It’s a paradox–a mystery–an unpleasant fact in the center of our lives that we manage not to think about much: The Great American Education Machine is broken. We’ve been shoveling money into the schools’ locomotive firebox for more than a generation, but our train keeps falling behind.

Don’t take my word for it. Mild-mannered Herbert Walberg, a research professor of education at the University of Illinois at Chicago, has the figures cold:

“During the half-century between 1940 and 1990, inflation-adjusted per student spending rose more than 500 percent, from $878 to $5,292. The U.S. currently ranks first in the world among major industrialized countries in per student spending. Yet for several decades U.S. students usually have ranked near the bottom of surveys of student achievement in industrialized countries; on the 1990 International Assessment of Education Progress, for example, American 13-year-olds outscored only students from Jordan, Brazil, Mozambique, and Portugal in mathematics.”

That’s from a research report published by the Palatine-based Heartland Institute anad unveiled last November. In it, Walberg compared 37 states’ average scores on the eighth grade math test given as part of the National Assessment of Educational Progress (NAEP). He found that students from states with smaller school districts, smaller schools, and smaller state shares of school funding scored significantly higher on the test, even after he took account of other factors like minority population and per-pupil expenditures.

His point? For the last 50 years educators have been beavering away in exactly the opposite direction. They convinced us that bigger schools and centralized education funding would be good for kids. It’s beginning to look like they were wrong. Maybe their “success” is the reason more money has bought such mediocre results. Walberg, an educational psychologist with impeccable credentials, is using hard-core research results to poke the school establishment with a sharp stick.

“The thing I remember about Herb Walberg,” says Mark Musick, chair of the Southern Regional Education Board in Atlanta, “is several years ago when he and I were on the [NAEP] board. We were holding a hearing in Washington, DC. . . . The first person in line to testify began, ‘Dr. Walberg, it certainly is a pleasure to meet you after years of reading your work.’ The next person said the same thing. So did the one after that–basically came up and kissed his ring.

“I said, ‘Hey, what’s going on?’

“‘Mark,’ he replied, ‘I’m very well known in a very small circle.'”

Walberg’s 35-page resume lists 300 journal articles, 69 book chapters, and 49 books he has published. Most he wrote with other authors. “I treasure colleagueship,” he says. “You know what they say–if two people agree, one of them is superfluous.” Most of his writings deal with the causes, effects, and measurement of learning; institutional and personal productivity; and international comparisons of achievement.

Until fairly recently, Walberg was the kind of Chicago professor who is better known in the District of Columbia thaan on Michigan Avenue. Now he’s starting to pop up locally as a kind of all-purpose educational heretic. He advocates school choice–allowing students to choose between public and private schools (“the essential reform”). He likes local control, but doesn’t think that Chicago school reform goes far enough in that direction, or that it will improve most kids’ education. And–heresy of heresies–he is pretty sure that the last thing Illinois schools need is more money from the state.

These views have made him better known in a wider circle, but not necessarily better liked. School-reform advocate Don Moore of Designs for Change is blunt: “He has a very good reputation for his research nationally. In my view, he’s hiding behind that reputation to put out shoddy information about Chicago school reform and vouchers.”

“I come from families of schoolteachers,” says Walberg. “My mother had ten sisters and one brother. Most of them were schoolteachers in Indiana.” His own family moved to the south side of Chicago before he was born in 1937, and he grew up there. “It felt like I was always in the shadow of the University of Chicago,” but he didn’t arrive there easily. “My father died and I had to work a lot.” Through high school and college and graduate school he worked washing dishes, driving a tractor for the Illinois Central, playing guitar in clubs and coffeehouses, and substitute teaching. He studied education and psychology at Chicago State, got a master’s from the University of Illinois in counseling and guidance, and finally entered the University of Chicago. There he studied under Bruno Bettelheim, Benjamin Bloom, Allison Davis, and others, graduating in 1964 with a PhD in educational psychology.

Almost as useful as the doctorate was his experience on the cutting-edge computers of the day, vacuum tube mainframe monsters that intimidated many of his colleagues. To this day, on authorial teams Walberg tends to be the one who knows how the mathematics works. On the NAEP board, says Musick, “Herb had more expertise in statistics and evaluation than anyone else.” Bernard Gifford, formerly vice president of education at Apple Computer, once called him “the scorekeeper [of] the learning society.”

Walberg remembers the University of Chicago as “very rigorous and very stimulating. I found found Harvard tame by comparison.” After three years there as an assistant professor of education, Walberg joined UIC’s faculty in 1970. In 1966 the journal Psychology in the Schools published his first article, “Changes in the Self-Concept During Teacher Training.”

That title pretty well indicates the Herbert Walberg that Chester Finn met 15 years ago. Finn, an assistant U.S. secretary of education under William Bennett calls Walberg “a scholar’s scholar, a low-visibility academic” writing on topics that could only interest another educational psychologist. But the research tool he was incubating would change that.

“In education,” explains Finn, “people who want to prove a point seek out studies that back them up. The total number of studies in the professional literature is humongous, and no one of them is conclusive. Herb is one of the inventors and masters of an obscure research technique–meta-analysis or synthesis–that gives us a little light in this fog.” According to the 1994 edition of Educational Psychology, a textbook for teachers, “Walberg has shown the field how to review 2,000 and 3,000 studies and end up with meaningful conclusions. . . . One of the continuing complaints about educational psychologists in general is that if you know the theoretical framework of the researchers, you can safely predict their conclusions without even looking at the data. Walberg, however, stands apart from any specific tradition and lets the actual analysis of the findings determine [the results].”

Walberg makes meta-analysis sound easy. Search the literature for published, peer-reviewed studies–on the value of homework, for instance. Then take a vote count: how many of them found that assigning homework got good results?

“But we don’t use only vote counts. We also calculate the magnitude of the effect, calibrating all the studies to a common scale. Then we can say, not just that most studies of homework show a positive effect, but give an index of how much or how often it helped.” (Let’s put this mathematically: For each individual study he subtracts the mean of the experimental group from the mean of the control group, then divides the result by the standard deviation of the control group, giving a figure comparable across all studies. The means do not have to be standardized-test scores; they can be responses on a psychological attitude test, or teacher assessments of student interactions–anything numerical is grist for the meta-analytical mill.)

If you do this for different teaching methods, for homework, for different class sizes, etcetera, then over time, says Walberg, you get “a bunch of indices for different things that have been studied, and you can line them up in order of effect. That allows policymaakers to choose what approaches to try without having to read thousands of individual studies. I don’t mean they sould choose on this basis alone. Some things are very costly, or inappropriate. We’re trying to reduce this to facts, not opinions.”

You can see why Finn called him a “scholar’s scholar.” Walberg’s daily work is as disant from first-grade classroom as a professor of metallurgy is from the foundry at Inland Steel. Behind the millions of teachers are thousands of researchers who conceive hypotheses and test them, visiting classrooms and testing or talking to students. Behind the researchers are Walberg and a few dozen colleagues collating the researchers’ work and distilling it down to indices–the news of which may eventually filter back to classrooms.

Having synthesized research on particular factors of education, Walberg began to apaproach the Holy Grail. How do students learn? What factors go into it? Educational makes a difference, and what is window dressing? In May 1984, Educational Leadership published Walberg’s most-reprinted article, “Improving the Productivity of America’s Schools,” which pulled together the results of 3,000 different studies to try to answer these questions.

As the title suggests, Walberg treats schooling as a matter of science, not sentiment. He sees education as a business–and, in the case of the U.S., not a healthy one. “By measurable standards, U.S. educational productivity has not kept up even with that of U.S. smokestack industries such as steel, automobiles, and consumer electronics–which themselves are declining as world-class competitors in quality and costs. . . . If education proceeds by fads rather than cumulative research, it will fail to make the great advances in productivity that have characterized agriculture and industry in theis century.”

These are not throwaway lines. They are, for instance, anathema to believers in progressive education. They are anathema to those who believe teaching and learning are too complex and too unique to be explained by Walberg’s indices. And they are anathema to bureaucrats, who would rather be judged on money spent or time put in than on what they accomplish. For those in the know, in these lines Walberg has signified that his outlook is simultaneously conservative, scientific, and antiestablishment.

After this overture, Walberg enumerates nine key factors in learning, some of them obviously easier to change than others: ability, age, motivation, time spent learning, quality of instruction, home, classroom social group, out-of-school peer group, and use of out-of-school time. Each of these factors he breaks down in turn. “Quality of instruction,” for instance, includes 26 specific instructional methods. The most effective ones turned out to be (1) rewarding correct performance (“Skinnerian reinforcement”), (2) accelerating advanced students, (3) training students to adjust reading speed and techniques to different purposes (skimming, et cetera), and (4) giving cues and corrective feedback.

Sounds pretty obvious, right? Wrong. For one thing, class size and homogeneous grouping are much more commonly stressed by educators than these four methods, even though Walberg’s synthesis shows them to be one-tenth as effective. For another, teachers often don’t practice the methods: University of Michigan psychologist Harold Stevenson was surprised at how rarely Chicago teachers gave cues and corrective feedback to their students: “In nearly half of the 160 class periods we observed in the Chicago fifth grades,” he worte in Scientific American “teachers failed to offer any type of evaluation as the children worked alone at their desks. In striking contrast, lack of such acknowledgment was practically never observed in Sendai [Japan] and only infrequently in Taipei [Taiwan].”

Just reading all this in a newspaper can be misleading. By journalistic convention, these are all things that “Herb Walberg says.” But these are not his opinions (in fact, some of them may conflict with his opinions). Most are not even his own research findings. They are the collective harvest of decades of educational research across the board. They are necessarily imprecise generalizations standing on research that may even be mistaken or misconceived. But they do not stand just on Walberg’s say-so.

Three of the top findings that emerge from this mega-synthesis:

Poverty is not destiny. The effect of socioeconomic status on learning is only one-quarter as great as that of each of the four top instructional methods, and only one-third as great as that of regularly assigned and graded homework. Any number of dedicated inner-city teachers could say the same thing; these results confirm that they are not relying on wishful thinking or individual charisma.

Money spent to reduce average class size is probably better spent elsewhere. Class size has little effect on learning, showing less than one-tenth of the effect that each of the four top instructional methods has. Personalized instruction, by contrast, had six times the effect of class size. Walberg suspects that smaller classes don’t help students because most teachers teach the same way, whether they have 100 students or 15.

Efforts to restructure and reorganize schools are largely wasted. In 1984 Walberg found that “class size, financial expenditures per student, and private governance (independent or sectarian incontrast to public control of schools) correlate only weakly with learning, especially if the initial abilities of students are considered.” This finding may not square with Walberg’s own support for the reorganizational scheme known as school choice. But this winter, in another synthesizing article Walberg wrote with Margaret Wang and Geneva Haertel of the National Center on Education in the Inner Cities, the cumulative research results had only become stronger.

“Fifty years of research contradict educators’ current reliance on school restructuring and organizational variables as key components of school reform,” they wrote. “Unless reorganization and restructuring strongly affect the direct determinants of learning, they offer little hope of substantial improvement. Changing policies is unlikely to change practices in classrooms and homes, where learning actually takes place.”

Let’s see: this is a guy who calls education an “industry,” is friendly with Reagan-Bush administration alumni, and is hard-nosed about competing with other countries. It would be easy to peg Walberg as an ivory-tower conservative–someone who sees schools as factories and students as widgets–and leave it at that.

But that would be a mistake. For one thing, he frequently acknowledges the intangibles in education. In a recent article by Walberg and Chicago public school teacher Richard Niemiec, they noted that the U.S. Department of Education’s current practice of picking exemplary school programs without even visiting them is likely to rub teachers the wrong way. Teachers, they wrote, “may not prize social science methodology as highly as academics. Their judgments are more intuitive; their ways of knowing are different.”

Walberg also went out of his way in that 1984 synthesis to include data rescuing “open education” from the dustbin of history. It turns out that this early-70s fad didn’t hurt student achievement. In fact, open-schooled students were slightly to moderately better on outcomes like curiosity, cooperation, and critical thinking–which parents value more highly than test scores. No conservative ideologue would have given journal space to a progressive-education idea at the bottom of the boom-and-bust cycle of educational popularity.

Even Walberg’s most typically conservative view–that spending more money on schools is unlikely to do much good–unexpectedly crosses ideological lines. First, it is based on research, both his own syntheses and those of University of Rochester economist Eric Hanushek. The correlation between per-pupil spending and achievement just isn’t there. Second, and oddly, Walberg’s view stems from an idea fashionable in progressive educational circles these days–decentralization.

Walberg’s November 1993 study showed that smaller schools with less state funding tend to do better, but it didn’t explain why. Walberg suspects (this goes beyond the data) that a school gains more from local control than it does from a boost in state funding, with the outsider control that inevitably follows. (He does not shrink from the draconian implications: when asked at a news conference if the results of his study meant that Illinois schools might benefit from a reduction in state funding, he agreed that they might.)

Why is this research on school spending so little known? “Extremely powerful lobbies stand to benefit from increased education spending,” Walberg replies. “But somebody’s got to look at the facts. It’s similar to the health field. We spend more than other advanced countries and yet we’re near the bottom among them in infant mortality and life expectancy. It’s very important for researchers to challenge the health and education establishments.”

That year, 1984, was a good time to be summing up education research. The Nation at Risk report had come out the year before, and the general public was still fussing and fuming about the schools, much as in 1957 after the Soviets launched Sputnik. (A public brouhaha over education is like war–we get one every generaltion or so whether we need it or not.) Walberg and his colleagues had reason to hoep that their results would be appleid pronto.

Gradually they have discovered otherwise. More than ten years later, even the simplest changes that research suggests have not been made. Take Walberg’s fourth factor “instructional time.” But while U.S. schools reorganize themselves like musical chairs, they have yet to significantly lengthen either the school day or the school year, which Walberg notes remain the shortest in the industrialized world.

The reason for inaction might be money, or vested interests, or inertia, or even willful ignorance. There are influential professionals with no interest in research that violates progressive-education dogma. In his 1992 book Schools That Work, Ohio University education professor George Wood flatly asserts (without even the fig leaf of a footnote) that “there is no evidence” more time in class or more homework will increase learning.

At any rate, the passage of time has made Walberg less patient, and more willing to write up research results in more popular publications. “We have a lot of research that shows what works, but not much has been done to implement it. . . . A lot of what I do wouldn’t be necessary if educators were more responsive.”

Middle age has if anything made him less tolerant of the way educators evade accountability by talking about processes instead of about results. And that has everything to do with what he has called Chicago’s “failed reforms.”

Walberg, who attended Bowen High School, has taken a professional interest in the Chicago public schools, but usually from a certain distance. In 1975 he published an article with Jeanne Sigler of Chicago United on how the schools were failing to meet the needs of local businesses. “The public schools,” they wrote, “appear to be ineffective in preparing an increasingly large portion of students for careers, and . . . the administrative and teaching staffs are not held accountable for the quality of their product, the students.”

In 1988, during the months of agitation and organization that led up to the Chicago school reform law, he collaborated with Michael Bakalis (then dean of Loyola’s College of Education), Joseph Bast (executive director of the free-market-oriented Heartland Institute, on whose governing board Walberg serves), and Stephen Baer (then executive director of the conservative United Republican Fund) on a paperback book, We Can Save Our Children! The book supported the Chicagoans United to Reform Education plan (which was similar to the decentralizing reforms that became law) and in addition proposed a variant of school choice they called an “education rebate,” which was intended to maake private schools more competitive with public schools and force the latter to shake off their bureaucratic chains.

Walberg’s hopes for school reform in Chicago were always connected to some more radical proposal that the partial decentralization of power to elected Local School Councils. Last September he and Bast prepared a Heartland policy study entitled School Choice: The Essential Reform. They surveyed existing choice programs in Milwaukee, Minnesota, and Vermont: “Schools that must compete for the loyalty of students and parents seem to find ways to organize themselves for success.” (They also noted, with the evenhandedness characteristic of Walberg the researcher, that even though Catholic-school students learn more than public-school students of similar background, “the difference may not be enough to justify changing the organization of schools.”)

Nobody paid much attention to this sedate research. But a bit of hell broke loose when Walberg and Niemiec, under the auspices of the Illinois Educational Choice Coalition, on September 15 publicly listed the bottom-line results of school reform to date: standardized-test results down slightly; attendance, graduation, and dropout rates unchanged; Chicago at or near the bottom of the 47 largest U.S. school districts on these measures. “Three years into restructuring,” they wrote caustically, “the Chicago reforms were accomplishing little, but the new players–principals, teachers, and council members–thought well of their own work.” (A slightly toned-down version of this assessment was published as the last chapter of a 1993 anthology Walberg and Niemiec edited, Evaluating Chicago School Reform.)

Two prominent “downtown” school-reform groups replied that achievement-test numbers do not yet tell the whole story, and that Walberg had some of them worng anyway. On the moderate side, John Easton and G. Alfed Hess Jr., of the Chicago Panel on Public School Policy and Finance called Walberg’s critique “simplistic,” and “misinformation” that “appears to be an attempt to further the Choice Coalition’s provoucher agenda rather than to provide any real insight into the progress of school reform.” Donald Moore of Designs for Change was less tactful. He said the paper was a “clumsy propaganda effort” and a “bogus study.”

In substance, they hit Walberg high and they hit him low: his report misquoted some Chicago test scores, compared noncomparable big-city districts, and didn’t pay enough attention to the “most important study of Chicago reform ever conducted”: the Consortium of Chicago School Research’s July 1993 report, A View From the Elementary Schools. Easton and Hess said that it is too early to look at test scores–and even if it wasn’t, they would be misleading anyhow, because a different form of the Iowa test is used every year, and beause now it is also being taken by bilingual students and students who in other years would have flunked a grade. (One reform goal was to “retain” fewer students than in the past.) Meanwhile, Moore has lately argued that test scores are showing improvements.

Walberg responded instantly and with evident anger (one participant later described the exchange as a “shouting match on paper”). Describing the reformers as “status quo advocates,” he questioned their ability to judge Chicago school reform at all. “The agencies are themselves advocacy groups,” he wrote in a five-page polemic he sent right back to them. “And the authors themselves are hardly independent scholars. They are beneficiaries of reform spending; their incomes depend on contributions by foundations and public agencies to defend and promote current failed reforms.”

Yes, he wrote, he had read the Consortium’s July report. It found that about one-third of the city’s low-achieving elementary schools had developed “strong democratic participation” and were planning in ways that seemed likely to lead eventually to improved achievement. According to Walberg, such a finding was beside the point: “In research, it is far better to measure real effects than presumed causes of effects.”

And why, he wondered sarcastically, is it so terrible to try to measure the effects of school reform? How long would you wait if your kids’ education was at stake? “The advocates say we are premature in evaluating reforms. They say we should wait five years. Five more years of inferior education for inner-city children may not try the advocates’ patience (they are after all financially benefitting from reform spending). But we think that neigher the legislature or citizens should wait any longer, given clear failure of the reforms to meet reasonable legislative goals (even aside from the twenty-year saga of strikes, budget deficit, and catastrophic mismanagement that is highly visible to the rest of the nation, if not to the Board of Education and status quo advocates. The advocates themselves point out that even in the rosiest case, only one-tird of the low-achieving elementary schools are making systematic improvements. If we wait five years for each third of the schools, it might keep the advocates in a lucrative business for fifteen more years. Then, they may feel ready for an evaluation of progress. But what about children now in school?” This was enough for the dean of UIC’s College of Education, who intervened, asking Hess not to respond. Hess, disappointed in Walberg’s ad hominem tone, had already decided not to.

Walberg has the backing of the 1988 state law, which does call for every Chicago school to reach national norms this year. But is that goal “reasonable,” as he says? It is not so clear that he has the backing of educational research. The Consortium report says the legislature’s deadline “is simply not defensible . . . as a realistic timetable for renewing a major urban school system with over 400,000 students and 25,000 teachers. . . . To the best of our knowledge, no similar aims have been attained anywhere, by anyone. Against this standard, virtually any plan for a major urban school system reform would likely fail. . . . Even the best cases of significant individual school improvement often required five or more years to culminate in a comprehensive restructuring, and typically did not show measurable change in student achievement until the later phases.”

Walberg has few regrets about the exchange. “It was somewhat personal. But I’d done those guys a lot of favors over the years, and now all of a sudden they were challenging my authority to say anything.” He doesn’t claim detailed knowledge of city schools: “Moore knows this local stuff better than I do. I dabble in it when something needs to be said.”

But on the substance, he is not even close to backing down. Is it too early to evaluate reform, as Easton and Hess contend? No way, he insists. “People who run programs don’t like evaluation. But taxpayers have a right to know. Corporations don’t wait five years. They have quarterly financial reports and watch these things very carefully.”

Walberg is not very sympathetic about the grass-roots porblems that would dog any reform. For instance, what about the problem of student mobility? Is it fair to evaluate a Chicago elementary school based on test scores, when (on average) three-eighths of the students in a given year just transferred in? Well, yes: “I can see why a principal would prefer to test only the more stable kids. That’s guaranteed to raise scores. But to evaluate the whole system we need them all.” The suspicion lingers that he may be extra censorious of current Chicago school reform because he believes school choice offers better; but Walberg says he would judge a school-choice program on the same fast timetable.

When Superintendent Argie Johnson announced in mid-February that test scores at the vast majority of city elementary schools had stagnated or declined, Walberg took it as vindicating his skepticla position. He fired off a letter commending her “rare candor” to the Tribune: “Others, who remain a part of the Pershing Road Bureaucracy or outside groups such as Designs for Change that are profiting from the reforms, want to paint a far rosier picture. But as Aristotle warned, consider the source.”

When I mentioned to widely published education scholar Diane (The Schools We Deserve) Ravitch that Walberg had been involved in a heated local controversy, she was astonished. “In my 20-plus years in education, I’ve never known him to be involved in a controversy.

“He’s very thorough and respected for being very balanced. He is not one of the combatants in the culture wars.”

Maybe he just joined up.

Art accompanying story in printed newspaper (not available in this archive): photo/Robert Drea.