What this covers

Angus Deaton of Princeton, speaking with EconTalk host Russ Roberts, traces the historical improvements in human health and living standards back through centuries of economic and social development. The conversation moves between broad patterns—the rise in life expectancy and material prosperity across rich and poor nations alike—and the measurement challenges that complicate any comparison of wellbeing across eras. Deaton acknowledges that living standards have risen dramatically, even as he notes that assessing the magnitude requires careful reasoning about what counts as progress when the goods themselves have transformed so radically.

The discussion spans distinct explanations for how health and wealth improved. Early gains came from germ theory, sanitation projects, and vaccination rather than from money alone. The post-war population surge—children surviving where they would have died—marks a one-time transition, not a permanent explosion. State capacity emerges as crucial: poor countries lack not funds but the institutional machinery to deliver services and regulate private sectors effectively. This reframes the foreign-aid problem. Deaton argues that direct transfers to poor-country governments can backfire by severing the social contract that makes states function—when revenue flows from donors rather than citizens, governments lose incentive to provide services in return. The conversation also addresses assortative mating and wage dynamics, the gap between measured and actual compensation in health-care-heavy employment, and pushback against the lump fallacy in population debates. Throughout, Deaton distinguishes between where economists have solid ground and where competing accounts remain plausible.

Sharpest takeaway

Deaton argues that the dramatic improvements in human health and wealth—the 'Great Escape'—were driven by knowledge, state capacity, and institutions rather than money alone, which is why foreign aid that bypasses those institutions largely fails to help the poor.

  • Health gains in poor countries are limited by lack of state capacity, not lack of money
  • Material prosperity has risen enormously over centuries despite measurement difficulties
  • Aid given directly to poor-country governments undermines the social contract that makes states work

The argument · threads8 threads · 38 claims
0.80

Material living standards and health have improved dramatically across generations despite measurement challenges.

5 pointscentrality 5/5
  • Comparing living standards across centuries is nearly impossible because goods change so radically (comparing a Toyota to a horse, or modern electronics to nothing), but regardless of method the direction is unambiguous: each generation is dramatically better off than the one before, so over the long run there has been tremendous material progress.

    “it's clear, however you do this, that there's no question about the direction of change over the very, very long period. We are way better off than our grandparents were, and our grandparents were way better off than their grandparents were.”

  • On average the world is doing a lot better than it ever has on health, which is terrific news, though many people have been left behind.

    “The world is doing a lot better than it ever has in the past, at least on average. And that's terrific news. Not so good news is that a lot of people got left behind”

  • A white, middle-class, well-heeled girl born in the United States today has roughly a 50-50 chance of reaching 100 years old, contingent on continued medical progress over the coming decades.

    “That's a guess but it's not an unreasonable guess... Curtis Simon is talking about white middle class, well-heeled girl born in the United States today.”

  • The world's highest female life expectancy has risen at a constant rate since 1840—roughly one year gained every four years—and repeated past predictions of an imminent ceiling have been wrong, suggesting there may be no natural limit to longevity.

    “that chart often comes with various projections that people made in the past, saying, you know, 10 years from now we'll be here; but the line kept on going up.”

  • Statisticians fail to fully capture the value of new and improved goods—free internet entertainment, Google, iPads, cars that no longer rust out after a decade—so measured growth likely understates true improvements in living standards, though it is unclear how large the correction would be.

    “there's a good case that we're not really adequately taking into account many of the new goods that come along and make our lives much better. Things like iPads.”

0.75

State capacity to enforce regulation and organize services, not funding levels, is the binding constraint on health outcomes.

4 pointscentrality 5/5
  • A regulatory ban is only meaningful where a state can enforce it—if the FDA bans trans fats they disappear, but if the government of India banned them it would make no difference because there is no one to ensure compliance, illustrating that India's health care is mostly an unregulated private sector staffed by people with no medical qualifications, plagued by clinic absenteeism.

    “If the government of India banned trans fats it really wouldn't make any difference because there's no one there to actually make sure that it happens. So, that's the sort of thing I mean by state capacity.”

  • Development assistance fails when treated as an engineering problem—counting doses of vaccine needed without regard to how it works in practice—because development is fundamentally a political problem; like McNamara's whiz kids who solved technical problems at Ford but failed in Vietnam and at the World Bank, technical solutions don't transfer to political realities.

    “Development is a political problem. It's not a financial or an engineering problem. And I think that's the big mistake that gets made there.”

  • Giving large sums directly to poor-country governments is harmful because when all government revenue comes from external donors, the social contract breaks: there is no need to tax citizens and provide services in return, so governments have no incentive to deliver health care or education and effectively use their populations as hostages to extract money from rich countries.

    “People need some sort of contract between the government and those who are governed, in which one set of people pay taxes and the government gives people back something in exchange”

  • The crucial missing link for health care in poor countries is not money but state capacity—a government capable of organizing services and regulating the private sector—because putting more money into the existing systems does not make things better; India's infant mortality is now lower than Edinburgh's was in 1945 despite India being far poorer.

    “I think the crucial matter for health care in poor countries is really the lack of a state, of a government that's capable of organizing these things. If you like, it's state capacity that seems to be the crucial missing link, not just money.”

0.71

Health improvements came from germ theory, sanitation, vaccination, and nutrition, but malnutrition reflects multiple causes beyond food shortage.

6 pointscentrality 4/5
  • The mid-20th-century population explosion was a one-time event caused by antibiotics and germ theory reaching the developing world, so children who would have died survived and life expectancy in some countries rose more than a decade per decade; this was temporary because parents eventually realized they didn't need as many births to have as many surviving children and cut fertility.

    “there was an enormous population explosion which was clearly a one-time event because antibiotics, the germ-theory of disease, had come to what was then called the Third World. That meant a lot of children who otherwise would have died, survived.”

  • Roughly a third of the historical improvement in health may be attributable to better nutrition and material wellbeing, though most of that effect is loaded early in the century when malnutrition was widespread, and early-life undernutrition may predispose people to bad health outcomes decades later.

    “I think maybe a third might not be a bad guess. Though most of it would be loaded up early in the century when there was [?] malnutrition around.”

  • Malnutrition should be understood as the body failing to grow properly—being too short or too skinny—which can result from too little food but also from disease or backbreaking manual labor consuming calories, so the policy focus on simply delivering food has been excessive relative to sanitation, disease control, and reducing physical labor.

    “food intake is not the only part of what we like to call malnutrition, which is how much your body actually absorbs and is capable of growing and getting bigger. And that distinction is extremely important”

  • Height differences across populations may be substantially driven by disease environment rather than food intake—widespread open defecation in India spreads bacteria that, through disease burden, helps explain why Indians are shorter on average, a problem fixable by sanitation or rising incomes enabling indoor plumbing.

    “the argument really is that this is an important part. And also, you have to be a little bit careful about using the word 'nutrition,' because it's too easy to equate nutrition with eating food.”

  • Height is used as a health proxy because it is easy to measure and correlated with things we care about; in populations it may be a marker for cognitive development, since children who don't get enough to eat or who live amid open defecation may suffer long-term effects on brain function, not just stature.

    “it may be a marker in populations at least for cognitive development. So that if kids don't get enough to eat or they are running around where people are openly defecating, this may have long term effects, not just on their height but on their brain function.”

  • Much of the health improvement in the early 20th century came from cleaning up water supplies, sanitation projects, and vaccinations, all enabled by a revolution in understanding germ theory of disease—practices like changing hotel sheets between guests that are now taken for granted did not exist in 1910.

    “A lot of that in the early part of the 20th century was cleaning up the water supplies--sanitation projects--extending vaccinations; all these things that we sort of take for granted now were not around in 1910.”

0.56

Aid success is real for specific interventions but often undermines local institutions and serves donor interests over recipient welfare.

3 pointscentrality 4/5
  • The clearest aid success is that many HIV-positive people alive on antiretroviral therapy today would otherwise be dead, largely funded by PEPFAR and the global fund, and that must count for a great deal; but the undermining of local institutions persists because trying to design and run another country's health care system from outside is neither sustainable nor desirable.

    “there are many people alive on antiretroviral therapy today who would otherwise be dead, because they are HIV-positive. And those lives have been saved, and much of that has come from PEPFAR”

  • Much of the aid enterprise is not actually run for the benefit of poor people but for the donors—evidenced by rich countries simultaneously maintaining cotton subsidies that bankrupt African farmers and selling arms to poor countries—so donors should stop pretending, while genuinely helpful actions like NIH research into diseases of the poor are not usually classified as aid.

    “we have cotton subsidies in the United States which are bankrupting farmers in Africa. And we sell arms to poor people around the world. I mean, what do we think we are doing if we are really trying to help them by selling them guns?”

  • Contrary to the view that policies like ending harmful subsidies or NIH research on poor-country diseases are politically impossible, if well-intentioned activists redirected their lobbying energy toward those reforms there is a reasonably good chance they would happen—whereas their current advocacy, with the best intentions, is actively hurting people in poor countries.

    “If those good people who are doing those, screaming and yelling in Washington, and creating a lobby against those things or for those things, I think there's a reasonably good chance that they would happen.”

0.47

Income inequality operates through different mechanisms at different distribution levels and should not be reduced to a single metric.

4 pointscentrality 3/5
  • In the 1950s-60s high-earning men married highly educated women who did not work, which dampened family inequality by pairing high earners with non-earners; over 60 years this completely reversed, so high-earning men and women now marry each other, powering up inequality at the top while single-adult households cluster at the bottom.

    “in the 1950s that dampened down inequality because the rich people were paired off with people who weren't earning so much or weren't earning anything at all, and now you are pairing, you know, a couple of lawyers who are earning $300 grand each, and that's sort of powering up the top.”

  • Inequality should not be thought of as a single number; one must parse the income distribution into different parts—what is happening at the top, the middle, and the bottom—because the picture is complex and a single summary statistic obscures the distinct mechanisms operating at different levels.

    “it is very important we shouldn't think of an inequality in terms of a single number. You just have to parse it out in different parts of the income distribution.”

  • Economists who add the imputed value of government-provided health care (Medicare, Medicaid) back into income show less rise in inequality, but this calculation is misleading: you cannot eat health care for breakfast or feed your kids on it, and if it bankrupts you, valuing a large unwanted quantity of medical care as income overstates wellbeing.

    “you can't eat the health care for breakfast; you can't feed[?] your kids on it. And so it may be worth something but at some point it becomes not worth very much if it's bankrupting you at the same time.”

  • Rising inequality from technical change is fine insofar as social and private rewards are aligned—no one regrets Steve Jobs getting rich for Apple's products—but it is a problem when they are misaligned, as with Wall Street bankers profiting from derivatives at others' expense or health-care players selling overpriced drugs and devices that often do little good.

    “provided the social rewards and the private rewards are reasonably well aligned with one another... I don't think anyone regrets that Steve Jobs got rich by giving us all the stuff that came out of Apple.”

0.38

Slower wage growth masks rising total compensation from tax-favored health insurance, preventing workers from demanding reform.

2 pointscentrality 3/5
  • Rising health care costs are largely paid for through lower wages in workplaces where employers provide tax-favored insurance subsidies, so measured wages fall (or rise less) even as total compensation rises; this is a disaster because workers don't perceive the cost, think health care is free, and so feel no pressure to reform an absurd system.

    “people don't perceive the cost of health care and they don't understand that it's making their wages lower than they otherwise would. And so they think this health care is coming for free, and so that takes a lot of the pressure off to reform it”

  • In India, more expensive doctors tend to deliver better care, though doctors are not very expensive and there is no insurance, so patients pay out of pocket and only get pseudo-insurance via state-covered hospital care when seriously ill.

    “it certainly is true that the more expensive doctors tend to do better. So there is some work that shows that. But you know, doctors are not very expensive in India.”

0.34

Future health gains in wealthy countries will come from expensive cancer treatments, while overpopulation fears rest on the lump fallacy.

4 pointscentrality 2/5
  • Even if one doubts claims of zero progress for the median earner, the real decline in the growth rate matters because fighting over a pie that is not growing, or growing more slowly, makes politics much harder.

    “I think the fact that the cornucopia seems to be drying up a little bit does make politics a lot harder. Because you are fighting over a pie that is not growing, or not growing as fast as it was.”

  • Future longevity gains in rich countries will come among older people, with the war against cancer the great hope—cancer may yield over the next 50 years as cardiovascular disease did over the past 50—but cancer treatments are extremely expensive, raising serious funding problems.

    “the war against cancer may actually be showing some results at last. So that's the great hope for the future, that cancer will yield in the next 50 years in the way that cardiovascular disease has yielded over the past 50 years.”

  • Despite a century of hard work, economists understand economic growth poorly—they know innovation and technical change matter but little beyond that—so sensible people can hold radically different views on why growth slows, and economics can rule out silly explanations but cannot say which competing account is right.

    “I don't think economists are very good at understanding economic growth, in spite of, you know, a century of hard work on the topic.”

  • The standard overpopulation worry rests on the 'lump fallacy'—the belief in a fixed quantity of stuff so that more people means less per person—but people arrive not just with mouths but with hands and brains, so as Julian Simon argued more people can be good on a per capita basis because they bring ideas and help, like dinner guests who also bring food and know how to grow more.

    “the 'lump fallacy'--not just I who call it that--which is that there is a fixed lump of stuff and if you have more people sharing it there is going to be less stuff for each”

0.20

Gender mortality gaps narrowed because women adopted smoking later, and life expectancy measures can mislead about actual lifespans.

4 pointscentrality 1/5
  • The gap between male and female mortality is smaller now than it has been for a very long time, primarily because women took up smoking later and have been slower to quit, so men's lung cancer mortality has been falling while women's has only recently started or, in some European countries, not yet.

    “that's because women took up smoking later, and they've been slower to quit. So the men's mortality rates from lung cancer have been falling for some time now. And for women, that's much later.”

  • Published life expectancy is not how long a child born today will actually live; it is calculated as how long that child would live if they faced today's age-specific mortality rates at each age—making it a conservative underestimate when health is improving and an over-optimistic figure when health is worsening.

    “the way they calculate it is they say, Ok, how long would that child live if they faced today's mortality rates at each age? So, in a world in which health is getting better, it's a very conservative number.”

  • Despite a large increase in obesity, mortality rates keep falling, which may be because obesity's penalties are now smaller—doctors are far better at prescribing anti-hypertensives, and since one route by which obesity killed was high blood pressure, daily medication may substantially reduce obesity's mortality cost.

    “doctors are very much better at prescribing anti-hypertensives than used to be the case, and it may be that one of the routes that obesity killed people was through high blood pressure, and if you are taking daily meds that prevent that from happening, the penalty from obesity may be somewhat less than we used to think”

  • Across countries the Gallup data clearly show people who are better off think their lives are much better—people in Denmark rate their lives terrific while people in Togo rate theirs terrible—so on at least one concept of happiness, wellbeing rises with income, though whether the Danes are happier than their own ancestors remains unresolved because comparison standards shift.

    “if you look at the Gallup data, people in Togo think their lives are terrible; and people in Denmark think their lives are terrific.”