What this covers

Ed Leamer and Russ Roberts examine how technological change has reshaped labor markets and income distribution in America. The conversation centers on a historical divide: industrial-era tools like the forklift equalized workers by making brute strength irrelevant and raising wages broadly, while computers and the internet amplify individual talent, concentrating rewards among the exceptionally skilled. Leamer traces how manufacturing's collapse since the 1960s—declining from roughly 35% of jobs to 8–9%—eliminated the high school diploma as a reliable ticket to middle-class stability, forcing workers into precarious alternatives like part-time service work. The speakers also examine why measured inequality has grown faster than hourly wage gaps alone would suggest: high-skill workers are choosing to work significantly more hours per year while low-skill workers work fewer, a pattern Leamer argues reflects rational responses to capital costs and falling real wages.

The discussion moves through several distinct territories. They distinguish Detroit vs Hollywood labor markets—the egalitarian manufacturing economy versus the winner-take-all intellectual services now emerging in architecture, law, and software development—and explore how Factor Substitution fails in knowledge work: you cannot deploy ten mediocre programmers where one talented one is required and get partial output. They separate Inequality vs Poverty as distinct moral problems, and challenge the framing of higher education as a public good, with Leamer skeptical that subsidizing demand will help and emphasizing that real education requires Teacher as Coach—Experience-Based Education where students discover under guidance rather than passively receive instruction. Throughout, they acknowledge that neither education nor retraining will solve the core problem: if raw talent now determines success in most growing fields, and talent is distributed unevenly, broad prosperity becomes harder to achieve regardless of schooling policy.

Sharpest takeaway

Leamer argues that a new kind of technological innovation—the computer as 'microphone' rather than 'forklift'—is converting Detroit-style egalitarian labor markets into Hollywood-style winner-take-all ones, where raw talent (not just effort or education) determines outcomes, threatening both economic equality and democratic stability unless education and cultural attitudes change.

  • Manufacturing decline since the 1960s destroyed the value of the high school degree as a path to middle-class jobs
  • The computer amplifies talent rather than equalizing it, producing extreme income inequality in intellectual services
  • Measured inequality is partly driven by high-skill workers choosing longer hours, not just higher hourly wages

The argument · threads4 threads · 35 claims
0.85

High-skill creative work lacks substitutability, permanently displacing lower-skill workers unable to transition to available alternatives.

5 pointscentrality 5/5
  • Unlike past creative destruction where displaced cheap labor could be redeployed (e.g. ten workers doing a job more slowly but cheaply), modern high-skill tasks like commercial software coding lack substitutability: you cannot substitute ten less-skilled people for one skilled coder because you get zero usable output, not slower output, so the displaced cannot be redeployed at lower productivity.

    “You can't get 10 people to do the coding, because you get none. It's not that they'll just do it at a slower pace or not as quite as effectively, but it's cheap, so, you don't mind. It's that: There's no substitution there.”

  • Where mechanical innovation once eliminated physical strength as an advantage and made brains newly valuable, we may now have reached the end of that process, where merely having some brains is no longer enough and only a particular kind of creative, innovative talent succeeds, leaving the rest to struggle in a bifurcated economy with winners and losers unlike anything in the 20th century.

    “we've reached sort of the end of that where even having some brains is not enough. You really have to have a particular kind of brain with a particular kind of creative, innovative talent; and the rest of the folks are going to struggle.”

  • The alternatives to manufacturing for displaced low-skill workers—construction, truck driving, taxi driving—require mainly reliability and honesty rather than technical ability, offer limited reward because they use little capital and do not make the worker much more productive, and there are far fewer such jobs in total since much of the team-based assembly line has been automated.

    “There are things like construction, truck driving, taxi driving, the things that require limited technical ability and mainly require reliability, honesty, intangible skills... but they're limited when they're being applied to skills that don't use a lot of capital and doesn't make the worker a lot more productive.”

  • Assembly-line team production amplifies the pressure to work hard for two reasons: the expensive equipment must not sit idle, and a worker who fails to match the pace of others effectively idles the whole team since it is a joint product, whereas solo work with cheap tools (an artisan or a writer at home) imposes neither penalty because the worker neither helps nor hinders anyone else.

    “people are exhorted to work hard for two reasons. One, they've got an expensive piece of equipment; you don't want to have it sit idle. And secondly is that, if you're not working at a pace that matches the other people, it's essentially sitting idle, because it's a team product.”

  • On the output side, manufacturing has been largely stagnant over the last 30-40 years except for the computing sector, so the apparent health of manufacturing output is driven overwhelmingly by computing.

    “We recently had Susan Houseman on the program where she suggested that on the output side, it's actually pretty stagnant except for computing: that a lot of the apparent health of manufacturing on the output side is driven overwhelmingly by the computing sector.”

0.59

Measured inequality between high and low-skill workers is significantly inflated by differences in hours worked and earners per household.

3 pointscentrality 4/5
  • Low-skill workers now work fewer hours per year while higher-skill workers work more per year, so a significant portion of measured inequality between 1980 and 2016 is driven by differences in effort (hours worked) rather than hourly wages alone, with high-skill workers working harder and low-skill workers working less.

    “low-skill workers now work less per year; higher-skill workers work more per year; and that's expanded the income differences across skill over time, just because of the effort involved.”

  • Household income inequality is exaggerated by differences in the number of earners per household: the top 20% averaged about $198,000 in 2018 versus $11,000 for the bottom 20% (a 17-fold gap), but the top quintile averages more than two workers per household while the bottom averages 0.5, so on a per-earner basis the gap shrinks to roughly threefold rather than 17-fold.

    “the top 20%, average income before taxes, is about 17 times... the average salary was $198,000. In the bottom 20%, it's $11,000.”

  • The ratio of average annual earnings of workers with advanced degrees to those with only a high school degree grew by 60% from 1980 to 2016, rising from 1.8 times higher in 1980 to 2.8 times higher in 2016; but if only full-time workers are considered, the gap grew only 34%, meaning the hours-worked component roughly doubled the measured gap.

    “In 1980, average annual earnings of those with advanced degrees was 1.8 times higher, and by 2016, it was 2.8 times higher. So, it got a lot bigger. That's the 60% increase.”

0.59

Absolute living standards matter more than income ratios, and most lower-income people are materially better off than in 1980 despite inequality.

3 pointscentrality 4/5
  • Most people in the bottom half of the income distribution today would not want to return to 1980 despite its more equal income distribution, because their absolute standard of living—access to food, health, housing, and things they enjoy—has improved enough that they are better off even though they remain far behind high-skill workers.

    “when I look at the data, I don't think most of the people in the bottom half of the income distribution today would want to go back to 1980.”

  • Inequality and poverty are two different issues often conflated: one can care about the ratio of incomes between top and bottom regardless of absolute levels (an emotional issue), while separately caring about absolute living standards; a world of 99 poor barbers and one rich programmer is very different from a world where all live well but the programmer lives better.

    “there are two issues that often get conflated: which is inequality and poverty.”

  • Comparing 2016 to 1980, workers with a college degree or greater do dramatically better than workers of the same educational level did in 1980, while workers without those degrees do worse in absolute terms.

    “the workers who have a college degree or greater, do dramatically better than they did in 1980 of the same educational level, but the workers who don't have those degrees do worse in absolute terms.”

0.38

Education's real value lies in transfer and application of learning, while economists wrongly oversell its social spillover benefits.

4 pointscentrality 2/5
  • The real test of education is the ability to bring a lesson learned in one context and apply it in a completely different one—like a student who, after a summer of research, could see her own intellectual life experience reflected in a father and son fishing at the pier—a capacity that traditional fact-based testing cannot capture or grade.

    “the ability to bring the lesson that you got from over here, to bring it over there, is what real education is.”

  • Economists repeatedly argue education is a great public good deserving subsidy because its benefits supposedly spill over to society beyond the students themselves, but this is a mistake and economists have a conflict of interest because they are direct beneficiaries of the resulting subsidies.

    “economists over and over again, talk about what a great public good education is. As if it deserves a subsidy... I think that's a mistake. And we've been the beneficiaries. We have a conflict of interest in that.”

  • Interactive teaching where students who disagree on a question are told to argue and convince each other often moves the class toward the correct answer, because real education—like graduate problem sets among bright peers—happens through open-ended argument over subtle questions rather than passive lecture, with the teacher then adding probing questions.

    “Find someone next to you or nearby who disagrees with you and try to convince them that you're right and they're wrong... then you say, 'Okay, let's do the poll again'... And often the answer has moved toward the correct answer.”

  • One should not walk past a homeless person in silence but at minimum greet them and recognize their humanity, ideally giving money and engaging them in conversation about their lives, because human contact and respect matter beyond the material gift.

    “the first thing is, don't walk by that homeless person and say nothing. At a minimum, smile and greet them, realize that they are a human being going through perhaps an unbearable struggle.”