Getty
This terrific “working paper” by distinguished AEI demographer Nicholas Eberstadt concludes with a fascinating (and worrying) appraisal of educational performance in China—perhaps no surprise, since he and his team wrote it with support from the Pentagon’s Office of Net Assessment.
Along the way, however, it’s also a tour-de-force of analysis of education-related factors that are associated with the prosperity of many countries—and the other factors that play a role, particularly in explaining and augmenting the education factors.
The project was undertaken to “examine ‘knowledge capital,’ the economically productive knowledge and skills of national populations: how such potential differs between nations; how it affects levels of national productivity; and the determinants of that potential internationally.” It relies on a host of data sources—especially, for our purposes, international assessments such as PISA and TIMSS—and deploys a number of analytic methods (some of them involving a fair bit of conjecture or assumption).
Among Eberstadt’s key findings—which draw on multiple sources—before getting to the China part:
- “[A] country’s student test scores provide highly meaningful information about both current differences in national economic potential (per capita productivity) and also national economic potential ten years in the future.” In other words, much as Hanushek and others have been demonstrating, a country’s prosperity tomorrow is linked to how its students do today.
- “[A] difference of 100 points on a country’s mean scores in student achievement tracks with about a 25-percentage point difference in national per capita productivity levels ten years hence.” In other words, the linkage mentioned above is strong and durable.
- “[T]aken together, measured academic achievement and measured educational attainment can explain close to thirty percent of the overall productivity disparity between the world’s most productive and its least productive economies.” In other words, combining test scores and the amount of education that the people of a country get is a 30-percent window into whether it has a strong or weak economy.
- “Two major trends are currently transforming the global terrain of knowledge capital—and both will shape the global economic balance. The first of course is the continuing expansion of education….This will mean the skills and knowledge of working age populations that power national economies on the whole are set to increase in the decades immediately ahead. But the second trend—less familiar but no less real—is skills decay in adult populations.” Almost nobody pays attention to how skills age, decay, and become obsolete.
- “Socioeconomic factors—levels of national schooling, income, urbanization, and ‘business climate’—could account for up to 70 percent of inter-country differences in mean achievement scores. But…by adding two additional factors, our models turned out to predict up to 90 percent of the academic achievement differences between countries. The first of these was geography….The second was cognitive ability. International differences in measured IQ levels provide meaningful independent predictive information about differences in national academic achievement, even after socioeconomic factors and geography are taken into account.”
That’s a lot, as the author pivots from how educational achievement bears on prosperity to the factors that account for a country’s educational achievement. Note that they’re not looking at the inner workings of a country’s education system but rather at how much education a country’s people get, gauged both by years of it and test scores. So this paper isn’t about how to fix education but about why it matters and what is associated with achievement.
Finally—the last quarter of a sixty-page paper—Eberstadt returns to “the mystery of China’s performance in academic achievement tests.” Why does China do so well on PISA—and can those data be trusted?
This is a gnarly topic.[1] The bottom line is that the results reported by OECD for China are profoundly misleading. That’s mostly because the only Chinese fifteen-year-olds who participate in PISA live in just four cities and do not represent their age cohort across that vast land, but it’s also because China is secretive and controlling such that it’s impossible even to know whether the test-takers are representative of those four big cities. Eberstadt’s “modelling” of how China as a whole would compare with other countries places it between Kenya and Cambodia based on 2009-15 data. He notes that education in China, as in many places, has been improving and its performance is now more likely on par with places like Turkey and Malaysia.
Be warned: “The true contours of knowledge and skills in the Chinese population will have a direct and powerful influence on the balance of power over the coming generation. But the cloud of empirical uncertainty surrounding China’s true terrain of knowledge capital has not yet been dispelled—indeed, as yet it has barely been pierced.”
SOURCE: Nicholas Eberstadt, “Knowledge, Skills, and the Global Balance of Power: What International Standardized Achievement Tests Tell Us about National Economic Potential and Prospects,” American Enterprise Institute (May 2025).
[1] A companion paper by Eberstadt and colleagues digs even deeper into the idiosyncrasies and mysteries of China’s PISA scores.