What this covers

Krugman, speaking with Tyler Cowen in 2018, ranges across economics, politics, and technology while grounding his analysis in two main claims: that rising income inequality has driven partisan polarization by dragging one party rightward and collapsing the political center, and that standard economic models—from labor markets to monetary policy to trade—obscure rather than illuminate reality. The conversation moves from big structural questions about American politics and inequality to specific puzzles about immigration, interest rates, tech platforms, and the future of work.

On inequality and politics, Krugman argues that the old Dimensional Analysis of Politics with separate economic and racial axes has collapsed into a single dimension, eliminating the bipartisan crossover that once gave the political system flexibility. On immigration, he acknowledges changing his view: evidence now suggests workers with high school education or less do not compete head-to-head with similarly educated natives, making the wage-distribution effects much smaller than simpler models predicted. On monetary policy, he contends the Federal Reserve should resist raising rates and abandon the 2% inflation target, since allowing somewhat higher inflation would preserve room to cut rates during the next recession—a case he grounds in secular stagnation and falling population growth. On technology and markets, Krugman examines how increasing returns, imperfect competition, and spillovers at tech platforms like Google and Facebook create Market Failure, yet notes the traditional economy still dominates, making abandoned antitrust enforcement a genuine contributor to dysfunction. Throughout, he emphasizes intellectual curiosity over complacency, drawing on Charlie Stross and reflecting on why people remain poorly informed despite information abundance.

Sharpest takeaway

Krugman argues that rising income inequality is the root driver of America's partisan polarization and political dysfunction, and that many economic phenomena—from immigration to trade to monetary policy—are better understood by abandoning oversimplified Econ 101 models in favor of evidence and richer frameworks.

  • Income inequality strongly correlates with partisan polarization and has dragged one party rightward, collapsing the political center
  • Evidence shows immigration's effects on native wages are smaller than simple labor-supply models predicted
  • Persistent low interest rates and secular stagnation mean a 2% inflation target is too low and conventional monetary frameworks are inadequate

The argument · threads20 threads · 40 claims
0.85

Rising income inequality has collapsed American politics into single-dimension partisan tribalism, eliminating deliberative center.

5 pointscentrality 5/5
  • There is a strong empirical correlation between income inequality and partisan polarization: the growing weight of the right tail of the income distribution has dragged one party in its direction, opening a large gap so the political center could not hold, eliminating prior economic-issue overlap between the parties.

    “there’s a pretty strong correlation between income inequality and partisan polarization. Political scientists who measure polarization have found that’s a really strong correlation.”

  • The American political system was designed around representatives who deliberate and exercise independent judgment, but in an extreme partisan environment only party affiliation matters, so deliberation has effectively ended—e.g., when the president and Senate majority leader are of different parties, presidential appointments don't even get hearings.

    “All that matters on most things is whether there’s an R or a D”

  • There was once a time when voters cared about a senate candidate seeming senatorial—having gravity and strength of character—but increasing partisanship and tribalism have made character matter less and less, so that in many places a candidate's party label outweighs anything they say.

    “Over time, with increasing partisanship and tribalism, that matters less and less.”

  • American politics used to have two dimensions—race and left/right economics—and the existence of populist Southern senators who were economically left but racist provided part of the political center; those two dimensions have now collapsed into a single dimension, eliminating that source of bipartisan crossover.

    “We used to have two dimensions in American politics, one of which was race — this is what the political science modelers tell us — one of which was race, the other of which was essentially left/right on economics. Those have collapsed into a single dimension now.”

  • The role of money in politics has created an environment where many political figures spend their entire lives inside a partisan ecosystem, where any display of independent conscience is career-destroying, so the selection process produces cynical or bad people in office.

    “a lot of people in political life now have spent their entire lives within a partisan ecosystem, never really having to step outside. Any independence in judgment, any demonstration of an independent conscience is actually career destroying.”

0.67

Tech platforms violate market efficiency through network effects, imperfect competition, and misaligned incentives.

2 pointscentrality 5/5
  • Tech companies, especially networked platforms, violate the principles that make markets efficient because they combine increasing returns, imperfect competition, and spillovers—so there is no reason to think Facebook, Twitter, or Google produce socially optimal outcomes, even though the specific market failure is hard to isolate.

    “Everything about tech, certainly everything about networks, is a violation of the principles that say that a market should be efficient.”

  • Zero-price platforms like Google and Facebook are not optimizing for users' welfare but for selling advertiser access, so the indirect, ad-driven business model gives even stronger reasons to expect non-optimal outcomes—exemplified by YouTube algorithms that push users toward more extreme content because that enhances the platform's ability to market its position.

    “Everything about tech, certainly everything about networks, is a violation of the principles that say that a market should be efficient.”

0.59

Demographic decline and low interest rates constrain monetary policy capacity to fight future recessions.

3 pointscentrality 4/5
  • Falling population growth makes it hard to generate enough investment demand to absorb desired savings, so interest rates stay low even in good times and hit the zero lower bound too often; Japan, where declining working-age population kicked in early, was the first dress rehearsal for the difficulties the advanced world has faced since 2008.

    “with falling population growth, we have a hard time generating enough investment demand to use the desired savings, so that interest rates tend to be low even when times are good and tend to hit the zero-bound way too often”

  • The next recession will be a 'smorgasbord recession'—a mix of many smaller weaknesses (corporate debt, China, Turkey-Argentina) reflecting the Minsky cycle in which complacency after long calm leads people to take on risky leverage; what makes it dangerous is that the Fed has only about 200 basis points to cut versus the roughly 600 basis points it typically cuts, leaving monetary policy ineffective from a starting position of low rates.

    “I’ve been arguing that the next recession will be a smorgasbord recession. It will be a mix of all of these.”

  • The Fed should not raise rates now because allowing somewhat higher inflation would let real interest rates fall lower during the next slump, giving more room to fight it; central bankers resist this for social rather than intellectual reasons, being wedded to the sacrosanct 2 percent target, but the economics favoring 4 percent is strong while the case that 2 percent was enough has been proven wrong over the last decade.

    “don’t raise rates because, actually, a somewhat higher inflation rate would be extremely useful... you’ll have room to get real interest rates lower during the next slump”

0.59

Structural economic problems require deepening inequality reduction, not merely procedural or institutional tinkering.

3 pointscentrality 4/5
  • The deepest fix for American political dysfunction is to recreate the kind of society that supported a more reasonable political process, which means reducing income inequality, rather than mere procedural reforms.

    “what we really need to do, if we can, is to recreate the kind of society that supported a more reasonable political process, which means reducing income inequality.”

  • Reparations for slavery would be just given that the country was built in large part on an incredible injustice, but they would be a one-time thing that would not change the structure of inequality, and Krugman finds it hard to believe they will happen.

    “our country was built, in large part, on an incredible injustice. But that’s not going to change the structure. That would be a one-time thing.”

  • The progressive movement arose during a period of great income inequality and corporate dominance of politics, and because we are again in such an era, critiques of conventional wisdom naturally sound the same as those earlier progressive historians.

    “the progressive movement arose during a period of great income inequality and corporate dominance of a lot of politics. And here we are, in an era of great economic inequality and corporate dominance of politics.”

0.53

Reduced antitrust enforcement and failure to address market concentration contribute to economic dysfunction and inequality.

2 pointscentrality 4/5
  • Most of the economy is still traditional, non-zero-marginal-cost business, so the fact that the US largely stopped enforcing old-fashioned antitrust is itself a real problem that contributes to economic dysfunction and rising inequality, even if the magnitude is hard to pin down.

    “the fact that we largely stopped enforcing good old-fashioned antitrust is actually a problem. It is part of the problem with our economy and part of the problem with rising inequality.”

  • The claim that cutting top marginal tax rates produces huge positive growth effects persists despite zero evidence because it benefits the wealthy, and it takes relatively little money relative to society's total wealth to build a whole structure of institutions that reward people who will endorse supply-side economics.

    “why do people continue to claim that cutting top marginal tax rates has huge positive effects on growth? A proposition for which there is zero evidence.”

0.44

Technological change disrupts long-standing economic patterns like Baumol's cost disease and education's wage premium.

5 pointscentrality 2/5
  • Although college degrees still carry a significant wage premium, in the longer run AI and machine learning can do much of the high-skill cognitive work we recently thought required years of schooling, while robot plumbers remain far off because the physical world is full of unpredictable, hard-to-model features—so it is a live possibility that education reverts to being a status symbol for the elite rather than an economic necessity.

    “a lot of the things we think of as being high skill, requiring lots of education, are actually the kinds of things that computers can, in the end, do.”

  • Contrary to predictions that the internet would make distance irrelevant and disperse activity to cheap-land low-traffic areas, agglomeration economies have become more important, so economic activity migrates toward already-rich areas—an outcome that wasn't clear ex ante and that no one had enough insight to predict.

    “It really looks as if agglomeration economies have become more important. We really do see a migration of economic activity to the areas that are already rich”

  • In the Krugman-Venables economic geography model, falling transport costs produce U-shaped behavior: starting from high transport costs with little globalization, reducing costs first differentiates the world into a rich advanced region and a poor periphery, and reducing them further lets the lower-cost periphery make a comeback at the center's expense—a nice but probably too-simple story for what actually happened.

    “Reduce them some, and the world differentiates into an advanced region and a peripheral region. Then reduce them further, and the peripheral region, with its lower costs, makes a comeback.”

  • Despite science fiction writers imagining something like Bitcoin enabling long-range interstellar intellectual-property trade, Krugman doubts something like Bitcoin is even going to work here on Earth, and real interstellar IP trade would more likely resemble government-to-government exchanges.

    “some science fiction writers are imagining that something like Bitcoin would make it possible to do these long-range . . . I don’t think something like Bitcoin is even going to work here.”

  • Baumol's cost disease is true of some things but the old version—that service costs always rise relative to goods—isn't clearly correct anymore, since much of the 1995-2005 productivity surge was in services and machine learning has already made formerly cost-disease-afflicted service sectors startlingly cheap, so cost disease afflicts different sectors as technology changes.

    “The old Baumol hypothesis was very much, service costs were always going to rise relative to goods production. But that’s not at all clear more recently”

0.40

Immigration involves genuine trade-offs between open borders and social safety nets, requiring awkward compromise.

2 pointscentrality 3/5
  • Anyone who is not at least somewhat conflicted about immigration has something wrong with them, because a strong social safety net conflicts with completely open borders—the perception that lots of people are coming to take benefits is a real problem—while shutting everyone out is terrible for global welfare since letting people move to where they're more productive is one of the best ways to improve lives, so an awkward compromise of restricted-but-open immigration is the right place to be.

    “If you want to have a strong social safety net, which I do and most people do if actually pressed on it, then completely open borders is going to get in the way.”

  • Krugman changed his view: 20 years ago a simple labor-force model implied low-education immigration would widen inequality, but he has been mostly convinced by evidence that immigrants with a high-school education or less do not compete head-to-head with similarly-educated native-born workers, so the income-distribution effects of immigration are much smaller than previously thought.

    “If you’d asked me about it 20 years ago, I basically went with a simple model of the labor force being workers with different levels of education. Immigration of low-education workers would widen income inequality.”

0.34

Voter information scarcity and polarized media environment prevent public recognition of policy substantive differences.

1 pointcentrality 3/5
  • Ordinary people with real jobs are understandably more poorly informed than professional commentators appreciate, and the news environment is information-scarce—e.g., a three-month analysis of 2004 network news on healthcare found not one story about the actual content of Bush's versus Kerry's healthcare programs—so we shouldn't expect voters to recognize bad ideas.

    “people who aren’t paid to think about abstract policy issues are, understandably really, more poorly informed that those of us who push around words and symbols for a living can easily appreciate.”

0.32

American infrastructure has become inexplicably expensive due to cross-cutting interests and opposing bureaucracies.

3 pointscentrality 2/5
  • America makes it far too hard to build housing where people want to live, pricing out many who should be part of cities; while some zoning restrictions are justified, they have little to do with why we don't build, and on this issue many liberals are wrong and the de facto pro-building policies of conservative places are right.

    “We make it way too hard to build housing for people who want to come here, and the result is to price out a lot of people who really should be part of this experience.”

  • New York City is dangerously dependent on public transportation that is not actually under the city's control and has been massively neglected; purely local control of the subway would internalize the gains and make complaints felt sooner, though the subway has wastefully spent on prestige projects rather than improving the daily ride.

    “it is crazy that you have a city that is so dependent on public transportation, and yet the public transportation is not actually under the city’s control and has clearly been massively neglected.”

  • American infrastructure has become inexplicably expensive—a mile of anything costs at least a billion dollars—and the cause is not mainly labor unions but some combination of cross-cutting interests and opposing bureaucracies; this is an everywhere-in-America problem, and the US that was once the great builder of the world can no longer build cheaply.

    “It’s not mostly the labor unions. It does seem to be something about some combination of opposing bureaucracies.”

0.26

Global trade surged then plateaued; regional supply chains may prove more durable than complex international logistics.

2 pointscentrality 2/5
  • A good guess is that regional trading agreements will prove more robust in the end than broader arrangements, because business has so much of a stake in maintaining them, though the outcome of current trade conflicts is genuinely unpredictable.

    “A good guess would be that the regional trading agreements are more robust in the end, just because business has so much of a stake in maintaining them. But who knows?”

  • Global trade soared from about 1990 to 2010 and then leveled off, appearing to be a one-time surge driven by emerging-market trade liberalization plus reduced transportation and transaction costs; there was also some overreach as businesses built overly complex international logistics chains vulnerable to disruption, so some retrenchment toward regional supply chains would have occurred anyway.

    “global trade really soared from about 1990 to 2010 and then sort of leveled off. It does look as if it was a one-time thing”

0.18

Economic productivity requires curiosity and serious engagement when standard explanations seem incomplete or contradictory.

3 pointscentrality 1/5
  • The key to intellectual productivity is curiosity rather than intellectual complacency: be interested in many things, but when something really grabs your attention, work at it seriously—if the standard explanations of a phenomenon don't quite hold together for you, worry at that itch and try to figure it out rather than letting it go.

    “be interested in a lot of things, but when you find something that really grabs your attention, work at it seriously. Figure stuff out.”

  • Krugman advises today's budding research economists not to do what he did—heavily theory-oriented simple models—but to exploit the golden age of empirical work: find data not heavily worked over, look for anomalies or natural experiments that shed light on disputed issues, and step beyond traditional economic methodology, since methods like ethnography could yield insights the standard paradigm misses.

    “don’t do what I did, which was very heavily theory-oriented, simple models... now we are in this golden age of empirical work.”

  • The maxim 'do what you love and the money will follow' is garbage, but the converse holds: once the money is already there and you're paid a nice salary, you can do what you love—reaching the 'end of ambition' with no more notches to add can produce a post-achievement depression resolved by finding things you're genuinely interested in.

    “You know that whole business about “Do what you love and the money will follow”? That, of course, is garbage. But if the money is already there, you’re being paid a nice salary, then you can do what you love.”

0.06

Constitutional reform like eighteen-year Supreme Court terms could reduce politicization while maintaining democratic accountability.

1 pointcentrality 1/5
  • Replacing lifetime Supreme Court appointments with 18-year terms would insulate justices from short-term political pressure while reducing the extreme drama of vacancies and giving every president a chance to nominate two justices.

    “instead of lifetime appointments to the Supreme Court, we should do 18 years, which is probably enough to insulate from short-term political pressures but ends this . . .”

0.06

Confederate monument removal matters symbolically because they were erected during Jim Crow to enforce racial subordination.

1 pointcentrality 1/5
  • Removing Confederate monuments matters as symbolic action because the monuments were not erected in the immediate aftermath of the Civil War to honor the dead, but during the Jim Crow era to remind people of the importance of keeping black people in their place, so they endorse violent action on behalf of slavery.

    “They weren’t actually put in in the immediate aftermath of the Civil War in memory to the heroes. They were put in during the Jim Crow era to remind people of the importance of keeping those people in their place.”

0.06

Universal basic income's automaticity advantage is outweighed by the expense of making it generous enough to help.

1 pointcentrality 1/5
  • Universal basic income's chief virtue is its automaticity—like Social Security and Medicare, it's just there with no means-testing—but it faces a dilemma: making it generous enough to help those who really need it is very expensive, while keeping it cheap makes it worse than the existing means-tested programs, which are more generous to those at the bottom than a skinny UBI would be.

    “The pro is that its automaticity is a big thing. It matters a lot that Social Security and Medicare are just there. No one asks, “Do you need it? Do you deserve it?””

0.06

Economic analysis succeeds on simple motivations like getting and spending but fails at modeling complex human behavior.

1 pointcentrality 1/5
  • The reason economics can get as far as it does (the dream of psychohistory) is that it concerns the simplest, crudest aspects of human motivation—getting and spending—where we at least have a good idea of what people are trying to do; beyond that, modeling human behavior gets dramatically harder.

    “economics is about the simplest, crudest aspects of human motivation and behavior — getting and spending, basically. Even there we don’t do too great, but at least we have a pretty good idea of what it is people are trying to do.”

0.06

Moderate inflation around three to four percent is acceptable and public resistance stems from central banker preferences.

1 pointcentrality 1/5
  • We shouldn't invent a public objection to moderate inflation without evidence: it is central bankers, not the public, who insist 3-4 percent is unacceptable, and the second half of the 1980s was a 4 percent inflation economy in which people did not complain bitterly—it felt OK—while the service-economy/fewer-unions structure may make wages more flexible, not less.

    “I am actually old enough to remember a 4 percent inflation economy, which is the second half of the 1980s, and I don’t remember people complaining bitterly about inflation during those years. It felt OK.”

0.06

Returns on capital historically fall below growth rates, invalidating perpetual-wealth fantasies like cryonics or time travel.

1 pointcentrality 1/5
  • The old story of putting a dollar in a perpetual trust and eventually owning the economy doesn't work because R has historically been less than G on average, so neither cryonics nor a near-light-speed Rip Van Winkle scheme would let you reclaim invested money and come back enormously rich—returns aren't high enough for that to be the binding constraint.

    “in fact, R has been less than G on average, historically, so that doesn’t actually happen.”

0.06

Technology access alone does not produce advanced society; genuine social development requires structural institutional change.

1 pointcentrality 1/5
  • Drawing on Charlie Stross's Merchant Princes novels—where world-walkers with access to our advanced timeline remain a medieval, backward society and merely use the ability to smuggle drugs—the lesson is a development-economics one: having access to advanced technology doesn't make you an advanced society, and genuine social change is very hard to achieve.

    “having access to advanced technology doesn’t make you an advanced society, and how hard it is actually to achieve social change.”

0.06

Space has limited economic utility for manufacturing or colonization and offers no frontier analogy to terrestrial expansion.

1 pointcentrality 1/5
  • Space is already extremely useful for communications and could become useful for some manufacturing, but it's hard to see that any of these uses require people to be up there, and the frontier analogy fails because space offers no rich land waiting to be exploited—only an extremely hostile environment, as evidenced by our not having colonized Antarctica either.

    “it’s hard to see that any of these things require that there be people up there.”

0.06

Column writing requires engaging hooks and strong conclusions because scarcity of reader attention demands stylistic excellence.

1 pointcentrality 1/5
  • Writing for a general audience requires that, since nobody has to read your column, you give a hook at the beginning and a zinger at the end—style matters precisely because people's time is scarce and you must give them a reason to spend it with you.

    “nobody has to read your column, so you have to have a hook at the beginning, and you have to have a zinger at the end.”