Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

Teacher Turnover and Teacher Quality

by David Grissmer & Sheila Nataraj Kirby - 1997

Draws on recent data from surveys and research conducted by the U.S. Department of Education to sketch the outline of the approaching changes in the teacher labor market and to comment on the issue of teacher quality. Enduring teacher quality requires institutional reform in the teacher preparation and compensation system. (Source: ERIC)

Perhaps the most dramatic change that has occurred over the ten-year period since we began research involving teacher supply and demand is the improvement in the quality of the data and the associated research. In the early 1980s, it was impossible to address such basic issues as:

  • How many teachers do not continue teaching from one year to the next?
  • How do attrition rates vary by teacher characteristics or subject taught?
  • How many teachers who leave will return and when would they be likely to return?
  • What are the sources of supply of new teachers?
  • How long do new teachers stay in teaching after entry?

At that time a plausible case could be made that a general teaching shortage might occur under certain conditions if then-quoted teacher attrition rates were accurate (Darling-Hammond, 1984). However, the conditions necessary to produce general shortages have not materialized, partly because then-nationally-published attrition rates were very inaccurate since they did not distinguish between permanent and temporary attrition. However, the report warning of a potential shortage helped highlight the need for, and spurred the production, of better data and research on teacher supply and demand. What followed over the next ten years was the production of some of the best data available on any occupational group in the nation and associated research that has revealed the complex patterns inherent in the teacher labor market.

The Department of Education's Schools and Staffing Survey (SASS), together with their Baccalaureate and Beyond Longitudinal Survey and some superb state data on teachers, has supported research that addresses most of the important questions concerning the flows of teachers into, out of, and back into teaching.1 More importantly, these data can provide the basis for describing the dramatic changes approaching in the teacher labor market and lay the basis for the planning and policies necessary to navigate through these changes.

The dramatic improvement in the data produced by the Department of Education on teachers, as well as on students, student assessment, and education issues generally, is an unheralded success story-especially given the beleaguered status of the department throughout much of this period. Recent annual volumes of The Condition of Education are the more visible manifestation of a climactic shift in the quality of data and associated research. Until such a shift occurred, the historical pattern of cycles of "hit or miss" educational reform was inevitable because research could not offer definitive guidance (Sarason, 1990; Tyack & Cuban, 1995). However, better quality research in education can begin to provide tested initiatives and accurate empirical evaluation of reform so that a continual cycle of self-sustaining improvement is engendered in our educational system (Wilson & Davis, 1994).

In this article, we want to draw on this new data and research to sketch the outline of the approaching changes in the teacher labor market and then comment on the critical issue of teacher quality.


The labor market demand for entry-level teachers has undergone dramatic shifts over the last 30 years in response mainly to changing enrollments.2 During the period of the 1960s and early 1970s, the baby-boom generation increased school enrollments by 25 percent, and demand for entry-level teachers increased dramatically. Figure 8 shows that new, first-year teachers numbered about 115,000 in 1961 but increased to 190,000 by 1971. As the baby-boom generation passed out of schools and enrollments declined in the late 1970s and early 1980s, the demand for entry-level teachers plummeted and fell to about 50,000 by the early 1980s. However, increasing enrollments in the early 1990s have pushed the demand higher to about 75,000. These data illustrate how sensitive the demand for new teachers can be to changes in enrollments and other factors. While enrollments only increased and declined by about 25 percent, the demand for new teachers experienced far larger changes-from a low of 50,000 to a high of 190,000.

Enrollments in colleges of education changed dramatically in response to these changes in demand for new teachers. Figure 9 shows how the production of bachelor's level graduates from departments of education shifted over this period in an attempt to accommodate the changing demand. The number of graduates nearly doubled from 1959/1960 to 1972/1973 as demand increased, and then plummeted back to 1959/1960 levels by the mid-1980s as demand dropped. These large cyclical changes in the demand for new teachers made it difficult to plan university

[insert Figure 8 here]
You must download the PDF document to view Figure 8

resources and faculty levels, maintain high standards for admission to education programs, maintain quality teacher preparation programs, and prevent large numbers of uncertified teachers from entering the classroom. It is only by being able to anticipate such cycles in advance that these problems can be avoided.

Schools of education are going to face another dramatic increase in the demand for entry-level teachers in the next fifteen years. Although the fact that the demand will increase is easily predictable from the combination of increased teacher retirements and increased enrollments, there are currently no good estimates of either the magnitude of increase in this demand or its precise timing. In addition, several factors that can dramatically affect the timing and magnitude of the demand have not been taken into account. More complex models are needed to incorporate all of the factors and produce good estimates of the number of new teachers needed by year and the sensitivity of estimates to changes in key parameters. Lack of anticipation and planning for the approaching increased demand could result in failure of a new national agenda to improve the quality of the teaching force (National Commission on Teaching and America's Future, 1996). Such an agenda is difficult to implement in stable labor market conditions; if departments of education have to respond unexpectedly to large increases in demand, it will be much more difficult to implement this agenda to, and make the critical changes called for by, this National Commission's report.

Another recommended direction of reform is the teacher compensation system. Some analysis has suggested the structure of the current compensation system and its associated lack of differentiation by quality and/or subject

[insert Figure 9 here]
You must download the PDF document to view Figure 9

as an important issue in reform (Ballou & Podgursky, 1997; Grissmer & Kirby, 1992). Compensation reform would ideally precede the large, increased demand so that the new teachers would start their careers under a new system of compensation incentives. Perhaps more importantly, new compensation incentives might influence the quality of individuals choosing to be teachers.

The future increased demand is much more complicated than the enrollment-driven demand of the 1960-1970s. Six factors will probably change the timing and magnitude of the demand for entry-level teachers between 1998 and 2013: rising enrollments, lower pupil/teacher ratios, rising teacher attrition rates, early retirement plans, a decline in the size of the teacher "reserve pool," and the current queue of teacher entrants.

Enrollments began to rise again in the late 1980s and will be rising until approximately 2005, with small declines between 2005 and 2013. The total increase is approximately 20 percent. This rising enrollment has generated additional demand for teachers in the early 1990s and will continue to generate more demand through 2005 if pupil/teacher ratios remain at current levels. Pupil/teacher ratios also strongly drive demand for new teachers. Historically, this ratio has fallen from around 26:1 in 1960 to 17:1 in 1996. However, the ratio has remained relatively stable over the last eight years. If the historical trend for student/teacher ratios continues to fall, an additional demand for new teachers will be generated. On the other hand, rising student/teacher ratios (as occurred in some states) would lower demand for new teachers.

The third factor driving demand for entry-level teachers higher will be highly predictable increases in teacher attrition rates caused by two factors: increased retirements and increased numbers of young teachers. Teacher attrition rates are highest for retirement-eligible teachers and for young teachers-two groups that will grow disproportionately between 1998 and 2013. Figure 10 shows annual attrition for teachers by years of teaching experience.

Teacher attrition rates follow a U-shaped curve, high for teachers early in their career, very low during mid-career, and high again for retirement- eligible teachers (Grissmer & Kirby, 1987, 1992). Currently about two- thirds of the teaching force is between ages thirty and fifty-mid-career- and we have a small percentage of inexperienced teachers (as indicated by Figure 8), showing that the percentage of teachers who are teaching for the first time has remained low since 1981. This distribution of teachers with fewer young teachers and mostly mid-career teachers is the main reason the attrition level is very low for teachers. For instance, the attrition rate for teachers in Indiana fell from over 13 percent in the late 1960s to under 5 percent in the late 1980s (Grissmer & Kirby, 1992).

This picture will change dramatically over the next fifteen years as the growing group of teachers with more than twenty years of experience move into retirement eligibility (the "graying" of the teaching force) and are replaced predominately by those with no experience. Figure 11 shows the change in the experience distribution of teachers from 1987 to 1994. It shows that the group with over twenty years of experience has grown rapidly from 20 to 30 percent of the teaching force. Thus the two fastest growing parts of the teaching force will be retirement-eligible teachers and young teachers-the two highest attrition groups.

[insert Figure 10 here]
You must download the PDF document to view Figure 10

[insert Figure 11 here]
You must download the PDF document to view Figure 11

Teacher attrition rates have fallen from high levels in the late 1960s, when there were large numbers of young teachers, to very low current levels and will begin to rise again during the next fifteen years. A key factor determining how fast this rise occurs is the structure of retirement programs. The current structure of teacher retirement systems provides strong incentives to continue working until eligible for retirement (Aureimma, Cooper, & Smith, 1992; McLoone, 1987). Thus, most of the bulge of teachers currently at mid-career will stay until retirement eligibility. Initial eligibility usually occurs between fifty-five and sixty years of age, but many teachers stay until age sixty-five or beyond. Little research has been done modeling the timing of the retirement decisions of teachers-but the timing of teacher retirements is a critical variable in determining the timing of the demand for new teachers. The fourth factor driving new teacher demand is the presence of enhanced early retirement programs for teachers which can have a dramatic impact on the timing of the increasing demand for teachers. If most teachers decide to retire closer to age fifty-five rather than age 65, one component of the demand for new teachers will occur ten years earlier pushing demand for new teachers up much sooner. It is uncertain currently how many teachers will retire as early as age fifty-five when retirement eligibility usually begins or wait until sixty-five or later. Some states and school districts have offered early retirement programs in an effort to reduce the burden of the immediate and near-term costs of a highly paid teaching staff (Aureimma et al., 1992). Many states and districts will probably be implementing early retirement pro-grams for teachers to reduce educational budgets. Many of the older teachers are at the top of the salary scale and replacement with younger teachers with salary reductions of 50 percent or more is an appealing option in periods of austere budgets, which many states may be facing. Understanding the size and scope of these early retirement programs will be important in estimating the demand for new teachers.

Another equally important component of rising attrition rates is the very high attrition of younger teachers. As we replace older, retiring teachers with younger teachers, younger teachers will become a greater portion of the teaching force, and their higher attrition rates (see Figure 10) will increase overall attrition also. It will be important to accurately model attrition of younger teachers as they become a larger potion of the teaching force. Research (Grissmer & Kirby, 1992; Murnane & Olsen, 1989, 1990) has shown that in addition to the age of the teacher, early attrition can depend on teacher pay levels, pay and opportunity for competitive jobs outside teaching, the subject and level taught, and the highest level of pay in the system. Consideration of these variables in modeling early attrition will be necessary.

A fifth source of increased demand for entry-level teachers will be the decline in the reserve pool of experienced teachers. A primary source of satisfying demand for teachers has been the reserve pool of teachers, teachers who have previously taught but are not currently teaching. If more of these experienced teachers are available, fewer entry-level teachers are needed. These returning teachers were approximately 40 percent of all entering teachers in the late 1980s (Grissmer & Kirby, 1992). However, this source of teachers will be declining between 1998 and 2010. Figure 12 shows that this decline has already begun. It compares the sources of supply for 1987-1988 to 1990-1991 and divides sources of supply into four categories. Two levels of entry-level teachers are shown-those graduating in the last year and those delayed entrants who graduated from college more than one year ago. Two sources of experienced teachers are shown-those transferring from private schools and those reentering (reserve pool). These data show that entry-level teachers (new graduates +delayed entrants) have increased their proportion of supply from 39 to 54 percent.

This trend will continue due to the continued shrinking of the reserve pool. The reserve pool is shrinking because teachers usually drop out of teaching in their late twenties or thirties due to family formation or other factors, but many return within a few years. However, few teachers over forty drop out and return to teaching. Thus, the size of the reserve pool will be large if most teachers are twenty-eight to thirty-eight, but will be smaller if most teachers are older. With today's graying teacher force, the size of the reserve pool has shrunk, and a smaller portion of the demand will be satisfied from this pool -and a larger proportion of the new demand will have to come from entry-level teachers.

[insert Figure 12 here]
You must download the PDF document to view Figure 12

Finally, analysis of the path of college graduates into the labor force (Henke, Geis, & Knepper, 1996) indicates that about one-third of education graduates do not teach in the next year, and that a significant number of new teachers have majors other than education. A critical question in projecting education majors will be the extent to which a queue exists now for new teachers (a reserve pool that has never taught), and whether education departments will increase or decrease their share of new teacher entrants.

Overall, the picture is for stronger demand due to higher enrollments and perhaps lower class size, and rising attrition rates, meaning more replacements (new teachers or reserve pool teachers) are needed. However, reserve-pool teachers are declining so that more of the increased demand must come from new, inexperienced teachers. It is easy to generate scenarios that would more than double enrollments in teacher education programs at some point in the next fifteen years. But much more precision is needed to predict the timing and magnitude of the increase.

The shifts in the teaching labor force from an older to a younger force will dramatically impact education budgets. Since teacher salaries and fringe benefits are usually 50 to 66 percent of education budgets, policy-makers can look forward to significant reductions in their salary and fringe benefits as older teachers are replaced by younger teachers. Planning for this reduction and even shifting it forward through enhanced early-retirement offers are key policy issues in the coming years.

However, another important issue that needs to be addressed is the projected minority content of the teaching force over the next fifteen years. While minorities will become a much larger proportion of students in the next fifteen years, the proportion of teachers that are minority has not been increasing (see Figure 13). The proportion of black teachers fell from 8 to 7 percent between 1990-1991 and 1993-1994, while the proportion of Hispanics has risen from 3 to over 4 percent. However, these pro-portions are far below the projected 37 percent of students expected to be black or Hispanic in future years.

It will take dramatic new policies to address this issue.


As opposed to the key role of data and research in unraveling the dynamics of changes in teacher supply and demand, the issue of teacher quality is not one primarily of better data and research, but one of structural and institutional reform in the teacher preparation and compensation system. Even if research could tell us the characteristics of the best teachers, little could be done with that knowledge to improve the quality of teachers with-out major reform of teacher compensation and the institutions preparing and providing continuing education for teachers.

In any profession, the process of ensuring the quality of professionals is fairly simple. It involves setting high standards for entrance into training in the field and a fairly demanding course of study with periods of testing and apprenticeship prior to full-fledged acceptance. After acceptance into the field, a compensation system (usually market driven) usually rewards those who are of "higher quality" and thus encourages them to stay in the field, whereas those of "lower quality" are sorted out by lower levels of compensation, which discourages continuance in the field. The profession of teaching does not yet have many of these features.3

Perhaps the most important of these issues is the failure of the system to

[insert Figure 13 here]
You must download the PDF document to view Figure 13

provide differential pay for outstanding teachers and prune ineffective teachers early in their careers. Basically we would like to achieve lower attrition for better-quality teachers and higher attrition for lower-quality teachers. The higher salaries for better teachers not only encourage more of them to stay, but will bring higher-quality applicants into the profession. If we can enlarge the pool of higher-quality applicants, then higher standards for admissions into the field are possible and harder courses of study can be given without fear of higher attrition rates in school. More high-quality applicants and higher continuance rates for high-quality teachers once in the profession also allow more pruning of lower performers. So higher pay for outstanding teaching is a prerequisite for implementing many of the other needed reforms.

Currently salary is differentiated mainly by years of experience, academic degree, and type of school district. Regardless of quality, teachers within school districts are given equal increments for experience and degrees. Other than through experience and degrees, the primary way teachers pursue higher pay is through migrating to school districts that have the best combination of pay and working conditions. Thus, although pay alone is often higher in some inner-city districts, the combination of pay and working conditions often means that better teachers will migrate to school districts with good pay and better students. If quality screening occurs prior to entrance to these more preferred districts, there will be some rewards given to better teachers. However, this also means that districts with poorer students will bear the brunt of a disproportionate share of lower-quality teachers. If compensation differentials were based strictly on quality, then a more proportionate distribution of quality teachers would occur across different types of school districts.

So failure to find a fair way to differentially compensate higher-quality teachers robs the profession of an enlarged pool of high-quality entrants, the opportunity for a more rigorous course of study, a higher proportion of high-quality teachers within the profession, and a more equitable distribution of high-quality teachers across school districts. No additional data or research is likely to change this situation until the institutions involved in preparing and compensating teachers solve the structural issues of differential compensation and the associated issues of higher entrance standards, more rigorous courses of study, and pruning of unsatisfactory performers.


1 This research would include-but not be limited to-the following: Ballou & Podgursky, 1997; Grissmer & Kirby, 1987, 1992; Haagstrom, Darling-Hammond, & Grissmer, 1988; Kirby, Grissmer, & Hudson, 1991; Murnane & Olsen, 1989, 1990; Murnane & Schwinden, 1989; Murnane et al., 1991; National Center for Education Statistics, 1993; 1994; 1995; 1996.

2 Large declines in attrition of teachers, falling pupil/teacher ratios, and changes in the size of the reserve pool were also factors affecting new teacher demand between the 1960s and late 1980s (Grissmer & Kirby, 1992).

3 Ballou & Podgursky (1997) make many of these arguments. For instance, their

research shows that principals in private and public schools rate the quality of inexperienced teachers about the same, whereas public school principals give much lower evaluations to experienced teachers than do private school principals. One explanation for this is that private schools can more easily fire low performers and reward higher performers, thereby improving the quality of more experienced teachers over time.


Auriemma, F. V., Cooper, B. S. , & Smith, S. C. (1992). Graying teachers: A report on state pension systems and school district early retirement incentives. (Eugene, OR: ERIC Clearinghouse on Educational Management, University of Oregon, ED 347 620).

Ballou, D., & Podgursky, M. (1997). Teacher pay and teacher quality. Kalamazoo, MI: W. E. Upjohn Institute for Employment Research.

Darling-Hammond, L. (1984). Beyond the commission reports: The coming crisis in teaching. Santa Monica, CA: RAND.

Grissmer, D. W., & Kirby, S. (1987) Teacher attrition: The uphill climb to staff the nation's schools. Santa Monica, CA: RAND.

Grissmer, D. W., and Kirby, S., (1992). Patterns of attrition among Indiana teachers. Santa Monica, CA: RAND.

Haagstrom, G., Darling-Hammond, L., & Grissmer D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND.

Henke, R. R., Geis, S., & Knepper, P. (1996). Out of the lecture hall and into the classroom: 1992-93 college graduates and elementary/secondary school teaching (NCES 96-899). Washington, DC: U.S. Department of Education.

Kirby, S., Grissmer, D. W., & Hudson L. (1991). Entering and reentering teachers in Indiana: Source of supply. Santa Monica, CA: Rand.

McCloone, E. P. (1987). The changing conditions of teacher retirement systems: A certain past, a tenuous present, an uncertain future. In K. Alexander & D. Monk (Eds.), Attracting and compensating America's teachers. Eighth Annual Yearbook of the American Education Finance Association. New York: Ballinger.

Murnane, R. J., & Olsen, R. J. (1989). The effects of salaries and opportunity costs on length of stay in teaching: Evidence from Michigan. The Review of Economics and Statistics, pp. 347-352.

Murnane, R. J., & Olsen, R. J. (1990). The effects of salaries and opportunity costs on length of stay in teaching: Evidence from North Carolina. The Journal of Human Resources, 25 (1), 106-124.

Murnane, R. J., & Schwinden, M. (1989). Race, gender, and opportunity: Supply and demand for new teachers in North Carolina, 1975-1985, Educational Evaluation and Policy Analysis, 11(2), 93-108.

Murnane, R. J., et al., (1991). Who will teach? Policies that matter. Cambridge: Harvard University Press, National Center for Education Statistics. (1993). Teacher supply in the United States: Sources of newly hired teachers in public and private schools (NCES 93-424). Washington, DC.: U.S. Department of Education.

National Center for Education Statistics. (1994). Qualifications of the public school teacher work -force: 1988 and 1991(NCES 95-665). Washington, DC.: U.S. Department of Education.

National Center for Education Statistics. (1995). Migration and attrition of public and private schools teachers, 1995. Washington, DC: U.S. Department of Education.

National Center for Education Statistics. (1995). Teacher supply in the United States: Sources of newly hired teachers in public and private schools, 1988-1991 (NCES 95-348). Washington, DC.: U.S. Department of Education.

National Center for Education Statistics. (1996). Digest of educational statistics. Washington, DC: U.S. Department of Education.

National Commission on Teaching and America's Future. (1996). What matters most: Teaching for America's future. New York: Author.

Sarason, S. B. (1990). The predictable failure of educational reform: Can we change course before it's too late? San Francisco, CA: Jossey-Bass.

Tyack D., & Cuban L. (1995). Tinkering toward utopia: A century of public school reform. Cambridge: Harvard University Press.

Wilson, K. G., & Davis, B. (1994). Redesigning education . New York: Henry Holt.

Cite This Article as: Teachers College Record Volume 99 Number 1, 1997, p. 45-56
https://www.tcrecord.org ID Number: 10381, Date Accessed: 10/26/2021 12:59:55 AM

Purchase Reprint Rights for this article or review
Article Tools
Related Articles

Related Discussion
Post a Comment | Read All

About the Author
  • David Grissmer
    RAND Corporation

  • Sheila Kirby
    RAND Corporation

Member Center
In Print
This Month's Issue