Postmortem on Sachs Episode
By Russ Roberts
After the Nina Munk episode, I considered inviting Jeffrey Sachs to respond. But I decided against it–he has plenty of outlets to get himself heard and I didn’t think an episode devoted to responding to another episode would be very interesting.
But shortly after the Munk episode, Sachs contacted me and asked if he could appear on the program. Despite my earlier decision, the fact that he asked made it hard for me to say no. My only condition was that he agree to discuss the Millennium Villages Project more generally. Along the way, I would ask him about the Munk episode and give him a chance to respond.
It ended up being a very challenging and rewarding interview as you can tell if you’ve listened. And the Munk episode came up early and often. Not my plan but that’s the way these things work.
It should be obvious that Sachs and I have a different perspective on development and on his work. But this wasn’t a debate. I asked my questions and he gave his answers. I occasionally pressed him with a follow-up. But the way the interview played out captured how I think about EconTalk and you, the audience. I respect you. I expect you to make your own judgments about the quality of the arguments that are made. EconTalk is a ridiculously long format by modern standards. But I still don’t have time to make every point or ask every question. I often hear from listeners who complain that I let a guest “get away” with something. But I can’t argue or challenge every “mistake” or better–every statement that a guest makes that I don’t agree with. That’s your job. That’s where some of the learning takes place–you digesting the responses and thinking about them. I’m hoping some of these EconTalk Extras will help with that process.
Listener Scott Morgan writes in a comment:
Normally Russ is skeptical about the accuracy and meaning of measurements. At the end of the podcast he seems to insist that measurements should be the basis for evaluating this project. Sachs, in this case, is suggesting that results are difficult to measure, not least because of the lack of a control group.
Everyone should be skeptical about the accuracy and meaning of empirical claims. But that does not mean that evidence or facts are irrelevant. If there are no data at all (or no data that confirm Sachs’s claims), Sachs has a problem. He’s just a storyteller. Nina Munk has a similar problem. That’s why I pressed her to explain why she thought the projects are a failure in the villages she saw. Her answer was that she saw no signs of a job market emerging–no employment, no industry. That is damning. But it is not decisive. Maybe that development will happen later. Maybe she missed it. But Sachs has the same problem. Sachs’s claims that everything is going swimmingly has little content without something to back it up. Unfortunately, there are no controls. Some data of some kind are needed to understand whether Sachs’s approach works.
Readers interested in these empirical challenges can find a nice summary of links here, compiled by Gabriel Demombynes. If you want to know more about the claims of Sachs’s group about child mortality, and the finding about mortality falling more slowly in the Millennium Villages than in the countries where they are located, scroll down toward the bottom and you’ll see material on articles in The Lancet, the one by Sachs’s group that had to be corrected and the article that challenged it leading to the correction.
At the end of the interview, Sachs talked about how the project was passing the market test because various governments were interested in adopting the initiatives. This took me by surprise. I had assumed that a successful MVP would encourage additional private donations to help other villages. I wish I had asked Sachs why he thought government would be able to implement various initiatives effectively. This essay by Lant Pritchett gives some idea of the challenges governments have in implementing policy.
I had planned to ask Sachs the following: “What have you learned from the experience? What would you do differently?” Unfortunately we ran out of time. If he comes back in 2016, I’ll still be interested in the answer.
Finally, a number of listeners feel I should apologize or be ashamed of my quote from the Munk interview about Sachs and his hubris smashing people’s dreams. Commenter Jerm pointed out that smashing people’s dreams isn’t the “cruelest thing.” I agree. I should have said “very cruel” not “cruelest.” I got carried away.
Sachs did catch me flat-footed with his challenge for me to retract the claim. I shouldn’t have tried to debate the conditionality of the statement but I had not remembered the context of my statement and went into debate mode which is usually not my style. I should have simply stood by the statement as I eventually did, pointing out that my statement was based on Nina Munk’s research and book. Read the book (or listen to the EconTalk episode). Munk describes a tragedy. Maybe she is wrong. I gave Jeffrey Sachs the chance to give his own view. Listeners and readers must make their own call on who is more convincing.