He wrote: “The microfoundations approach proves its value virtually every day.”

Sociology too has its micro-macro debates. Macro sociology examines the large structures and systems of society; micro sociology focuses more on individual actors and action. Max Weber and Talcott Parsons tried to integrate the two.

My first conscious introduction to macro-micro was discovering a dusty book on Public Choice languishing neglected in the university library. At a time when the teachers were keen to focus on grassroots interventions on one hand and broad categories like ‘state’ and ‘class’ on the other, it was an eye-opener to encounter theory that looked *inside* the state. Public choice seemed to break The State open conceptually into its individual components and motivations and dynamics (e.g. rent seeking and budget maximization). Very useful if you are preparing to go study public policy in a developing country.

Micro and macro are dangerous in isolation. Perhaps the problem I have with the Duflo and Banerjee approach only begins at the point where the micro-macro link moves, or does not move, from understanding into prescription. It is commonly claimed that macro dimensions of development are more impervious to intervention. This is to some extent true. However, until the *big* questions in development are answered, the answers to the *smaller* questions will be partial, piecemeal, quite poor, very slow.

I don’t want to disparage the work of Esther Duflo. My gripe is narrowly the effect on the whole field of development of a grandly self-proclaimed paradigm shift towards empirical trials (RCTs) in an area already overstocked with microlevel and aid-oriented research.

There are a number of well known criticisms of RCTs (including problems in replicating results, and possible confirmation bias in relation to results unfavorable to aid or incentives to study ineffective projects), but little doubt that RCTs really *can* usefully reveal the economic impacts and social cost-benefit of marginal interventions like clean water or dams for irrigation, and the positive or negative effects of aid financing.

Yet arguably the nearest RCTs come to making a truly transformational contribution to country-level development is the evidence they provide that business enterprise (all sizes) is a force for development. This insight should be obvious, but its empirical confirmation apparently does serve a purpose in convincing hesitant investors that there will be a profitable return.

The great limitation of today’s RCTs is the inability of the randomized experiments paradigm to say anything about 200-year experiments with the markets, technologies, and institutions, which eradicate poverty and modernize societies across-the-board. Developing countries need this knowledge more than anything. Do they get given it in comprehensible bite-size pieces?

Even on the ‘big’ questions, it is the proponents of activist states (rather than the market institutional nexus) who have had the greater disproportionate influence in the declining share of macrolevel research in development economics.

You can see now that the title of the post (Esther’s baby) was simply shameless entrapment. What I really wanted to talk about is the micro-macro link in development. Micro is the detail of units; macro is about whole systems. Obviously, Esther’s baby and Charles Murray’s fondness for playing poker are micro details.

When a House of Commons committee investigates micro entrapment tactics at the seedy end of the media spectrum they might not necessarily ask about the macro institutional or cultural reasons why journalists who are all employees of the same press baron behave differently in Australia and the UK.

You can be sure they will not raise questions about perfectly legal lubricated lunches method of entrapment in the highbrow press. Esteemed older white male academics with watery eyes and politically incorrect memories -- beware. Like investing in risk, you only have yourself to blame when it goes wrong.

After prolonged immersion in detail it is wise to step back to fundamental knowledge. The novelist A.S. Byatt said this month that reading biographies of great writers has tended to destroy her relationship with those writers. She goes back to the books the novelists wrote in order to get to know them all over again. Jeez… I know what she means!

After Duflo’s fidgety cuffs and Murray’s fondness for truffles, better go back to their books. When the fieldwork is done, back to theory. And so it goes on.;

Cookies and Privacy

We use cookies to improve your experience on our website. To find out more, read our updated cookie policy and privacy policy.