Here’s what looks like a policy dilemma. To attain the economic growth that it desperately needs, the United States must improve its schools and train a workforce capable of competing in the global economy. Economists Eric Hanushek, Dean Jamison, Eliot Jamison, and Ludger Woessmann estimate that improving student achievement by half of one standard deviation—roughly the current difference between the United States and Finland—would increase U.S. GDP growth by about a full percentage point annually. Yet states and the federal government face severe budgetary constraints these days; how are policymakers supposed to improve student achievement while reducing school funding?

In reality, that task is far from impossible. The story of American education over the last three decades is one not of insufficient funds but of inefficient schools. Billions of new dollars have gone into the system, to little effect. Luckily, Americans are starting to recognize that we can improve schooling without paying an additional dime. In fact, by unleashing the power of educational choice, we might even save money while getting better results and helping the economy’s long-term prospects.

Over the last four decades, public education spending has increased rapidly in the United States. According to the Department of Education, public schools spent, on average, $12,922 per pupil in 2008, the most recent year for which data are available. Adjusting for inflation, that’s more than double the $6,402 per student that public schools spent in 1975.

Despite that doubling of funds, just about every measure of educational outcomes has remained stagnant since 1975, though some have finally begun to inch upward over the last few years. Student scores on the National Assessment of Educational Progress (NAEP)—the only consistently observed measure of student math and reading achievement over the period—have remained relatively flat since the mid-1970s. High school graduation rates haven’t budged much over the last 40 years, either.

Graph by Alberto Mena
Graphs by Alberto Mena

For further evidence that hiking spending produces few educational outcomes, look at how private schools compare with public ones. That $12,922, remember, is a national average; spending in urban public school systems is often far higher. The Cato Institute’s Adam Schaeffer recently calculated total expenditures per pupil for public school systems in America’s five largest metropolitan areas and Washington, D.C. Washington spent the most—an average of $28,000 per public school student, which was more than the maximum tuition charged to attend such prestigious private schools as Lowell School ($25,120), Sheridan School ($24,700), and Georgetown Visitation School ($20,600), and only slightly below the maximum tuition charged at St. Albans ($31,428), National Cathedral School ($30,700), and Georgetown Day ($29,607). Does the handsome funding of urban public schools produce results? Not according to the NAEP, which shows, for instance, that more than 25 percent of public school eighth-graders are reading below the “basic” level, compared with only 8 percent of private school students.

Obviously, it’s misleading simply to compare the performance of private and public school students without adjusting for the type of student enrolled in each sector. A student whose parents can afford to pay private school tuition is likely to score higher on standardized tests than the average public school student, regardless of the quality of the school.

But there’s another way to prove that public schools don’t get as much for their dollars as private schools do: research on school voucher programs, which pay tuition for students (usually low-income) to attend private schools. The best studies use a random design similar to what’s used in medical trials and broadly accepted as a “gold standard” methodology. They take advantage of the fact that when more students apply for a program than there are vouchers available, the program awards vouchers randomly, ensuring that the only difference between the subsequent performance of those who received vouchers and those who didn’t is whether they wound up going to the private school or a public school. A researcher can thus compare the achievement of these two sets of students and determine which setting, public or private, does a better job.

The nearly uniform finding from this research is that students benefit academically when they attend private school, rather than the public school that they would otherwise have attended. Some disagreement persists about how large the private schools’ impact is and about whether it affects all students or only those from particular backgrounds—but not even the harshest critics claim that attending a private school harms students.

Of particular interest to budget-strapped state and local governments is that the cost of the vouchers in these studies—and even the total tuition charged by the private schools, if it’s greater than the cost of the voucher—is well below what the public schools would spend to educate the same child. For instance, economist Robert Costrell found that by paying tuition to send 18,500 public school kids to private schools, Milwaukee saved taxpayers $31.9 million in 2008.

The data on charter schools are more mixed, but the general lesson is similar. Charter schools are publicly funded schools that operate essentially as their own school districts, free of the rules that bind regular public schools. Like voucher programs, charter schools usually admit students by lottery when there are more applicants than available seats. Here again, studies using the random-assignment approach have found that charter schools in New York City, Boston, and Chicago produce better educational outcomes than the local public schools that students would have attended. Further, state funding for charter schools is, at most, equal to—and usually less than—the funding for traditional public schools. The bottom line: a substantial body of research shows that at worst, students perform as well in private and charter schools as they would have in regular public schools, and at a lower cost.

Public schools are inefficient for many of the same reasons that the Department of Motor Vehicles and other government bureaucracies are. In her book Educational Economics, University of Washington researcher Marguerite Roza shows that public school inefficiencies are largely the product of burdensome regulations imposed by a top-down organizational model. School districts collect money and allocate it from a central base according to a variety of bureaucratic rules, only some of which make sense. Schools themselves have little discretion over how to use their resources.

Consider the way public schools spend money on their most important asset: teachers. According to the Department of Education, teacher salaries and benefits account for about 54 percent of public school budgets, which surely suggests that they should be structured in a way that maximizes those dollars. Instead, teacher salaries depend entirely on two criteria that, the evidence shows, bear little or no connection to a teacher’s effectiveness: years of experience and number of advanced degrees. As a result, schools must pay higher salaries to teachers who may not be more effective than teachers lacking advanced degrees or with fewer years on the job. A more efficient system, of course, would direct capital to the teachers whom the school most wants in the classroom, regardless of what their résumés look like.

In most districts, public schools aren’t even allowed to decide which teachers to employ, since tenure ensures that principals can’t remove the least effective teachers. Most collective bargaining agreements also allow more senior teachers to push their way into job openings, regardless of whether the principal thinks they’re right for the job. Nor can schools make their own decisions about whom to keep when they’re laying teachers off: either by state law or by collective bargaining agreement, most school systems require that layoffs be carried out strictly according to seniority, without any consideration of teachers’ value. Thus, when budget cuts arrive, schools not only face staff reductions; they often lose their best young teachers. And since pay is based on seniority, the schools are simultaneously dismissing their least expensive teachers.

Policies designed to solve these problems often lead to more inefficiencies. New York City eliminated seniority-based transfers in an attempt to give principals more control over who taught in their schools. The new policy created a group of teachers who, having lost their jobs in one school, could no longer use their seniority to push their way into another school. In a normal system, that wouldn’t be a problem: those teachers could simply be fired. But the collective bargaining agreement prevented the school district from doing so. Instead, the teachers entered what the city calls its Absent Teacher Reserve (ATR), members of which are paid full salaries and benefits while continuing to move up the pay scale. The nonprofit New Teacher Project calculates that the ATR costs city taxpayers about $74 million annually.

Notwithstanding public schools’ high spending and poor results, state lawmakers and courts keep pushing for even more spending. Over the last decade, high courts in several states have ruled that public school spending in certain urban systems violates state constitutional requirements to spend enough on public schools to produce “adequate” results. These rulings have been influenced by so-called adequacy studies, which use statistical models to estimate the minimum expenditure that a school district with certain characteristics—for instance, a particular percentage of students who are low-income—must incur to produce a desired academic result.

As Costrell, Eric Hanushek, and Susanna Loeb have shown, however, these studies suffer from a variety of conceptual and methodological problems. The chart below, which is reproduced from an article by the three economists in the Peabody Journal of Education, illustrates a simplified version of the adequacy approach. Each dot on the chart represents a Missouri school district, with district spending per pupil plotted on the vertical axis and test-score performance on Missouri’s state exam, the Missouri Assessment Program, on the horizontal. The line running through the middle of the chart shows the average amount that a district spends to achieve a particular outcome. For example, districts with 40 percent of their students scoring in the top two categories on the test spend an average of $7,000 per student.

Graph by Alberto Mena

Advocates of the adequacy approach point to the line as their estimate of how much a district needs to spend to achieve a particular outcome; they would say, for instance, that if a district wants 40 percent of its students to score in the top two test categories, the district should spend $7,000 per student. But that approach is wrongheaded on many levels. First, the line shows us the average amount spent to achieve a particular result, which is not the same as showing the minimum amount necessary to achieve that result—the only meaningful definition of “adequacy” in this context. Notice the dot just to the right of the 40 percent marker but also well below the line. That dot represents an actual district spending just over $5,000 per student to achieve the 40 percent mark. That district must have adequate resources to produce the desired outcome; after all, it is actually producing that outcome with those resources! The chart shows, in other words, that the adequate funding necessary to achieve the desired outcome is $5,000 per pupil, not $7,000.

For that matter, what about the districts above the expenditure line? According to the models, those districts are spending more than they should to produce their results; that is, they are inefficient. Strangely, the authors of the adequacy studies don’t argue that money should be taken away from those districts.

Also, it’s obvious from the chart that there is great disparity in the achievement of various districts spending the same amount. Just look at how widely the dots are dispersed, and you’ll realize that the relationship between spending and achievement is far from straightforward. So the line necessarily includes an enormous measurement error. Using the line to determine how many dollars a district must spend to get certain test-score results is irresponsible, to say the least.

A final flaw of the adequacy approach is that it evaluates school spending under the current system, when it is precisely that system’s structure that leads to widespread inefficiency. Perhaps public schools don’t have adequate resources to succeed under the terrible rules governing their allocation of dollars. The answer to that problem isn’t to give even more money to them; it’s to change the system and find ways to allocate dollars more productively.

Schools don’t need more funds; they need the freedom to use their funds as they see best. That can happen only if the restrictions of the current system no longer bind them. A better system—one that the United States should begin moving toward—would be a taxpayer-funded one of relatively autonomous schools. Every school would become, in effect, a charter school. Districts would still have a role in this kind of system, imposing performance standards that schools would have to meet to keep their doors open. But it would be each school’s responsibility to adopt sound policies and use its resources wisely.

Such a system of autonomous schools isn’t as far-fetched as it once seemed. In some places, the charter sector is beginning to rival the traditional public school system. For instance, about a third of all public school students in Washington, D.C., attend charter schools. Though just 3 percent of New York City’s public school kids are in charters, certain neighborhoods post better numbers—Harlem, in particular, where the fraction is about 15 percent. School voucher programs have also surged, though far less rapidly. As I wrote recently in these pages, 2011 was the “Year of the Voucher,” with legislatures in 12 states either adopting new school voucher policies or meaningfully expanding existing ones.

As more students use public dollars to attend schools outside the traditional public school sector, student achievement will probably improve, and expenditures will certainly decline. That’s an outcome that should interest lawmakers in these fiscally troubled times.


City Journal is a publication of the Manhattan Institute for Policy Research (MI), a leading free-market think tank. Are you interested in supporting the magazine? As a 501(c)(3) nonprofit, donations in support of MI and City Journal are fully tax-deductible as provided by law (EIN #13-2912529).

Further Reading

Up Next