Jeffrey Sachs
What this covers
Sachs and Tyler Cowen examine why development economics too often reduces complex realities to single explanatory variables—above all, institutions—when the actual work of lifting countries out of poverty requires something closer to Differential Diagnosis (Clinical Economics)). Sachs walks through the evidence against institutional determinism: cross-country regressions from 1960 fail to predict which countries actually grew, and the lived history of financial crises, geographic constraints, and resource dependencies tells a messier story than any one-factor model allows. The conversation moves through manufacturing, agriculture, East Asian development, and climate to show how the same mechanism—say, resource wealth or corruption—produces vastly different outcomes depending on context, institutions, and timing.
The discussion pivots repeatedly on a core distinction: Catching Up vs Forging Ahead, which Sachs argues requires fundamentally different policies and institutions than economists' models usually acknowledge. Where Daron Acemoglu and James Robinson emphasize innovation-driving institutions, Sachs contends that successful East Asian economies—China especially—have grown rapidly through deliberate policy coordination and technological adoption without matching the free-market template. He contrasts this with Russia's 1991 crisis, which he treats as a distinct catastrophic pathology requiring emergency stabilization, not standard reform advice. The conversation also covers Paul Krugman's mishandling of the 2008 financial crisis, the resource curse as context-dependent rather than mechanical, how climate change threatens vulnerable geographies, and why Sachs believes economics needs retraining—more fieldwork alongside engineers and doctors, less timeless modeling. He closes on China's 21st-century ascent as perhaps the largest material improvement in human welfare in history, one his peers have consistently underestimated.
Sachs argues that development cannot be reduced to a single master variable (especially 'institutions'); it requires context-specific 'differential diagnosis' that takes geography, resource base, financial crises, and the distinct logic of 'catching up' seriously alongside institutions.
- Cross-country regressions in 1960 show no measure of institutions reliably predicts which countries succeed, undermining single-cause institutional theories
- Catching up is a fundamentally different phenomenon from frontier innovation, requiring different institutions and active policy
- Financial crises are a distinct pathology (Russia 1991) that explain divergent outcomes more than ideology or advice quality
Institutions matter, but geographic blindness from overemphasizing institutions ignores how geography, resources, and institutions interact contextually.
- Sachs has never claimed institutions don't matter; he argues that the 'institutions matter' idea has been taken to an extreme that causes 'geographic blindness,' when in fact geography, resource base, and institutions all matter, and the art of good economics is figuring out what is important in which context.
“This "institution matters" idea is one of those that I've found taken to extremes by causing a kind of geographic blindness, if I could say it.”
- There is a real natural resource curse for many resource-rich countries, but looking back Sachs would now put more weight on political economy factors—including vulnerability to invasion and to internal and geopolitical disruption—than on the pure market-driven 'Dutch disease' mechanism he emphasized 25 years ago.
“At this point, looking back, I would put more weight on the political economy aspects probably than the pure, market-driven "Dutch disease" aspects which I talked about a quarter century ago.”
- Whether resource wealth is harmful depends heavily on context: stable advanced economies like Norway, Australia, and Canada will adjust to oil price shocks (their currencies weaken, they find other things to do), while chronically dependent and vulnerable economies like Venezuela or the Middle East fall into deeper crisis—as the Soviet Union did when oil prices stayed low.
“Norway will do well, whether the oil price is high or low. On the other hand, last time the world oil price was persistently low, the Soviet Union collapsed.”
- Corruption's impact on growth is context-dependent: some places are so corrupt they are driven into the ground, while other long-corrupt places (New York, Washington, Lagos, Beijing) achieve substantial economic progress—China grows rapidly despite massive corruption and the US has prospered over 200 years despite a huge amount of corruption.
“Other places have been corrupt for a long time — I'm thinking of New York, Washington, Lagos, Beijing, and other places — and you get a lot of economic progress.”
- While the idea behind cities doing something different is welcome, Paul Romer's charter cities concept—an extraterritorial place not governed locally but by an external board—never struck Sachs as politically realistic, and it did not turn out to be.
“I'm not happy to say, "We're going to take an extraterritorial place in your country. It's not going to be governed locally. It's going to have an external board."”
Catching-up economies need different institutions and policies than technology-frontier leaders, enabling rapid growth through export orientation and technology diffusion.
- Because of the catching-up phenomenon, a lagging country can achieve very rapid advance (aiming for ~10% growth rather than the ~1.8% per capita the US sustained for 200 years), but this requires concerted policy to close gaps and a 'big push' on many fronts—not government simply stepping back and letting things happen.
“Through pure government just stepping back and letting things happen? No. Through a concerted policy to close gaps. That's the difference.”
- Productivity converges quickly in manufacturing but only slowly in agriculture because manufacturing uses standardized technologies embedded in international value chains, while agriculture depends on local soil, pests, climate, and crop varieties, making technological diffusion much more complex and context-specific—as shown by the Green Revolution requiring local experimentation before Sonoran Mexican seeds worked in the Indian Punjab.
“The reason you can get manufacturing convergence of productivity easily is that those are standardized technologies, usually part of international value chains.”
- Acemoglu and Robinson's single-cause focus on innovation misses that development has at least two fundamental drivers: forging ahead at the technology frontier (innovation) and 'catching up.' The institutions required for catching up are quite different from those needed to be the technology leader, which explains why successful East Asian economies—and China especially—do not resemble the textbook free-market model.
“The phenomenon of catching up is very different from the phenomenon of forging ahead at the front of the technology horizon.”
- There is no special magic to manufacturing; what developing countries crucially need is to export, because exporting earns the means to import the technology (99.9% of which comes from outside) needed to develop—so import-substituting manufacturing failed where pursued to excess, export-oriented manufacturing worked, and export-oriented services, mining, or agribusiness could substitute.
“I don't think that there's any magic to manufacturing. What there is crucially is a need for all developing countries to export. You need to export because you need to import technology.”
- The Barro and Sala-i-Martin convergence finding—that a country loses about two percentage points of growth each time it closes by half the distance to the technology frontier—is about right, implying China's high growth rate will gradually decline as it converges.
“I think that the general idea of Robert Barro [and Xavier Sala-i-Martin] from 20 years ago that every time you approach by half the way to the frontier you lose about two percentage points of growth is about right.”
Geography and resource endowment fundamentally constrain development prospects for landlocked and disease-burdened economies.
- Every country has a comparative advantage, but if a country has nothing it can actually sell in world markets, comparative advantage doesn't make it prosperous—so countries' real constraint is the balance of payments, and landlocked, dry, disease-burdened economies like Chad, Niger, and Mali with no easy resource base and ports too far for assembly operations face extraordinarily difficult development.
“Everybody, every place has a comparative advantage but it doesn't necessarily make you prosperous.”
- Africa's demographic trajectory—from one billion today in sub-Saharan Africa to roughly two billion by mid-century and about four billion on the UN medium-variant fertility path—makes sustainable development in those terms infeasible given the existing difficulties of disease, tropical environments, and landlocked geography.
“Africa's going to have two billion people before the middle of the century on the trajectory. It's going to have around four billion people on the so-called medium variant fertility trajectory that the UN population division puts forward.”
- Africa is better off having to catch up today than 25 years ago because the basic tools for catching up—leapfrogging technologies and the early-stage information revolution (still on the first slope of the S-curve)—are stronger now, even though demography and climate change create severe headwinds.
“I'd say better off now, because the basic tools for catching up with the information age are stronger now than before.”
- International capital markets underperform their role of financing international flows because of inherent shortcomings and institutional barriers—especially the difficulty of getting repaid on cross-border loans—so institutions like the Export-Import Bank that facilitate capital flow from rich-country savings to capital-scarce developing economies are valuable and too few in number.
“I believe that in general, our international capital markets underperform their role of financing international flows. I think we need institutions to overcome the inherent shortcomings of international finance.”
Economics relies on outdated static models and should adopt clinical case-based methods, biologist-style structural analysis, and interdisciplinary field engagement.
- Economics' core weakness is epistemological: it relies on static, timeless, general models while the real world has undergone 230 years of relentless technological, structural, and cultural change, so economists avoid studying specifics for fear of becoming dated and end up always behind the curve—whereas they should work like biologists who study the actual structure (e.g. four base pairs, double helix) rather than abstract assumptions.
“Yet our economics models are basically static, meant to be timeless. If we really want to understand the world, we need to go deep into understanding what Baxter is doing — Baxter the Robot — or Watson, or what really is changing technologically.”
- Economics should operate like clinical medicine through 'differential diagnosis': rather than applying one explanation to all cases, the economist should ask sequenced questions down a decision tree to identify the core causes for a proper prescription while minimizing serious risk, recognizing that a problem seen in one place is not the same as a problem seen elsewhere.
“The purpose of a differential diagnosis is two things. First, it is of course to try to get to the core reasons so that you can make a proper prescription based on a proper diagnosis. Second, it's done in a way that you're minimizing serious risk.”
- Economics relies on an 18th-century theory of humanity in a world that now has 20th-century evolutionary psychology, neuroscience, and a century of psychological understanding, so the field needs a new 'anthropology'—a better theory of man—and should value the thick descriptive understanding of societies that anthropology provides.
“We have an 18th-century theory of humanity in a world of 20th-century evolutionary psychology, neuroscience, and a century of psychological understanding. We need a new anthropology in our field.”
- Economists should be trained the way doctors are—grappling with real problems in real places, working alongside engineers, public health and medical professionals—because the complexity of interacting physical, technological, political, economic, and natural systems defines the right response, and any dataset is only a tiny fraction of what can be known about a problem, so the right epistemological approach is to get out into the world rather than to mine data.
“I wouldn't want to train doctors without the medical students walking the wards with their mentors. I don't like training economists without them grappling with real problems in real places.”
Clear focal points and shared knowledge enable cooperation in prisoner's dilemmas and multi-actor coordination problems like climate change.
- In messy, multi-actor noncooperative problems like climate change, the way to make progress is to identify and clearly articulate a Schelling-style focal point—a concrete description of where everyone could go that is an improvement on the status quo—and repeat it clearly enough that actors converge toward it, since there is no single leader who can impose a solution.
“It comes back to what [Thomas] Schelling called a focal point for bargaining. Here's where we could go.”
- Sachs holds a theory of change in which knowledge matters: greater clarity, evidence, and appropriate analysis make good outcomes more likely, and much of what poses as pure zero-sum struggle or harsh ideological conflict is often resolvable through good clear ideas, evidence, or a clear game plan among antagonistic actors.
“I tend to believe there's a way out of a crisis, and I tend to believe that a lot of what poses as either pure zero-sum struggle or harsh ideological conflict is often resolvable by good, clear ideas, or good, clear evidence, or a good, clear game plan.”
- A crucial structural feature of human life is repeatedly facing prisoner's dilemma situations where the logic of defection produces noncooperative outcomes worse for everybody; the Kennedy-Khrushchev resolution of the Cuban Missile Crisis and the nuclear test ban treaty exemplify finding the way to the cooperative outcome despite hard-liners on both sides insisting agreement is impossible.
“If I could put it in pure conceptual jargon, it's a story of finding a way to the cooperative outcome of a prisoner's dilemma.”
- Poland's successful transition worked because it had a clear, compelling focal point—'return to Europe' / becoming a normal European country—which gave coherence to all the rapid institutional reforms, like carving an elephant from ice by removing everything that doesn't look like a European normal economy.
“But the point I want to make, Tyler, is that the endpoint was clear for them. Everything else fits, once you have an idea of that compelling focal point.”
Automation and robots displace labor, creating dangerous inequality without government redistribution between capital owners and substituted workers.
- Robots and AI systems function as a kind of immigrant, competing with workers by replacing the services they provide, so automation produces a distributional effect similar to immigration debates—those who own the technology gain while substituted labor loses income.
“Yes, the robots are a kind of immigrant, so they're competing with the other workers that provide some services that are being replaced by the machines or by the artificial intelligence systems.”
- Smart machines have for 200 years allowed humanity to escape the heavy physical labor it took on with the Neolithic shift to sedentary agriculture roughly 10,000 years ago; in principle automating that labor is desirable, as people abandon backbreaking agricultural work whenever they can, but it becomes dangerous if it produces massive inequalities of power, income, and wealth or Orwellian mass surveillance.
“what smart machines have been doing for 200 years, is allowing us to ease that physical burden. We've had robots for a long time now.”
- The rapid ascent of information technology and robotics is displacing many jobs across manufacturing and increasingly services, which is one cause of low and stagnant real wages; while letting machines do the heavy lifting is in principle a great thing humanity has sought since the Neolithic, those who own the robots make fortunes while those whose labor is substituted see real income decline, so the distributional transition requires government intervention (redistribution from capital owners to labor, old to young) to make everyone better off.
“I believe that it is displacing lots of jobs and it is one of the reasons for the low wages, the stagnation of real earnings in the economy.”
East Asia's rise and China's success disprove models assuming institutional prerequisites for development; geopolitical power, not economics alone, shapes policy support.
- China will be a great and successful country in the 21st century; Summers, Lant Pritchett, and Krugman have been wrong about China for 20 years because, like Acemoglu, they assume that because China doesn't fit their model it must collapse—but the East Asian upswing of one and a half to two billion people is happening and is the most significant scaled improvement of material conditions in human history.
“Summers and Lant Pritchett have been wrong about China for 20 years. And they continue to be. And Krugman, too.”
- For the first 200 years of modern economic growth, the North Atlantic (US and Europe) generated roughly 90% of the technological advance to which the rest of the world responded, but because of Japanese, Korean, Taiwanese, and especially post-1978 Chinese development, East Asia has become a third (arguably second) transformative global growth pole of fundamental geopolitical, historical, and economic significance.
“What's happened of fundamental significance for the world is that East Asia has become the third growth pole in the world, or arguably the second.”
- Mexico is roughly accurately rated as underperforming, stuck in middle-income stasis for 100 million people, partly because of the paradox that the United States created as many problems as opportunities—including a militarized narcotics trafficking and war on drugs that has been devastating for modern Mexico.
“The paradox: that somehow the US has created as many problems as it's created opportunities. It has created a militarized narcotics trafficking and war on drugs, for example, which has been devastating for Mexico in modern times.”
- The same reform advice that the US strongly supported for Poland in 1989 was flatly rejected by Washington and the IMF for Russia in 1991—not because the economics differed, but because the US wanted a unipolar world and viewed Poland as a NATO bulwark while Russia remained on the other side, so the quality of advice was irrelevant to whether it was backed.
“We didn't want to help Russia in 1991. We wanted our unipolar world.”
- Climate change will intensify water crises globally—drying the American Southwest, Great Plains, and many descending-Hadley-cell zones while faster evapotranspiration dries soils—and these ecological crises can turn into social catastrophes, as Syria's record drought from 2006 to 2010 contributed to the social ramifications behind the 2011 explosion of violence; we should not assume crisis automatically produces innovation rather than catastrophe.
“Syria had a huge drought, the biggest in its modern history, from 2006 to '10. It led to many social ramifications that contributed to the explosion of violence starting in 2011.”
Financial crises are distinct pathologies requiring emergency-stabilization treatment, not stimulus; China's absence of acute crisis in 1978 enabled divergence from Russia.
- A deep financial crisis is a distinct category of pathology—fulminant and capable of ripping a society apart quickly, like a meningitis epidemic—that requires its own specific treatment (debt cancellation, payment standstill, emergency stabilization fund); Russia in 1991 had this acute crisis from collapsed oil prices and Gorbachev's depleted reserves while China in 1978 did not, which is a key reason their transition paths diverged.
“The first point is financial crisis is in and of itself a distinct category of pathology. Russia had it, China, thank goodness for them, didn't have it.”
- Krugman both misdiagnosed and mis-prescribed the 2008 crisis: it was not a Keynesian animal-spirits collapse needing fiscal stimulus, but a Hayekian credit-boom displacement that became a crisis only through the Lehman Brothers failure triggering a financial panic; the correct treatment was unlocking the panic via Bernanke's liquidity flood and QE, not the 2009 fiscal stimulus, which diverted political capital and public investment from longer-term priorities.
“It was two things, I would say. If you want to classify it, it was a Hayekian displacement from a credit boom, but even that wasn't enough to lead to a crisis. It was Lehman Brothers failing in September 14, 2008, that drove a financial panic.”
- Krugman's vitriolic criticism of the UK Cameron government's economic record is ideologically driven rather than fact-driven, because placing the UK recovery next to the US recovery shows they look almost identical, yet Krugman praises the Obama recovery while condemning the UK's.
“What I did was just put the UK recovery next to the US recovery, and they look almost identical. He loves the Obama recovery, and he hates the UK. He says one's awful, one's great. It's ideology. It's not fact-driven.”