Lant Pritchett is a development economist from Idaho. Having now thrice retired, he is currently a Visiting Professor at the London School of Economics in the School of Public Policy and the co-founder and Research Director of Labor Mobility Partnerships (LaMP).
In this week’s conversation, Yascha Mounk and Lant Pritchett discuss why the traditional foreign aid approach to development is fundamentally misguided, how countries actually achieve prosperity through organic national transformation, and whether the classic path to development remains viable in the 21st century.
Will you be in London on Sunday, September 6? I’ll be interviewing Francis Fukuyama about his life and thought to mark the publication of his memoir In the Realm of the Last Man at the Sekforde at 5pm. Find out more and book tickets here. Paying subscribers can access a code for free tickets here. —Yascha
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: When I talk to people who care a lot about economic development, they love to talk about inclusive development, about sustainable development. They tend to focus on how much more affluent countries should donate to less affluent countries, perhaps figuring out what the best intervention is—whether foreign aid should be spent on this kind of thing or on that kind of thing. Not to simplify, but you think that whole approach is basically wrong. Why is that?
Lant Pritchett: What I think of when I think of development is what we call “getting to Denmark.” There was a historical process whereby many countries—not just Western countries, but other countries—had a fourfold transformation. They had a transformation from a low productivity to a high productivity economy, and it was mostly broad-based. That is economic growth, and inclusive economic growth, if you want to add the adjective. They also went through a transformation of acquiring state capability—the ability of the public sector to do things that needed to be done, like regulation and providing certain services. They also went through a transformation from subject to citizen, to a polity that was based on responsiveness to the needs and wishes of the citizen rather than vice versa. They also went through this harder-to-describe transition of equality under the rule of law, whereby kith and kin and other identities became reduced in importance and everybody was treated equally.
That is what development meant in the post-decolonization era, after World War II, as countries became independent from their colonial overlords. It meant this big fourfold transformation. Foreign aid can be modestly helpful with that, but it is not very central to it. The more you think about donors and what donations should be and what the right interventions are, the more you lose the plot.
In my papers, I show that if you get to what I call national development—this fourfold process—that is a machinery for producing good things. We are worried about whether people have clean water, decent housing, and all of these material things that are really good for people to have. But if you get to national development, you get to that, and vice versa.
There is a big question of whether what we talk about as foreign aid is aid to the process of national development, or whether it is aid to mitigating the worst consequences for human well-being of that lack of national development. The latter is not a strategy for the former. By and large, a lot of the development agencies lost the plot completely and became essentially charity organizations focused on mitigating the worst consequences of the fact that many countries had not acquired national development.
There is nothing wrong with that. Mitigating terrible things that happen to human beings because they live in a country that is underdeveloped is a good thing. But it is not development. Too much attention to the latter detracts attention from the core issue.
Mounk: Is the concern here simply that this money is being spent on things that don’t actually help to solve a problem in the long term—that the most effective use of that money would be to invest in things that actually help solve these underlying problems, and therefore make the country more affluent so there is not as much poverty or human desperation that you need to buffer with these donations? Or is it more profound than that—that those donor dollars in some ways make it harder for that process to take place, that they might in some complicated way backfire? Is the concern mostly about efforts wasted, or efforts that might in some complicated way impede the country that is supposedly being helped from solving its long-term problems?
Pritchett: The big problem is more about ideas than it is about the concrete use of money. I have what I call the “bird on the elephant” theory of development. Development agencies are spending all this money, financing projects—some of them economic projects like roads and bridges and power plants, and some of them charity-like projects funding health interventions and the like. But that is in some sense secondary to the fact that this creates a global discourse about development and about how countries can do development.
I am much more worried about the waste of effort on ideas. We have geniuses—truly stunning geniuses—devoting themselves to charity work as opposed to thinking about development strategy. Ideas are supremely important to the fate of nations, and the ideas that get transmitted via a global discourse of research and practice, to government officials, to people in power, to people who have influence—that is a huge deal. Obviously the most consequential thing that has happened in the last 50 years is the leadership of China changing its mind about what to do in China and how to make China a better place to live. That was, fundamentally, a change of ideas. By losing the plot on national development in favor of mitigation, we also draw the discourse, the research, and the ideas away from big questions: How do we get states to be more capable? What is the right sequencing of state capability and democracy? How does democracy interact with the creation of economic growth—does it impede it or not? The number of people in the world who can produce new, original, and correct ideas is very few, and drawing those people onto small issues is a huge loss.
I don’t think aid is often counterproductive. Some of my friends—Bill Easterly, and Angus Deaton—think that aid can sometimes foster mentalities and practices that impede development. I think the elephant is mostly neutral. But if the bird, who sits on top of the elephant and sees what is going on, warns the herd of impending dangers, and can provide a vision—if that gets messed up, then the whole elephant is diminished.
Mounk: Let’s take a step back. One of the intuitive things about what you’re saying is that the United States, the United Kingdom, and France didn’t become rich because some much richer countries said, we’re going to give you a bunch of development aid. These were mostly internal processes, obviously with an international component—trade with each other and so on. So it stands to reason that if we want to think about how countries like India or Kenya might become rich in the future, it would probably be by following some of the same kind of processes.
Now, a certain kind of progressive critic would respond by saying that there is an economic structure in the world and that these countries are in some ways interdependent. If you’re one of the most developed countries in the world, you can specialize in high-return services, for example. If you are much poorer than those countries, then you can’t follow the same development path as those that developed historically, because you occupy a different niche in the economy. The other difference, of course, is just the stage of development. A lot of the countries that grew rich in the 19th and early 20th century did so through industry and manufacturing. But nowadays most wealth is not created in factories—it’s created in the knowledge sector, in the service industries. Perhaps the same kind of path to development just isn’t open anymore. When I was in India, a lot of people were worried about whether the traditional path to development is still available in the 21st century.
What do you think about that? Can you broadly follow the same playbook that made the rich countries of the world rich in the past? Or do we actually need to look for a different path, either because of a change in the nature of the world economy or because of the different relative standing of the poorest countries in the world within it?
Pritchett: Your question brought up two very different strands that we shouldn’t conflate. One is the path being organic and being driven by a country’s own dynamic. If you take, say, Dieter Mikloski’s work, the key to all of this was creating an environment in which you can give it a go. That is distinct from the question of specifically what, economically, a country is going to do. The question of whether the path that Denmark took is open to Kenya has two very different components.
We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
If you have any questions or issues setting up the full podcast feed on a third-party app, please email leonora.barclay@persuasion.community
One is the deeper question: can Kenya develop an organic dynamic within its polity, society, and economy such that the actors find a path for Kenya to become prosperous? That is the true endogenous process. The second question is, when Kenya organizes itself to find that path, will it look like the specific economic path that Denmark followed? The answer to the latter is certainly no—and no in both a good way and a more challenging way.
The first way the answer is no: what happened—and I wrote a paper a long time ago called “Divergence Big Time” that emphasized this—is that the rich countries in 1870 were not that advanced relative to the most lagging countries, because there is a floor. There is only so poor you can be. What has happened in the world is that the rich countries have done this through exponential growth. They have created the basic hockey stick graph.
The rich countries collectively grew at 2% a year for 120 years, and the power of compound exponential growth means they are manifold richer than they were in 1870. But in the process, they invented and discovered a great many things—in science, in practice, and in all kinds of domains—that weren’t available in the world in 1870 but are available in 1970 or 2026. This means the countries that enter this process later have the potential to grow incredibly faster than any rich country grew.
None of the countries that were rich in 1970 or 1980 were rich because they grew fast. They were rich because they grew steadily. But that created the possibility that later countries could accelerate their growth enormously. The fast-growing countries in the world are growing incredibly faster than they did historically. China, Korea, Vietnam—growing at 6% per capita wasn’t an option for Denmark, which just had to stay on 2%. That is the good news. Countries that, in whatever way, manage this endogenous process that creates an organic drive for prosperity can discover ways to grow incredibly faster than was available to the old countries that had to make their own way.
We have seen incredibly good news, and we don’t want to lose the plot on that. The years since World War II have been the best years for improvement of the material condition of humankind by a multiple of any prior period. I work some on education. The average person in the developing world had roughly two years of education in 1950 and has eight now. From whoever your mythic forebears were—Adam and Eve, let’s say—to 1950, humanity had accumulated two years of education. In just 60 years, it added six. Three times more education was added in 60 years than in all of human history combined. That is true of health. That is true of a whole range of things. Just fantastic progress.
Take child mortality. There were all kinds of countries in the world 50 years ago where basically one in five children was dying before the age of five—a rate of around 200 per thousand. Worldwide, that is now down to around 30. Many people in the world in 1950 were living in material conditions not that much different from ancient Greece or ancient Egypt. Even today, if you compare Egypt in AD 1—when Joseph and Mary went to visit—to some of the poorest African countries today, they are at roughly the same level of GDP. For a very long period of human history there was very little progress, and then it accelerated dramatically. From all of human history, we got to still one in five children dying around 1960, and now that figure has fallen enormously. If you look at the people with access to electric power, people with access to water and sanitation—it has just been an amazingly good run.
These questions always start from the assumption that the global order is preventing progress, and that is just surreal when you look at the world over the last 60 years. There has been amazing progress in many places on many things, and this isn’t just a matter of measuring GDP.
Mounk: One illustration of that from my own life: when I was an undergraduate in England in the early 2000s, there were big debates about the World Trade Organization. The shape of the debate, and I remember this vividly, was always whether this was going to screw over China and India—whether the World Trade Organization was just a smart way for the rich countries of the world to keep the poor countries from developing. Today, when you look at the critiques of letting China into the WTO, it is exactly the opposite. The question is whether America screwed itself over, or screwed its working class over, by letting China in—which led to a huge increase in wealth in China and rapid deindustrialization in parts of the US.
You can take seriously the idea that there were some mistakes made in how that was done and what impact it had on the working class in Michigan, Pennsylvania, and other places. But if you have to choose, from the standpoint of humanity, between those two scenarios, we have ended up with the much better one. We have lifted hundreds of millions of people out of poverty to genuine middle-class status in China. Perhaps we could have done that a little bit better, with less disruptive consequences for the already pretty affluent people in the United States. Perhaps that’s a red herring. But it is striking to what extent the progressive case against things like the WTO in 2000 has completely flipped on its head in terms of how we talk about it today.
Pritchett: One of my favorite colleagues and friends at the Kennedy School was Dani Rodrik. If you look at Dani Rodrik’s intellectual trajectory, his most recent book is about shared prosperity, which is hugely concerned about the prosperity of the middle class in America and how the global system hasn’t been good for it—versus his earlier concerns about the WTO and whether its trade structure was truly open to facilitating the growth of poor countries. It turns out we had the opposite problem from the one we thought we were going to have.
Absolute poverty in the world has just amazingly declined. Every discussion of the world and how it’s going needs to start from that factual basis. Hans Rosling wrote a book called Factfulness—the fact is, on nearly every measure of material well-being, things are just fantastically better. In many dimensions, by the way, better than we would have expected even from the economic growth that we got. Sometimes people say, you economists focus just on economics, and we really should have focused on these other things. But if you look at the expansion of schooling, for instance, schooling expanded not by less than we would have expected from the economic growth we got, but way more. Just getting kids into school was one of the most phenomenally transformative things in history, and it happened more than we would have expected given the growth, not less. All of development was focused on expanding schooling. So you cannot say that economists focused on growth at the expense of these other things. That is just not true.
Mounk: Let me push you on that—it’s a question I was going to pose in any case. There are a lot of people who are very skeptical of GDP as a metric. They think we can be incredibly rich and yet people are miserable. Society can be very rich in the aggregate, but only because a few people are incredibly rich and everybody else is incredibly poor. There are these standard examples that come up in conversation all the time. If I’m stuck in a traffic jam burning gas and not getting anywhere, that’s increasing GDP—so GDP is a really bad metric.
You believe that GDP is in fact a very good metric of human well-being and that it correlates very strongly with things that we care about more directly. When listeners next encounter somebody who says GDP is a terrible metric we shouldn’t care about, what should they respond?
Pritchett: I want to be clear: I don’t think GDP is a good metric of human well-being. GDP is a very good metric of the net production of a society, and that production creates the material basis for human well-being. GDP per capita is just factually very highly correlated with nearly everything we care about in terms of material well-being.
But here is the response you should make. The relationship between GDP and most things we care about in terms of material well-being is concave—meaning those things get better as GDP gets better, starting from poor levels to middle levels. By the time you reach a GDP per capita of roughly $40,000, you have met most of the basics of material well-being, and hence the relationship flattens out. It doesn’t go away necessarily, but it does flatten out.
The real danger in the world discourse is that people unhappy sitting in traffic in Luxembourg—which I use as an example because the only time I went to Luxembourg I got caught in traffic—are right that their overall well-being is not that highly correlated with GDP, because their GDP is so high in the first place. But we shouldn’t extrapolate that back.
Mounk: To put this point a little more polemically: it is very easy—and I grew up in that kind of milieu, not particularly affluent, but among artists and musicians in Germany living in pleasant towns—for people to say that the important thing in life isn’t to drive a big car and go on a fancy holiday, that there are things much more important than material well-being. Of course, if the floor of your society is that you have a decent apartment with heating and running water and you eat three meals a day, then that is probably true. But if you are sitting in a mud hut in a developing country, that is most assuredly untrue.
Pritchett: One of the fundamental insights of economics is declining marginal utility—the more you have of something, the less important it is to you. If you say the West is suffering a paucity of purpose and people are unhappy because they don’t have some driving purpose, well, that is because they have already satisfied the purpose of having material things like a hot shower, a heated home, and the ability to go 300 miles in a few hours. Whereas most of humankind, for most of history and even today, is nowhere near that.
I am happy for German artists to have angst about what their purpose in life is and whether they are really happy and whether more material goods would make them happy—and it probably won’t matter that much to them. But don’t project that back onto Africa, or conclude that India doesn’t need growth, or Bangladesh doesn’t need growth, or that we don’t really need to worry about whether those countries grow.
The right response to the claim that GDP is a bad metric for improving the human condition is to ask: where are you on this spectrum of existing progress? Since you are likely having this conversation with people who are materially fantastically well off relative to any period in human history and relative to most of humanity, they should be sensitive to the fact that yes, that may be true for them, but it is not true for six billion other people on the planet. Don’t project your life and concerns onto theirs, because you have diminishing marginal utility precisely because you have so much.
Mounk: Let’s go to those poorer people in the world. You are pointing out that GDP per capita is very strongly correlated with the other things we care about—child mortality, life expectancy, education, and so on. What does that tell us about what we should aim for? Does it suggest that we should just aim to increase GDP per capita and assume the other things are likely to follow? Or could the association go the other way around—that what we need to do is have all of those specialized interventions to improve the local hospital and the local school, and that is what will then correlate with increases in GDP? Which way around do we read the correlation?
Pritchett: This depends on where you are in the spectrum. I have done a paper where I try to disentangle this question. One of the things that GDP does, by the way, is create a broader tax base. People often get engaged in a very strange discussion as if focusing on GDP growth means ignoring the need for government services. But you cannot have government services unless governments have revenue, and government revenues are mostly tax buoyant—meaning they grow more than proportionately with growth.
Mounk: So GDP growth actually is what enables you to have all those government services?
Pritchett: At one point I did the calculation: Ethiopia’s government spending per head is around $300 per person per year. What can you do with $300 per person per year? Ethiopia’s GDP per capita is so small that if the government started taking bigger proportions of it, they would be eating into expenditures on food. Sixty percent of a poor household’s budget is on food. You cannot just say Ethiopia’s government should fund all these specific interventions—how are they supposed to do it?
When I argue in favor of GDP per capita, it is not because GDP per capita exclusively funds private goods. It is also the basis for funding public goods and public service provision. Getting that mix right is complicated, but economists are not out there saying there should be no government and everything will take care of itself if we just have high GDP per capita. The argument is that high GDP per capita reflects a high productivity economy, and a high productivity economy creates the material basis for both private and public expenditures.
National development is about the process of creating both the material possibilities and the mechanisms for doing that—hence my fourfold definition of development. You could call me a growth fanatic, but I am not a growth-only person. You also need a decent government. In all of the empirical work, you can look at Equatorial Guinea—a few kleptocrats dominating a bunch of oil, with the rest of the economy completely disarticulated from that. Can you get to high GDP per capita and still have low levels of living? Yes, if all of that is being captured by relatively few people.
Mounk: You probably need a natural resource for that as well. It would be very hard to imagine a case where 0.1% of people have an incredibly productive company that doesn’t rely on natural resources and doesn’t rely on broader education and so on. It probably takes a somewhat special case—like kleptocratic control over natural resources.
Pritchett: The way I like to describe the goal—the way we should think about how countries can make progress—is what I call “inclusion into productivity.” The reason human beings are this amazing species is that we have learned how to cooperate to create value. A large part of the development process is getting more and more sophisticated ways in which people can cooperate over time and space to create value. When I look at a great big corporation with tens of thousands of employees, this is a mechanism of cooperation to create value—a cooperation across people with all kinds of different skills and contributions.
The process of development is getting more and more people out of being engaged in activities where they work a plot of land and attempt to eke out a living in low cooperation, and into more and more sophisticated value chains. That is inclusion into productivity. What I am actually concerned about is the productivity of individuals—but they are going to be more productive not by being more separated from a sophisticated modern economy, but by being more embedded into it. That dynamic, which leads to indicators of inclusive growth, is how I think about the fundamental dynamic on the growth side. Growth should be a process of more and more of the population engaged in higher and higher productivity, with more and more people included into these mechanisms.
Mounk: To go back to those two different challenges I posed earlier: one is whether poor countries can still grow at the same speed that others did, and your answer is yes—in fact, a lot of them are growing more quickly. There is a strange thing in the premise of that argument, which somehow implies that we haven’t seen examples of poor countries developing very quickly. That is partially because people who think about this tend to drop successful countries out of their sample. They don’t remember that China in 1980 was an incredibly poor country. Today we no longer think about China as being part of that sample because it has grown out of it. That in itself is evidence that rapid development is possible. If China was able to do that, there shouldn’t be an in-principle reason why Kenya or India couldn’t achieve the same feat.
The other question is what that looks like in the 21st century, when some of the historical development path is no longer available. My understanding is that your answer is going to center on those four development factors. Why are those so hard to implement in places where they are not in place? Ideas are really important, as you said earlier. If only the right people adopted the right ideas, they should be able to put those four things into place. But clearly it is more complicated than that. It is not just that the rulers of these countries have never had the right ideas. There must be obstacles beyond that.
Pritchett: One of the obstacles is that the process of growth of inclusive productivity is a transformational process. A transformational process requires winners and relative losers. In many places, economic, political, and bureaucratic structures congeal. They congeal because there is a certain way in which a country produces value, and that way of producing value often produces certain ways in which the government extracts value. That can lock into an equilibrium that resists change rather than encourages it.
There are low-level equilibrium traps—ways in which countries get stuck. We already alluded to one of those. In resource economies, if you are relatively well endowed with certain point-source resources—and by point-source I mean not land, which is diffuse and requires a geographically distributed population, but oil or diamonds—you can see how the people who mine the diamonds and the people who control the country get embedded in a relationship where, as long as that elite bargain can survive off the extraction of value from diamonds, it is not looking for anything else. It is not looking to solve the problem we are talking about, which is how a country comes to be embedded in more and more sophisticated value chains. If an elite can generate a bargain that sustains itself politically over time with just diamonds, it is not looking to do anything else, and that becomes an obstacle.
Economies and countries get stuck in an elite bargain that is more worried about the threats from new industries creating new power structures than about stagnation around the parts of the productive structure they are already in. You can easily get stuck in the dynamic between economy, polity, and bureaucracy in which an elite bargain undermines the rule of law and undermines the expansion of new opportunities, in the interest of playing defense around where they are. Finding countries that can handle this dynamic of change is very hard. That is why we see so few transformational successes, even though the global order makes them possible.
Mounk: Is there some way to change that, to contribute to that? If economists at the moment are focused far too much on whether giving people in this village one kind of intervention or people in that village another kind of intervention will lead to a little bit more growth—and you think that is fundamentally the wrong question to ask—do we have better answers to how we can help a country like Nigeria, with a very fractious political system and an elite bargain that is very bad for the country, get out of that trap?
Pritchett: Many of your questions have two sophisticated halves to them. Let me dwell on the first half—the premise that people are wasting their time. I want to emphasize that they are not just focused on the wrong question. They are focused on ontologically the wrong question.
The word ontology is not one I like to use, but what it means here, at its fundamental roots, is this: the problem with the world is not poor people. The problem is that people are in poor places. If you are studying the dynamics of how to make people less poor, you have to be ontologically studying the characteristics of the system, not the people. The methods being deployed by economists to study how to make people better off are focused as if the person were the unit at which we should be studying this. That is ontologically wrong.
Mounk: Let’s stay with this half of the question before moving on to the other half. Take a step back and explain to us what the dominant paradigm in development economics has been for the last 20 years and how that dominant paradigm is particularly vulnerable to the critique you’re making. I take it you’re mostly talking about what are called RCTs. What is an RCT? How did it become so dominant? Why do you think that’s the wrong way to ask this question?
Pritchett: If you characterize what development actors are doing as carrying out interventions, then you can get obsessed with understanding whether the particular intervention you’re undertaking has a truly causal effect. We can’t just evaluate a project on a before-and-after basis—we need to really separate out the causal impact. For separating out the causal impact, doing a randomized controlled trial is the best way to do that.
The problem is, if I’m trying to do a randomized controlled trial, I need some group of individuals to receive the treatment and some individuals to be the control group. That means I have already ontologically assumed that the important causal effects are individuated, as opposed to being country-wide or market-wide phenomena. This is as plain as I can make it.
Mounk: Let me try to put this in plain terms to see whether I understand correctly. The idea is: I study Village A and Village B, I give Village A deworming medicine and Village B nothing, or perhaps some other kind of treatment. If what I’m trying to figure out is how to spend a hundred million dollars for a charity, that is a very reasonable question to ask. The problem you are pointing to is that what actually explains why both Village A and Village B are very poor—versus Village A and Village B in England being relatively affluent—is these country-level characteristics: in particular, whether they have managed to figure out the rule of law and all the other things we have been talking about. You are defining the interesting stuff out of the question if you are just comparing two villages within the same country that are meant to be as similar as possible, and then seeing whether this or that kind of treatment is going to marginally improve the lives of people in each of them.
Pritchett: Let me take what I consider a paradigm example—one that, when you encounter it, should make you realize something has gone badly wrong. There was a paper in Nature—the most highly reputable scientific journal in the world—about carrying out an experiment in Niger where a cash transfer was given to some people and not to others. Bundled with that cash transfer was a psychosocial intervention, and the study then looked at whether this psychosocial intervention independently and causally caused individuals in Niger to see their incomes rise.
This was Nature magazine’s characterization of what development economics was doing, with 11 prominent development economists as co-authors. When you see that, you should immediately think: this is madness. People in Niger are poor because they are in Niger. Niger, on every indicator of national development, is a basket case. If you are not fixing Niger, thinking that you are doing good in the world by making tiny tweaks to psychosocial interventions at the individual level, you are assuming that a large part of the low standard of living there must be a characteristic of the people in Niger. That is just wrong—wrong by orders of magnitude. People in Niger are poor because they are in Niger. If you allow a person in Niger to move to France, their income will converge to that of people in France almost instantaneously, because France is a high productivity place and Niger is a low productivity place.
If Niger as a country had national development at the level of France, all of these problems would be addressed, because national development is a machinery for endogenously identifying and solving problems. If you don’t have that, attempting to solve these problems at the individuated level is wildly, ontologically wrong. You are not taking into account that the outcomes individuals have are determined not by their own characteristics, but by the characteristics of the political, organizational, economic, and social system they live in.
Mounk: The other thing that strikes me about this study is the theory of action lurking in the background. I have no doubt that people who encounter psychosocial problems are less productive, and I am moderately optimistic that the right kind of psychosocial intervention might reduce those problems and therefore make people more productive. But how on earth are you going to deliver psychosocial intervention at scale in one of the poorest countries in the world? It is not going to happen because an army of therapists is flown in from Brooklyn, New York to treat everyone.
How do you actually reduce the amount of psychosocial problems that people have? How do you increase access to therapy and other support for people who do have serious psychosocial problems? By making the country a lot more affluent. If the country is a lot more affluent, perhaps your child doesn’t die at age three and you have fewer psychosocial problems to begin with. You are less depressed about the fact that your child just died. You will have the money to afford a therapist, or perhaps the government will have the money to fund a mental health service. You are asking the question from the wrong end.
Pritchett: Then you get obsessed with cost effectiveness, which means reducing both the numerator and the denominator. In this study, per capita income went from $1.80 a day to $1.85 a day. If you look at Vietnam, in 1990 it had a per capita income of about $1.80 a day—similar to where Niger is now—and it has since gone up to around $7.50 a day. That is development. None of that in Vietnam came from people becoming psychosocially more capable first, with everything else following from that.
Mounk: It’s not that USAID or the Ford Foundation sent an army of psychologists to Vietnam.
Pritchett: It was because the development actor and discourse helped Vietnam change its mind about its development strategy. Vietnam said: we can embody Vietnamese labor into global value chains in a way that is going to radically change the productive possibilities and the inclusion into productivity of Vietnamese people. They did it successfully, and extreme poverty disappeared in Vietnam right before our eyes through a process that involved almost no direct anti-poverty interventions.
So that is the first point: this isn’t just wrong, it is ontologically wrong. Ontologically wrong means you cannot fix it by doing slightly better experiments on this or that. It is looking at the wrong set of ways in which prosperity actually happens. Prosperity happens through the inclusion of individuals into productivity—a cooperative endeavor involving artifacts and institutions whereby large numbers of people can cooperate to create value: markets, governments, bureaucracies, and large organizations. Not whether I am a more go-get-em person. That was the first part.
Mounk: The second part then is: if RCTs are the wrong thing to look at, if that is not how we are going to make change, if we shouldn’t be obsessed with whether it is this intervention or that intervention at the level of the village or the town—but rather should be asking how we help countries choose successful development paths like Vietnam, paths that put in place the institutions needed to actually develop—how can development economists or other social scientists help with that?
Is the problem fundamentally that the people in Nigeria just don’t get it? Or is it, as you were saying earlier, that the people in charge have their own interests and want to preserve the privileges they have? They are worried that in the process of development they might be displaced from power, and so they would rather continue to be the kleptocratic elite in a relatively poor country than be displaced in a much more affluent one. Is development economics, or more broadly the field of social science, actually going to be able to help in that process? Or is that a hopeless endeavor?
Pritchett: It is not a hopeless endeavor, because after all, it has happened. The strange thing about saying it is a hopeless endeavor is that it has to ignore that Korea happened, and Taiwan happened, and Vietnam happened, and China happened, and Indonesia happened, and India happened.
Mounk: But did it happen because those places, for whatever reasons, had more favorable circumstances where they were more able to challenge their elites—or did we just get lucky that they put a few of the right reforms in place? Or did it happen in some way because social scientists helped guide them?
Pritchett: Yes, social scientists helped guide them, because social scientists created the data—like GDP per capita—to show that some countries have had really rapid growth by doing certain things and other countries have not. Deng Xiaoping didn’t just decide on moving towards a path of unleashing incentives for people to engage in private enterprise—perhaps too strong a word for what happened in China in 1978—in an intellectual vacuum. He looked around the world, brought in experts who had studied economics and the relative performance of economies, and was convinced that there was an alternative path from where China was to where China wanted to be.
The idea that ideas are irrelevant and that social science hasn’t had an impact is surreal. In the case of India, which I know relatively well—I have been going back and forth to India since 1992 and have spent time living in different cities there—there was a debate in India in 1991 about how to respond to an incipient macroeconomic crisis. That debate drew on decades of social science about the relative importance of market versus government-led development, and of import substitution versus export orientation. It drew on a body of empirical science that had been generated over time, and it led the country to do different things. You cannot act as if India decided what it did in the reform period in an intellectual vacuum. It is surreal to say that social science cannot in principle affect the way governments act, because it has demonstrably done so in specific instances. Maybe it has not always been as successful as one would hope, and let’s admit that oftentimes social science is pushing against direct material and political interests.
But let me answer your question. The answer is what I call full Trinity growth diagnostics. If development economics were focusing its time, effort, and capabilities on the development of full Trinity growth diagnostics, I think we would be in a radically better place to help countries with their organic, country-level development strategies.
What do I mean by full Trinity? For something to be successful, it has to be technically correct—it has to have a correct causal model of how, if actors undertake this action, working perhaps through a complex adaptive system, a particular outcome will result. It is not direct cause and effect, because markets are complicated and complex adaptive systems. But the technical question is: if I do this, will this be the result?
The second element of the Trinity is organizational. If I am asking a government to undertake actions to promote a given outcome, does the government have organizations capable of doing that? If I say I am going to do industrial policy to promote high-tech industry, do I have government organizations capable of identifying and promoting high-tech industry?
The third element is that it has to be politically supportable. I have to be able to assemble and sustain a political coalition to support the implementation of these actions.
The first step of this is a growth diagnostic—a process that I and colleagues at the Kennedy School developed over time—of asking, in a given place, what are the binding constraints to having more rapid growth than we are currently experiencing. The binding constraints matter because when we ask what is good for growth, we tend to produce a very long list. If you say “get to Denmark,” you can say Denmark has this and Denmark has that. But you cannot do everything that Denmark now does. You have to do the right thing. Identifying the right thing involves an analytical process of prioritizing among the many things you could do—asking which would have the biggest impact on the prospects for a rapid, sustained, and inclusive episode of growth. We have mechanisms for trying to adjudicate among contesting claims about what the binding constraint is.
But then we have to supplement that by asking: of the things that should be priorities and would have a big impact on economic growth, which can we actually do? Saying that if you had Korea’s current capabilities you could do a certain thing is not very helpful if you have Niger’s capabilities. We have to ask which of the binding constraints can realistically be addressed. Then you have to have a realistic, positive model of the politics—which of these can the existing elite be persuaded to engage in? None of those questions poses an intellectual challenge anything like reconciling general relativity with quantum mechanics.
Mounk: But it’s an art rather than a science, I imagine. What might that look like concretely? Let’s say there is a new government in some country—genuinely reforming in certain ways, perhaps a little corrupt in others. They have just come to power, perhaps through election, perhaps in some other way. They can make some changes, but they cannot make a million changes at the same time. It is hard to drive change across sectors simultaneously, and there is a limited amount of money to invest because it is a very poor country. Should they invest in education or in the judiciary? Should the first big push be to reform the judiciary—so that who wins a dispute is driven more by the facts and creates stable expectations, rather than by bribes or family relations? Or should they start with a big push on schools, making sure teachers are competent and actually show up? I imagine that is the kind of choice a government might realistically face. How should you go about making that decision?
Pritchett: There are analytical tools for addressing this, and this is where we get into the mechanics of what a growth diagnostic would do. Partly what a growth diagnostic does is ask: if it were the case that a proposed action were the binding constraint to growth, what should we observe about the economy?
There are four or five things we look for. First, if we think X is a binding constraint—say, corruption—then when we see relaxations or improvements in corruption, we should see more growth. If I have my arm in a sling and the movement of my arm is constrained by that sling, then if I remove the sling, I should be able to move it a lot more. If I can’t, maybe it was the shoulder injury all along and there is a different cause. Changes in the constraint should cause changes in outcomes.
The second is what we call bypassing. If finance is a constraint and I could otherwise be productive and profitable, we should see firms actively engaged in creative ways of raising finance. We should see enterprises actively adapting around the constraint.
The third involves what we have given the quirky name of camels and hippos. Different industries in an economy are more or less intensive in a particular proposed constraint. If we say water is a constraint, then we should observe camels—animals well adapted to a lack of water—and we should not observe hippos. If we look at the economy and the industries that are thriving are ones that economize on the proposed binding constraint, that is evidence in favor of its being the binding constraint.
These are plausible, sophisticated ways of analyzing the current situation of a country. They can produce answers like: putting more kids through school really isn’t a binding constraint here, because we don’t see changes in that producing changes in growth, we don’t see firms desperately engaged in training because they are short on skills, and we don’t see labor-intensive, non-skill-intensive industries thriving because they are economizing on the lack of skill. That means we need other explanations.
This technique has been deployed and it produces interesting results—not clean RCT-looking results, nor should one expect that. To me, the key test of a growth diagnostic is whether it comes to different answers for different countries. If you come to the same answer for every country, you have an ideology. If every country you go to produces the answer that reforming the trade system would accelerate growth, you are an ideologist. Maybe you are right sometimes, but you are not always going to be right, because there are going to be societies in which that is not the most important binding constraint. When we do these growth diagnostics in different countries, we come to demonstrably different answers. The data and the evidence line up around different things.
This is a promising technique, but it has been radically under-invested in—because the resources in the world have been devoted to deworming studies rather than to creating a global community of practice that does technical growth diagnostics. More support to a global community of practice focused on honing the art and science of growth diagnostics would be a tremendous way the world could support the organic process of countries finding their own development paths. It would allow you to engage with countries and actors within countries in a radically different way than coming to them and saying: you lack these 50 things, all of which would be good if you did them—when you know they are going to be able to do two, maybe three.
In the rest of this conversation, Yascha and Lant discuss migration, how to build a fair asylum system, and what the destruction of USAID means for the future of development. This part of the conversation is reserved for paying subscribers…
