May 2003
Are Teachers Paid Too Little . . . or Too Much?

If pressed to answer the question "are teachers overpaid or underpaid," I would respond that one-third of teachers are underpaid; one-third are paid approximately what they should be; and one-third are overpaid. If one were to look at the aggregate data on teacher salaries and compare it to aggregate data on salaries for other professional occupations, one could make a fairly decent case that teachers are paid roughly what they should be given market conditions in other disciplines. I would also add, however, that teachers are paid more than they would be if we lived in a world in which we had competitive schools operating in an environment where funding would be student- rather than school-centered.

Research on teacher salaries is a prime example of where one has to be careful in using statistics. Looking at annual data on teacher salaries shows that teachers are generally paid less than those in other professions. But in examining hourly wage figures from the Bureau of Labor Statistics, factoring in the shorter work year and generous fringe-benefits, it appears that teachers earn as much as or more than those in many other occupations. Annual measures do not account for the summer vacation or the pension/benefit plans that public school teachers enjoy. I estimate that once those are included and wages are calculated on an hourly basis, teachers earn about 10-15 percent more per hour than the average wage per hour of all professional workers in the American economy.

There is also less risk involved in a teaching job. It is unlikely that veteran teachers will lose their jobs or be forced to take a pay cut during a recession. Consequently, they also earn more on a risk-adjusted hourly wage scale when compared to other professionals. If an individual is given the opportunity to earn an average of $50,000 a year over a period of time but with a lot of volatility versus $48,000 consistently each year, the smart person would take the second option because it is less risky.

Some teachers are underpaid. It is often hard to retain teachers because many leave the profession to pursue more lucrative careers, and that is very frustrating for education leaders. Subjects such as math, science, and special education, should be able to offer more money to their new hires because schools face a constant shortage of qualified personnel in these disciplines compared to English and social studies. Every time a suburban high school advertises for a new English teacher, it probably receives fifty to one-hundred applicants, and there are probably more than one or two that are of decent quality. Those teachers in suburban districts that teach subjects that have a deep applicant pool are the ones that I would identify as overpaid. That is, if they went to get a job in an alternative sector with the same skill set, they would be forced to work for $5,000-10,000 less a year.

A telling statistic is that private school teachers earn only 60 percent of the average public school teacher’s salary and yet they arguably produce the same product. This is the case for two reasons. First, the teacher unions have been incredibly successful at extracting what economists call "economic rents." These organizations have secured higher salaries for their membership than what it takes to hire them and get them to do the job. The other reason is the substantial amount of combat pay involved in teacher salaries. Private school teachers are much more likely than public school teachers to report satisfaction with working conditions. Public school teachers complain that the parents are not supportive, greater discipline problems exist, and that they have little freedom to choose the curriculum. Because of this dissatisfaction, it becomes necessary to pay teachers more than their private school counterparts to entice them to work under apparently subsatisfactory working conditions. It is difficult to separate union power and combat pay arguments when analyzing the teacher salary problem.

The real problem is not the level of pay, but the distribution of pay. The salary schedule is shamefully rigid; merit pay is largely nonexistent; specialization differentials do not exist; and there are no incentives for cost-reducing practices. Many barriers to entering the labor force also exist, as well as a tenure system that keeps ineffective teachers around for too long. With these obstacles in place, it is impossible to get true labor market outcomes in teaching, and the profession is suffering as a result.

One way to look at the question of teacher pay is to examine whether or not there is an adequate supply of applicants given the current compensation rates. The late 1990s were terrible in terms of teacher shortages. A combination of rising enrollment, nationwide class-size reduction, and post-World War II lows in unemployment created a teacher hiring frenzy that will not be replicated. The numbers from the late 1990s show a teacher hiring rate that far exceeds the growth rate of student enrollments. This period saw high levels of teacher turnover, with districts bringing in many new young teachers-who have high turnover rates and who often left the profession for opportunities in other sectors. Even with this "teacher shortage," in the National Center for Education Statistics’ 1999-2000 Schools and Staffing Survey, 68 percent of respondents said it was easy to recruit elementary school teachers. Clearly, the high turnover rate of the late 1990s drove hiring practices until the recession hit in 2001, causing it to decline sharply and create a surplus. Clearly, even in a time of shortage, the supply of teacher labor was adequate, and we are now entering a period where there is going to be a surplus of potential teachers.

The other way to determine the adequacy of teacher pay is to compare it to other professions. The vast majority of teachers work a 190 days each year, with 180 days of classroom time and ten days of professional development, parent-teacher conferences, etc. The required workday on site is also relatively short compared to other professions. The survey data shows and average workweek of thirty-eight hours for teachers. If you combine a shorter work-year and a shorter workweek, it is my estimate that teachers are working an average of 30 percent fewer hours per year. It is important to factor this difference into any comparison of teacher pay with other sectors. Looking at the Bureau of Labor Statistics hourly wage numbers, which take into account the shorter work-year for educators, teachers come in consistently higher than registered nurses and even slightly higher than executives, administrators, and managers in the private sector.

It is not just the quantity of hours that makes teaching an attractive option, however; there is also a lot of flexibility in the hours that teachers work. It is much easier to take time off for personal reasons in teaching than in other occupations. The Schools and Staffing Survey reports an absentee rate of 5.2 percent for teachers, which translates to 9.4 days in a 180-day school year, a high level of personal leave. Teaching is a very family-friendly profession, and examining fertility rates for female college graduates shows that teachers, on average, have more kids than non-teachers.

In addition to salaries and flexible work schedules, the fringe benefits that teachers enjoy are considerable. The vast majority of teachers are enrolled in defined benefit programs, which means they do not have to contribute to their retirement fund, but will receive a pension of almost full pay upon retirement if they stay a certain number of years. This is a backloaded benefit, and older teachers benefit from the high turnover that occurs among younger teachers. The health insurance packages that teachers enjoy are also, on average, more generous than in the private sector.

In short, teacher compensation compares favorably to other professions. Particularly interesting is how public school teachers compare to nonsectarian private school teachers. Because a lot of private schools have a religious orientation, many teachers that choose to teach at them do so because of their faith and they arguably produce a different product than the public schools. To combat this criticism, I eliminated all sectarian private schools from the data set and examined how nonsectarian private school teacher pay compares to public school teacher pay. If one looks at average nonsectarian private school teacher pay as a percentage of average public school teacher pay, first year private school teachers earn about 78 percent of what beginning public school teachers make. Private schools narrow the gap after about twelve years of experience and then fall off again. One argument against this comparison is that private schools self-select; that is, they get easier children to deal with; they do not have to spend as much time on discipline; and they enjoy parental involvement. To account for this, I looked at only suburban school districts with low poverty rates and compared them to nonsectarian private schools, and I still found that private school teachers make about 76 percent of the public wage at the start, rise to 87 percent after a decade of experience, and then decline again. If private schools are taken as the benchmark, it seems that public school pay is adequate compared to schools that are producing the same product.

Some critics still insist that we are comparing apples to oranges. To counter this critique, I looked at trends in relative pay. For men that choose to enter the teaching profession, the general trend line for relative pay is flat. For female college graduates, the trend in relative teacher pay has admittedly been downward. This should not, however, lead us to believe that teacher pay for female graduates is inadequate; in 1960, 58 percent of women graduates that were not teachers were secretaries, and only 13 percent were in managerial positions. By 2000, these numbers had reversed themselves and women were making more money because they had jobs with greater responsibilities and longer hours. Therefore, while the pay of female teachers declined relative to all other women, it is largely because of a demographic shift and not because of stagnant teacher wages.

The bottom line is that we need more data on teacher salaries. The problem is that much of the debate over teacher pay is framed by the teacher unions. We need arm’s length data on teacher compensation and benefits that is produced by a neutral third party, as the Bureau of Labor Statistics does for the public and private sectors.

One must look at many different comparisons when examining wages. The problem with the Bureau of Labor Statistics hourly data is that it assumes that full-time workers work every weekday in the year. For workers that work all year round, the number of days they work includes all the vacation days and paid holidays, which are subtracted later as benefits. It is important to note that the 190-day teacher work year does not include any holidays or paid vacation.

Teachers also have a very long workweek, though some of it is not included in the average thirty-seven hours that are required on-site. This notion of required time, which increased 14 percent between 1994 and 2000 from thirty-three hours to thirty-seven hours for teachers, is not exactly accurate. There is a lot of work that goes on outside of the required time and regardless of the method of collection. It is difficult to get an accurate reading on how many additional hours teachers work off-site. Our estimates put the teacher workweek at over fifty hours. Only a few occupations total that many hours.

Both articles in Education Next attacked the union data and although some of the criticism was apt, some of it was unfounded. We emphasize trends over time, and from 1992 to 2001, pay for accountants increased 42 percent, pay for computer analysts 33 percent, and pay for engineers increased 40 percent compared to teacher pay, which only increased 22 percent. Granted, this is the most drastic time period to compare salaries, a period of economic boom, but the upward trend was obviously less evident in teacher salary rates.

We also compare teacher salaries to that of the average worker, and we receive a lot of grief for that because the average worker does not have a college degree. But the measure has been in there for two decades, and we find that in the year 2001, teachers made 8 percent more than the average worker, down from 18 percent in 1992. In terms of per-capita GDP, in 1992 teachers made 38 percent more than that figure, but by 2001 that had declined to 21 percent more.

As far as benefits go, the percentage of compensation that goes to benefits for teachers is 24 percent, which makes them the lowest among all occupational groups and below the professional average of 27.9 percent. With regard to sick days and personal leave, it is important to realize that arranging for a substitute is a definite annoyance. Nothing gets accomplished under a substitute; the material has to be revisited the next day; and extra time must be spent laying out the teaching instructions the day before. These difficulties do not facilitate excessive personal leave.

The comparison between public schools and private schools is problematic. Comparing private schooling to public schooling is not like comparing other private sector occupations to public sector occupations because there is no real for-profit private schooling sector in this country. In reality the private schooling sector is characterized by private nonprofit funding for education, which is unlike private sector for-profit corporations. Because private schools do not behave exactly like for-profit private sector entities, the use of them as a benchmark of the private sector is not viable.

The union is not as opposed to alternative pay arrangements as some may be lead to believe. Salary schedules are important to the American Federation of Teachers, but they exist in the states that do not have unions and they existed before the unions ever came along. With regard to merit pay, the idea of extra pay for extra work is a much better way to turn teachers on to performance based incentives than any concept that uses evaluation techniques that may not be fair or accurate.

Even I would admit that today the teacher shortage appears to be over and that economic conditions have crushed the demand for teachers, but the shortage was also caused by a nonmarket force: conservative state legislatures. In addition to the conditions that Mr. Podgursky discussed, the fact that in the early 1990s many conservative state legislatures said they would only increase teacher salaries at the rate of inflation did not help the teacher supply. I believe these policies are the root of the shortage which exploded in the late 1990s.