Joseph Stiglitz
What this covers
Joseph Stiglitz engages Tyler Cowen in a wide-ranging conversation rooted in a single organizing principle: that imperfect and asymmetric information fundamentally reshape how markets and institutions actually work, upending the textbook models taught in most economics courses. Rather than defending a single thesis, Stiglitz moves across decades of his own research—from sharecropping to financial markets to climate policy—showing how institutions that appear wasteful or puzzling become rational once you account for the Principal-Agent Problem and Risk-Incentive Tradeoff. The conversation is conversational and wide-roaming, touching on his time working in government, his skepticism of deregulation, and his evolving thinking about when centralized hierarchies help or harm decision-making.
The substance divides into several clusters. On information itself, Stiglitz explains why the Grossman-Stiglitz Paradox—that informationally efficient markets would destroy the incentive to gather information—undermines the efficient-markets view, and traces how the same logic applies to AI scraping value from creative work. On labor and credit, he argues that unemployment and credit-allocation failures are not departures from optimal equilibrium but symptoms of deeper monitoring problems. On policy, he challenges the notion that shareholder value maximization serves the broader economy, contends that the 2008 crisis stemmed from poor credit allocation rather than excess credit, and frames climate action not as a drag on developing economies but as pro-growth given falling renewable costs. He also defends land-value taxation, modern industrial policy including the CHIPS Act, and zoning rules through the lens of One Person's Freedom Is Another's Unfreedom—arguing that unfettered individual choice often generates Externality that legitimates collective rules. Throughout, he contests the assumption that unfettered markets self-correct.
Stiglitz argues that markets are pervasively shaped by imperfect and asymmetric information, so unfettered freedom and standard competitive models mislead; efficient outcomes often require attention to incentives, risk-sharing, credit allocation, and the recognition that one actor's freedom constrains another's.
- Sharecropping and information economics show seemingly inefficient institutions arise rationally from risk-incentive tradeoffs under imperfect information
- Informationally efficient markets are impossible because they would destroy the incentive to gather information
- One person's freedom is another's unfreedom, justifying zoning, climate policy, and a rules-based order
Shareholder value maximization does not produce socially optimal outcomes when firms ignore externalities and market failures.
- If all firms maximize shareholder value, this does not in general lead to welfare maximization for society.
“The result of that was that shareholder value maximization did not, in general, lead to welfare maximization.”
- Freedom is a person's opportunity set, and when one person exerts an externality through his freedom he constrains the freedom of others; this is why unfettered building without zoning is undesirable, since a tall building can deprive a neighbor of light, create noise, or place a brothel next door.
“one person's freedom is another person's unfreedom... when one person exerts an externality on another by exerting his freedom, he's constraining the freedom of others.”
- Rent-seeking has become an increasingly important issue creating a divergence between what gives rise to the wealth of nations and what gives rise to the wealth of particular individuals, a distinction central to understanding 21st-century wealth.
“there's a big divergence of what gives rise to the wealth of nations, and what gives rise to the wealth of particular individuals.”
- Hugo Chávez and especially his successor inflicted enormous costs on Venezuelan society and the wider Western Hemisphere, with the resulting flood of migrants creating migration problems that have affected the politics of much of North and South America.
“the cost to his society, particularly even more of his successor, has been enormous... because of the flood of migrants from Venezuela... affected the politics of much of North and South America.”
Information is fundamentally costly and its gathering cannot coexist with efficient markets; trading motives and market microstructure shape real behavior.
- Much modern trading is computers trading with other computers to make a micro-cent off another trader by eliciting information, rather than informed people trading with uninformed people or trading to hold a position.
“computers trading with other computers, so it's not informed people trading with uninformed people. It's really computers with one body of data trying to make a micro cent from another trader... It's not trading for holding a position, it's trading to elicit information.”
- If markets were perfectly informationally efficient and transmitted all information costlessly from informed to uninformed traders, no one would have any incentive to gather information; therefore informationally efficient markets are impossible, contrary to the efficient-markets view.
“if that were the case, there would be no incentive for anybody to gather information. So the market might be transmitting information, but it would be all free information. It would be information that nobody had done any work to collect.”
- Google and AI scraping information from newspapers, podcasts, and other sources without paying appropriates the value of others' knowledge, which will decrease the incentive for others to produce high-quality, valuable information, mirroring the Grossman-Stiglitz information-incentive interaction.
“with Google and AI scraping so much information... they're trying to appropriate the value of the knowledge that's been created by other people without paying for it. If they succeed in doing that, of course, that will decrease the incentives for others to produce information of high quality and of value.”
Strategic government policies addressing learning, dynamics, and resilience—not just market power—justify industrial policy despite trade rule tensions.
- The critical issue in strategic trade policy today is not the market-power rationale Krugman drew from the Dixit-Stiglitz monopolistic competition model, but dynamics, learning, and resilience; markets fail to price defense and resilience risks, justifying policies like the CHIPS Act despite their being near-violations of WTO rules.
“the more critical issues in strategic trade policy that we're facing today are not those that Paul Krugman argued for... It's really about dynamics, learning, and resilience.”
- Modern green industrial policies like the IRA grab jobs for the US at the expense of developing countries and even Europe that lack resources to compete, breaking the old trade model where everyone benefits and thereby requiring a rules-based order.
“we're trying to grab more jobs for us in this green transition, but the developing countries and emerging markets don't have the resources to engage in that kind of policy... our success in some of these areas comes at the expense of others.”
- Climate change is a global public good and developing countries, disproportionately located in the tropics, will be most adversely affected, so curbing emissions is in their interest; falling renewable prices (down 90%+ in 15 years) and the decentralizability of renewables mean an early green transition is actually pro-growth for them, making the growth-versus-green tension a misframing.
“Climate is a public good. It's a global public good from which all of us will benefit, and most especially those in the developing world and emerging markets, which disproportionately are located along the tropics”
Credit allocation matters far more than money supply; financial deregulation and weak supervision caused the 2008 crisis and prolonged recovery.
- Measured liquidity is misleading because much of it comes from fast/flash traders who exit instantly; what seems like abundant liquidity dries up precisely during big events, with markets sometimes ceasing to function and trades not actually executing, so the self-correcting capital argument fails at critical moments.
“It seems like there's a lot of liquidity, but then all of a sudden, when you have a big event, that liquidity dries up.”
- The 2008 crisis was not caused by too much credit per se but by poor credit allocation; deregulation and low rates meant no one watched where credit went, supervisors failed to monitor risk, and banks committed fraud by passing off mortgages they falsely claimed to have inspected. Rapid credit growth in one area (housing) is a warning sign.
“the issue isn't the amount of credit. It was the allocation of credit. If they had used that credit for productive uses, how much better our economy would have been.”
- What matters in monetary policy is not the money supply or the interest rate but credit availability; in 2008 providing massive liquidity to banks did little because banks were reluctant to lend, producing a slow recovery, so fiscal policy would have been more effective than monetary policy.
“I think what matters is not the money supply, not the interest rate; it's the credit availability of monetary policy... providing so much liquidity to the banking system didn't help that much, that the banks were very reluctant to lend out that money.”
Decentralized decision-making and institutional learning outperform hierarchical approval; personal background shapes economic priorities and policy choices.
- The conditions under which all shareholders would agree on firm decisions regardless of their beliefs or perceptions are very, very restrictive.
“That theorem said, what were the conditions under which all shareholders would agree with each other, regardless of what their beliefs were or perceptions, just like if they knew what prices were? The answer was very, very restrictive conditions.”
- In political campaigns you may not succeed while you are present, but sowing the seeds of discussion and debate can lead to winning the argument 5, 10, 15, or 20 years later.
“Sometimes you begin a campaign knowing that in the next year, two years... you may not succeed, but sowing the seeds of discussion, debate, maybe in 5, sometimes 10, sometimes 15, 20 years, things turn out and you wind up winning the debate.”
- Since all humans are fallible, decision systems trade off rejecting good projects against approving bad ones; hierarchies (sequential approval) versus polyarchies (many decentralized decision-makers) screen errors differently, and living in real Washington hierarchies made Stiglitz appreciate decentralization more, since hierarchies screen out too many good ideas, especially when leaders are poor decision-makers.
“Underneath this was the idea that all humans are fallible. There are going to be some cases of disapproving good projects and some cases of approving bad projects. How do you balance the two?”
- Coming from Gary, Indiana, the epitome of US de-industrialization, gave Stiglitz empathy for those who did not start at the top and made him far more concerned with inequality than his colleagues, a thread running throughout his career.
“coming from Gary, Indiana, gave me empathy for those who didn't start at the top in life... I was more concerned about inequality than most of my colleagues.”
- Poland's success was not due to shock therapy, which had a negative macroeconomic effect, but to abandoning it quickly and adopting a gradualist policy of building institutional infrastructure, aided by EU membership providing a legal framework and by returning migrants bringing skills and capital.
“it wasn't the shock therapy that had such a negative macroeconomic effect. It was the fact that after that moment of shock, they began a gradualistic policy of reform, of creating the institutional infrastructure that is the basis of the market economy.”
- Although advanced countries created the historical carbon problem since the Industrial Revolution, developing and emerging market countries are now the larger emitters of greenhouse gases, so they bear responsibility and climate change cannot be addressed without their participation.
“right now, actually, the developing countries in emerging markets are the larger emitter of greenhouse gases, so they really have a responsibility, and we're not going to address climate change unless they're on board.”
- Fraternities at a small college are divisive because they segregate students by membership and class, undermining the goal of forming one cohesive community-wide network rather than the efficiency-enhancing networks an outsider might assume.
“They were just, I thought, very divisive in a small community... our objective was to have a network of all the kids working together.”
- Debating is a valuable intellectual skill because students must argue both sides of a topic, forcing them to understand the opposing argument well enough to present it themselves.
“you have to take both sides... you have to be able to think, to understand the other side so well that you can even present their argument. That's a really important skill.”
- The colonial heritage in newly independent Kenya was very present because the colonial masters had failed to train enough civil servants to run a well-functioning government.
“the fact that there weren't enough civil servants, people trained to run the government, that the "colonial masters" had failed in their duty to enable it to be a well-functioning government”
- Lecturing is a learning experience because questions stimulate new developments of an idea and help the speaker understand the listener's mindset, allowing future presentations to be adapted to how the audience sees the world.
“No, every lecture is a learning experience for me. I learned that often, questioning will stimulate me to think of other developments of the idea, other aspects I hadn't thought of.”
Information asymmetries and incentive problems make interlinked markets and risk-sharing arrangements rational and efficient under imperfect monitoring.
- In the presence of incentive problems, it is efficient for a landlord to integrate credit, seed, fertilizer, and milling (interlinked markets) rather than have separate markets, because subsidizing the right seed and fertilizer elicits more effort from a tenant who otherwise underworks due to effectively being taxed 50%.
“in the presence of these incentive problems, it turned out to be efficient to have this integration of these various activities. For instance, by providing the right kind of seed, subsidizing the seed... the landlord could elicit more effort on the part of the tenant.”
- Sharecropping persisted for thousands of years despite imposing an effective 50-67% tax because, under imperfect information and high risk, farmers cannot bear the risk of full land tenancy; sharing the crop shares the weather, disease, and other fluctuation risk with the landlord, which is the foundational principal-agent risk-incentive tradeoff.
“in the absence of perfect information and the presence of a lot of risk, farmers couldn't bear the risk of land ownership... With sharecropping, they divided that risk, and a lot of the risk was borne by the landlord. That was a model of what came to be called the principal-agent problem”
- Because employers cannot perfectly monitor workers, workers have an incentive to shirk; if being fired meant instant rehiring under full employment there would be no deterrent, so equilibrium requires some unemployment as an incentive device, refuting the standard view that labor demand equals labor supply with no unemployment.
“if you couldn't monitor labor at every moment of time, workers would have an incentive to shirk... if they were fired, with no unemployment, they'd be hired the next moment, so there'd be no incentive to work. So, part of that equilibrium, there had to be some unemployment to induce people not to shirk.”
Land and natural resources are systematically undertaxed despite their role in generating windfall gains and optimal public finance.
- Land ownership remains one of the most important bases of taxation and is almost surely undertaxed, as shown when public investments like a new subway create windfall land-value gains for nearby owners paid for by all citizens.
“the ownership of land still provides one of the most important bases of taxation, and we almost surely do not tax it as much as we should. When the government, say, in New York City, builds a subway, those near the subway have an enormous increase windfall gain from the value of their land.”
- There is a theorem that for optimal-size cities, taxing land rents at 100 percent would yield exactly the right amount to finance the optimal level of public goods, capturing Henry George's insight that land rent taxation is the most efficient way to raise revenue.
“if you have optimal-size cities and you tax the rents 100 percent, that would be exactly the right amount to finance the optimal amount of public goods.”