Paul Krugman: America's Cruelty Is a Choice, Not a Law of Technology

Asked what an economic utopia would look like, Paul Krugman declines the word *perfect*. Someone once described his ideal as slightly imaginary Sweden: a high safety net, serious environmental rules, and the smaller defects of well-run countries actually fixed. Total equality is neither possible nor especially desirable, he says, but nobody should be hurting, and nobody should be so wealthy that they no longer live in the same material universe as everyone else.

Competition, in his view, is excellent where it can work. He is old enough to remember a single phone company and limited choice; multiple carriers were a genuine improvement, and the same pattern holds across many industries. It is not one size fits all. Health care is the example he reaches for first, because rivalry there does not deliver what textbooks promise.

The diagnosis is not new. Krugman cites Kenneth Arrow's 1963 paper, which he says still holds up, and walks down the list of conditions a market needs: both sides of a transaction well informed, able to make intelligent decisions, and able to understand what is going on. Health care fails on every dimension, and so does health insurance. The idea that you can comparison-shop for major surgery makes him wonder whether the speaker is living in the same world he is.

Fridman presses on how hard it is to be well informed about anything. There is more information at your fingertips than at any point in history, Krugman replies, but some of it is false and some of it is simply too hard to understand. Buying a car, you can check *Consumer Reports* and get a decent picture. Surgery often arrives without a chance to plan, and medical school takes many years; searching the internet is not a substitute.

Disagreement about the good society, Krugman tells students, comes in two layers. The first is about justice itself. One camp, associated with John Rawls, asks which society you would choose if you did not know who you were going to be. The other focuses on process: a just society is one with no coercion except where it is absolutely necessary. There is, he says, no objective way to pick between those views. He is pretty much a Rawlsian, and he thinks many people are too.

The second layer is about what actually works. Any card-carrying economist will say incentives matter; the live question is how much. How far does a higher tax rate deter work? How far does a stronger safety net make people lazy? Krugman has a strong view that the evidence points considerably to the left of where most American politicians stand, but he grants legitimate room for disagreement on the magnitudes.

He is not a fan of the Gini coefficient as a headline metric. It is popular because it is a single number and usually tracks other measures, yet it has no natural interpretation. A Gini of 0.25 is Denmark and 0.45 is Brazil, but there is no easy mapping of the space in between onto a lived society. He watches median family income, the income of the top one percent, poverty under various definitions, health, life expectancy, and life satisfaction, which he prefers to the squishier word *happiness*.

Polling, he notes, finds Danes fairly satisfied with their lives and Americans less so. Sweden does not, contrary to the joke, win every time; Denmark and Norway have been taking the prizes. None of the measures is perfect. His fallback is almost a pornography test: how do you know a decent society? You kind of know it when you see it.

America has virtues, Krugman says, and also a level of harshness that lets someone with bad luck fall off the edge. That should not happen in a country this rich. "We have somehow managed to produce a crueller society than almost any other wealthy country for no good reason." Every other advanced country has some universal guarantee of adequate health care. The United States is the only place where citizens can fail to get basic care because they cannot afford it. The country has gotten a little better, he adds, but the gap remains a big deal.

The holes in the safety net go beyond insurance. Support for children is remarkably weak; parents of young children get much more help elsewhere and often get nothing in the United States. Long-term care for the elderly is hit-and-miss. The really large failures, as he puts it, are that the country does not take care of children who make the mistake of having the wrong parents, or of people who make the mistake of getting sick. Those are things rich countries should be doing.

Fridman asks whether caring for the unlucky is only a matter of justice, or whether it is also good for the economic health of the society. By and large, Krugman answers, doing the right thing is also the right economics. A world of tax rates that kill all incentives, and a safety net so generous that nobody bothers to work, could be a problem; he does not actually know any society that looks like that. Even in European countries with very generous nets, people work, strive, and innovate.

The other side of the ledger is the damage from scarcity. Children who grow up without adequate health care or nutrition are developmentally challenged and do not live up to their potential as adults. The United States, he argues, pays a price. "We're harsh, we're cruel, and we actually make ourselves poorer by as a society, not just the individuals, by being so harsh and cruel."

Krugman has no interest in the government running steel mills or owning farms. Markets are a pretty effective way of aligning incentives. Farmers do not grow crops because they want to feed people; they grow them because they can make money, and that turns out to be a good way of getting food produced. The invisible hand is important. There is nothing mystical about it. It is a mechanism for organizing economic activity that works well given a bunch of preconditions, which agriculture, manufacturing, and many services satisfy. Health care and education do not.

The mix he likes is roughly three-quarters invisible hand and one-quarter visible hand. Even Adam Smith, he notes, has a section on why it is important to regulate banks. Fridman, who grew up in the Soviet Union, still finds the supermarket almost magical: bananas, piles of fruit, no central planner. Krugman grants the wonder and then quotes an old movie line: sometimes the magic works and sometimes it doesn't, and you have to have some idea of when it doesn't.

A lot of money has been spent, he says, on making Americans hate the government. The reality is that government does some things pretty well. It does health insurance well enough that people hold signs saying not to let the government get its hands on Medicare, a program they love far more than private coverage. Local public high schools are, in his view, the right place to train students; for-profit education is by and large a nightmare of rip-offs and grift. Even the DMV, the punchline of anti-government talk, has given him positive experiences, though he jokes that he may just be going to the right one.

Some of this can be argued from first principles, as with the logic of why markets mishandle health care. Some of it is just experience. Countries that nationalized steel industries did not do well. The United States tried privatizing education, and that did not go well either. You find out what works. On tech, he wants more regulation in social media, where competitive forces and self-policing are both failing, but he is a skeptic of the claim that we are living through greater technological change than ever before. Since the Industrial Revolution, work has been repeatedly remade; it is not clear that the present shock is bigger or harder to cope with.

Fridman puts the popular claim on the table: machine learning, data, and software are about to wipe out work on a transformational scale. Krugman does not deny local hits. Demand for translators has fallen because machine translation is imperfect but not bad. Isolated jobs go away all the time. What he does not see is an economy-wide counterpart: the ability to produce stuff with many fewer workers than before. Overall productivity growth has been slow, slower than in some past periods.

New York once had thousands of longshoremen loading and unloading ships. They are almost all gone. Giant cranes now move containers in Elizabeth, New Jersey. That is not robots and it does not sound high-tech, but it pretty much destroyed an occupation. It was not fun for the longshoremen. Society coped and moved on. Farmers tell the same story at national scale: the United States used to be a nation of farmers; there are now very few, not because people stopped eating, but because farming became so efficient. "The idea that there's something qualitatively different about what's happening now so far isn't true."

He leaves a door open for a singularity, a moment when machines get smarter than we are. That future is not visible in anything we are seeing now. The test he wants people to use is productivity: real GDP, a construct that adds apples and oranges using market prices adjusted for inflation, divided by hours worked. If technological progress were truly rapid, that ratio would rise at a rapid clip. It did, at about two percent a year, for the generation after World War II. It dropped, recovered for a decade from the mid-1990s to the mid-2000s, and then dropped again. "It's not impressive right now."

Krugman had recently sat through a conference on lagging wages and soaring inequality in which participants simply assumed that robots were a big part of the problem. It was not even presented as a hypothesis; it was treated as what everyone knows. He offers three reasons the story spreads. First, the technologies advancing now are visible to the chattering class. When containerization wiped out longshoremen, not a lot of college professors were close friends with them. Today's shocks hit writers, translators, and customer-service workers, people intellectuals actually see.

Second, the country had just come through a severe financial crisis and a long stretch of high unemployment that had finally come down. That unemployment, he says, was about macroeconomics: a failure of demand. Macroeconomics is deeply unintuitive. People have a hard time believing that something as trivial as "people just aren't spending enough" can produce the mass misery of the 1930s, or the less severe but still serious misery after 2008. There is always a tendency to reach for something big, such as technological change that means we no longer need workers. The 1930s had the same reflex.

Third, a large part of what has happened to wages is political: the collapse of the union movement and policies that squeezed workers' bargaining power. A lot of influential people do not want to hear that story. They want it to be an inevitable force of nature, technology making middle-class wages impossible. Technological determinism has a receptive audience because it erases responsibility. Economy and politics, he adds, each affect the other in at least two ways, a rule he was taught in economics.

Look at unions. The American story is that the world changed and unions no longer have a role. Two-thirds of workers in Denmark are unionized, and Denmark has the same technology and faces the same global economy. The difference is political choice. That is the message of the welfare-state course he teaches at CUNY: all advanced countries have roughly equal competence and make very different choices. America's retirement system is one of the better ones, he notes; the country is not always wrong. The shape of the economy is still, to a huge degree, a political artifact. He uses the old, demonized term *welfare state* a little to thumb his nose, and reminds students that Ronald Reagan warned in 1961 that Medicare would destroy American freedom. It sort of didn't happen.

Every safety-net design, Krugman says, trades off universality against targeting. Universality is clean, but it sends money to people who do not necessarily need it. Targeting stretches limited resources toward the sharpest problems, and it comes with incentive problems plus political and even psychological costs. The great thing about Social Security and Medicare is that they ask no questions. You do not have to prove you need them. Krugman is on Medicare, run through his *New York Times* health insurance; he never filed an application with the Medicare office. It happened when he turned 65. That is good for dignity, and it is good for political support, because everybody gets Medicare.

A guaranteed income comfortable enough to live on is, by contrast, a lot of money. Unemployment insurance already exists for spells of job loss; he thinks American benefits are too short-lived and too stingy, and would rather make them more comprehensive. Universal basic income either sets the bar too low to live on or becomes an enormously expensive program. At this point he would rather build on what exists: food stamps, the earned income tax credit, and a lot more family-support policy. Those tools, he argues, can do more to diminish misery.

UBI, in his telling, travels hand in hand with the belief that robots are going to take all the jobs. If that were really happening, he might reconsider. He does not see it happening. The popularity of a check for everyone is, on this view, a cousin of the productivity story the data have not confirmed.

Public discourse, Krugman says, is more fragmented than ever. The world you inhabit if Fox News is your principal source is very different from the one you get from the *New York Times*. Mainstream political and policy reporting is not great, but it is better than when he first became a pundit, when lots of things never got covered and everything that did was both-sides theater. His line from the 2000 campaign still stings: if a candidate said the earth was flat, the headline would be that views differ on the shape of the planet. That habit is less dominant now, and more people writing on policy actually understand it. How informed the typical voter is remains unclear. He hoped the Democratic debate stage would shrink from something like twenty-seven people; the debates were reasonably substantive, if still soaked in theater criticism.

Fridman asks the awkward open-mindedness question: people on the left and right often view the others as complete idiots; is it possible the other side is correct, and is the food fight productive? Krugman will not paper over the problem. A lot of stuff out there is simply false, and it tracks political allegiance. He is happy to listen respectfully when a view is worthy of respect. A lot of stuff is not. In economics he knows he has been wrong plenty of times, and he still draws a line between views inside the realm of interesting discussion and crank doctrines, or ideas disseminated because someone is paid to disseminate them.

There are plenty of serious center-right economists he is happy to talk to. None of them, he says, had any role in the Trump administration, and Republicans in Congress by and large only wanted to listen to cranks. Pretending otherwise would be dishonest with his readers. There is no way to reach people who think reading Ayn Rand novels is how you learn monetary economics. Radical ideas are fine if they have a basis. If a claim has been refuted again and again and people keep repeating it, it is a zombie idea, and there comes a point when it is not worth faking respect. Flat-earth talk growing in popularity does not make it a debate he owes his audience.

Compared with physics, economics studies a complex system in which you mostly do not get to run experiments. It will never be quantum mechanics. That said, natural experiments arrive, and rival doctrines get tested. In the immediate aftermath of the financial crisis, one macroeconomic tradition going back to Keynes made a few sharp predictions: under those circumstances, printing money would not be inflationary, big budget deficits would not drive interest rates up, and austerity would lead to depressions if tried. Other people predicted the opposite.

The test, Krugman says, was fairly robust. Interest rates stayed low. Inflation stayed low. Countries that implemented harsh austerity suffered severe downturns. You do not get much clearer than that, and it will not be true on everything. He likes that younger economists are heavily data-based; that is the way to go. Among serious researchers there is also a lot less disagreement than the public imagines. Chicago Booth runs an ideologically diverse panel and regularly poses questions; on most things the consensus is remarkable. The illusion of dispute, he argues, is often fed by political forces.

True unknowns remain. Effective ways to regulate technology industries are one; "we really don't know the answers there." On the minimum wage, he sees pretty overwhelming evidence that a modest increase from current levels would not have a noticeable adverse effect on jobs. Twelve dollars seemed pretty safe given what was known at the time of the conversation; fifteen dollars was a place of legitimate disagreement. Where the line is, he truly does not know. You are going out of sample, and there is good reason to believe the relationship is nonlinear: past a point, the world stops behaving the way it has.

Something like 70 to 80 percent of the growth in per-capita income, Krugman says, is the advance of knowledge, not crude accumulation of capital. A lot of that technology is prosaic. He likes to talk about containerization, and earlier the invention of the flat-pack cardboard box, without which Amazon deliveries would not exist. The web matters too, but what would we do without cardboard boxes? The same instinct should cover bridges, tunnels, and an EZ Pass-style improvement that will never get a biopic and still boosts productivity. Why, then, has the country not built another rail tunnel under the Hudson River, for which the need is overwhelmingly obvious?

His theory is that a values dispute about the welfare state contaminates everything else. Politicians believe, probably rightly, in a halo effect: if government is seen succeeding at one kind of thing, the public will think more favorably of it doing other things. Enhanced Social Security benefits are really very different from a Hudson tunnel, yet conservatives tend to oppose any increase in non-military spending, no matter how obviously a good idea, because they fear it is the thin end of the wedge for bigger government. Liberals tend to favor such spending partly as proof that government can do things well, and therefore can turn to broader social goals. A discussion that should have been technocratic, about research and infrastructure, is poisoned by the word *government*.

Fridman asks, half in jest, what political and economic system should run a society started from scratch on Mars. Krugman first doubts humans will actually do that, then plays along. He wants representative democracy, not government by referendum. People do not have time to master every issue; California's ballot measures show how easily votes conflict. Churchill's line still fits: the worst system except for all the others. Starting from scratch, you would not want a Senate in which 60 percent of the population holds half the seats. Parliamentary and presidential systems all work when times are good and all have different modes of breakdown. He is a fan of *The Expanse*, and finds it gratifying that Mars in the show is the Martian Congressional Republic.

International trade looks mysterious next to two neighboring farmers, but Krugman says it is not really that different, only vastly more complex and with many more players. Countries trade for the same reasons individuals do: they are different, and they gain by concentrating on what they do relatively well. Economies of scale matter too. You probably should not try to be both a surgeon and an accountant; countries benefit from specializing for the same reason. A "made in China" iPhone is assembled from components made all over the world. Multi-stage production does not change the fundamentals much.

The dirty little secret of trade conflict, as he puts it, is that conflicts among countries are not that important. Most trade is beneficial to both sides. The large effects land on the distribution of income *within* countries. The growth of U.S. trade with China made both countries richer and was pretty bad for people in the North Carolina furniture industry, whose jobs were displaced by a wave of imports. That is where the political complexity lives. The United States has real problems with China, he says, more about respect for intellectual property than about trade as such, and it is not clear those problems have much to do with the trade war then underway.

The war, in his account, was driven by a fundamentally wrong notion: selling goods to China is good and buying goods from China is bad. That misunderstands the point. Trade in both directions is a good thing; Americans would be poorer without it. Any economic change that makes a country richer also hurts some people, the way a new technology does. He would not undo what has happened. He does wish the United States had a better policy for supporting and compensating the losers from that growth.

Fridman asks whether Krugman, tweeting and writing strong opinions in the *Times* through a wave of radicalization, ever hesitates or feels fear. He feels fear all the time. It is not hard to imagine scenarios in which he personally ends up in the crosshairs. He is, he says, the king of hate mail, and he gets the most amazing correspondence. It affected him when he started. These days he has a thick skin. If a column does not draw a wave of hate mail, he probably wasted the day.

His advice to other economists is mostly not to follow him. Do your research. Not everyone can or should be a public intellectual. He is glad he did not get into that business until his late forties; the decades of greatest intellectual flexibility are better spent on deep questions than on confronting Twitter mobs. When you do write for the public, the rule is to dance like no one is watching. Write what you think is right, make it comprehensible and persuasive, and do not get intimidated because some people will say nasty things. You cannot do the job if you are worried about criticism.