Lant Pritchett on Poverty, Growth, and Experiments
May 22 2017

cash%20chicken.jpg How should we think about growth and poverty? How important is the goal of reducing the proportion of the world's population living on less than a dollar a day? Does poverty persist because people lack skills or because they live in economic systems where skills are not rewarded? What is the role of experimental methods in understanding what reduces poverty? Author and economist Lant Pritchett of Harvard University talks with EconTalk host Russ Roberts about these questions and more in a wide-ranging discussion of how best to help the world's poorest people.

Angus Deaton on Inequality, Trade, and the Robin Hood Principle
Nobel Laureate in Economics Angus Deaton of Princeton University talks with EconTalk host Russ Roberts about the economics of trade and aid. Deaton wonders if economists should re-think the widely-held view that redistribution from rich nations to poor nations makes...
Richard Davies on Extreme Economies
Economist and author Richard Davies talks about his book Extreme Economies with EconTalk host Russ Roberts. The conversation explores economic life in extreme situations. Examples discussed are the Angola State Penitentiary in Louisiana, two Syrian refugee camps in Jordan, the...
Explore audio transcript, further reading that will help you delve deeper into this week’s episode, and vigorous conversations in the form of our comments section below.


May 22 2017 at 8:03am

First, you should add Blattman’s response to Pritchett’s response.

But aside from that, I agree that economic growth is the most important thing to alleviate poverty, but in the meantime countries still have to (for the most part) fund basic social welfare programs, no? That’s where I see the chickens vs. cash argument being more relevant. You’re not going to generate growth from giving people cash or chickens or wheelbarrows or whatever (not at a macro level), but surely you’re going to make individual households more resilient to economic shocks, which has real value, especially if institutional changes might take 10-20 years or more to really have a major impact.

[Note: Chris Blattman’s response was already featured toward the top of our reading list. Just sayin’. Nothing wrong with your re-linking to it, of course.–Econlib Ed.]

john penfold
May 22 2017 at 3:57pm

Thanks, another wonderful exchange. I hope Econ Talks is getting into economics department seminars. An educator told me his approach to fix our schools was to eliminate schools of education; We could make the same wisecrack about development economics. We know what gives rise to economic growth, we don’t know how we foreigners can cause countries to get there from where they are. I do not agree with your guest that the average development of the world’s post war economies attests to positive returns to development efforts. By giving money and pursuing development fads we may have diverted pressures, alleviated acute problems, rewarded corrupt leaders and we certainly sustained the development industry by giving them new fads to try that hadn’t yet been proven ineffective. And yet some studies were important to shaping views. Bhagwati and Kruger’s studies on import substation, and McKinnon’s on money and capital I know personally influenced Latin American economists that got into positions of power and moved their countries in the right direction. Mexican President Salinas de Gotari had studied a year of economics at Harvard. When US Senator Brock, to put pressure on EC negotiators in GATT talks, threatened to expand our bi-lateral with Canada to include Mexico, Salinas seized on it to give him leverage to liberalize the Mexican economy. What could USTR Brock do? thus NAFTA was born. Then there are all those young foreign students who left elite US schools and the London School of economics as socialists and set their countries back decades. Still something happened in the post war world, hopes,the notion that nothing was fixed, that ideas matter and I think we helped spread that optimism even if in the course of doing so we stumbled around a lot .

May 22 2017 at 7:22pm

PETA aside, Papa Gates is free to use his own cash, chickens or whatever other personal surplus to test his ideas.

However, misguided Paternalism is a rather big problem and should have been part of this discussion. I am guilty of that too, as is Papa Roberts and Papa Pritchett.

May 22 2017 at 11:29pm

“Gates had suggested that, if we want to reduce poverty, we should give sub-Saharan Africans chickens….And Chris Blattman thought, ‘Hmmm. Maybe we just give them money.'”

Both Gates and Blattman miss the point. Giving will not have any long term effects. Investment is the most likely solution to poverty.

A problem is that these developing countries are not market economies. Transactions are through force (e.g. central planning, violence) rather than free exchange.

Change is unlikely as long as those in power are profiting from their current system.

May 23 2017 at 3:36am

[Comment removed for supplying false email address. Email the to request restoring this comment and your comment privileges. A valid email address is required to post comments on EconLog and EconTalk.–Econlib Ed.]

May 23 2017 at 11:50am

How do we tell which of the many piles of research are the best or most effective?

Pritchett illuminates a hard and interesting problem: How do we distinguish between large piles that accumulate because researchers pile on (not to pun, of course)–perhaps quite reasonably drawn to high-productivity or high-payoff ideas, self-interested or funded or not–versus piles of equal quality that may be smaller? Or between equally-sized piles that may be in opposition?

Innovative ideas that branch off to create new piles are, merely because of the shorter life of being newer, often unable to create piles as large as a long-established pile. A new pile starting with the same quality as an older pile started with can’t compete with the size of an established, long-growing pile–by definition.

That’s not even talking about how large piles often have support of government-funded agencies such as the World Bank or internal government funding. So smaller piles may have trouble competing. Many fewer individuals.

There have been many large piles of research, well funded historically, that have not worked out to the best. Alchemy was a major such research pile–a major accumulation of years of research time put in by the greatest and highest-quality researchers of their day. Massive number of articles and research. And not just researchers who were fly-by-night. Then, sometimes just one individual says something at odds, or a different pile accumulates; and it changes the whole tenor of even a very reasonable flock.

How do we know that this or that idea promoted–be it by a pile or an individual or something in between–is not the analogy of alchemy?

I do not know the answer to these questions. My best answer is that we have to think for ourselves. All the time. I worry about being persuaded merely by this or that of the many piles of research in any one direction. And, I worry about someone saying that the size or existence or creation of a pile is in itself convincing. That there is a pile is not in and of itself convincing. Interesting, but not convincing. More interesting is the argument made by the brunt of the pile.

Glenn Mercer
May 23 2017 at 1:04pm

I would like to make a comment “across” multiple ET episodes, sparked by this one but applicable IMHO across the board. There are two aspects to my comment, one of effectiveness, one of efficiency.

Effectiveness (that is, the value of the econtalk podcast). I’ll phrase it as a question: what does Russ believe? In so many episodes Russ reverts to his familiar themes of “most stuff is not knowable,” “correlation is not causation,” “what worked before may not work in the future,” on and on. But if he believes all these things, why does he host this podcast? Are the only thinkers he trusts Hayek and Smith, clutching to them the way a deconstructionist professor of literature, after smugly proving to his students that no writer has ever had an independent thought, grudgingly admits that maybe Shakespeare was original? In which case, why not skip interviewing people, and just read chapters of Smith on the air?

Efficiency (how efficiently the podcast delivers to listeners the points the interviewee wants to make). I think it would be much more efficient if Russ were to start each interview with a declaration of all his familiar priors, then say “Let’s now set all my epistemological terrors aside, and let the interviewee make her or his case freely and completely. Then AFTER that, we’ll dive into issues with the case.” This would be a more efficient approach to the discussion. Instead, Russ seems to spend less time discussing the merits of a case, and more time waiting eagerly for the interviewee to step on a Robertsian landmine, e.g. by saying “the evidence shows” or “we know that…” or “history tells us” — triggering an explosion of well-worn retorts such as Hayek’s “The curious task of economics is to demonstrate to men how little they really know…” or some extension of the “epidemiology is a cesspool of intellectual failure” conceit to sociology or economics or Lord knows even physics! It’s like watching a college basketball game where the play-by-play guy says “Nice foul shot but should these kids even be playing sports, when they would be better off studying skills they can use in the job market?” Good question, but can we GET BACK TO THE GAME? And then, when the post-game show is on, raise the point at THAT time?

I probably have been listening to Econtalk too long, such that the Robertsian tics (I was going to write “Russian,” but you can see the problem there) are driving me crazy. It is still an incredibly valuable program, and I am grateful for its existence (and yes, I would PAY to listen), and I listen to every episode, but jeez Russ, if nothing is reliably knowable since March 23, 1992, why do the show?

Thus endeth my own rant. I of course have no foibles at all….

May 23 2017 at 1:59pm

Okay, so the bottom lines are…:

1. Don’t give them chickens or cash, it doesn’t work.

2. The best thing to do would be to give them luggage so they could come to our magic dirt in the US, western Europe, Australia, Canada, etc., where they will thrive. (How many you may ask? Well, we’ll get back to you on that. But hey, if we import a billion to the US, maybe we can get that 5% gdp growth! Sounds great!!)

3. The likely best politically acceptable alternative…. encourage them to change their culture and economic systems at a macro level. Replace their dictators with a smart well-meaning free market loving capitalist.

Hmm, maybe old fashioned British and French imperialism wasn’t so bad?

And so the cycles repeat……

May 23 2017 at 6:21pm

[Comment removed. Please consult our comment policies and check your email for explanation.–Econlib Ed.]

Steve Hardy
May 24 2017 at 1:15am

How about supporting organizations that are promoting free markets, rule of law, private property, less regulation, free trade and individual liberty around the world? One such organization is the Atlas Network that has helped establish and partners with 462 free market think tanks in 96 countries.

David Zetland
May 24 2017 at 4:18am

(1) Glenn’s comment has increasing relevance, as I have also detected a little too much of Russ’s hobby horses in recent years (e.g., the big aside on regulation with Cass last week)

(2) I really liked this podcast, as Lant’s point was exactly correct: Big development requires functioning institutions, markets, etc., not micro-interventions with chickens or worms. His aside on economist’s incentives (do things you can get recognized/published for) deserved another 20 minutes of its own.

Over the years, I’ve tended to disagree with Russ when he’s on topics related to development or the environment, both of which are not is his background, and both of which fall out of the excludable goods paradigm where markets are indeed “the answer”. Maybe we’ll talk one day about the commons, etc (hell knows, I’ve offered!)

Russ Roberts
May 24 2017 at 1:27pm

Gary Mercer, David Zetland (and all of my listeners),

I will always have a bee in my bonnet about the proper role for government vs. decentralized ways to solve problems. I do not think markets are anything close to perfect. I am not an anarchist. Government has many roles to play in letting bottom-up solutions thrive. And there are many cases where top-down government solutions outperform markets.

I will always challenge guests who notice market failure but ignore government failure and what is called public choice.

I will always point out what we don’t know. Reminding people of the limits of knowledge is a great source of satisfaction to me and one of many reasons I enjoy creating EconTalk episodes each week.

I will not always use the same theme music for EconTalk. But I might.

Greg Alder
May 25 2017 at 1:53am

David Zetland,
And what is your experience with development in Africa?

I spent four years in a rural, poor part of Africa. And in every one of the EconTalk podcasts about development that I’ve listened to, I find your perspective on it more accurate according to my first-hand experience than your guests’, who apparently consider themselves experts.

May 26 2017 at 9:32am


Do people actually think you’re an anarchist? I can understand why you may attempt to preempt them but a discussion on Anarchism would be an interesting show – you could argue why some top down solutions work. Ed Stringham, Walter Block, … (you know the names) would make an interesting episode.


With respect to this episode, I found Lant over confident at best. I also found his attempt to switch the burden of proof onto sceptics of the development Econ profession self-serving. What needs to be done is to define what growth actually is (IMO GDP is poor in the extreme) and then analyse what causes growth. Only with these tools can we investigate in history what caused what and consider counter-factual situations – this burden is on all sides in the debate.

Another issue not discussed in regards development is persistent world wide IQ disparities – even if it is entirely environmental (extremely doubtful), in the West attempts to bring average IQ for races together has failed. Further, IQ and general prosperity have a more than a weak positive correlation. On that Charles Murray would be an interesting guest as wel.

John Christopher
May 26 2017 at 9:38am

The conclusion I draw from this podcast and from Lant Pritchett’s ideas is that the best way an outsider could help counties with lots of poor people is not give them chickens or money but publish millions of books like “Wealth of Nations”, “Road to Serfdom”, and “Capitalism and Freedom” in the native languages of the countries and distribute them freely in those countries and over time increase the probability that the leaders and institutions in those countries would be influenced by this information and begin to develop, implement, and grow the institutions and policies prerequisite for growth. Anybody else get that?

May 26 2017 at 2:44pm

The link on “5 stages of team development” doesn’t seem related to what Lant means by that term.

Lauren Landsburg, Econlib Editor
May 26 2017 at 3:42pm


I agree; but have you got a better link to suggest? I’d be happy to replace it. I was looking for something on “team development,” which term I’d never heard before. The link does at least clarify the term a little bit. Perhaps the issue is one of whether you were developing a team of people to accomplish something versus picking a team of people to make recommendations about developing a country? Lant seems to mean the latter, but the former seems to be the more common usage online. In both cases, it seems to be about coming to a consensus decision. So, maybe Lant is talking about team-development development. 🙂

Chris Douglas
Jun 1 2017 at 1:22pm

Pritchett is right about 1950-2010 being a time of poverty reduction. But his argument misses:

(a) the fundamental changes of the 20th to the 21st century economy – fossil fuels and heavy industry transitioning to technology, automation, internet/communication, knowledge economy, where top-level intervention is often too slow with too many harmful externalities

(b) the specific geographic and temporal contexts of economic development for countries like China and India – former empires with favorable populations, infrastructure, geography, global trade, etc.

(c) the new reality of Development/Aid, with more private sector and citizen engagement, “poverty traps” instead of countries, and situations where direct chicken/cash interventions are the difference between life and death (see earlier EconTalk interviews with Easterly, Jerven, and Deaton)

As a result of these factors, those countries most in need of real economic growth are those that can benefit the most from the kind of “programmatic” economic engagement that Blattman’s chicken/cash experiment describes.

The opportunities for growth in those countries at the lowest end of the Human Development Index are opportunities to build 21st century businesses around communication, transportation, agriculture, healthcare, banking, education, fashion, entertainment, etc. – and you don’t need World Bank-sanctioned top-down policies to create those businesses.

Today’s development challenges mean a new formula for real economic growth and connectivity. That’s the reality of Team Development that Pritchett talks about. So there are extremely good reasons to look at new ways of using resources to fight poverty, like investing in cash transfers.

1950-2010 was a period of more people becoming more free to trade more goods/services/ideas in freer markets. Continuing that pattern requires smarter, leaner, and more entrepreneurial interventions in Development/Aid – they’re way more consequential than Pritchett seems to recognize.

Jun 5 2017 at 3:30pm

[Comment removed. Please consult our comment policies and check your email for explanation.–Econlib Ed.]

Jun 5 2017 at 8:36pm

I disagree with Lant that simply moving sub-saharan blacks to another country will suddenly make them protective. There are about 200 million in Nigeria alone. If all of them moved to France or to the USA, I think most people’s common sense would tell them that before long the conditions in the United States would look like what these PhDs are trying to fix right now. I would have liked Russ to bring up the examples of Zimbabwe and South Africa as examples of this phenomena.

Jun 17 2017 at 6:25am

If only the banks were not bailed out in ’08. There would have been a redistribution of wealth. When failure is not allowed, the glass will always be half empty

Comments are closed.


EconTalk Extra, conversation starters for this podcast episode:

This week's guest:

This week's focus:

Additional ideas and people mentioned in this podcast episode:

A few more readings and background resources:

A few more EconTalk podcast episodes:



Podcast Episode Highlights

Intro. [Recording date: May 5, 2017.]

Russ Roberts: Our topic for today comes from an open letter that Chris Blattman, another former EconTalk guest--an open letter that Chris wrote to Bill Gates. Gates had suggested that, if we want to reduce poverty, we should give sub-Saharan Africans chickens. Right now I think the number is about 5% of sub-Saharan Africans--and they are very poor people in general--raise chickens. We get that number up to 33% we could really make a dent into poverty. And Chris Blattman thought, 'Hmmm. Maybe we just give them money.' Which Chris has been working on; and we did an EconTalk about that with Chris. And he proposed an experiment to find out which is better, giving poor people chickens or money. And here's what Chris wrote:

Despite the suggestive research that I've cited here, no one has run the race between chickens and cash programs. No one has asked whether the expensive training or supervision that often goes along with these things is worth it.... It would be straightforward to run a study with a few thousand people in six countries, and eight or 12 variations, to understand which combination works best, where, and with whom. To me--[this is again, Chris Blattman writing Bill Gates]--that answer is the best investment we could make to fight world poverty.

That meaning that if we could just figure this out, we could really make some progress. And that's the end of the quote. And then you, Lant Pritchett, wrote a lengthy response to this idea--and we'll link to your response; we'll link to Chris's open letter, etc. But you are very critical of Blattman's proposal, and more generally critical of, I would say, Blattman's perspective of development and poverty generally. So, what's wrong with the idea of running an experiment to settle this question of whether it's better to give people money, or, say, chickens?

Lant Pritchett: There's nothing wrong with that idea. It's just, if I had to rank best investments to address world poverty, it wouldn't be first. It might be 999th. And, the fundamental issue is, when we say the words, 'world poverty,' what do we mean? And, I am a big detractor of a phenomena that's been going on for a couple of decades now, in which development--generally the development institutions like the World Bank and Development Economics--has been moving away from a broad, encompassing definition of world poverty to mean people escaping poverty into prosperity, to a very narrow, what I call 'kinky' definition where we are going to pursue narrow, low-bar, specific targets like 'dollar-a-day poverty.' So, dollar-a-day poverty, as a concept, was invented in 1991; and it has slowly eaten its way into the conception of what people mean when they say 'world poverty.' And I just think it's wrong in every conceivable way.

Russ Roberts: Well, I didn't know it was invented in 1991. That's interesting. Is there a person who is associated with that invention? Or an organization?

Lant Pritchett: Oh, absolutely. The World Bank wrote a World Development Report, which is the World Bank's annual flagship, you know, external publication. And they wanted to write it on world poverty and how to address world poverty. And they wanted a headline number of like, 'Okay, if we are going to talk about world poverty, how many poor people are there?' And that's kind of a super-hard question to answer. But the proposal was: Let's call dollar-a-day in consumption expenditures, per person per day--suitably adjusted for purchasing power, obviously, across countries--let's call that the poverty line. And let's measure how many people are below that line. And I was in the room--somewhat--but Martin Ravallian, who is now a professor I think at Georgetown, was in some sense the creator of this number argument as to why it should be $1-a-day. And it was incorporated into this report. So, it was--before then I don't there had been in some sense even any quasi-consensus global poverty line.


Russ Roberts: So, what's wrong--obviously it's arbitrary; poverty in general is somewhat arbitrary: there's questions of relative or absolute poverty, of course. We would all agree that if you have less than $1 a day, your life is challenging relative to a much higher standard of living, say, in the United States. I guess the question would be: Is there any meaning to the statement, 'Well, now that they're above a dollar a day, they are not poor any more?' I guess that would be one problem with it. But I think you have other problems. So, why does it bother you to create that arbitrary cut-off?

Lant Pritchett: Well, I think the issue you raised is first and foremost the big issue. Which is: If we are going to divide the world into two groups--there's going to be Type I error and Type II error, right?

Russ Roberts: Explain. For non-statisticians [?] in the crowd[?]

Lant Pritchett: Sorry. Type I and Type II is jargony. But there's going to be: Who do we say is poor who really isn't? And, who do we say is not poor who really is? And basically, the choice of dollar-a-day poverty, the dollar-a-day standard, unambiguously if you are below a dollar a day, you can be characterized as poor. No one's going to dispute that. But, if you are at $1.01 a day, or $1.02 a day, or even $2.00 a day, is there any meaningful sense in which you can be said to be not poor? And I think that's--so, you can imagine, we are going to have a line that in some sense for sure is the lowest--we draw it so incredibly low and so incredibly penurious that no one can dispute that everyone below this line is in fact poor. But, at the same time, by doing that, you exclude a whole bunch of people who legitimately could be called poor. And so, my argument is, basically, 'Let's have a super-low-bar poverty line, and let's call that destitution, or let's call that extreme poverty. And let's leave that $1-a-day or some super-low number like that--and by the way, the whole $1-a-day thing has been subject to inflation, and now it's $1.85-a-day, and we still call it $1-a-day. But okay. And then, but let's also say: What's the conceptual, symmetrical counterpart of that? What's the line above which we're pretty comfortable saying that if you are above that line, you are not poor? Right? And we'll minimize the mistake of calling someone who is poor: we'll sort of, instead of minimizing the odds that we're wrong that we said you were poor and you weren't, let's minimize the mistake that we said you weren't poor and you are. And my argument is: Rich countries have a poverty line. And my argument is: People below the poverty line in rich countries are legitimately called poor. The people who are below the poverty line in the United States are legitimately poor, in my mind. And those lines, drawn by rich countries, tend to be more like, not $3/day or $4/day. They tend to be more like $15/day. And so my argument is: Let's just make sure in the world, that we acknowledge that that is in fact a legitimate measure of poverty. And that we apply that measure of poverty to the whole world, too. So, we have world--and we could call escaping that poverty line 'prosperity.' We could have all the people in the world who aren't in prosperity; and that's world poverty, by one definition. And all the people that are in true destitution or extreme poverty--and that's another number. But, by only creating statistics around the one number, we bias attention towards exclusively benefitting that group in a way that just has no fundamental legitimacy in economics; no fundamental legitimacy in politics. Just no fundamental legitimacy.


Russ Roberts: So, when we look at the last 20 years or so, a lot of people would point out that due to the growth of economies in India and China, there's been a huge amount of progress in "fighting poverty." And they would use the measure you are critiquing.

Lant Pritchett: Right.

Russ Roberts: They would point to the fact that many, many fewer people now in India and China, and elsewhere, are earning less than a dollar a day, are living on less than a dollar a day. And that's great. We celebrate that. So, that's okay, right?

Lant Pritchett: That's all good, sure.

Russ Roberts: So, why is that a problem?

Lant Pritchett: Well, because the problem is, is that if you move most of the reason--nearly all of the reason why we have in fact in the world made progress on reducing dollar-a-day poverty is that people moved from low-growth economies to high-growth economies, which was mostly associated with a switch toward more market-oriented reforms at the general level--which increased the overall productivity of the economy, which increased the productivity of the poor, which increased their wages, which led them to escape poverty. So, economic growth is in fact the key to reducing even low-bar poverty. And unambiguously, unquestionably the only way to reduce high-bar poverty. So, economic growth works for both. And, when we talk about why people in China escaped poverty, it was because the Chinese embarked on, you know, moved from, you know, an exclusively Marxist system towards more market, more incentives, more allowance of private-property-like mechanisms at first; and then moving increasingly towards private property. But, when you focus on low-bar poverty, you get tempted to focus on a programmatic approach--like chickens. You say, 'Oh, if we have a low-bar poverty line and we are identifying people below that low-bar poverty line, you can imagine that the solution to that low-bar problem can be addressed programmatically. I can pick out that individual; I can do something for or to that individual. And that individual will cross the poverty line.' And treating world poverty as if it can address programmatically biases attention away from the real solution to poverty, which is: Having higher productivity economies.


Russ Roberts: So, that's a very deep point. And I have a feeling for listeners out there, I have a feeling this is going to be a theme through a few episodes. In particular, I can't help but note that growth has gotten a bad name recently in the United States. Mistakenly, in my opinion. Because I do think that gains to growth have been much more widely spread than some numbers suggest. But, focusing on this issue of poor countries: One way to think about this is, you know, if a person has a tumor, you'd want to operate on them, remove the tumor; and they would be back, we hope, into a state of health. If you live in a country that generates tumors, going person by person is probably not the right way to do it. So, one way to think about this chicken issue is, you are suggesting that focusing on chickens is a bandaid. It's not a systemic way to make progress.

Lant Pritchett: Exactly. And I think I hadn't thought of that--actually it's a pretty good analogy. Because, I think one of the many downsides of measuring this low-bar poverty or destitution is it leads you into a mentality in which you think the problem with the poor person is the poor person. But mostly, high levels of poverty are not a reflection of personal pathology. They are a reflection of systemic pathology. Mostly in the world there aren't poor people. There are people in poor places. So, you know, if you say, 'Why are there high levels of poverty in Nigeria?' it has nothing to do with the personal characteristics of the Nigerians who we would measure as being below the poverty line. It's not like they are uniquely endowed with some poverty-creating characteristic that needs to be alleviated. They are in Nigeria. And Nigeria is a low-productivity place. And everything we have learned about economic growth suggests that the factor we call A in our equations as economists, or just the general factor that determines what the productivity of all factors in the economy is, from capital to raw labor to, you know, human capital, is affected by this general productivity term. And people are poor in Nigeria because they are applying their labor and their skills and their capital in a low-productivity environment.

Russ Roberts: So, I mean, that's extremely interesting; and I think plausibly true. You could imagine an experiment--and we run this experiment, in a very bad form, but people come from poor countries to the United States, and they earn more than $1-a-day. They may have gone from a world where they were earning a dollar a day. When they come to the United States, they earn a lot more than a dollar a day. And it's not--I emphasize--not because we have a higher-than minimum wage. It's because those skills, in the American economy, that person's skills, are much higher. Now, of course, there's a problem that we don't get a random assortment of people who come here. We get people who have, tend to be higher ambition and drive, etc.

Lant Pritchett: But the problem is that's not at all the issue.

Russ Roberts: I agree. No, I agree--

Lant Pritchett: Michael Clemens and I have a paper, as you might know, that is titled "The Place Premium," in which we do exactly the numbers you are talking about. We say, 'Let's take the U.S. Census and look at Nigerians born in Nigeria, educated in Nigeria, working in the United States. And let's take a Labor Force Census of Nigerians born in Nigeria, educated in Nigeria, working in Nigeria.' And let's just say: What's the wage of the observationally equivalent person in Nigeria and the observationally equivalent person in the United States? And those numbers, you know--that wage, for a low-skilled person, somebody with less than high-school education, it's 16 times. So, your wage goes up 16-fold. Not 16%, not [?]--

Russ Roberts: I believe that. I believe that number is in the ballpark. And it's a big enough number that it's probably reliable.

Lant Pritchett: And then, by the way, we have done every conceivable correction for how much this is biased by, 'Yes'--so, first of all, we've corrected for observational equivalence.

Russ Roberts: Yeah--the problem is non-observational--the intangible things--

Lant Pritchett: Exactly--

Russ Roberts: like ambition, passion, drive, reliability, etc.

Lant Pritchett: Exactly. So, we have pursued, like 6 separate ways of thinking about the magnitude of that: How much biased that number is? The consensus is at most about 25% of that is selection of--selection, particularly, we are talking about low-skill. Right? If we are talking about who is at Harvard University or, you know, clearly that's massively selected. But we are talking about people with less than a high school education. So, we are talking about Nigerians who work in the United States with less than a high school education. By and large, when we look at the occupations undertaken by immigrants in the United States with less than a high school education, none of them are at Harvard; none of them are doing heart surgery at Massachusetts General. They are in low-skilled occupations. So, the idea--

Russ Roberts: They are not making--they might be poor by the United States's standards, but they'd be doing pretty well in Nigeria.

Lant Pritchett: Super--again, 16 times as much as they would have made in Nigeria.

Russ Roberts: Not in Nigeria, but the equivalent of a Nigerian.

Lant Pritchett: Exactly. And so, the issue of selection is an important conceptual of issue. But, that would reduce the ratio from 16 to maybe 14, so it's still--you know, this is like making tiny adjustments to a huge number. It's like we might be off on the distance from the United States--if we are comparing the distance from the United States to Europe and the United States to the moon, we can mis-estimate the distance of the United States to the moon by 10 or 20 percent and it would be a trivial adjustment in the comparison. And that's what we're facing here. The adjustment for selection for wage gains to low-skilled workers is a tiny adjustment on a massive number. And like, 6 different methods--methodologically we have the snails.

Russ Roberts: Well, I hope you're right. And I'm willing to accept it as a truth. And I'm certainly willing to accept it as a truth for now. So, let's accept it as true--

Lant Pritchett: And, by the way, that does precisely illustrate the point: that it's not Nigerians. It's Nigeria. Because this is exactly the thought experiment: Let's take this person who would be poor in their country and just put them in a high-productivity environment and see whether in absolute terms they get massively more productive, and absolutely. And by the way, we can do the calculation of, like, how many people from Haiti in the world are not poor? Well, 82% of all Haitians who aren't poor are living in the United States. So, the route out of poverty for a nation is to get out of Haiti into a high-productivity environment like the United States and they relative to any reasonable--even a global high-bar poverty they escape poverty; and all of them escape low-bar poverty.


Russ Roberts: And this gets us back to the remark I make about every seven episodes on EconTalk, which is: the biggest way to reduce poverty is luggage--to get people, make it easier for people to leave. The problem with that joke--it's not a funny joke, actually--it's a clever joke but it's not actually funny. The problem is, of course, they need somewhere to go. And I'm going to channel my inner Chris Blattman right now and suggest that, 'Okay, there's no doubt that people in sub-Saharan Africa right now are poor because sub-Saharan Africa is a tough place to be productive. And it doesn't have economies of scale. And it doesn't have a lot of capital. And it's a tough place to live. And even with chickens, it's not so easy. But, they don't do very well there; and so, given that they are probably not moving to the United States or France, tomorrow, what do we do for them? And then the question is, I think Chris would say: It's better to give them money or chickens than do nothing. It's better to give them something to reduce their personal straits rather than just say, 'Well, it's not their fault. It's the system they were born into.' Which is of course probably true. But now what? And I think Chris would say we do have to decide, if we are going to help them, what way is most effective? And then we might want to know if it's giving them something like chickens, some tool--a cow, a wheelbarrow, some kind of capital, or a chicken--versus cash. And Chris says, 'I think we should give cash.' And there's some evidence that that's a good idea. Other people would say, 'No,' like Bill Gates, 'let's give them chickens.' So Chris has proposed: let's try to figure out whether this is a horse race we can decide; let's do an experiment. So, is that a bad idea?

Lant Pritchett: Again, it's not a bad idea; but it's not a good idea. Because, there are really three options. Right? One is, we could--and I have a book titled, Let Their People Come. So, we could let their people come; unambiguously it's the largest, most reliable way to attack global poverty. Nobody's mostly talking about it for political reasons--

Russ Roberts: Yeah. A lot of people are against it--

Lant Pritchett: and they think it's not going to happen. We could have a whole session on that. Second, though, is we could mitigate the consequences of Nigeria's low productivity through a variety of programmatic approaches to attack targeted individuals with targeted interventions that would raise their incomes. And we could debate whether that should be cash or chickens. But the third option is that we could devote money, resources, time, talent, research effort to figuring out how to get Nigeria as an economy, society, and polity to become more developed, to become more productive, the government more responsive, the administration more capable across the board. So, we could focus on systemic interventions. So, let me just--so, India--I went to India in 1991. India in 1991 was on the verge of an incipient macro crisis that had affected many Latin American states--where there was going to be debt problems, where there was an overvalued exchange rate that was causing balance of payments problems. All the symptoms that had let ultimately to the decades of lost growth in Latin America. India, you know, ultimately acted quite decisively to shift their fundamental economic strategy from a more inward, pro-government, regulated economy towards a more market oriented economy. I have done work on, research on economic growth in which we look at specific episodes of economic growth: when they began, how long they lasted, and what was the total value of the episode? Right? When you look, India has had two growth accelerations, according to this econometric procedure of dating when episodes started, one starting in 1993--kind of after the year of reforms--and one in 2002. Those two reforms combined added at least $2 trillion of GDP (Gross Domestic Product) that has been produced in India today that wouldn't have been produced, counterfactually, had they maintained business-as-usual growth, instead of the growth they got from the reforms--or from whatever happened. Right? And I'm not so much arguing what the reforms were; but they did do something. Everybody acknowledged that there was a fundamental shift in strategy. And the consequence was, we observe in some econometric procedure applied to all countries in the world, there was a growth acceleration in 1993; there was another subsequent growth acceleration in 2002. The combination of those two growth accelerations produced $2 trillion dollars in additional GDP, in India.

Russ Roberts: And that GDP of course went to more than just the top 1%, say.

Lant Pritchett: Absolutely. You started by saying India and China have had very rapid expansions in growth and very rapid reduction in poverty--so, there has been just an extraordinarily large cumulative reduction in poverty in India over precisely that period. And there's a huge debate about the magnitude of that--which we can come back to. But, no one doubts that India is night and day a different country for across the board than it was in 1991. So, my argument is: How much money should you have invested in creating the conditions in which India did the thing that created $2 trillion dollars of GDP, versus how much money should you have invested in giving Indians chickens in 1991? Well, I think it's a no-brainer that anything we could have done to increase the probability that India did the right thing rather than the wrong thing and descend into some Latin American decades of lost growth from a macro crisis, was ex post fantastically a high return in investment. Because it augmented the productivity of all factors in India; and also induced people to, you know, invest more in creating all kinds of factors. So, in my mind, investing in creating the conditions for countries to adopt development-promoting policies is enormously higher returns than investing money in the question of cash versus chickens.


Russ Roberts: Yeah; I'm tempted to agree with you, but I can't. So let me disagree with you; and you tell me why I'm wrong.

Lant Pritchett: Sure.

Russ Roberts: I want to start by saying we may not have the time to come back to it--I don't think $15 million dollars spent on trying to figure out whether chickens or cash is money well-spent, either. That's a different debate. I think you are making a much richer argument against that. And if I have time I'll say why I believe that. But I want to get back to this richer argument you are making, which I find extremely interesting. So, here's what I find troubling about it. We can think of it in a couple of different ways. This is the Lant Pritchett open letter to Bill Gates: 'Bill, don't spend that $15 million on chickens. Take the $15 million, pay some really bad leader in some sub-Saharan African country, say, $12 million, to leave. And you run the country for a while. And just hire with the other $3 million, hire Lant Pritchett and a bunch of his friends to tell you what to do right. And then you'll solve the problem.' And the reason--that's a little bit of a parody--but we've spent a lot of money through the World Bank trying to figure out how to make things better. Some people would say, 'Oh, we know how to do it.' Others would say, 'Boy, it's kind of subtle and difficult; and the really key parts are endogenous. And they are organic. And you can't impose them from the outside. And really there's not much we can do.' And again, 'Let's just at least give them chickens.' If I really thought that we could really, that we could significantly influence the lousy, poorly-governed countries of the world to liberate their citizens' talents and let them flourish, I'd be all for it. I just don't think we know that well how to do that.

Lant Pritchett: But this--I find this general condition we're in just really supremely puzzling. Which is: At the end of WWII, we created the--and in the subsequent events--we created a self-conscious notion that we, as a global system, was going to be structured in order to promote the development of the newly de-colonialized sovereign states. And so, development as a fundamental idea was born roughly as part of the post-WWII shift in mindset, shift in the global political order. The years from 1950-2010 have, on every single measure of human wellbeing, been the best 60 years in the history of man by a factor of multiple.

Russ Roberts: Agreed.

Lant Pritchett: Education has gone from 2 years on average in developing countries to 7 years on average in developing countries. Infant mortality has plummeted in nearly every country in the world. Economic growth has chugged along at 2%--so people worry it hasn't been a converging rate of growth--chugged along on average in the world at 2%. Which means the world is enormously richer than it was 60 years ago. Many countries, in various ways, have not just had average growth of 2% but have managed long, sustained episodes. Absolute poverty has absolutely plummeted during this period. Everyone--team development [?] won. Team[? teen?] development is the most successful team in improving human wellbeing in the history of many by a factor multiple.

Russ Roberts: What is?

Lant Pritchett: Team development. Think of team development.

Russ Roberts: What is that? What do you mean by that?

Lant Pritchett: Team development is we have this group of--we have a mentality that we are going to self-consciously promote development and that's going to involve some development organization; it's going to involve some research into development economics; it's going to involve some research into things; it's going to involve a global order that's open to trade; it's going to involve a whole bunch of things. Right? But the point is: The world is massively more on every single conceivable measure of human wellbeing, night and day better off than it was before people self-consciously said, 'Let's promote development.'

Russ Roberts: Well, it's [?]--

Lant Pritchett: So, during the development--

Russ Roberts: Correlation isn't causation.

Lant Pritchett: Yeah, yeah, yeah. Absolutely. Believe me. I understand that.

Russ Roberts: I know you do.

Lant Pritchett: Right. But, the point is, we don't want to get into the--so, the point is, but let's at least start from that. We are not trying to parse out why there is failure. We are trying to parse out--I mean, correlation isn't causation, but at least let's get the facts right. The facts are: There was massive success. And then, you can object, 'Yeah, but the fact there wasn't massive success doesn't mean that development aid played any role in it.' I agree. But we are trying to parse out attribution of massive success, not attribution of failure. Whereas, oftentimes, people start talking about foreign aid as if it is obvious somehow that it was ineffective and a failure. Well, if in fact, the world had in fact been a failure, if in fact there hadn't been progress for 50 years, then its attribution of failure. But, we're in attribution of success. And then the question is: Correlation isn't causation, but let's not get obsessed with the counterfactual, over the factual. The factual is, things are enormously better by factor multiples than any previous episode in human history; and it just so happens, in this episode of human history, all of these things happened, including groups of people self-consciously promoting development. So, it's not at all obvious that they failed. If development had been a failure, in the sense that there hadn't been massive improvements in human wellbeing, then we could be discussing, you know, the anatomy and the pathology of failure. But we shouldn't, you know, act as if poverty has like fallen massively in the world. So, it's not as if we have failed in some sense to reduce poverty. Now--

Russ Roberts: Well, the question is: Who is 'we'?

Lant Pritchett: Who is we, and how did it happen? Right?

Russ Roberts: Yeah.

Lant Pritchett: But again, let's run the thought experiment of, you know--let's run the thought experiment. In 1978, right, Deng Xiao Ping decides that China is going to pursue a fundamentally different economic model. Right? You can imagine that that decision was completely independent of a discipline called 'economics.' Completely independent of people having demonstrated that differential growth rates in GDP per capita of Korea and Taiwan, versus what was happening in them. You can imagine that the ideas and concepts of a more market-oriented economy were completely irrelevant to its decision. But that seems pretty implausible to me. It seems to me he was probably influenced by evidence, by discussion, by conceptual advances, by rhetoric, by arguments. And that paid off, in the sense that--forget the $3 trillion or $2 trillion added in India: China has added $11 trillion dollars in additional GDP, relative to the counterfactual of what we would have predicted for China had China continued. So, you can say, 'Ahhh, I don't know how much credit we can take for that $11 trillion.' But let's keep in mind the question is how to apportion the attribution of the $11 trillion dollars in success. Suppose the development industry, in creating a mindset that made the idea of a development [?] from possible contributed one tenth of one percent to that $11 trillion gain in China. It was still worth it. By an order of magnitude.


Russ Roberts: Okay. So, I think we have a vocabulary issue here; and then I think a conceptual issue. When you talked about development aid making a difference, I think most people thought of it as I do when you said it--you are talking about--

Lant Pritchett: No, no, I tried never to say that.

Russ Roberts: I thought I heard you say that, which to me means giving people money.

Lant Pritchett: No, no.

Russ Roberts: But what you are saying is that--

Lant Pritchett: No, no, no--

Russ Roberts: we've learned some stuff--

Lant Pritchett: I'm saying team development.

Russ Roberts: Right. We've learned some stuff--

Lant Pritchett: Everything we didn't promote development. Some of which was giving people money. I don't think the money given via aid was particularly the causal mechanism that caused success.

Russ Roberts: It's hard to argue that, given that--

Lant Pritchett: Exactly--

Russ Roberts: China and India didn't get a lot of it. The countries that did, didn't do nearly as well. Etc., etc.

Lant Pritchett: Right.

Russ Roberts: But let me now critique your broader claim.

Lant Pritchett: Right.

Russ Roberts: And in doing so, I'm going to make myself squirm a little bit, because I really like your claim; but I have to concede when I see the rest of the evidence that maybe it's a little more complicated. So, you may have an answer for it, or you may be squirming a little bit yourself. You'll let me know. So, here's the thing that I notice. I look at--I don't know how you look at the world exactly, but I tend to be a bottom-up rather than a top-down person, recognizing of course that there are really important things from the top down that matter--rule of law, contracts that get enforced, private property, etc., etc. So, I'm not an anarchist. I do think it's really important to allow a modicum of flourishing that certain things come from the top. And like you I believe very strongly that don't allow individuals to flourish are ones that are governed badly. However, when I then ask: What is the set of key institutions, etc., that make, allow countries to escape poverty dramatically, we did another type of experiment in the last 40 years. We tore down the Iron Curtain--

Lant Pritchett: Yeah.

Russ Roberts: And Communism disappeared. And a lot of us would have thought, and did think, that once Russia and other Eastern European countries became more market-oriented, they were going to just explode in their flourishing. And we would also argue, we could also argue--and that didn't happen--and then we have to then say that's 'because'--and we have a lot of hemming and hawing. And then we look at the countries, like a country like China where it's true it went to a more market oriented process, but it did it with a very centralized authority at the top that's been steering lots of things for a long time. It may blow up in their face as they build these cities in the middle of nowhere that nobody's living in. But, they seem to be doing pretty well, for running development from the top down rather than the bottom up. And then, given those two things, how does someone like me, with my viewpoint, tell a poor country what they are supposed to do? It seems to me, again, it's not obvious. So, I come back to this question of--you mentioned India in the early 1990s, early 2000s--they did the "right thing." What's the lesson, then, for Nigeria? What's the lesson for Sudan? What's the lesson for Peru? For Argentina? Etc. It's easy to say what the lesson is: They need better property rights; they need less corruption; they need a more representative government; they need more transparency; they need less red tape. I don't know how to get there from here. I don't know how to invest $15 million dollars, or even $150, to make that more likely.

Lant Pritchett: Ahhh--I think you do. I think you invest in people in those countries who are researching, acting, and investigating, and talking about precisely those questions. I think investing $15 million in a group in Nigeria who was asking in an evidence- and experience-based way how to reduce corruption in Nigeria--I think that seems like a pretty plausible way to get those questions answered. Not investing--

Russ Roberts: The World Bank does that every day. Right? The World Bank--I have friends who work there. They travel to a country; they write these great studies. They tell them what they should do. And the study gets put in a drawer.

Lant Pritchett: I think that is a complete and total caricature of what the World Bank does. And I think they are missing the point. In the sense that--so, I'm a huge believer in Keynes's sort of idea that what drives the practical men of affairs of the world is, you know, 'the ideas of some long-dead scribbler'. Right?

Russ Roberts: I'd like to believe that, Lant. As a scribbler, I can see why that appeals to you and me.

Lant Pritchett: Exactly. What could be a more self-serving belief about the world.

Russ Roberts: I know it appeals to you; but keep going.

Lant Pritchett: Okay.

Russ Roberts: It appeals to me, too. That's why I'm doing this program. I'm not doing it because I think it's--

Lant Pritchett: You're trying to get people out in the world to listen to ideas--

Russ Roberts: To learn something.

Lant Pritchett: To learn something. Okay. Now. So--and let me just--and I first heard this view, and I'll give credit where credit is due--I first heard this view from Paul Krugman. Paul Krugman said, 'What's the value of economic research?' And he said, 'There's this micro-macro paradox. You look at each individual paper published in economics and it seems stupid and silly and trivial.' And so why are we doing it? His view is: You do all of this kind of work not because you expect the individual report that's written to make a difference, but the overall milieu and the professional consensus that arises about general directions creates ideas about the world that themselves become influential. But they only become influential because they are built on this kind of mound of stuff. It's like a mound of gravel: each rock, you could pick it up and go, 'Ehhh, this is nothing. This is stupid. Who needs this?' and you could pick up the next piece of gravel and say, 'Ewww, this is stupid.' But a mound of gravel is a mound of gravel. So, I think your friends would say they go to Nigeria, they write a report, nobody reads it. They don't see that they are a mound of gravel. They are a mound of gravel. We've changed what the world believes on all kinds of things. I have lived through one of the most powerful social movements of the 20th century, and the social movement was called 'free trade.' There were a group of us that were committed to free trade, that thought developing countries having imposed and adopted an inward-looking, trade-restricting strategy was harming world poverty. And we consciously used every lever at our disposal to change people's ideas about free trade. And we won. We swept the board. We were fantastically successful, and I continue to believe--

Russ Roberts: For now.

Lant Pritchett: Ehhh, believe me: the revisions to the free trade consensus that are going on are moving back 2% of where the world was 30 years ago. Thirty years ago, countries--all kinds of countries essentially banned all imports of consumer products. Like, in India, every single consumer product was banned. Just banned. It wasn't even a tariff. It was just banned, right? Countries maintained lists of, you know, imports--there were imports that were banned; imports that were under some scheme where you needed a permit for. And that was the regime. Right? You couldn't import anything without a permit. That's gone. Never to come back, right? We might reintroduce some tariffs. We might go back on some deep commitments to integrated areas. But relative to where the world was, in the late 1980s free trade won. And, how did free trade win? I could point to each individual study on the impact of free trade and say, 'That study had no impact. That study had no impact. That study had no impact.' And yet somehow there was a huge impact. So I think trying to--I think that we're suffering from the perils of partial attribution. We're trying to ease out this causation question by looking at each individual piece of gravel, and ignore that there's a pile of gravel there. So, I think, you know, devoting time, energy, resources, to creating the ideas that motivate, help Nigerians construct a reality within Nigeria which Nigerians promote--a more prosperous economy, a more responsive polity, a more capable state--is super-important. And we could look again--and this is a problem--donors said, 'Oh, no, no. What was the impact of this study that I financed, versus giving somebody a chicken?' Right? That's the mindset that I'm attacking. Because you are ignoring the fact that the study is contributing to what might be a mound of gravel that will be enormously larger. Whereas there's no possibility that giving somebody a chicken creates a larger pile of gravel. It is what it is. Full stop; that's all you've done. So, I think investing in the pieces of gravel that are going to make the mound that are going to change the world is enormously higher returns than saying, 'We'll ignore building the pound of gravel. We'll take for granted there is no pound of gravel. It can't be built because each individual piece is so small. And focus on giving chickens away because these countries are bound to be poor. Let's do philanthropy on an individuated basis.' Just wrong. Wrong. Wrong.


Russ Roberts: That's really interesting. It reminded me of an old line of Milton Friedman's. He would say that the sum of negligible effects need not be negligible. That is, things accumulate into that pile of gravel you are talking about. Now, I don't agree with about half of what you said; I don't think that--and unfortunately I have to disagree with Paul Krugman. I don't think that the problem with economic research is that it's just a little pile of gravel that eventually builds. I don't think, a lot of it I don't think builds very well and doesn't even deserve to be in the pile. That's a separate point. But the deeper one--

Lant Pritchett: That's a separate issue, yep. And I agree. A lot of--and there is a lot of contesting: some people are building a pile in the wrong place.

Russ Roberts: Yeah. And they are building a pile that actually is counter-productive. But I think the deeper point--I'm going to give you the benefit of the doubt and try to take your claim as far as I can see it, because I think it's a really interesting claim. You are saying there is a cultural zeitgeist, a milieu, a--I can use two pretentious words in one sentence there. There's something in the air. It's again that Keynesian idea, almost. There's something in the air that makes freedom or trade or incentives the default. I like the way Dan Klein puts it. He said, 'Adam Smith had a presumption of liberty.' Meaning, he wasn't a libertarian or a hard-core anarchist or anything. His view of the world was: Freedom tends to be a thing that generally works. There are exceptions. It may not work well every time. But the burden of proof is on those who would take it away, not those who would impose it. And I think that cultural view--which was true in Smith's time, and it's been true through much of the United States, and is definitely not true in many places of the world and in many times--that has a value. What I hear what you're saying is: That has a value that's really quite extraordinarily large. And once you lose that--which I think to some extent we are at risk of losing it right now, and have been for the last 20 or 30 years in the United States--once the default, once the burden of proof is on those who would argue for liberty, then we may lose something very valuable that wasn't obviously created by any one book, any one paper. I was once at a meeting--I don't think I've told this story on the air; if I have, I apologize. But, I was once at a meeting where somebody said, 'You know, we just need a book that explains how markets work. And then just give that to people!' And my view was, 'Well, we need about a hundred books like that. We have quite a few, actually. We have more than one. And most people don't read them. Or don't find them compelling.' So, that's not the solution. But the idea that you need a hundred, that you need different ways of telling the story and of convincing the body politic, to the extent it's existing in ideally in a free-ish country--that, that matters a lot. Even though, it's not like there's a study that comes out and everyone goes, 'Oh, I guess free trade is good. So, where I agree, that's a fascinating idea. Where we disagree is I think a lot of it came from Adam Smith and we haven't added that much to the pile. I think--you know, I think about Maggie Thatcher who supposedly carried around The Wealth of Nations in her pocketbook. I don't know if that's a true story or not. But when she moved to a more market set of policies in England and per capita income in England grew dramatically relative to other countries in Europe, maybe that was due to the free market approach she took. And to the extent that it is, I don't think it was due to a piece in Econometrica. I think it was due to the fact that she had that book in her pocketbook.

Lant Pritchett: But, I don't think you can--I don't think you can maintain the power of those ideas without people actively working on them.

Russ Roberts: Fair enough.

Lant Pritchett: And to the extent that the field of Development Economics, and smart, genius, capable people like Chris Blattman aren't fundamentally seen as the perpetuators of a fundamental idea that roughly market-oriented economies are the path to increasing productivity that will reduce poverty--and get distracted into, 'Let's ignore that and focus just on mitigating the consequences of low productivity'--I think that's a loss to the world. I think re-allocating people from addressing the big question, even if it's hard to see how they do it, even if it's hard to see the impact, and moving to the directly-attributable, I think is a huge mistake. An analogy I came up with, sort of thinking about what I want to call the Perils of Partial Attribution, which is this obsession with what caused what, ignores--like I say, there is a huge mound of gravel, now. Human well-being is enormously better off than it was before people started, kind of self-consciously promoted, through a field called among other things, Development Economics. And we shouldn't just ignore that that mound of gravel has in fact been built. And maybe it wasn't this study or that study, or maybe, who knows? But it is there. We have to explain that it's there. Right? And it's mostly there. Most of poverty reduction in the world is because--you know, India, China, Vietnam, Indonesia in the 1960s--one after another, you point to when the poverty reduction began--it just so happened to begin in a dramatic fashion when countries moved towards more market-oriented, growth-oriented policies. And none of it--0% of Indonesia's poverty reduction was due to its pursuing more effective poverty programs. So, devoting--imagining that the high returns to poverty reduction come from investment in more effective programs when in fact none of the actual variation is explained by that seems an odd way to go about doing science. It's the ultimate drunk-under-the-headlight approach. We have a technique for providing, you know, attribution to the reduction in poverty to this; so we'll apply it to things that we know for sure don't explain the aggregate reductions in poverty. We know, for sure, 100% that none of the poverty reduction in Indonesia from 1965-1996 was poverty programs. We know for sure that a very small fraction of the reduction of poverty in Vietnam for more than half of the population, to under 10%, was the result of programmatic[?] efforts. We know for sure that it's not the story of China. For sure. Right? So, why we know that these big facts for sure but are continue to say the highest return is investing enormously valuable resources of economic, trained people who are like massively intelligent and massively capable, to these tiny little questions--I just don't see how people get to that conclusion. As opposed to having them focus on continuing to build the mound of gravel one small piece at a time--even though we can't, ex post, do the attribution. So, my analogy is: Certain players in basketball, they score a lot of points. And since we keep track of individual points, it draws your attention to, 'Oh, this player must have been a good NBA (National Basketball Association) player because he scored a lot of points.' But when they started actually doing, did they continue to their team winning? It turns out, their players--and I've been told by someone who knows this industry that there's a player who played for Houston, Kelly Martin, who scored a lot of points but just was terrible on defense.

Russ Roberts: Only half the game. Yeah.

Lant Pritchett: Exactly. Only half the game.

Russ Roberts: But we don't notice it.

Lant Pritchett: People said, 'But he just scores so many points; it must be the case that he's contributing to our victory. Because, look we keep track, and we know this is how many points he has.' And, we didn't actually keep track--it's harder to attribute causality on defense. But it turns out, once you looked at it, he was negative value added. He actually reduced wins above replacement relative to the average NBA player because his defense was so truly spectacularly awful that it actually--so, this super-high scorer would have been regarded as a star under attribution model that was incomplete in attributing all of its effects was actually a below-average contribution to wins for his team. So, I worry that moving too much to the attribution model--what am I doing for poverty that I can attribute directly to the actions of this individual?--is like focusing exclusively on points scored of an individual player and not thinking about, 'What are you contributing to the team wins?' And I think team wins, team development, are going to come when we keep our eyes focused on the big picture. The default to where, you know--the fact that we can't answer precisely the question what we can do to promote economic growth doesn't mean we should have nearly everybody working on that question.


Russ Roberts: So, I hope Chris Blattman can come on and speak for himself. I want to move away from Chris, per se, and just talk about the general phenomenon that there are a lot of people doing these programmatic changes. And there's certainly an incentive issue there for those scholars, especially when they are younger, to find things that will get them tenure rather than big-picture things that might be a dead end. So, we understand that problem in economics; and we'll put that to the side. I think the more interesting--

Lant Pritchett: I don't know why. Whoa, whoa, whoa. Why are we putting that to the side? That seems pretty relevant.

Russ Roberts: We're putting it to the side for this conversation. I think it's huge. I'm just saying I don't have anything more to say about it; and it would be interesting to hear what Chris and others like Chris would say about it. I want to ask a different question: I want to defend the programmatic approach even though I certainly agree with you in some dimension--but I want to defend the programmatic approach for a minute, that the small changes in poverty--I want to come back to that argument. Which is: There was a study done by Michael Kremer and some others back in, I forget what year--that maybe it was useful to deworm children. This electrified the world, because the impacts were so large. It got a lot of people to devote resources to deworming. We've talked about that with Will MacAskill, on effective altruism, an episode on that. And since that enthusiasm got started there's been an enormous push-back on that finding. It's still up in the air. There's still a debate, I would say. But a lot of people have said, 'Well, you know, when we tried to do that experiment in a larger scale, it didn't scale. A lot of the magnitudes were smaller. And what you found out was that in that village maybe it helped, but it doesn't mean it's true for all of poor people in the world. Deworming doesn't have the return we thought.' Now, so, my view on this is that it's still an open question. Obviously the people in favor of deworming have argued that those studies questioning their findings are wrong and we'll go back and forth--maybe we'll make some progress on that question. But put that to the side. Let's assume it is true--again, I understand it may not be true--but let's assume it's true that deworming of children, reduction of parasites, has an enormous ability of their ability to sit in school and do better. That would be important. I wouldn't call that a little tiny thing. It's true that if you are trying to sit in school in a country that doesn't have a very good economy, sitting in school doesn't have as big a return as it would have if the economy was better--which is your point. But I wouldn't want to argue--I don't think we want to argue that it's trivial. And I might not even want to argue with small. It might be somewhat big. I don't know.

Lant Pritchett: No. It's not. And that's a good example. Because this is something we actually know facts about. And facts are good, right? So, the world does international comparable assessments of student learning. And they are roughly normed so that the mean is 500 in OECD (Organization of Economic Cooperation and Development) countries. And the student standard deviation across all countries is 100. Which means, sort of, only 16% of, you know, OECD students are below 400. So, that's kind of the range of performance. As we try and chain-link to discover what the poor countries who don't participate in PISA (Programme for International Student Assessment) score like, the scores in the two Indian states that participated in PISA, the scores were around 320. Which means--you know--just unbelievably low number. Literally, you would put, the average student in India in 8th grade would be in the bottom 1-percentile if they were suddenly put in a Danish or U.S. school. Unbelievably low, right? And low at 320 and not a huge variance, actually. So, the number of people, even scouring above 400, is very, very small.

Russ Roberts: When was this? When are we talking about?

Lant Pritchett: This was 2009.

Russ Roberts: In India.

Lant Pritchett: In India. PISA exam. They did a PISA exam in two states, and India chose what they thought would be the best two states. So that probably is an overstatement--it must be a pretty substantial overstatement of the performance, India-wide, because they did it in two states that everybody thought would be the best. With good reason. [?] production, [?]. So, now we say, 'Okay. We have a gap of 180 points between India and the United States.' How many of those points are we going to get from deworming? Suppose we achieve universal, not-a-single kid is suffering from worms.

Russ Roberts: I have no idea. I don't know.

Lant Pritchett: My guess is 1. We'll go from 320 to 321. It's just--

Russ Roberts: Well, why?

Lant Pritchett: Because it's just not that big a deal. It's not that big.

Russ Roberts: Well maybe [?] because they don't go to school, they can't concentrate, and if you deworm them, they can get up to 450.

Lant Pritchett: No. No. It's just--nothing about the empirical estimate suggests anything like that is true.

Russ Roberts: Okay. Perhaps.

Lant Pritchett: It's just not that big a deal. Because, after all--you know, it's the--the net impact is going to be fraction of kids actually suffering from worms, times reduction in fraction of kids suffering from worms from deworming, times the incremental score of an Indian kid with no worms over an Indian kid with worms. That's the kind of causal chain. And I'm pretty sure if you multiply out those numbers, in India you are talking about, you know, a number that's going to be more like 1 than it's going to be like 10, and certainly not going to be like 100. There's no way, shape, or form. In fact, in no way, shape, or form these effects are on the order of 100. I don't even think--and I haven't done this calculation--but from reading the original study, you get incremental score improvements and the key thing was there were externalities, so it kind of improved everybody's scores to address this in another question. But it's just not a first-order effect. So, I am doing research on the question of how do we get a country like India from 320 to 500. And more research on deworming just isn't helpful. It's fundamental features of the system: How it works, how teachers are hired, how teachers are assessed, and how teachers perform. How is the curriculum relative to their ability to implement it? Just all kinds of features are huge first-order features. But, they are very difficult to do precise, attributable studies of. And so the research is massively biased towards the questions we know are important, towards the questions we know can have attributable research done.

Russ Roberts: Yeah, I guess-

Lant Pritchett: And we just know that that's not the answer.

Russ Roberts: I guess the parallel way to think about this might be malaria.

Lant Pritchett: Yeah. Malaria is a different story, by the way.

Russ Roberts: Well, that's why I want to hear why. Because you could argue we could give people bed nets. They might not use them; they might not use them correctly. [?] we should try to eradicate mosquitoes, or things that allow mosquitoes--

Lant Pritchett: Absolutely--

Russ Roberts: to flourish. And, again, the question would be: Do we know how to do those two things in ways that are effective? And if we only know how to do one that is effective we might do that one even though it's not as good as the other. I don't--I wouldn't want to minimize the impact of, say, if only 5% of India suffers from a worm problem and those kids actually can, they can bump up to a much higher number--

Lant Pritchett: And again, I think you are overstating the effect.

Russ Roberts: It sounds like it--

Lant Pritchett: Nobody claims it's a much higher number. It's a higher number but --

Russ Roberts: Well, I think that--

Lant Pritchett: nobody thinks it's a much higher number.

Russ Roberts: I think that your point about the standard deviation within India is the right point. If there are bunch of people there who don't have the worm problem, it could be they have other things that are associated with the worm problem--anyway, it's a complicated thing.


Russ Roberts: We're almost out of time. Let's--

Lant Pritchett: But the question should be, how do we get countries from 320 to 500? That's the question. And: Is it the case that research aimed at only those things that are narrowly and easily attributable is the right research? Or, are there in fact likely large, systemic effects that explain these huge differences? And, even if we can't do absolutely precise attributable research, we can have a presumption in certain directions. Like a presumption for learning performance. And within that presumption for learning performance we can point to some specific directions like teaching at the right level. And that those--like Adam Smith's presumption for liberty, those presumptions towards certain things actually are going to cumulative, huge effects. More so than had we devoted equivalent amounts of money exclusively on the premise that what's attributable is what's important. I think that's the huge mistake: is we're running into causal, attributability as the issue versus success as the issue. Success is the issue. We should be focused on success, not on being the high scorer.