T he conventional wisdom is now firmly established: American students can't hold their own against their peers in other nations. They can't read, they can't do math, they are abysmally ignorant of science. That has been the message of countless stories in the media, supposedly backed up by international data. And this poor performance, we have been told, is responsible for the economic woes the United States has experienced in recent decades.
But global comparisons show no such thing: American students look better in international tests than the critics would have us believe, and the schools have little to do with the "competitiveness" of the economy. For decades, the media have uncritically reported unfavorable comparisons of educational performance, often based on dubious research, and have slighted more positive findings. The result is that an inaccurate picture of total national failure dominates educational policy and politics. American schools do need improvement. But the crux of the problem lies among the lower third of schools and requires a far more targeted and discriminating approach than the heralds of educational apocalypse have called for.
OLD WHINE, NEW BATTLES
The fretting over American schools' international performance became a national pastime during the 1950s, when there was a real source of anxiety: the space and weapons races with the Soviet Union. Some cold warriors were famous educational worriers, such as Admiral Hyman Rickover, who looked at European schools and without a lot of evidence declared them more rigorous than our own. More serious for Rickover were the numbers supplied to him by CIA director Allen Dulles showing that the Soviet Union would produce far more scientists, engineers, and mathematicians than the United States would. Rickover repeatedly admonished his audiences, "Let us never forget that there can be no second place in a contest with Russia and that there will be no second chance if we lose."
The Russians' launch of Sputnik in October 1957 seemed to confirm what the critics had been saying. The following March, Life magazine published a five-part series on the "Crisis in Education," prominently contrasting a stern-faced Russian student conducting optics experiments in his school lab with a happy-go-lucky American in typing class ("I type about a word a minute," he says). A large photo shows the American boy laughing as he returns to his seat after "struggling" with a geometry problem at the blackboard. The text reads, "Stephen amused class with wisecracks about his ineptitude." Obviously, the Russians were going to "bury" us, as Nikita Khruschev would soon tell Richard Nixon.
The 1980s saw a replay of the same alarm, this time with "competitiveness" as the worry and Japan, Germany, and Korea playing the role of educational heavies. In 1983 the widely publicized report A Nation At Risk put the schools in an unremittingly harsh light and announced a virtual state of national siege. "If an unfriendly foreign power had attempted to impose on America the mediocre educational performance that exists today, we might well have viewed it as an act of war." In the 1990s, the going wisdom persists that our schools are awful. American students "come in last or next to last in virtually every international comparison," wrote Louis V. Gerstner, IBM's chief executive, in 1994. In October 1997, Chicago Tribune columnist Joan Beck was so certain of the outcome of national testing that she declared, "Testing fourth graders in reading and eighth graders in math will only tell us what we already know. The United States lags behind most industrial nations in educational achievement."
W hat do the data actually say about American kids in relation to their peers abroad? It depends on what's tested. In the major comparative study of reading, conducted in 1992, American students finished second in a comparison of 31 nations. The only students who did better came from Finland, a small, homogeneous country that taxes its citizens at a far higher level. And the Finns, of course, have no immigrant population that needs to be taught Finnish as a second language, which might be a daunting task. The top 10 percent, 5 percent, and 1 percent of American students were the best in the world at both ages tested, 9 and 14. In other words, our best readers outscored the best readers in all other nations that participated in the test, even the Finns.
One reason why Americans believe their children compare poorly to foreign students is that for 12 years, the Reagan and Bush administrations promoted a conservative agenda hostile to the public schools and gave bad news about education far more publicity than good news. The treatment of the reading study by Bush's Department of Education illustrates the point. Although the department had held a press conference only a few months earlier to publicize negative findings on American students' performance in science and mathematics, it held no press conference to announce the results in reading. And no one noticed the study. It even took Education Week, the industry's newspaper of record, two months to discover the report; USA Today then carried front-page coverage featuring a quote from a Bush administration official dismissing the study as irrelevant. No other media outlet thought the story newsworthy.
Indeed, the study was so neglected that in June 1996 Secretary of Education Richard Riley re-released the report. USA Today once again put the news on page one. A few other papers ran a story by Josh Greenberg of the Los Angeles Times. When I asked Greenberg why his paper paid attention to a study that was four years old, he replied, "We were very suspicious about the story, but when we checked around we found that no one knew about it, so it was still news." By that criterion, it still is.
FAILING MATH AND SCIENCE?
Mathematics and science are generally considered disaster zones in American schools. Many people have heard, for example, that only the top 1 percent of American students score as high in math as the average student in Japan. This statistic comes from research conducted by Harold Stevenson of the University of Michigan and has been widely disseminated by respected journalists. But the publicity given Stevenson's work illustrates how data showing America's schools in a poor light are accepted less critically than are favorable data. Stevenson's methods violate two cardinal principles of research: The samples of students must be representative of the nations being compared, and they must be comparable to each other. Stevenson's samples meet neither of these criteria. (His American sample contains a large number of poor families, 20 percent of whom did not speak English at home, while his Japanese sample contained many more well-educated parents than the country as a whole.) If American students had finished ahead of Japanese students, the study's methodological flaws would probably have been quickly spotted and the research never published. (I am convinced that if a study comparing American and Japanese students found the Americans finished ahead, the headlines would read: "Japanese Students Second; Americans Next to Last.")
There is no doubt that Japanese children do better in mathematics than do Americans at the same ages, but Stevenson's data exaggerate the gap. Three larger, more sophisticated mathematics studies provide a more reliable picture of international differences in math and science: the 1996 Third International Mathematics and Science Study (TIMSS), the 1992 Second International Assessment of Educational Progress (IAEP-2), and the 1989 Second International Mathematics Study (SIMS).
I emphasize TIMSS here because it is not only the most recent international study, but the largest and the best controlled methodologically. Some nations are obsessed with appearing in a positive light in international comparisons and, accordingly, do not provide a sample with a proportionate number of low-performing schools. The TIMSS report notes which countries failed to meet the sampling criteria. About 50 countries began the TIMSS and 41 completed it at the eighth-grade level, with 26 countries also testing fourth graders. TIMSS has not yet provided data on twelfth graders, nor have its directors clarified how they will handle the methodological problems that those data will pose. Countries differ enormously in the proportion of students who remain in school through grade 12, the proportion at that level who still take mathematics and science courses, and the number of courses that different groups of students have taken. Paradoxically, a country with a high dropout rate may appear to perform better because the students who would have scored lowest don't show up in the sample.
In math, American eighth graders finished slightly below average among the 40 nations. They got 53 percent of the items right, while the international average was 55 percent. American fourth graders, on the other hand, finished above average, ranking twelfth of 26 nations. In science, American eighth graders were slightly above average, scoring 58 percent correct compared to an international average of 56 percent. At the fourth-grade level in science, American students finished third among the 26 countries. However, only about 15 percent of American students scored as high on TIMSS math as the average Japanese student, while about 39 percent of American students scored as well as 50 percent of the Japanese students in science.
Overall, then, American students are near the top in reading, just below average in math, and just above average in science. (In a small international comparison in geography, American students finished in the middle of the pack.) These results are comparable to earlier studies that found American students scoring high in reading and near the average in other subjects. For instance, among 20 nations in SIMS, American eighth graders were tenth in arithmetic and thirteenth in algebra. The algebra result is interesting since most American students don't take algebra in the eighth grade. American eighth graders taking algebra or pre-algebra-about 20 percent of the total-did nearly as well as the Japanese students, who had the highest average score of any nation; even a comparison of those 20 percent of American students to the top 20 percent of Japanese students found the scores to be quite close.
A sian nations have regularly occupied the top ranks in these international math and science tests, but that may not chiefly result from differences in schools. A number of powerful extra-school influences affect students in Asian societies, who work very hard in the middle and high school years. For Asian teenagers, getting into the right high school and then the right college are life-determining events. Kazuo Ishizaka, president of the Japanese Council on Global Education, observes, "Japanese society tends to judge people on the basis of the schools they attended, rather than their ability and skills." Children in Japan often come home from public school at 3:30 in the afternoon, eat, and go on to a private school or tutor. They attend school on Saturdays, and many go on Sundays as well. And an article on Korean schools claims that "today's South Korean students make the famously intense Japanese students look easygoing."
Americans might worry that students who followed an Asian-style regimen were missing valuable experiences. American parents expect their children to become involved in extracurricular activities, to date, and to take after-school jobs. Short of a cultural revolution, it is not clear that American schools, however re formed, could produce the test results that extreme social pressures generate in Asian students. Moreover, American higher education, which is more widely available than in Asian countries, seems to make up for a less intense pace at the primary and secondary levels.
Among the countries that appeared to beat the United States in math and science was Singapore, but the reasons may have nothing to do with the superiority of its schools. Many poor people cross into Singapore each day from Malaysia, do the low-level service jobs, and return home, sparing Singapore the task of educating their children. Longer-term "guest workers" from the Philippines and Indonesia also leave their families behind. In addition, some Singapore families of means whose children are not doing well in the Singapore educational system send their children to school in Malaysia, while some Malaysian children who score well on tests are admitted to the Singapore schools. The relevant numbers aren't available; these are not the kind of statistics that the dictator of Singapore, Lee Kuan Yew, likes to see made public. But Singapore may well get its high scores by exporting low-achieving students, while importing high-achieving students.
Aside from the four Asian nations at the top and a slightly larger number of developing countries at the bottom, the remaining roughly 30 countries (including all the developed countries of the West) look very much alike in their TIMSS mathematics scores. Students from 18 countries-including Israel, Sweden, England, Norway, Germany, and Denmark-score within five percentage points of American students. The TIMSS science scores also fall within a narrow band. When scores are so compressed, small differences in the percentage of correct answers make huge differences in rank, but such rankings may be meaningless.
Once again the media coverage is instructive. When American fourth graders scored third in the TIMSS science tests, the results hardly received any notice, but the eighth graders' lower rank got plenty of attention. The media have been treating the average scores and rankings in international tests as if they prefigured the fate of the nation, but the averages may be entirely the wrong focus.
Focusing on the average scores of the United States (or any nation) obscures the variability of performance with a nation. The differences among American students are enormous compared to the variability among countries. For instance, in IAEP-2, the top third of American schools had average scores as high as the average scores of the top two nations, Taiwan and Korea. The lowest third of American schools, though, did not have scores as high as the lowest nation, Jordan. Disadvantaged urban students in American schools had even lower averages.
These results support an alternative conception of the educational landscape: The top third of American schools are world-class (however defined), the next third are okay, and the bottom third are in terrible shape. This view of our schools leads to a different approach to educational reform than has been customary since A Nation At Risk appeared in 1983. The dominant interpretation has assumed that the typical school-indeed the whole system-is "broken," as Gerstner put it. The data from IAEP-2, though, argue for a reform effort more focused on schools that generally have the least resources and the most difficult social environments.
THE MYTHOLOGY OF "COMPETITIVENESS"
Overall, then, American students have not shown the miserable performance ascribed to them by the speakers cited earlier. Does their performance still put us at risk in the global marketplace? In a word, no.
In the 1980s, when A Nation At Risk argued that schools were responsible for our economic maladies, the economic trends seemed to lend credence to that position. Now the American economy has improved, while Germany, Japan, and Korea-the countries whose schools are often held up as models for American educators-have been mired in long recessions or plunged into serious crises. But if, as critics continue to claim, American schools have not improved and Asian schools are better, what is the relevance of the schools to economic performance?
The answer, of course, is that, although their long-term contribution may be substantial, the schools are not responsible for the fluctuating state of the economy. As the educational historian Lawrence Cremin wrote in 1990 in his thoughtful and highly readable book, Popular Education and Its Discontents:
American economic competitiveness with Japan and other nations is to a considerable degree a function of monetary, trade, and industrial policy, and of decisions made by the President and Congress, the Federal Reserve Board, and the Federal Departments of the Treasury, Commerce and Labor. Therefore, to conclude that problems of international competitiveness can be solved by educational reform, especially educational reform defined solely as school reform, is not merely utopian and millennialist, it is at best a foolish and at worst a crass effort to direct attention away from those truly responsible for doing something about competitiveness and to lay the burden instead on the schools.
The World Economic Forum in Davos, Switzerland, ranks nations for international competitiveness. In 1994 and 1995, it ranked the United States first among 25 countries; in 1996, the forum changed its formula and the United States fell to fourth. (In the rankings produced by another Swiss organization, the International Institute for Management, the U.S. stayed in first place.) Eighteen of the 25 nations ranked by the forum also participated in TIMSS. The rank-order correlation coefficient between the forum's competitiveness ranking and the TIMSS math rank is very close to zero, meaning that there is no relationship.
THE SCHOOL REFORM WE NEED
None of this means that things are fine in American education. Many schools, even the good ones, have serious problems. The current craze for charter schools reflects a widespread sense that school systems are too bureaucratic and unresponsive. High school standards are too low and could be raised without burning out the kids as happens in Asian nations.
International comparisons also have much to teach us about possible lines of improvement. A TIMSS study of curricula found, for example, that American textbooks are three times thicker than their European and Asian counterparts, leading teachers to confront children with three times as many topics. The math curriculum, as the cliché has it, is a mile wide and an inch deep. TIMSS also showed that American classrooms were interrupted about one-third of the time, whereas Japanese classrooms never suffered interruptions. Another TIMSS analysis showed that Japanese teachers were much more apt to give an elaborated explanation of a mathematics concept than American teachers were. The United States, TIMSS concluded, has one of the best-educated and most poorly trained teaching forces in the world: Many more of our teachers have advanced degrees than teachers elsewhere do, but other nations provide more internships and on- the-job training to prepare future teachers and to sustain them as professionals.
This list of differences could be extended with little effort. Even without considering the difficulties of children in poverty, there are plenty of problems to work on. But they can be worked on without the drumbeat of attacks on the schools that seem premised on the theory that "the beatings will continue until morale improves." These attacks have been accompanied in many quarters by a nostalgic effort to restore a golden age of American education. Terrel Bell, who served as Reagan's secretary of education, asks in his memoir, "How do we get back to being a nation of learners?"
Unfortunately, the era Bell refers to never existed. As Will Rogers put it, "The schools are not as good as they used to be and never were." But while the history of American education records no golden age, it reveals an astonishing accomplishment with the people whom the inscription on the Statue of Liberty calls the world's poor, huddled masses. Despite waves of immigrants and the inclusion of the minority poor, the level of educational attainment in the United States has steadily increased. Not only have secondary and higher education expanded enormously in this century, but, save for a decade between 1965 and 1975, the expansion has been accompanied by improved outcomes. We can continue to build on that achievement without false alarms about the Russians or the Japanese burying us in international competition. America can do better, and we can learn from other countries if we pay attention to what they actually do, but junking our whole system isn't the way.