Comparable worth: salary data fail to account for the shorter workday and work year in teaching. Once adjusted, teacher salaries look about right. (Forum).

IT IS AN ARTICLE OF FAITH AMONG MANY SUPPORTERS OF public education that teachers are underpaid. As Gayla Hudson, a former National Education Association official, once put it, "Until you start paying teachers at the level that other professions receive, recruiting will be a problem." Some respected academics have argued similarly. For instance, Peter Temin of the Massachusetts Institute of Technology (see "Low Pay, Low Quality" on page 8) claims that the opening of new job opportunities for women has drained the most promising candidates from the pool of potential teachers. In this view, higher pay is necessary to lure high-quality applicants away from more lucrative alternatives,

Most new teaching positions typically attract numerous applicants, especially in elementary schools, in subjects like English and social studies, and in the suburbs, where school districts frequently draw 100 applications for a single slot. But the mere ability to fill positions is a poor measure of whether pay is adequate. The question is whether salaries are high enough to draw applicants of the caliber parents and policymakers desire. Furthermore, the benefits of higher salaries must be weighed against the benefits of other reforms that the money could be put toward. For instance, say a school district could raise its average test scores by one point by spending an extra $1 million on teachers' salaries. But what if the school district could induce a two-point increase in scores by spending the same amount on nonprofessionals to tutor struggling students? Or what if reforms outside education, such as better prenatal care, incentive-driven changes to public assistance programs, or job training, provided mor e bang for the buck in raising student achievement? In either case, public funds would be better spent on reforms other than increasing teacher salaries.

The Evidence on Teacher Pay

A substantial body of evidence implies that teachers are not underpaid relative to other professionals. Using data on household median earnings from the U.S, Department of Labor, I compared teachers with seven other professional occupations: accountants, biological and life scientists, registered nurses, social workers, lawyers and judges, artists, and editors and reporters. Weekly pay for teachers in 2001 was about the same (within 10 percent) as for accountants, biological and life scientists, registered nurses, and editors and reporters, while teachers earned significantly more than social workers and artists. Only lawyers and judges earned significantly more than teachers--as one would expect, given that the educational training to become a lawyer is longer and more demanding.

Teachers, moreover, enjoy longer vacations and work far fewer days per year than most professional workers. Consider data from the National Compensation Survey of the Bureau of Labor Statistics, which computes hourly earnings per worker. The average hourly wage for all workers in the category" professional specialty" was $27.49 in 2000. Meanwhile, elementary-school teachers earned $28.79 per hour; secondary-school teachers earned $29.14 per hour; and special-education teachers earned $29.97 per hour. The average earnings for all three categories of teachers exceeded the average for all professional workers. Indeed, the average hourly wage for teachers even topped that of the highest-paid major category of workers, those whose jobs are described as "executive, administrative, and managerial." Teachers earned more per hour than architects, civil engineers, mechanical engineers, statisticians, biological and life scientists, atmospheric and space scientists, registered nurses, physical therapists, university-lev el foreign-language teachers, librarians, technical writers, musicians, artists, and editors and reporters. Note that a majority of these occupations requires as much or even more educational training as does K-12 reaching.

Government data on wages and salaries also exclude fringe benefits. Typically, teachers' retirement and health insurance benefits are more generous than the average professional's, particularly those who work in the private sector. Federal data suggest that, on average, teachers receive a package of benefits valued at more than 26 percent of their salaries. By contrast, the average for "all domestic industries" is about 19 percent; for private industries it is even less, below 17 percent. Take health insurance. Federal data suggest that about one-half of teachers pay nothing for single coverage (the employer pays everything), whereas the proportion of private" professional and technical" workers who pay nothing is only one-fourth. If direct hourly compensation averages perhaps 5 to 8 percent more for teachers than for all professional workers, and fringe benefits are perhaps 5 percent more, all told, teachers' average hourly compensation plus benefits exceeds the average for all professional workers by roughl y 10 to 15 percent. In addition, teachers experience more job security, rarely suffering layoffs or firings. An architect's income falls with recession-induced declines in construction spending, while the economy is currently littered with unemployed computer programmers, whose jobs disappeared when the dotcom bubble popped. Most risk-averse persons would gladly accept some decline in average annual income in order to avoid unexpected adverse shocks to their standard of living. Thus, in a "risk-adjusted" sense, teacher pay is often substantially higher than that of comparable occupations.

Shortages

Of course, these are simple averages. It is possible that the pay for some kinds of teachers is relatively low. Others may receive salaries higher than they would earn in other professions. In my view, three groups of public school teachers can make a good claim for not being typically paid above-market compensation: unusually effective teachers, teachers in some geographic areas, and teachers with specialties that are in great demand.

In almost all private-sector jobs, pay is proportional to performance--to the estimated revenue that a worker's efforts bring to the firm. This theory of how wages are determined lies at the core of standard microeconomic theory and has been verified empirically in literally hundreds of studies. Workers are paid on the basis of outcomes. In public school teaching, by contrast, teachers are usually paid on the basis of inputs-the amount of schooling and experience that they possess. Qualitative differences in teaching ability are almost always ignored. Thus it is possible that teachers of average quality may be "overpaid" in the sense that they are earning more than workers of comparable training and productivity in other professions. Likewise, highly effective teachers might be "underpaid," earning less than equally skilled workers in other professions.

Critics of "merit pay" argue that it is impossible to evaluate a teacher's ability with any precision. Thus, allowing administrators to exercise subjective judgment could introduce major inequities in pay, since they will give merit raises to their friends and punish their enemies. Yet merit pay is commonplace in universities, and professors' teaching is evaluated. In a high-quality university, salaries among professors with the same formal education and years of experience might differ by as much as 100 percent or more, and differentials of 50 percent are not uncommon within the same academic department. Even in schools of more modest repute, top-flight faculty members may make 50 percent more than those perceived as less productive. In K-12 schools, a whole variety of indicators, such as test scores, enrollment figures, and college placement statistics, could be used to evaluate teachers, yet this is seldom done, By the way, this problem has a long pedigree. Writing in 1776 in the Wealth of Nations, Adam Smith complained," The usual reward of the eminent teacher bears no proportion to that of the lawyer or physician:' He attributed this to the move away from market-based funding of education, noting that in ancient Greece, where all education was private and financed by tuition payments, great teachers were often extremely well paid relative to their less-talented counterparts.

Average teacher salaries vary enormously in most states by area, tending to be higher in suburban and urban districts and much lower in rural areas. In my state of Ohio, for example, the average teacher salary in the district with the highest pay is about double that of the district with the lowest pay, a differential that cannot be explained entirely (or even mainly) by differences in teachers' characteristics. Applicants are often abundant in high-achieving, high-salary suburbs, while school principals and human-resource managers in poor, relatively isolated rural school districts may face genuine teacher shortages. If their salaries were higher, they might be able to attract teachers from places that are more desirable to live.

Interstate variations in average salary are likewise substantial. In 2000 the average public school teacher salary reported by the National Education Association was less than $30,000 in North and South Dakota, but more than $50,000 in New York. Yet these differentials are often justified by local labor-market conditions. The average teacher made $51,800 in Connecticut, but only $35,000 in West Virginia. Yet, in a relative sense, the West Virginia teacher earned a higher salary. Teachers in West Virginia earn 30 percent more than the average workers in West Virginia, while teachers in Connecticut earn only 11 percent more than the average worker in their state. Generalizations about geographic inequalities in teacher salaries typically ignore these profound differences in labor markets.

Shortages have developed in fields like mathematics, the sciences, and special education in large part because school districts make no distinctions among teachers. According to the rigid salary schedules typically negotiated by the teacher union, those who are in high-demand specialties must earn the same as teachers in fields where there is a surplus of applicants. Thus schools that have no trouble finding English, history, or elementary teachers may have to scramble to hire special-education or science teachers. At a typical university these days, a new assistant professor in my field, economics, will earn 50 to 100 percent more than a new assistant professor of English or history. By contrast, in a typical public school district there is no differential, and new English teachers may be paid perhaps $5,000 or more than their "reservation wage" or the lowest wage for which they will work. Meanwhile, the district may have trouble finding any special-education teachers, indicating that the going rate is below the reservation wage of competent potential employees.

In short, even where there are documented instances of teacher shortages, the problem is seldom that teacher salaries overall are too low, but rather that inefficient practices ignore the realities of the market. Without changing average teacher salaries a penny, most districts could ease their recruitment problems by paying some teachers (good ones, those in high-demand specialties) more than others (less-competent ones in low-demand specialties). The problem is not the level of pay, but rather its distribution.

Public versus Private

The most compelling evidence that teachers are not on average underpaid comes from the world of private schools. The last comprehensive analysis, performed during the mid-1990s, indicated that average private school salaries were slightly less than 60 percent of average salaries in the public schools (though the gap between public and private schools narrows substantially when religious private schools are excluded from the analysis. See Michael Podgursky's article, "Fringe Benefits" on p.71 of this issue). There is no reason to believe that this differential has changed dramatically in the past few years. In other words, a public school teacher might make more than $40,000 in annual salary, while a private school teacher with similar levels of experience and education would earn $25,000.

Are public school teachers any more effective than those in private schools? On the contrary. In general, children in private schools perform better on standardized tests, even after adjusting the data to reflect differences in other factors that are relevant to learning, such as family income. The public school crowd retorts that public school teachers are certified by their state, indicating that they have taken the requisite number of courses in pedagogy, while most private school teachers, especially at the secondary-school level, lack certification. But there is little to no evidence that certification improves the quality of instruction. The absence of certification requirements is actually a virtue of the private school market, since it gives private schools access to a larger pool of qualified applicants.

Teacher salaries are set by local school boards, which compete with one another for talent. This suggests that salaries should be the going rate. So how can private schools get high. quality teachers for 40 percent less pay than the public schools can? For one thing, it is important to acknowledge that many religiously motivated teachers will sacrifice higher pay to work in a private school whose mission comports with their core beliefs. Secular private schools tend to pay more than religious schools. However, at least part of the answer lies with the powerful influence of the teacher unions. Through negotiations, political pressure, and the threat of strikes, the unions have been able to secure wages that are higher than would be dictated solely by market considerations. From this perspective, the unions' persistent quest to raise teacher salaries, carried on under the banner of improving education, looks more like a ploy to redistribute income from one group of adults--the general taxpayer--to another-teach ers.

It is probably no accident that the differential between teachers' pay and the pay of workers in general accelerated during the 1960s, the decade when unionization among teachers took hold. Some scholarly studies suggest that unionization has raised teacher salaries (and per-pupil costs) by perhaps 10 percent. If accurate, the average pay of the 3 million public school teachers may be $5,000 or so higher annually because of unionization, or $15 billion for the entire teacher corps.

It may be that the premium paid to public school teachers reflects both the influence of their unions and the more difficult environments they work in. Public school teachers enjoy far less freedom than their private school counterparts; their actions are severely circumscribed by laws, court decisions, and administrative dictates from remote bureaucracies. Unruly students cannot be expelled or adequately disciplined. Teachers often cannot use the instructional materials that they prefer. They have little or no say in the hiring of colleagues, the establishment of curricular goals, or the disciplining of students. There is a sense of frustration, a lack of trust and of teacher empowerment. Teachers in private schools work in an environment that is comparatively less bureaucratic. They enjoy school communities where collegiality is high, goals are well stated, and success is rewarded.

In other words, the salary differential between public and private school teachers may be explained partially by what is often called "combat pay" There is some evidence to support this. In surveys of private and public school teachers taken during the 1990s, public school teachers consistently showed less satisfaction with their work conditions than their private school counterparts. For example, on a 1993 survey, 21 percent of public school teachers did not agree with the statement, "The school administration's behavior toward the staff is supportive." Only 12 percent of private school teachers did not agree with the statement. Some 23 percent of public school teachers did not agree that "there is a great deal of cooperative effort among staff," compared with only 10 percent in private schools. Nearly half of the public school teachers thought that they did not receive a great deal of parental support, compared with only about 12 percent of their private school counterparts. "Student disrespect for teachers " was a serious problem for 19 percent of public school teachers, compared with only 3 percent of private school teachers.

Yet when asked about their salaries in the 1993 survey, teachers in both private and public schools reacted similarly. Despite earning far less, teachers in private schools were only slightly less satisfied with their salary than their public school counterparts. This is highly consistent with the idea that some of the premium paid to public school teachers is due to less desirable working conditions.

Teacher Pay and Education Policy

I suspect that the failure to link teacher pay to individual teachers' productivity has driven the education policy debate in unexpected ways. Consider the move to reduce class size. Teachers and their unions have been major proponents of this "reform." But would that be the case if teacher pay were tied to productivity? I doubt it.

Suppose a high school has two teachers of junior and senior classes in history, each with 100 students in five classes averaging 20 students each. Suppose both teachers make $50,000 a year, including fringe benefits. Then suppose that the less effective of the two teachers retires. Typically, the school would hire a replacement, costing perhaps $40,000 a year, because he has less experience. Now, suppose that instead the school simply asks the more effective history teacher to take on the retiring teacher's load, thereby doubling his class sizes to 40 students. The teacher would typically howl in protest. But what if the school district offered the teacher $15,000 in additional pay and fringe benefits, plus a part-time teacher's aide paid $10,000 a year to assist in grading and record keeping? The teacher would probably accept the assignment. The district would save $15,000 and probably maintain the level of instruction. The excellent teacher is likely to be better at teaching 40 students per class than an un known, less-experienced teacher would be at teaching 20 students per class. Currently, there is no premium associated with taking on larger classes, so teacher unions fight for smaller classes. But sharing some of the cost savings of larger classes with teachers may be one way of breaking down teachers' resistance to changes in the status quo. The resistance to larger classes is not just among teachers. Private schools typically have relatively small class sizes--and affluent parents seem to want them even if it adds significantly to tuition costs.

Market-based reforms would also force changes in the system of teacher compensation. Earlier I argued that public school teachers receive a pay premium for two reasons: the influence of their unions and the combat pay necessary to compensate them for their less than ideal working conditions. Market-based reforms would reduce the premiums associated with both aspects of the current public school system. Expanding the use of publicly funded vouchers and increasing the number of charter schools would reduce the power of the unions as more and more students moved to nonunionized private and charter schools. In turn, more and more teachers would move to private and charter schools, where their work environments would probably improve because of the stronger sense of school community that comes with having chosen a school. If taxpayers are paying a huge premium to public school teachers merely because they belong to powerful unions and work in unpleasant environments, then the costs of failing to reform the system are indeed substantial.

Richard Vedder is a professor of economics at Ohio University and a senior fellow at the Independent Institute.