Global Lessons for Improving U.S. Education

The middling performance of U.S. students on international achievement tests is by now familiar, so the overall results of the latest Program for International Student Assessment (PISA) study, released in late 2010, came as no surprise. Among the 34 developed democracies that are members of the Organization for Economic Cooperation and Development (OECD), 15-year-olds in the United States ranked 14th in reading, 17th in science, and no better than 25th in mathematics. The new wrinkle in the data was the participation of China’s Shanghai province, whose students took top honors in all three subjects, besting U.S. students by the equivalent of multiple grade levels in each. Home to the nation’s wealthiest city and a magnet for its most ambitious and talented citizens, Shanghai’s results are hardly representative of China as a whole. Yet its students’ eye-popping performance seemed to highlight new challenges facing the U.S. economy in an age of unprecedented global trade.

The notion that educational competition threatens the future prosperity of the United States had already been a recurrent theme in many quarters. The Obama administration, for example, cited a link between education and national economic competitiveness in making the case for the education funding allocated through the American Recovery and Reinvestment Act, for the state-level policy changes incentivized by the Race to the Top grant competition, and for increased federal support of early childhood education.

However, the relationship between education and international competitiveness remains “a subject rife with myth and misunderstanding,” as even Arne Duncan, secretary of the U.S. Department of Education, has noted. This confusion may stem from the fact that the concept of international competitiveness is notoriously difficult to pin down. Academic economists, for example, have long criticized the view that countries in a globalized economy are engaged in a zero-sum game in which only some can emerge as winners and others will inevitably lose out. All countries can in the-ory benefit from international trade by specializing in those activities in which they have a comparative advantage. In what sense, then, does it make sense to talk about national economies competing?

These general lessons seem doubly true for education, where the mechanisms by which gains abroad would undermine U.S. prosperity are altogether unclear. Educational improvements in other countries enhance the productivity of their workforces, which in turn should reduce the costs of imports to the United States, benefitting all U.S. residents except perhaps those who compete directly in producing the same goods. At the top end of the education spectrum, growth in the number of graduate degrees awarded in fields such as science and engineering fosters technological advances from which the United States can benefit regardless of where key discoveries are made. For these and other reasons, developments such as Shanghai’s performance on the PISA, although at first glance startling, may in fact represent good news.

This is not to say that the very real educational challenges facing the United States are irrelevant to its future economic performance. On the contrary, the evidence that the quality of a nation’s education system is a key determinant of the future growth of its economy is increasingly strong. Therefore, the United States may benefit by examining past and ongoing research on educational performance across countries and considering the actions that higher-performing nations have taken that have helped their students to succeed.

Hard comparisons

Launched in 2000 as a project of the OECD, the PISA is administered every three years to nationally representative samples of students in each OECD country and in a growing number of partner countries and subnational units such as Shanghai. The 74 education systems that participated in the latest PISA study, conducted during 2009, represented more than 85% of the global economy and included virtually all of the United States’ major trading partners, making it a particularly useful source of information on U.S. students’ relative standing.

U.S. students performed well below the OECD average in math and essentially matched the average in science. In math, the United States trailed 17 OECD countries by a statistically significant margin, its performance was indistinguishable from that of 11 countries, and it significantly outperformed only five countries. In science, the United States significantly trailed 12 countries and outperformed nine. Countries scoring at similar levels to the United States in both subjects include Austria, the Czech Republic, Hungary, Ireland, Poland, Portugal, and Sweden.

The gap in average math and science achievement between the United States and the top-performing national school systems is dramatic. In math, the average U.S. student by age 15 was at least a full year behind the average student in six countries, including Canada, Japan, and the Netherlands. Students in six additional countries, including Australia, Belgium, Estonia, and Germany, outperformed U.S. students by more than half a year.

The second-rate performance of U.S. students is particularly striking given the level of resources the nation devotes to elementary and secondary education. Data on cumulative expenditures per student in public and private schools between ages 6 and 15 confirm that the United States spends more than any other OECD country except Luxembourg. Most of the higher-performing countries spend between $60,000 and $80,000 per student, compared with nearly $105,000 in the United States.

Some observers have speculated that despite the modest performance of its average students, the U.S. education system is characterized by pockets of excellence that can be expected to meet the needs of the knowledge economy. However, there is no clear evidence that educating a subset of students to very high levels is more important for national economic success than raising average achievement levels. Moreover, the United States in fact fares no better in comparisons of the share of students performing at exceptionally high levels. For example, only 9.9% of U.S. students taking the PISA math test achieved at level 5 or 6, the two top performance categories, which, according to test administrators, indicate that students are capable of complex mathematical tasks requiring broad, well-developed thinking and reasoning skills. Twenty-four countries outranked the United States by this metric. The share of students achieving level 5 or 6 exceeded 20% in five countries and exceeded 15% in another 10. In Shanghai, 50.4% of students surpassed this benchmark, more than five times the level in the United States.

Another common response to the disappointing performance of U.S. students has been to emphasize the relative diversity of U.S. students and the wide variation in their socioeconomic status. Family background characteristics and other out-of-school factors clearly have a profound influence on students’ academic achievement. The available international assessments, all of which offer only a snapshot of how students have learned at a single point in time rather than evidence on how much progress they are making from one year to the next, are therefore best viewed as measuring the combined effects of differences in school quality and differences in these contextual factors. The latter are poorly measured across countries, making it difficult to pin down their relative import.

Even so, it is difficult to attribute the relative ranking of U.S. students to out-of-school factors alone. The share of U.S. students with college-educated parents, a key predictor of school success, actually ranks 8th among the OECD countries. The typical U.S. student is also well above the OECD average, according to PISA’s preferred measure of students’ socioeconomic status.

A record of poor results

U.S. students, however, have never fared well in various international comparisons of student achievement. The United States ranked 11th out of 12 countries participating in the first major international study of student achievement, conducted in 1964, and its math and science scores on the 2009 PISA actually reflected modest improvements from the previous test. The United States’ traditional reputation as the world’s educational leader stems instead from the fact that it experienced a far earlier spread of mass secondary education than did most other nations.

In the first half of the 20th century, demand for secondary schooling in the United States surged as technological changes increased the wages available to workers who could follow written instructions, decipher blueprints, and perform basic calculations. The nation’s highly decentralized school system, in which local communities could vote independently to support the creation of a high school, provided a uniquely favorable mechanism to drive increased public investment in schooling. As economic historian Claudia Goldin of Harvard University has documented, by 1955 almost 80% of 15- to 19-year-olds were enrolled full-time in general secondary schooling, more than double the share in any European country.

The United States’ historical advantage in terms of educational attainment has long since eroded, however. U.S. high-school graduation rates peaked in 1970 at roughly 80% and have declined slightly since, a trend often masked in official statistics by the growing number of students receiving alternative credentials, such as a General Educational Development (GED) certificate. Although the share of students enrolling in college has continued to climb, the share completing a college degree has hardly budged. As this pattern suggests, both the time students are taking to complete college degrees and dropout rates among students enrolling in college have increased sharply. This trend seems especially puzzling in light of the fact that the economic returns from completing a postsecondary degree—and the economic costs of dropping out of high school—have grown substantially over the same period.

Meanwhile, other developed countries have continued to see steady increases in educational attainment and, in many cases, now have postsecondary completion rates that exceed those in the United States. The U.S. high-school graduation rate now trails the average for European Union countries and ranks no better than 18th among the 26 OECD countries for which comparable data are available. On average across the OECD, postsecondary completion rates have increased steadily from one age cohort to the next. Although only 20% of those aged 55 to 64 have a postsecondary degree, the share among those aged 25 to 34 is up to 35%. The postsecondary completion rate of U.S. residents aged 25 to 34 remains above the OECD average at 42%, but this reflects a decline of one percentage point relative to those aged 35 to 44 and is only marginally higher than the rate registered by older cohorts.

To be sure, in many respects the U.S. higher education system remains the envy of the world. Despite recent concerns about rapidly increasing costs, declining degree completion rates, and the quality of instruction available to undergraduate students, U.S. universities continue to dominate world rankings of research productivity. The 2011 Academic Rankings of World Universities, an annual publication of the Shanghai Jiao Tong University, placed eight U.S. universities within the global top 10, 17 within the top 20, and 151 within the top 500. A 2008 RAND study commissioned by the U.S. Department of Defense found that 63% of the world’s most highly cited academic papers in science and technology were produced by researchers based in the United States. Moreover, the United States remains the top destination for graduate students studying outside of their own countries, attracting 19% of all foreign students in 2008. This rate is nine percentage points higher than the rate of the closest U.S. competitor, the United Kingdom.

Yet surely the most dramatic educational development in recent decades has been the rapid global expansion of higher education. Harvard economist Richard Freeman has estimated that the U.S. share of the total number of postsecondary students worldwide fell from 29% in 1970 to just 12% in 2006, a 60% decline. A portion of this decline reflects the progress of developed countries, but the more important factor by far has been the spectacular expansion of higher education in emerging economies, such as China and India. In China alone, postsecondary enrollments exploded from fewer than 100,000 students in 1970 to 23.4 million in 2006. The increase over the same period in India was from 2.5 million to 12.9 million students. In comparison, just 17.5 million U.S. students were enrolled in postsecondary degree programs in 2006.

Although these enrollment numbers reflect China and India’s sheer size and say nothing about the quality of instruction students receive, several recent reports have nonetheless concluded that the rapidly shifting landscape of higher education threatens the United States’ continued dominance in strategically important fields such as science and technology. Perhaps best known is the 2007 National Academies report Rising Above the Gathering Storm, which warned that “the scientific and technological building blocks critical to our economic leadership are eroding at a time when many other nations are gathering strength.” A follow-up report issued in 2010 by some of the authors of Gathering Storm warned that the storm was “approaching category five.” Although critics claim that the reports exaggerated the degree to which the research coming out of emerging economies is comparable to that produced by scholars based in the United States, it seems safe to conclude that in the future the nation will occupy a much smaller share of a rapidly expanding academic marketplace.

Costs of low-quality education

How concerned should the United States be about these developments? And is it the improvement in educational outcomes abroad that should motivate concern?

After all, until very recently the performance of the U.S. economy had far surpassed that of the industrialized world as a whole, despite the mediocre performance of U.S. students on international tests. Some observers have gone so far as to question the existence of a link between available measures of the performance of national education systems and economic success.

Such skepticism was not entirely misplaced. Economists as far back as Adam Smith have highlighted the theoretical importance of human capital as a source of national economic growth. For technologically advanced countries, highly educated workers are a source of innovations needed to further enhance labor productivity. For countries far from the frontier, education is necessary to enable workers to be able to adopt new technologies developed elsewhere. Because a given country is likely to be both near and far from the technological frontier in various industries at any given point in time, both of these mechanisms are likely to operate simultaneously. Yet rigorous empirical evidence supporting these common-sense propositions has been notoriously difficult to produce.

One key limitation of early research examining the relationship between education and economic growth is that it was based on crude measures of school enrollment ratios or the average years of schooling completed by the adult population. Although studies taking this approach tend to find a modest positive relationship between schooling and economic growth across countries, years of schooling is an incomplete and potentially quite misleading indicator of the performance of national education systems. Measures of educational attainment implicitly assume that a year of schooling is equally valuable regardless of where it is completed, despite the clear evidence from international assessments that the skills achieved by students of the same age vary widely across countries.

Economists Eric Hanushek of Stanford University and Ludger Woessmann of the University of Munich have addressed this limitation in an important series of papers published since 2008. Their key innovation is the use of 12 international assessments of math and science achievement conducted between 1964 and 2003 to construct a comparable measure of the cognitive skills of secondary school students for a large sample of countries. They went on to analyze the relationship between this measure and economic growth rates between 1960 and 2000 across all 50 countries for which cognitive skills and growth data are available and separately across 24 members of the OECD.

Their work has yielded several notable results. First, after controlling for both a country’s initial gross domestic product (GDP) per capita and the average years of schooling completed in 1960, they found that a one standard deviation increase in test scores is associated with an increase in annual growth rates of nearly 2%. Taken at face value, this implies that raising the performance of U.S. students in math and science to the level of that of a top-performing nation would increase the U.S. growth rate by more than a full percentage point over the long run; that is, once students educated to that level of academic accomplishment make up the entire national workforce. Second, they found that both the share of a country’s students performing at a very high level and the share performing above a very low level appear to contribute to economic growth in roughly equal amounts, suggesting that there is no clear economic rationale for policymakers to focus exclusively on improving performance at the top or the bottom of the ability distribution. Third, after controlling for their test-based measure of students’ cognitive skills, they found that the number of years of schooling completed by the average student is no longer predictive of growth rates. This suggests that policies intended to increase the quantity of schooling that students receive will bear economic fruit only if they are accompanied by measurable improvements in students’ cognitive skills.

Although these studies have offered a clear improvement over previous evidence, skeptics may wonder whether the pattern identified linking education quality and economic growth in fact reflects a causal relationship. It is of course possible that there are unidentified factors that enhance both the quality of national education systems and economic growth. Hanushek and Woessmann have performed a series of analyses intended to rule out these concerns. Although none of these tests of causation is definitive on its own, together they strongly suggest that policies that increase education quality would in fact generate a meaningful economic return.

Moreover, the magnitude of the relationship observed is so large that it would remain important even if a substantial portion of it were driven by other factors. Consider the results of a simulation in which it is assumed that the math achievement of U.S. students improves by 0.25 standard deviation gradually over 20 years. This increase would raise U.S. performance to roughly that of some mid-level OECD countries, such as New Zealand and the Netherlands, but not to that of the highest-performing OECD countries. Assuming that the past relationship between test scores and economic growth holds true in the future, the net present value of the resulting increment to GDP over an 80-year horizon would amount to almost $44 trillion. A parallel simulation of the consequences of bringing U.S. students up to the level of the top-performing countries suggests that doing so would yield benefits with a net present value approaching $112 trillion.

Yet despite ubiquitous rhetoric about education’s importance for countries competing in the global marketplace, there is no evidence that these potential gains would come at the expense of other nations. Put differently, there is no reason to suspect that U.S. residents are made worse off in absolute terms by the superior performance of students in places such as Finland, Korea, or even Shanghai. At the higher education level, U.S. universities clearly face growing competition in recruiting talented international students and faculty and will probably find it difficult to maintain their current dominance of world rankings. Yet as labor economist Richard Freeman of Harvard University has explained, “the globalization of higher education should benefit the U.S. and the rest of the world by accelerating the rate of technological advance associated with science and engineering and by speeding the adoption of best practices around the world, which will lower the costs of production and prices of goods.”

This is not to say that a continued decline in the relative standing of the U.S. education system would leave the nation’s economy entirely unaffected. As Caroline Hoxby, a labor and public economist at Stanford University, has noted, studies of the “factor content” of U.S. exports and economic growth have long documented their disproportionate reliance on human capital. This pattern suggests that the United States has traditionally had a comparative advantage in the production of goods that depend on skilled labor, which in turns reflects its historical edge in the efficient production of highly educated workers. In recent decades, U.S. companies have increasingly addressed labor shortages in technical fields by “importing” human capital in the form of highly educated immigrants, many of whom received their postsecondary training in the United States. This strategy cannot be a source of comparative advantage in the long run, however, because other countries are by definition able to import talented immigrants at the same cost. The decline in the relative performance of the U.S. educational system may therefore have adverse consequences for the high-tech sectors on which the nation has historically depended to generate overall growth. The ability of the nation’s economy as a whole to adapt in the face of such a disruption is, of course, an open question.

Policy lessons

In short, although there is little indication that education is an area in which countries are engaged in zero-sum global competition for scarce resources, education reform does provide a means to enhance economic growth and, in turn, the nation’s capacity to address its mounting fiscal challenges. Even if that were not the case, the moral argument for addressing the performance of the most dysfunctional U.S. school systems and the inequalities in social outcomes they produce would be overwhelming. What then are the lessons policymakers should draw from the growing body of evidence examining the performance of school systems across various countries?

The first and most straightforward lesson is simply that dramatic improvement is possible and that this is true even of the best-performing state school systems. Not only do many countries perform at markedly higher levels despite being at lower levels of economic development, but several of these countries have improved their performance substantially in the relatively short period since international tests were first widely administered. Nor do the international data suggest that countries face a stark tradeoff between excellence and equity when considering strategies to raise student achievement. In fact, the countries with the highest average test scores tend to exhibit less overall inequality in test scores and, in many cases, weaker dependence of achievement on family background characteristics.

A policy agenda centered on closing the global achievement gap between U.S. students and those in other developed countries would provide a complementary and arguably more encompassing rationale for education reform than one focused primarily on closing achievement gaps among subpopulations within the United States. The urgency of closing domestic achievement gaps is without question, but the current emphasis on this goal may well reinforce the perception among members of the middle class that their schools are performing at acceptable levels. The 2011 PDK-Gallup Poll shows that more than half of all U.S. residents currently assign the public schools in their local community an “A” or “B” grade, while only 17% assign one of those grades to public schools in the nation as a whole. This gap between local and national evaluations has widened considerably over the past decade, and similar data from the 2011 Education Next–PEPG Survey show that well-educated, affluent citizens are particularly likely to rate their local schools favorably. Reporting systems that make it possible to compare the performance of students in specific U.S. school districts to that in top-performing countries internationally could help to alter perceptions and broaden support for reform.

A second lesson is that reform efforts should aim to improve the quality of education available to U.S. students in elementary and secondary schools rather than merely increase the quantity of education they consume. The large economic return from the completion of college and especially graduate degrees suggests that there is considerable demand for workers who have been educated to those levels, and policymakers would be wise to address issues, such as the complexity of financial aid systems, that create obstacles to degree completion for academically prepared students. But increasing educational attainment should not be an end in and of itself. Doing so is unlikely to yield economic benefits without reforms to K-12 schooling that ensure that a growing number of students are equipped for the rigors of postsecondary work.

Another general lesson is that additional financial investment is neither necessary nor sufficient to improve the quality of elementary and secondary education. Current data clearly show that other developed countries have managed to achieve far greater productivity in their school systems, in many cases spending considerably less than the United States, to achieve superior results. Nor have countries that have increased spending levels in recent decades experienced gains in their performance on international assessments, a pattern that is consistent with the mixed track record of spending increases in producing improved student outcomes in the United States.

If countries with high-performing elementary and secondary education systems have not spent their way to the top, how have they managed to get there? Unfortunately, using international evidence to draw more specific policy guidance for the United States remains a challenge. Although it is straightforward to document correlations between a given policy and performance across countries, it is much harder to rule out the existence of other factors that could explain the relationship. The vast cultural and contextual differences from one country to the next also imply that policies and practices that work well in one setting may not do so in another. Even so, there are three broad areas in which the consistency of findings across studies using different international tests and country samples bears attention.

Exit exams. Perhaps the best-documented factor is that students perform at higher levels in countries (and in regions within countries) with externally administered, curriculum-based exams at the completion of secondary schooling that carry significant consequences for students of all ability levels. Although many states in the United States now require students to pass an exam in order to receive a high-school diploma, these tests are typically designed to assess minimum competency in math and reading and are all but irrelevant to students elsewhere in the performance distribution. In contrast, exit exams in many European and Asian countries cover a broader swath of the curriculum, play a central role in determining students’ postsecondary options, and carry significant weight in the labor market. As a result, these systems provide strong incentives for student effort and valuable information to parents and other stakeholders about the relative performance of secondary schools. The most rigorous available evidence indicates that math and science achievement is a full grade-level equivalent higher in countries with such an exam system in the relevant subject.

Private-school competition. Countries vary widely in the extent to which they make use of the private sector to provide public education. In countries such as Belgium, the Netherlands, and (more recently) Sweden, for example, private schools receive government subsidies for each student enrolled equivalent to the level of funding received by state-run schools. Because private schools in these countries are more heavily regulated than those in the United States, they more closely resemble U.S. charter schools, although they typically have a distinctive religious character. In theory, government funding for private schools can provide families of all income levels with a broader range of options and subject the state-run school system to increased competition from alternative providers. Rigorous studies confirm that students in countries that for historical reasons have a larger share of students in private schools perform at higher levels on international assessments while spending less on primary and secondary education. Such evidence suggests that competition can spur school productivity. In addition, the achievement gap between socioeconomically disadvantaged and advantaged students is reduced in countries in which private schools receive more government funds.

High-ability teachers. Much attention has recently been devoted to the fact that several of the highest-performing countries internationally draw their teachers disproportionately from the top third of all students completing college degrees. This contrasts sharply with recruitment patterns in the United States. Given the strong evidence that teacher effectiveness is the most important school-based determinant of student achievement, this factor probably plays a decisive role in the success of the highest-performing countries. Unfortunately, as education economist Dan Goldhaber of the University of Washington has pointed out, the differences in teacher policies across countries that have been documented to date “do not point toward a consensus about the types of policies—or even sets of policies—that might ensure a high-quality teacher workforce.”

Although increasing average salaries provides one potential mechanism to attract a more capable teaching workforce, there is no clear relationship between teacher salary levels and student performance among developed countries. Especially given the current strains on district and state budgets, any funds devoted to increasing teacher salaries should be targeted at subjects such as math and science, in which qualified candidates have stronger earnings opportunities in other industries, and at teachers who demonstrate themselves to be effective in the classroom. Intriguingly, the only available study on the latter topic shows that countries that allow teacher salaries to be adjusted based on their performance in the classroom perform at higher levels.

Vital national priority

During the past two decades, state and federal efforts to improve U.S. education have centered on the development of test-based accountability systems that reward and sanction schools based on their students’ performance on state assessments. The evidence is clear that the federal No Child Left Behind Act and its state-level predecessors have improved student achievement, particularly for students at the bottom of the performance distribution. Yet the progress made under these policies falls well short of their ambitious goals. Equally important, the progress appears to have been limited to a one-time increment in performance rather than launching schools on a trajectory of continuous improvement.

International evidence may not yet be capable of providing definitive guidance for closing the global achievement gap between students in the United States and those in the top-performing countries abroad. It does, however, indicate that holding students accountable for their performance, creating competition from alternative providers of schooling, and developing strategies to recruit and retain more capable teachers all have important roles to play in addressing what should be a vital national priority.