What this covers

Eric Hanushek of Stanford's Hoover Institution joins Russ Roberts on EconTalk to discuss his book Endangering Prosperity, making an economic case that American students' mediocre performance on international tests—ranking slightly below the OECD average on PISA—threatens the nation's future earnings. The conversation centers on a quantified claim: countries with higher average cognitive achievement in math and science grow faster economically over decades. Hanushek argues that if the U.S. raised student achievement to Canada's level over 20 years, the present value of average American paychecks over the next 80 years would increase by roughly 20 percent. The primary lever he identifies is not higher spending—quadrupled real per-pupil funding since 1960 has yielded no measurable gain in average performance—but rather improving human capital by removing the lowest-performing teachers. Replacing just the bottom 5 to 8 percent of the teaching workforce with merely average teachers could, he argues, achieve that Canada-level result.

The discussion moves across several connected claims. Teacher effectiveness is the single most important determinant of student learning, yet standard credentials—years of experience, Master's degrees, certification, professional development—correlate poorly with actual classroom results. Hanushek contends that teaching ability appears largely fixed rather than trainable, and that schools should hire and evaluate on demonstrated performance instead. He contrasts the U.S. approach, which rarely removes ineffective teachers, with high-performing systems internationally that use counseling and reassignment to keep weak teachers out of classrooms. The U.S. advantage in innovation has derived less from schools than from an entire economy rewarding entrepreneurship; meanwhile, countries like Korea produce high test scores but suppress workplace creativity through culture. Hanushek also examines policy levers: accountability systems paired with local autonomy, parental choice, and pay tied to performance. He addresses counterarguments about whether bad teachers can be identified (he argues they obviously can, within any school) and cautions that reform must align incentives across principals and administrators, not teachers alone.

Sharpest takeaway

Hanushek argues that the mediocre cognitive achievement of U.S. students relative to other countries imposes a large, quantifiable drag on future national prosperity, and that the single most effective lever to fix it is improving teacher quality—specifically by removing the worst teachers—rather than spending more money.

  • International test scores in math and science strongly predict national GDP growth over decades.
  • Replacing only the bottom 5-8% of U.S. teachers with average teachers could raise achievement to Canada's level and lift every worker's paycheck ~20%.
  • Decades of quadrupled real per-pupil spending have produced no average achievement gains, indicating misaligned incentives rather than insufficient funds.

The argument · threads7 threads · 31 claims
0.85

Teacher effectiveness is the single most important determinant of student learning outcomes.

6 pointscentrality 5/5
  • Teacher effectiveness is the single most important determinant of student learning: in an urban district study, the best teachers produced a year and a half of learning in one academic year while the worst produced only half a year, meaning the classroom assignment alone can create a full year's difference in learning in a single grade.

    “the best teachers in an urban district, all urban district, were getting a year and a half worth of learning in an academic year. The worst teachers were getting half a year of academic learning in a year.”

  • Standard measures of teacher background—years of experience, holding a Master's degree, full state certification, and professional development—are none of them closely related to actual classroom effectiveness, yet salary structures reward these credentials despite their lack of connection to student learning; there is no substitute for directly measuring student achievement.

    “It includes whether you have a Master's degree or not. It includes whether you are fully certified according to the state standards... None of these things are closely related to effectiveness in the classroom.”

  • Experiments offering sizeable financial incentives to teachers have not substantially increased student learning, suggesting that the vast majority of teachers are already trying as hard as they can to do a good job—they simply differ in innate ability—so the variation in effectiveness is more about fixed skill than motivation.

    “we offer fairly sizeable amounts of money to people if they do a better job. And what we've seen is that student learning hasn't increased that much.”

  • The bottom group of teachers is not merely below average but is actively harming children, and this fact is well recognized within schools, yet officials respond with resignation—shrugging that these teachers shouldn't be there but nothing can be done.

    “They are harming kids. There's a group that's harming kids. And they are well recognized. And when you ask officials in schools or other teachers they sort of shrug their shoulders and say, yeah, they shouldn't be here but what can you do?”

  • Because research cannot identify predictive teacher characteristics in advance, the rational staffing approach is to make a best guess at hiring (using interviews and practice lessons), then evaluate who actually performs well in the classroom, give large rewards to keep the effective ones, and counsel out or reassign the ineffective ones—essentially what a private school principal does and a public school principal typically does not.

    “I'd make my best guess at who would do well in the classroom... I would then evaluate who does well in the classroom and who doesn't, and I would give large rewards to those who did well in the classroom to try to keep them in the classroom. And I would make decisions that some people should be doing other jobs.”

  • Unlike a craft such as woodworking where a course can reliably teach the fundamentals, education schools do not have a good idea of how to create effective teachers, which raises the possibility that teaching ability is largely fixed and schools should hire on demonstrated ability rather than on training credentials.

    “If I were talking about woodworking, I would know how to put together a course that taught people how to do good woodworking... When we look at schools, education schools, we don't see that they have a good idea of what to do in order to create good and effective teachers.”

0.80

Replacing the lowest-performing teachers with average ones would raise U.S. achievement to competitive international levels.

5 pointscentrality 5/5
  • If the U.S. replaced just the bottom 5-8% of its 3.5 million teachers with merely average teachers—a one-time improvement, not annual culling—it could raise national achievement to Canada's level and capture the projected 20% paycheck increase, and these worst teachers are already well-known and consistently identified by principals, teachers, and parents alike.

    “if we could replace the bottom 5%-8% of our teachers with just an average teacher, we could beat Canada. We could get this 20% per year increase in everybody's paycheck for the next 80 years”

  • Using the same proficiency standard from the U.S. National Assessment of Educational Progress applied internationally, the U.S. math proficiency rate is 32% while Singapore's is 63%, indicating other countries demonstrate what is feasible far beyond current U.S. performance.

    “the U.S. proficiency rate is 32%; and Singapore has 63% proficiency by the standard set in the United States of what proficiency means on our national assessment of educational progress.”

  • The achievement gap within the United States is enormous: the average California student performs only at the 24th percentile of the Massachusetts student distribution in math, showing large within-country variation comparable to international gaps.

    “the average student in California is at the 24th percentile of a Massachusetts student.”

  • U.S. student achievement, as measured by international tests, is considerably below that of other developed countries and many developing countries, with the U.S. ranking slightly below the OECD average on the PISA test.

    “Our achievement levels, as measured by international tests, is considerably below what we see in other developed countries. And many developing countries.”

  • The U.S. achievement deficit is not merely a problem of inner-city schools or minority populations: white children and children of college-educated parents also fail to compete well internationally, and teacher effectiveness varies across the board even in the best suburban public schools, refuting the (racist and incorrect) argument that demographics explain America's poor international standing.

    “our kids, our white kids, don't compete very well internationally, our kids of college-educated parents don't compete well internationally.”

0.71

Schools must adopt accountability systems linking teacher pay and evaluation to student performance outcomes.

6 pointscentrality 4/5
  • Washington, D.C. demonstrates that serious teacher accountability is possible: under a contract instituted four years prior, its evaluation system tied to student performance and outside ratings gave bonuses to about 1000 teachers and fired about 300 over three years—dramatically different from history—though it was enabled by the school system's unusual governance under the U.S. Congress.

    “in the last 3 years, Washington has given bonuses to some 1000 teachers, and they've fired 300 teachers.”

  • Reform cannot target teachers alone; principals and administrators must operate under the same incentives, being rewarded and punished based on the same student-success outcomes, otherwise misaligned administrators could undermine the system (for example, by hiring relatives).

    “you can't have the principals of schools and the administrators operating with a different set of incentives than the teachers. So you have to reward and punish principals who get the same kind of success that we are looking for in teachers.”

  • Contrary to arguments against merit pay that good teachers can't be measured, identifying the worst teachers is obvious within any school; the genuine difficulty is designing an evaluation system viewed as reliable and fair enough to support personnel decisions—something only recently emerging as state legislatures demand improved teacher evaluations.

    “I think it's very obvious who they are. Now, designing a system that uses that information is what we have had trouble doing. What you need is an evaluation system that is viewed as reliable and fair and that you can use for these personnel decisions.”

  • While the U.S. nearly universally fails to fire obviously bad teachers, the best-performing school systems internationally—though not via explicit firing policies—use counseling, reassignment, and training to ensure bad teachers don't stay in the classroom long, and their teacher unions have become more accommodating to the idea that a quality product matters, unlike U.S. unions.

    “the best school systems, one way or another, don't let bad teachers stay in the classroom for very long.”

  • States that improved achievement did so through accountability systems—Massachusetts set high standards, tested them, and tied graduation to passing a tough test (while also spending more), whereas Florida achieved similar or faster gains emphasizing accountability and early-grade reading while spending virtually no more money in real terms—suggesting accountability rather than spending drives improvement.

    “The people in Massachusetts relate a lot of this to an improved accountability system where they had high standards set up and they tested these standards and held kids in school systems responsible for them.”

  • Research across a wide range of improving countries consistently identifies four levers: more parental choice over schools, pay-for-performance linking rewards to school/teacher performance, a good accountability system measuring performance and holding schools responsible, and more local decision-making—holding schools accountable for value added while letting them decide how to achieve it, mirroring how the rest of the U.S. economy operates.

    “One is having more choice... Another is pay for performance... Another is having a good accountability system... And a final, at least within the most developed countries, is having more local decision making.”

0.70

Higher student achievement directly produces measurable long-term economic gains for workers and nations.

3 pointscentrality 5/5
  • If the U.S. ramped up its student achievement to Canada's level (roughly Massachusetts's level) over a 20-year period, the historical relationship between achievement and growth projects that the present value of the average paycheck of all U.S. workers over the next 80 years would be 20% higher.

    “we ramp up to Canada over a 20-year period and then use the historical relationship to project out what that would mean for the U.S. economy. The answer is stunning. It says that the average paycheck of all workers in the U.S. economy over the next 80 years would be 20% higher.”

  • Over the last 50 years, countries whose populations have higher cognitive achievement in math and science grow faster economically, because a more skilled population enlarges GDP, with huge implications for future societal well-being.

    “countries where the population has higher achievement, knows more math and science, grow faster. And their GDP (Gross Domestic Product) gets larger because they have a more skilled population.”

  • Cross-country growth differences from 1960-2000 can be explained almost entirely by two factors—a country's starting GDP level and its achievement score—because countries that start far behind can grow fast simply by copying others, while countries at the frontier must innovate and invent to grow.

    “If you start way behind, all you have to do is copy somebody else in order to grow fast. If you start ahead you have to innovate and invent things.”

0.51

Increased school spending without aligned incentives has not improved student achievement over five decades.

2 pointscentrality 4/5
  • Since 1960, real inflation-adjusted spending per pupil in the U.S. has increased more than fourfold, yet on average produces no measurable gains in student performance, because current incentives in schools are not aligned with using funds productively—not because money can never matter.

    “Since 1960 till today, spending per pupil in real terms, inflation-adjusted... has increased more than fourfold.”

  • Technology and online resources (MOOCs, Khan Academy) can dramatically improve instruction and will substitute for the worst teachers—perhaps one in five doing a bad job—but the system has not yet figured out how to harness technology to make classroom teachers more effective, partly an incentive issue.

    “We do have technology that can dramatically improve over the instruction that we see in some classes... We haven't figured out how to get classroom teachers using technology to be more effective. I think this is partly an incentive issue.”

0.39

Educational systems must build deep conceptual knowledge and procedural fluency, and innovation requires both institutional incentives and classroom foundations.

5 pointscentrality 2/5
  • America's advantage in innovation stems less from schools and more from an entire economy that rewards people for innovating; in countries like Korea, students perform well on tests but the economy and workplace culture punish young people for questioning authority or proposing new ideas, suppressing creativity regardless of test scores.

    “we do know that their economy doesn't reward creativity in the same way. One of my best graduate students ever was a Korean who got into trouble when he went back to Korea to work and his boss told him the answer to a problem, and he basically said: I don't see that... And he got into serious trouble.”

  • Finland's achievement gains over 30 years are a single data point through which many competing explanatory lines can be drawn—no testing, late school start, teacher quality, teacher autonomy, or societal homogeneity—so its success cannot be cleanly attributed; Hanushek's own lean is that teaching being an unusually attractive profession there (despite modest salaries) is the key factor.

    “We have basically one data point, one observation. And we have about 15 lines that we can draw through that data point, and everybody chooses their favorite line.”

  • The new 'common core' curriculum emphasizing deep conceptual learning while de-emphasizing mechanical skills risks harming students, because genuine creativity and innovation must build on a foundation of known facts and procedures—you start at a high level and then think of something different, rather than expecting students to reinvent foundational knowledge from scratch.

    “my view of this is that you have to build on what you know. That creativity springs from starting at a high level and then thinking of something different from the high level. Not that we want 15-year-olds to reinvent Newton's calculus.”

  • There is no reason to expect countries to teach to international tests like PISA, because the tests are given to a small random sample of schools that don't know they will be tested; any apparent teaching-to-the-test simply reflects teaching a good curriculum well, which is the desired outcome.

    “I don't think there's any reason to expect countries to try to teach to these tests. It's taken by a relatively small portion of the population of a random set of schools, and schools don't know that they are preparing to be tested on these tests.”

  • Although the U.S. economy has thrived by letting market forces and accountability sort out who produces well and what consumers want, talking about applying market forces is treated as a pejorative term in education, even though importing voluntary choice and competition could improve schools.

    “talking about using market forces is a pejorative term in education, even though the U.S. economy has thrived by in fact having market forces making decisions about where things are produced and what's produced.”

0.38

Family background predicts achievement but effective schools with strong teachers can substantially overcome disadvantage.

2 pointscentrality 3/5
  • Parental education and income strongly predict student achievement (e.g., 12% proficiency if parents didn't finish high school versus 44% if parents attended college), but family background is not completely determinative—good schools with effective teachers can make up the difference, though it requires extra work, so the policy lever is improving schools rather than government intrusion into families.

    “goods schools can make up for the difference in family background. It takes extra work.”

  • The U.S. has performed economically better than its test scores would predict because of compensating advantages—earlier universal secondary schooling, good economic institutions (open markets, limited government, property rights), the best higher education in the world, and an open immigration policy that attracts and retains talent—but all of these advantages are now eroding as other countries emulate them.

    “back in 1960, the United States was practically the only country that had universal secondary schooling for its population... we've had free and open labor in product markets, we've had limited intrusion of government... we probably have the best higher education in the world... we allow a lot of good foreign people to come into the United States and work here.”