This paper addresses the teachers' salary grid employed in virtually all public schools in the United States. It will show how the salary grid causes the data on teachers salaries collected by the National Education Association to underestimate the rate at which these salaries have increased in the decade of the 1990s.
Since the grid recognizes only years of service and accumulated education credits, the incentives in the grid tell teachers not to worry about student performance and to spend their time accumulating education credits and degrees of no value to student performance. Further, if you have talent in math or science, or are ambitions and want to work hard to get ahead, you can do better outside of teaching where you will be rewarded for good performance.
The existence of the grid makes it very difficult and expensive to upgrade teacher quality. The existence of a rigid grid means that most of the money spent on salary increases goes to those already on the grid and further locks these teachers into place--the very ones you want to replace. Research shows that this is exactly what happened during the 1980s when teacher salaries generally were raised by about 20 percent with no discernable improvement in teacher quality. This is also what is happening to the Teach for America program for teacher reform. With the grid in place, all piecemeal attempts to improve teaching will fail. The grid has an insidious ability to thwart reform by wringing competence out of the public schools.
The grid survives and is popular because it has advantages for teachers, unions, administrators, and school boards in the political market in which teachers' pay is determined. Taxpayers and parents, the only ones who would count if education were market based, are out of this loop.
Recently, the National Education Association issued its annual review of public school teachers' salaries (see: http://www.nea.org/nr/nr020408.html). Predictably, it is accompanied by the usual posturing that these salaries are too low and need to be increased. This year, the story is that teachers' salaries have barely kept up with inflation: "Teachers Union Berates Salaries" was the headline in my local newspaper, and the headline continued, "NEA Says Pay Barely Keeps Pace With Inflation." The not-so-subliminal message is, as always, that teachers are under-paid and stingy legislators and school boards need to be prodded by indignant voters to raise taxes and spend more money on salaries.
Before the public buys into all this, a little straightforward analysis of both the teachers' market and the salary numbers is in order. In particular, the effect of the unified salary schedule under which virtually all public school teachers are paid needs to be exposed, as it affects published salary figures, influences the general market for teachers, and attempts to reform the public school system.
The Grid and Average Salaries
Let us look at the unified salary schedule--the grid--and see how it affects published average teacher salaries.
Like most government compensation systems, the grid typically puts all teachers into a rigid salary schedule of rows and columns, with years of service along one axis and amount of education, primarily in pedagogy, along the other--bachelor's degree, master's degree, master's plus 30 education credits, doctor's degree, etc. Over time, teachers automatically move up the grid as they accumulate years of service and various college credits.
Teacher attrition is concentrated in the early and late age/experience groups. Thus, over time individual teachers enter at the bottom of the grid, and if they do not leave in the first few years of teaching, they then ride the grid upward until retirement. And with the ranks of teachers growing, as they have over the past decade, there are more teachers coming into the bottom of the grid than leaving elsewhere.
These simple dynamics have implications for the behavior of reported "average" salaries over time. These averages are computed by adding up the salaries of different teachers, in different places, on different salary grids. A key thing to keep in mind is that the number and identity of persons in these different places--the mix--changes over time. Primary grade teachers usually have lower salaries than secondary teachers, possibly due to fewer years of experience and fewer education credits. If, over time, the proportion of primary teachers in the mix is increasing--either due to faster growing primary student enrollment or the sustained drive to lower class size in the primary grades, as in California--this lowers the computed average, even though no actual individual teacher's salary falls. If the age or years of experience of teachers are falling, then this fact, in and of itself, would lower the computed average. If the number of teachers is growing, more teachers are probably coming into the mix at the lower end of the salary grids, producing lower average salaries. In short, merely changing the mix in the salary grid in these ways can make the computed average come out lower.
On the national level, this is exactly what has been going on: the number of teachers has been growing faster than the number of students, as evidenced by falling pupil/teacher ratios, and is expected to continue into the future. The lowering of average salaries is artificial; it is not real and is not because any individual teacher experienced a stagnant or lower salary. Simply put, the "average teacher" is a statistical abstraction, not a person. The vast majority of real teachers who stayed in teaching inevitably accumulated more years-of-service and professional development education. These real teachers inexorably marched up through successively higher salary grids--grids where every element, every step, gets bigger over time. During the last decade, teachers ended up with a substantially higher salary in both nominal and real terms. No teacher ever moves down the grid, and no grid's elements ever get smaller over time.
If the NEA really wanted to track teachers' salaries over time, it would take a random (or total) sample of real teachers in every state in 1990, and record their progress up through the elements of the succeeding higher grids until 2000, and see what happened to their salaries annually. If random sampling is chosen, the NEA could also take a sample of teachers in 1991 and succeeding years. The result would be a table of the actual growth of teachers' salaries every year over this decade. The NEA, of course, does not do that, which gives credence to the speculation that the results would be much different than the simple reporting of artificial "average" salaries. This is another of the dogs that don't bark at the NEA, along with the computation of teacher benefits.
There are other problems with reported teachers' salaries, most notably the failure to take account of the generous benefits given to teachers relative to employees in the private sector. While these problems are oblique to our discussion of the salary grid, the interested reader can find an excellent review in Wynne and Watters (1991).
The Incentive Effects of the Grid
Aside from the effects of the grid on reported average salaries, the grid has other, more serious, economic effects on the teachers market.
The most important of these is its effect on teacher quality. All teachers in any school district are put into the same grid no matter how well or what they teach. Dolts get paid the same as superior teachers, and PE teachers are paid the same as physics teachers. Unlike in the real market place, teachers who perform well, or have valuable talents, are not paid more, and those who perform badly, or whose talents are commonplace, never make less. Thus, we have an artificial shortage of math and science teachers, and U.S. students' performance in these subjects is about the lowest in the world. We also have a surplus of P. E. and social science teachers, who then, sadly, often end up teaching math and science.
The NEA's spin on this is to emphasize the shortages of math and science teachers, but never mention the surplus of PE, social science and elementary teachers. One of the NEA's regular non sequiturs is to use the perennial shortage of math and science teachers as a lever to raise all salaries in the grid. One could just as logically argue that the common surplus of most other teachers means that all salaries in the grid should be lowered. But you never hear that.
There is also evidence that better college graduates, as measured by SAT and ACT scores, are less likely to enter teaching and more likely to become disillusioned when trapped by the public school inflexible salary grids (see: http:// www.edweek.org/sreports/qc00/templates/chart.cfm?slug=intro n-c3.htm; Barro. 1992, p. 157).
One could not consciously design a policy with worse incentives for attracting and keeping good talent and performance.
The salary grid would have some appeal if it were true that teachers' experience and education improved student performance. This is not supported by the best research (Walsh, 2001, pp. 57; Walsh and Podgursky, 2001; Ballou and Podgursky, 1997, p. 8; 2000[a], 2000[b]; Hanushek, 1986; Wenglinsky, 2000; Ehrenberg, et. al., 2001). The bottom line is that teacher quality is almost entirely determined by: (a) teacher cognitive ability--brains, (b) experience for the first few years of teaching, and (c) at the secondary level, possession of a master's degree in one's teaching subject. Everything else can't be found on the student performance radar screen. If you measure teacher quality in terms of student performance, then certification, licensing, and in-service training are beyond detection. The education establishment, of course, hotly denies this because it demolishes the raison d'etre for a very large fraction of the public school establishment.
The net result of all this is that teachers are being told by their pay incentives not to worry about student performance, to spend their time accumulating education credits and degrees of no value to student performance, and if you have talent in math or science, or are ambitious, you can probably do better outside of teaching. Yet, the NEA defends the grid tooth and nail.
In a recent study assessing student achievement, Grissmer, et. al. (2000, pp. xxvi-xxvii) summarized the effects of the teacher salary grid:
The current system rewards experience and education--but neither seems to be strongly related to producing higher achievement. If the system could distinguish and provide higher compensation for higher-quality teachers and those who are more effective with lower-scoring students, for whom there is more leverage for raising scores, one would expect a dollar of compensation to be more effective. However, in the current system, another dollar of compensation is used to reward experience and degrees and to raise all salaries--rewarding both high- and low-quality teachers--and teachers of both low- and high-scoring students. With such a compensation system, lower effects might be expected.
How the Grid Thwarts Public School Reform
The salary grid is responsible for much of the sclerosis in the whole lower public education system, a sclerosis so bad that it makes piecemeal reform virtually impossible.
Suppose you want to improve teaching, as measured by teaching performance. The evidence is that better teachers are smarter, and smarter people always have more opportunities outside the teaching profession. Thus, if you want to attract smarter people to the profession, you are going to have to pay new teachers more. But the only way to do this within the confines of the salary grid is to raise the whole grid. This, in turn, raises everyone's salary, which, in turn, further reduces teacher attrition and turnover rates. When combined with ironclad teacher tenure, this locks in place many of the teachers you are trying to replace. Lower turnover rates also mean less demand for new hires, which in turn will eventually have a depressing effect on the supply of those training for teaching due to reduced prospects of gaining employment. And this will greatly affect the career decisions of smarter, potential teachers who have better employment opportunities outside of teaching. This then reduces the relative quality of potential new hires. Aside from the wasteful aspect of simply paying windfalls to existing teachers--to those you wanted to upgrade--the reduced turnover rates also mean that it will take a long time before new, higher quality, hires arrive in sufficient numbers to make a difference. The net result is that--even assuming the administrators want and find better quality teachers, as measured by improved student performance--a general upgrade of the faculty will take a long time and waste a lot of money paying lower quality faculty to hang on until retirement.
In addition, unfortunately, there is no evidence that those running the existing schools really want to upgrade teacher quality, as measured by student performance, even when given the chance (Strauss, 1993; Ballou and Podgursky, 1997). This is contrary to what happens in private schools, which tend to hire teachers with talents that improve student performance (Ballou and Podgursky, 1997; Hoxby, 2002).
This is not simply an academic exercise. The following analysis from Ballou and Podgursky (1997) pulls this all together in the context of what happened in public education in the 1980s.
In the decade of the 1970s, real teacher salaries fell by more than 10 percent. Then, in response to general concerns about low teachers salaries and the Nation at Risk report (1983), and aided by the prosperous economy of the 1980s, real average teacher salaries nationwide rose by about 20 percent in the 1980s. This means that whole salary grids were raised by about 20 percent on the average, restoring competitiveness between teaching salaries and opportunities elsewhere. In this environment, schools should have been able to raise teacher quality, again as measured by student achievement.
Ballou and Podgursky found that they did not, largely because of the sclerosis caused by the prevalent salary grid. Their reasoning is as follows:
Teacher salaries are not differentiated on the basis of performance. When teacher pay rises, it rises for all teachers. As a result, quit rates fall and jobs become more difficult to find. The fact that prospective teachers must invest in occupation-specific training that has no value outside public education makes them sensitive to declining job prospects. This effect is greatest for those with the most attractive options outside teaching, who incur the greatest loss if they train to become teachers and cannot find a teaching job. By contrast, persons with no professional prospects outside education will scarcely be deterred by a decline in job opportunities (1997, p. 163).
Thus, the higher salaries offered during the 1980s "had little if any discernable impact on the quality of newly recruited teachers" (ibid.). It was also found that public school administrators are no more likely to hire better candidates than those with far fewer quality credentials. In short, in the culture of the public schools there is no incentive to hire better teachers even when given the opportunity to do so. This observation is also confirmed by Strauss' (1993) analysis. I would add that this culture is a result of the fact that there is little or no competition for students in the retail public school market, thereby insulating administrators from any incentive to improve student performance by hiring more competent teachers.
In the context of the salary grid and the general lack of incentives to hire better teachers even when given the chance, the prospects for improving public school student performance by hiring better quality teachers seems hopeless.
This is further illustrated by the following statement from Jane Liebbrand, Vice President for Communications at the National Council for Accreditation of Teacher Education (NCATE), a private organization that advocates further raising the barriers into public school teaching by increasing the pedagogy requirements for certification and licensing. The Teach for America (TFA) program she refers to is aimed at recruiting top graduates from selective colleges who do not major in education but are the kind of teachers that research has found to be the best at increasing student performance. Liebbrand says:
During the past decade, TFA has placed 7,000 recruits in classrooms--but only 2,000 to 3000 of these are still in the classroom. Retention is lower than among regularly trained teachers, creating constant turnover in those schools that choose TFA recruits. . . .
However, TFA is not the answer to the teacher shortage in America. . . . The select group of top college graduates that includes TFA members will not stay for long in jobs that are at the bottom of the pay and perk scale. . . . The vast majority did not attend Ivy League schools to earn less in a year than it costs for nine months of their undergraduate education. . . .
If the salaries and working conditions of teachers were raised to a level commensurate with those of other college graduates, the profession would begin to attract more of the best and the brightest into teacher preparation programs (Liebbrand, 2002).
Ignore the usual spin about a teacher shortage--there is none--and the typical failure to mention that teachers' benefits are way above those in the public sector.
I have shown above how the grid fails to reward performance. Good teacher performance is best predicted by the brains of the teacher, not training in a college of education as NCATE advocates. Smarter teachers, trapped by the salary grid that does not reward performance, are the ones least likely to enter and most likely to leave public school teaching early.
In this context, it is not surprising that TFA recruits display higher turnover. Their experience is simply another example of what happens to good teachers who are trapped by the grid. They leave teaching and leave the lower performing teachers behind.
The response of the public school establishment and NCATE is not to do the obvious--abolish the grid, or at least make it much more flexible. The strategy is to use the problem created by the grid--the failure of the public schools to retain the best teachers--as a lever to argue for higher salaries for all. This would result in rewarding lower quality teachers and further entrenching them in public school teaching.
Another example of how the grid might thwart piecemeal reform in the public schools comes from New Jersey's alternate teacher certification program. There, after a finding that "most undergraduate education courses are not useful" and that they displace academic courses, the state reduced the required education courses to three, and "gave all candidates with degrees in academic subjects the option of taking the three courses during their initial year of employment" (Klagholz, 2002). The obvious question to ask about this good faith effort to improve public school teaching is: what is going to happen to these new teachers once they get into teaching and are trapped by the salary grid? Won't the salary grid simply chew them up and spit them out as it does other quality teachers? Aren't all other piecemeal attempts to improve the system doomed to the same fate?
Given the lack of competition for students in the public school market, Liebbrand's and NCATE's suggestion that all salaries be raised will be fruitless, as it has in the past. As I have already indicated, the evidence shows that even when better-qualified teachers are available, public school administrators do not hire them. For them, it is easier to take the safe alternative and hire a "certified" teacher. There is no incentive to improve hiring in the monopolistic public school retail market where most parents are trapped. Contrast this with private schools which do hire better teachers because their retail market requires them to compete for students by doing so.
Whether teacher turnover is good or bad depends on the filtering criteria. With the salary grid, the best teachers are filtered out, and the worst filtered in. In a market oriented world, where administrators have incentives to hire and keep the best teachers by adopting a reward system consistent with this objective, the best teachers are filtered in and the worst out. This is impossible with a rigid salary grid.
In the context of a monopolistic public school system at the retail level, the teachers' salary grid has an insidious ability to wring competence out of the public schools.
Why is the Grid Popular?
If the teachers' salary grid is economically irrational, why is it so universal and popular in public schools? The general reason is: because teacher salaries generally are the product of a political, not a market, process. In such negotiations, both sides need a focal point, and in a non-market context, it is much easier to discuss pay via a grid that applies to all. In such discussions, neither side has to bother with messy variations among individuals. The exchange process has become collectivized and homogenized to the point where individual productivity and reward have become lost. This is much different than in other managerial and professional situations--even universities--where salaries are usually decided at the department level where relevant market conditions for individual fields are taken into account.
I have mentioned above how the NEA uses the grid as a lever to raise all salaries because of the shortage of math and science teachers, ironically, caused by the grid. Another successful tactic used by teachers and their unions to raise all salaries is to argue that the first step in the grid is too low, thereby, allegedly, making it difficult to hire new teachers. Since in general there are always more teachers available than there are job openings, from a market standpoint this assertion is nonsense (Ballou and Podgursky, 1997, pp. 56-57). This is why the argument then proceeds on moral, not market, grounds. Should we really entrust our children to someone making only $25,000 or $30,000 per year? Don't teachers "deserve" more than that? Shouldn't we have more "respect" for those who teach our children? This, of course, is the never-never land of the medieval search for the "just" wage, to which there is no answer, just political power, of which the NEA has plenty.
However, as much as unions may posture on behalf of the beginning teacher, they are really using this argument as a lever to raise the whole grid, thereby benefiting all teachers. They never argue, for example, that the first step of a grid be eliminated and new teachers be hired at step two. They want to raise the first step and all succeeding ones as well. One particularly potent way to do this is to exercise political muscle at the state level by getting legislators to mandate a higher, statewide, minimum salary. This mandate then forces all school districts in the state to raise all the elements in their grids. This is just another example of how teacher unions are able to improve their well-being by exercising their political muscle at the state level and make gains that they could never get at the school district level (Peltzman, 1993, 1996).
Worried about the quality of public school teachers? This can also be used to argue for raising the salaries of all in the grid. As I have indicated above, the grid has the additional benefit of being useful in turning back any piecemeal attempt at public school reform.
In a broader sense, in the political environment in which teachers' compensation is determined, and with a monopolistic public school retail market, the existence of the grid self-generates focal points for increasing salaries generally. Unions can always find one element that is, or is claimed to be, "too low", and then use that as a reason for raising the whole grid. Salaries for entry level, math and science, and special education positions are favorite candidates for this salary ratcheting technique.
As irrational as it is from a market standpoint, the salary grid is popular because it has something for all the actors in the political pay-determination process. For teachers and unions, it gives many of them an automatic, stealth, annual salary increase no matter how they perform--no performance review, no salary negotiations, and no newspaper headlines showing increased teacher pay. As indicated, the grid also gives them a convenient lever for raising all salaries by focusing on one element of the grid and then generalizing it to the whole grid. For union leaders and school administrators, the grid is easy to administer and makes their lives easier. It gives both something to hide behind as it relieves them of the burden of having to make, justify, and adjudicate, via grievance procedures, salary differentials based on such performance factors as merit and subject taught. And for School Boards, who are playing with taxpayers' money, not their own, the grid allows them to give automatic, low-profile, salary increases without specific negotiations that might get the public's and media's attention.
Unlike in the marketplace, the taxpayers who are paying the bill, and the parents whose children are being educated, are out of this loop. This is in sharp contrast to what would happen in a market environment where parents had a choice between schools where cost and product quality determined which schools survived.
From an economic standpoint, the only general question relevant to teachers' pay is: what mix of salary, benefits, and other job dimensions are necessary to hire competent individual teachers? Since these vary across states, school districts, schools, and teachers' talents and competencies, there is no rational one-size-fits-all answer, such as the grid.
* John T. Wenders is Professor of Economics, Emeritus, at the University of Idaho. His e-mail address is: firstname.lastname@example.org.
Ballou, Dale, and Michael Podgursky. 1997. Teacher Pay and Teacher Quality. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.
Ballou, Dale, and Michael Podgursky. 2000[a]. "Reforming Teacher Preparation and Licensing: What Is the Evidence?" Teachers College Record, 9/13/00.
Ballou, Dale, and Michael Podgursky. 2000[b]. "Reforming Teacher Preparation and Licensing: Continuing the Debate." Teachers College Record, 102, 1:5-27. http://www.tcrecord.org/Content.asp?ContentID=10524
Barro, Stephen M. 1992. "Models for Projecting Teacher Supply, Demand, and Quality: An Assessment of the State of the Art." In Teacher Supply, Demand, and Quality, Erling E. Boe and Dorothy M. Gifford, eds. Washington, DC: National Academy Press.
Ehrenberg, Ronald G., Dominic J. Brewer, Adam Gamoran and J. Douglas Willms. "The Class Size Controversy." Psychological Science in the Public Interest, May 2001. Draft: February 6, 2001 (http://www.ilr.cornell.edu/depts/cheri/work.html). A more popular version titled "Does Class Size Matter?" appeared in Scientific American, Nov., 2001.
Grissmer, David W., Ann Flanagan, Jennifer Kawata, and Stephanie Williamson. 2000. Improving Student Achievement: What State NAEP Test Scores Tell Us. Santa Monica: The RAND Corporation.
Hanushek, Eric A. 1986. "The Economics of Schooling: Production and Efficiency in Public Schools." Journal of Economic Literature, 24-3 (September):1141-1177.
Hoxby, Caroline M. 2002. "Would School Choice Change the Teaching Profession?"' Draft; January 2002.
Klagholz, Leo. 2002. "No alternative." Education Next. 2 (summer):6-7.
Liebbrand, Jane. 2002. "Quick fix." Education Next. 2 (summer):6.
National Commission on Excellence in Education. 1983. A Nation at Risk: The Imperative for Educational Reform. Washington DC: Government Printing Office.
Peltzman, Sam. 1993. "The Political Economy of the Decline of American Public Education." Journal of Law and Economics, 36 (April):331-70.
Peltzman, Sam. 1996. "Political Economy of Public Education: Non-College Bound Students." Journal of Law and Economics, 39 (April):73-120.
Strauss, R. P. 1993. "Who Should Teach in Pennsylvania's Public Schools." Pittsburgh, PA: Carnegie Mellon University Center for Public Financial Management.
Walsh, Kate. 2001. Teacher Certification Reconsidered: Stumbling for Quality. Baltimore, MD: The Abell Foundation. www.abell.org
Walsh, Kate, and Michael Podgursky. 2001. Teacher Certification Reconsidered: Stumbling for Quality: A Rejoinder. Baltimore, MD: The Abell Foundation. www.abell.org
Wenglinsky, Harold. 2000. How Teaching Matters. Princeton, N.J.: Educational Testing Service and The Milken Family Trust.
Wynne, David J., and Charles W. Watters. 1991. "Teacher Compensation: How it Compares with the Private Sector." Government Union Review, Vol. 12, No. 3 (Summer):31-43.