What should I ask Abhijit Banerjee?

I will be recording a Conversation with him, no associated public event.  So what should I ask him?

I thank you all in advance for the assistance.


Given his recent NYT op-ed, maybe ask him what explains all the findings that demonstrate the strength of financial incentives, e.g. the tax evasion/avoidance literature, health econ and the effects on physician and patient behavior, etc.

+1. That op-ed seemed highly ideologically-motivated and at odds with plenty of findings that they conveniently omitted.

Yeah. I even saw an article recently (in the NYT no less) about how even small increases in co-pays lead to reductions in patients' pharmaceutical usage (even life-saving drugs that treat leukemia!).


Tyler, the fawning over Banerjee and Duflo implicit in your decision to do podcasts with them right after they won the Nobel Prize is straightup at odds with the central thesis of Small Attachments. I really think you know that. I've read the responses to Lant Pritchett's piece but I think his main objections still stand. At the end of the day liberalization has had an effect on poverty that is multiple magnitudes larger than any RCT-inspired policy could ever any hope for. If the status of Nobel laureates needs to be relentlessly lowered in order to make that fact clear, then so be it.

*Stubborn Attachments, sorry

Perhaps of little consolation to you but both were booked well in advance of the Nobel announcement.

What? I am not in your field, but I kind of get your drift, however, it's a bit too derivative for my non-specialist taste. Are you signaling to TC that you know better than the master?

RTC-inspired policy I infer is the equivalent of a J. Sach's type "Big Push" to alleviate poverty. I take it you believe in the long-run beneficial effects of liberalization, pace however Jupiter's Red Spot and pace dynamic equilibrium (if you're into physics you'll see my drift, it's a bit derivative but so be it).

My own take on poverty is that it's a First World obsession. Most impoverished people are happy, and don't see themselves as somehow blighted. Recall the happiest people are the Nigerians followed by the Filipinos, and they are Third World. Are you, First World Do-Gooder, projecting your own angst onto more impoverished people?

Most impoverished people are happy, and don't see themselves as somehow blighted.

World Happiness Index, 1 to 10:

North America...........7.0
Western Europe.........6.6
Latin America.............6.2
Eastern Europe..........5.8
Russia, Ukraine, etc...5.5
East Asia......................5.3
S.E. Asia.......................5.2
Middle East.................5.0
North Africa................5.0
Sub-Sahara Africa......4.4
South Asia...................4.4

Least happy are between 3.0 and 3.5:

Haiti, Rwanda, Burundi, South Sudan, Botswana, Central African Republic, Tanzania

Bonus: There is no word for "happiness" in Russian.


Hi Ron!

Is there a word for "gullible" in Russian?


The same word also means "trusting"ю

World Happiness Report: "Some argue that questioning on overall life status leads humans to overweight income concerns, rather than happiness. For instance Colombia came 37th in the 2018 World Happiness Index but 1st by daily emotional experience." - ha ha.

That wasn't a quote from The World Happiness Report but by a guy who wrote in on Wikipedia.

One out of the six variables is economic: GDP per capita, social support, healthy life expectancy, freedom to make life choices,
generosity, and freedom from corruption.

If the question is how people see their lives, many of the same bottom 10 come up. South Sudan 3.3, Haiti 3.6, Burundi 2.9, Tanzani 3.4, etc. (Table 2)


Cowen is a contrarian. I admire and respect that. He challenges his students. And I'd like to believe that includes us.

Ask if he dated and married his student before the #MeToo era and whether his expected utility of doing so would be different today.

How would one evaluate the effectiveness of RCTs as a policy instrument? Presumably RCTs deliver useful knowledge that could drive better policy decisions. Could you test that by looking at areas where RCTs were run vs not run in India? Could you do a meta RCT where you actually randomize where the RCTs happen in a certain domain and compare if there was a noticeable difference in an important outcome?

Chicken or pork vindaloo?

What advice would you give a country looking to substantially increase its rate of economic growth (NOT only make poverty less miserable)? Where did the best evidence on this matter come from?

In example after example Banerjee and Duflo point out that influx of immigrants do not put pressure on wage rates. What does this say about the elasticity of demand for labor and minimum wage rate policy?

Ask him if he can think of anything else going on economically in Miami in 1980 to 1984 besides the Mariel Boatlift that might have affected wages relative to Atlanta and Houston?


As below, given the arguments both he and Duflo make in favor of immigration, to the effect that the addition/removal of immigrants has minimal impact on local wages, aren't they implicitly making the argument here that if we’d like to see more mechanization, automation, and the attendant positive impacts on productivity *over the long haul*, we should block low-skill immigration? After all, we have loads of cheap capital available and they're telling us that in the absence of immigrants the employers will tend to automate those positions.

Does he still think that "inequality of access to sex" is a significant problem that should be addressed with policy?

Tyler, consider pushing back on some of the claims made by Banerjee and Duflo in their recent NYT op-ed (https://www.nytimes.com/2019/10/26/opinion/sunday/duflo-banerjee-economic-incentives.htmL). They blame the labor market for workers’ not moving to where the jobs are and conclude (partly from this) that economic incentives don’t really matter:
"In the fantasy world where most economic policy conversations about trade shocks and technological innovations take place, people quickly adjust to those changes — workers move smoothly from making clothes in North Carolina to folding clothes in New York or selling clothes online. But in the real world, it is unreasonable to expect markets to always deliver outcomes that are just, acceptable — or even efficient."

"when jobs vanish and the local economy collapses, we cannot count on people’s desire to seek out a better life to smooth things out…When jobs disappeared in the counties that were producing toys, clothing or furniture, few people looked for jobs elsewhere. Nor did they demand help to move or to retrain — they stayed put and hoped things would improve.."

First, WHO is saying that “markets always deliver outcomes that are just, acceptable — or even efficient”? I've never heard this, not even from libertarians. Market failures are quite literally econ 101, and this is a straw man.

Second, there are at least four major government distortions *of* the labor market that make it harder for people to move to where the work is. Occupational licensing and expanded disability programs restrict the supply of eligible workers. Land-use policy like single-family zoning restricts the supply of houses and apartments. Heavily subsidized driving, roads, and parking restrict the supply of land to build on and ways to build, increase sprawl (building out rather than up), hinder public transit, and increase commute times – all at tax-payers' expense. These four things restrict labor mobility, but they are failures of *government*, not markets.

For workers looking to move, the benefit of a salary at a new job elsewhere isn’t just being compared to the cost of uprooting oneself and one’s family. It’s also weighed against the above-market cost of getting an often-needless license, paying artificially high rent, or wasting away in unnaturally congested traffic. These are economic disincentives, and they matter.

Third, another way that economic incentives matter is in the rising opportunity cost of joblessness. "Around 10 percent of prime-age workers who described themselves as disabled in 2016 had found a job by 2017. "

"Analysis of Current Population Survey data shows that around two-thirds of the new jobs taken by people who formerly described themselves as disabled have been full time. This group also appears to have been more likely than other job finders to find work in manufacturing and construction, and no more likely to be self-employed." (https://www.nytimes.com/2018/03/15/upshot/will-employment-keep-growing-disabled-workers-offer-a-clue.html). From disability to construction! Who would've thought that the cure for disability isn't diet or exercise; it's a good economy.

Fourth, one more way that economic incentives matter is from the presence of a new constraint: working women. In the 1950s, most women didn’t work at jobs and those who did weren’t paid much. Today, almost 47% of US workers are women (https://blog.dol.gov/2017/03/01/12-stats-about-working-women), and they actually make good money. Today’s couples therefore need two matches in the job market, not just one. It's only natural that earlier generations had better labor mobility.

Fifth, one non-market, non-government reason why people don't always move to where the jobs are is because labor isn't completely fungible. In human history, there have been three major reallocations of labor: foraging to farming, farming to manufacturing, and manufacturing to low-skill services. Each transition period has gotten increasingly short, and each new form of labor has gotten increasingly complex. The next major reallocation of labor, already underway via skill-biased technological change, will be from low-skill service sector jobs to high-skill ones. Some portion of the labor force will reach its cognitive limit, and all the government retraining programs in the world won't teach coal miners how to code.

In sum, the op-ed blames markets for what are really the failures of government (occupational licensing, expanded disability, restrictive zoning, subsidized driving), the rise of working women, and the limitations of human nature. Any analysis on the effect of incentives in the labor market that doesn’t consider these three things is incomplete at best, misleading at worst.

Since the 1980s, half a billion people have been lifted from poverty in China. Before that, rapid growth in South Korea lifted the majority of the population out of poverty. What do you say to people that claim the key to poverty alleviation is to be found in examples like these and not RCTs?

Thiago, hun. Give it up. For all our sakes.

These are pretty brutal...

That being said the op-ed really was heavily ideological...

Congrats to him!

1. Many RCTs fail to replicate in second RCTs done in new settings. USA examples include Family Rewards 2.0 (conditional cash transfer) or Tennessee pre-K. How much does he think about replicability?

2. Lots of failed experiments (including RCTs) never get published. For example, we tried to replicate Banerjee's "Teach To The Right Level" in Kenya and failed, nil result. (Though a friend did replicate it in Botswana!) How much does he worry about unpublished negative findings, of RCTs and other high quality research?

3. I hypothesize are 3 data tribes.
"I Hate Math and Distrust Research."
"I'm a Numbers Person and Love Data....If It Validates What I Already Believe."
"I'm Genuinely Intellectually Curious and Love RCTs Even When They Crush My Intervention Dreams."

Does that resonate?
What % of people in each tribe?
What are implications for his work?

This is very cool. I am wondering if we will be able to leverage the likes of Google and Facebook and the Mobile operators as platforms for many large scale RCTs in development across borders?


The same questions should be asked to every “expert”.

When was the last time you were incorrect about something?

How did you learn you were incorrect?

Upon learning this, did you question any other ideas/concepts/beliefs you held?


Ask why they are so concerned with poverty in other countries and not so with the 12% poor in the riches country in the world were they live. Do they believe that the nudges they try in their rcts will work equally well in the usa.

Ask him whether the so-called economists' grandiose theories have ever been successful to change the course of lives of common people, and thereby they themselves playing themselves into the hands of power-that-be?

Given the arguments both he and Duflo make in favor of immigration, to the effect that the addition/removal of illegal immigrants has minimal impact on local wages, aren't they implicitly making the argument here that if we’d like to see more mechanization, automation, and the attendant positive impacts on productivity *over the long haul*, we should block low-skill immigration? After all, we have loads of cheap capital available and they're telling us that in the absence of immigrants the employers will tend to automate those positions.

Also, you can ask AB why there are so many haters out there? It blows my mind when I read the comments to your posts TC, I can’t understand the folk that write hateful comments are so insecure etc. For my part, I think your blog is awesome and it brightens my day. Thank you

To solve global poverty, shouldn't we first learn from historical poverty reductions, and figure out if there's any way to replicate them? How do you justify focusing on small scale interventions instead of promoting growth? What is the optimal allocation of resources between alleviating poverty and promoting growth?

Comments for this post are closed