PHOTO: John Sheretz

We expect too much of education, and that is why we blame it for nearly everything that goes wrong. To every initiative touting education as a silver bullet, there comes failure and the inevitable school-bashing reaction. Economic growth often figures in the debate, for virtually every national development plan emphasizes education as an economic investment, and virtually every school-reform movement stirs up fears of “a nation at risk” to gain support for its program.

This rhetoric has escalated. It is no longer enough to educate more people. Our schools now must also meet world-class standards—above all, in the teaching of mathematics and science education. The reason, we are told, is that the better a country’s children do in these subjects, the faster its economy will grow. This assertion is the cornerstone of much educational policy throughout the world.

But it is simply not true. My colleagues and I have examined math and science achievement of students from many different countries and compared these variations with rates of economic growth. We have found no clear connection between them.

In a paper published in November 2006 in the American Journal of Education, Xiaowei Luo, Evan Schofer, John W. Meyer, and I examine the math and science education in 38 countries over two periods: between 1900 and 1970 and between 1980 and 2000. We also separate out the statistical influence of the “Asian Tigers”—Hong Kong, Singapore, South Korea, and Taiwan. We found that the expected, positive correlation between technical education and economic growth was much stronger in the first period than in the second, and that the difference could be entirely attributed to the fortunes of the four Asian Tigers.

This finding suggests that the observed correlation between achievement and growth derives from only a few cases. More important, the effect sharply diminishes in the second period, during which the regional economic recession known as the Asian flu removed the Tigers from the hopper; this was sufficient to break the link between achievement and economic growth. When we look at the results for the period between 1990 and 2000, we can identify no effect whatsoever. A finding that fluctuates that much from one period to another is clearly not reliable.

Ironically, it was the United States, a country well to the back of the class academically, that dazzled the world economy in the last seven years of the 20th century. This raises an obvious question: how did that country’s relatively underachieving school system produce such an overachieving economy? Our study provides a partial answer. It turns out that you can gain an economic payoff by moving from the bottom to the middle tier of academic achievers, but not by moving from the middle to the elite tiers.

What this means is that the United States should worry less about not having the achievement profile of South Korea or Singapore and care more about not sinking to the level of Nigeria or Thailand. Aspiring to have the best schools in the world may be great rhetoric, but it is not sound economic policy.

The takeaway point is not that science education is a bad economic investment, for indeed in other studies we have found that policies that expand the cadre of scientists and engineers do, in fact, promote economic development. It is, rather, that the overall technical competence of the population as a whole (at least insofar as that competence is measured by international achievement tests) does not matter.

What about education as a whole? Here the matter is much clearer: a growing number of researchers have concluded that the question is not whether it affects economic growth, but under what conditions it affects that growth. We agree. The same principle applies to the study of academic achievement and economic growth. Much more work is required to see if we can make sense of fluctuating findings.

We obviously need a moratorium on exaggerated “educational silver bullet” claims. These statements inevitably give rise to equally exaggerated, school-bashing indictments. If our study leads to a healthier and more research-oriented skepticism, we will have better served the policy process in the long run.