• Introduction
• The Newly Hired Teachers Model
• Data Sources and Assumptions
• Results for Public Schools
• Results for Private Schools
• Comparison of Results With Bureau of Labor Statistics Projections
• Conclusions
• References

An increased need for newly hired teachers is expected over the next decade. Depending on the assumptions made, for example, this report projects that from 1.7 million to 2.7 million newly hired public school teachers will be needed by 2008-09. The report examines a model for projecting the need for newly hired teachers in both public and private schools and discusses results based on the model.

Each year, over 150,000 public school teachers are hired to meet the ongoing demands of replacing teachers who retire or who have left the profession, filling new positions in growing school districts, or addressing special needs or meeting new requirements (table A). In addition to these extensive ongoing demands for additions to the teaching force, many schools and school districts have faced the prospect of a wave of retirements as the large numbers of teachers hired during the baby boom enrollment years approach retirement age.

As a group, elementary and secondary teachers are significantly older than the general labor force. The median age of public school teachers in 1993-94 was 44, compared with a median age of 38 for all workers in October 1993 (Bureau of the Census 1993). The burden of replacing large numbers of retiring teachers comes at a particularly challenging time, as enrollments in elementary and secondary schools are projected to set records each year well into the next decade (Gerald and Hussar 1998). Over the next 10 years, an unusually large need for newly hired teachers is expected, both to replace teachers as they retire and to meet the needs of increasing enrollments. These newly hired teachers will include both people who are new to the profession and those who are returning to teaching after some time away from the profession.

Table A.—Full-time-equivalent (FTE) teachers, newly hired FTE teachers, and the percentage of FTE teachers that are newly hired, by control: 1988-89, 1991-92, and 1994-95

Using an algebraic model based on teacher demographic data, this report examines the need for newly hired teachers for the period from 1998-99 to 2008-09. The model is used to predict the impact of the existing age distribution on the composition of the teaching force and to estimate the number of newly hired teachers that will be needed over the forecast period.

Several alternative projections are produced for the number of newly hired school teachers in both public and private schools at the national level. The alternative projections are based on differing assumptions concerning the rates at which teachers of various ages will continue teaching from one year to the next and the total number of teachers that will be needed each year.

One key assumption of this analysis is that continuation rates of teachers, by age group, remain constant over time. This assumption is required as there are not enough observations to develop an econometric model for continuation rates. A sensitivity analysis of this assumption was conducted by examining results using three different continuation rates. Similarly, the report examines results using three different scenarios for total number of teachers.

The report does not analyze the issue of supply relative to demand of teachers. Instead, it is assumed that there will be enough supply to meet the demand, which reflects historical precedent. However, the report does include some discussion of how supply and demand forces might affect the results.

The Newly Hired Teachers Model projects the total number of newly hired teachers that will be needed over time to replace teachers leaving the profession because of retirement and other reasons, as well as to instruct additional students that are expected to enter the system.

The key component of this model is the aging of the teacher force over time, based on the counts of teachers of each age from the 1993-94 Schools and Staffing Survey (SASS). The model estimates the number of continuing teachers, by age, through the use of age-specific continuation rates from SASS. Each year, the model brings just enough newly hired teachers into the teaching force so that the sum of the continuing teachers and the newly hired teachers equals a projected number for total teachers. Calculating the number of newly hired teachers (new teacher hires) summed over the forecast period is the focus of this study.

The Newly Hired Teachers Model requires four data items: (1) the number of teachers by age (age distribution) for a recent year; (2) the total number of teachers for each year under study, including both historical years and forecast years; (3) an estimate of the continuation rate for each age; and (4) an estimate of the age distribution of the newly hired teachers. The main sources for these data are the 1993-94 SASS and the 1994-95 Teacher Follow-up Survey (TFS), although other sources such as the Common Core of Data (CCD) and Projections of Education Statistics to 2008 (Gerald and Hussar 1998) are used as well.

The analysis was conducted at the national level only, as the TFS was not designed for state-level analysis. Thus,continuation rates for each state could not be calculated due to sample size.

The model requires an age distribution to use as a starting point for the aging of the teacher force over the forecast period. The total number of public and private school teachers, by age, was obtained from the 1993-94 SASS. The median age was 44 for all public school teachers and 42 for private school teachers. For the nation as a whole, there were more public school teachers age 47 than any other age (figure A).

The SASS age distribution is for a headcount of full- and part-time teachers. Because the number of teachers forecast for each of the later years is for full-time-equivalent (FTE) teachers, however, the number of FTE teachers by age for 1993-94 is required. For modeling purposes, the age distribution of FTE teachers was assumed to be the same as the age distribution of teachers using the headcount number. This assumption seems reasonable both because the age distributions of full-time teachers and part-time teachers were found to be similar in 1993-94 and because the relatively small number of part-time teachers would result in minimal impact on the model in any case.

Figure A.—Age distribution of full-time and part-time public school teachers: 1993-94

Three different assumptions were used to produce alternative scenarios for the numbers of public and private school teachers that will be needed for each year under study:

Scenario 1. For the first scenario, the pupil/teacher ratio was assumed to remain constant at 1995-96 values. The total number of teachers needed each year was estimated by dividing the appropriate enrollment projections by the pupil/teacher ratio. Greatest emphasis was given to the results from scenario 1, although results from the other scenarios were analyzed.

Scenario 2. For the second scenario, it was assumed that for each year from 1996-97 to 2008-09, the number of teachers would remain at 1995-96 levels despite increasing enrollments.

Scenario 3. For the third scenario, the national teacher projections from Projections of Education Statistics to 2008 (Gerald and Hussar 1998) were used. This method gave the highest figures for newly hired teachers needed because it included an assumption of some decline in the pupil/teacher ratio.

The model calls for a constant set of age-specific continuation rates to be applied to each year of the forecast period. The TFS provides three sets of continuation rates for recent years: from 1993-94 to 1994-95, from 1990-91 to 1991-92, and from 1987-88 to 1988-89. Each set includes separate continuation rates for teachers who continued teaching in public schools and for those who continued teaching in private schools.

The 1993-94 to 1994-95 continuation rates, obtained from the 1994-95 TFS, are the most recent available. Most of the results presented in this report were produced using the 1993-94 to 1994-95 rates, but the sensitivity of the model was examined by using the other sets of rates to produce alternative projections.

Comparisons of forecasts made using all three sets of continuation rates suggest that the model is sensitive to changing continuation rates. While there are few statistically significant differences in continuation rates over time, these rates apply to the entire count of teachers each year. It is not surprising, therefore, that continuation rates are by far the most sensitive facet of the model.

First-time teachers and returning teachers have lower continuation rates than those of the same age who had been teaching the previous year. If the proportion of new teachers in the teaching force grows over time, it will tend to push continuation rates downward.

The fourth type of data needed for the model is the age distribution of the newly hired teachers during each year under study. As with the continuation rates, usable data are available for both public school teachers and private school teachers. The most recent actual age distribution of newly hired teachers, obtained from the 1993-94 SASS, was used as the estimated distribution for each year.

An important assumption is that in the forecast period the age distribution of newly hired teachers remains similar to that in the 1993-94 SASS. Comparison of the 1993-94 SASS with the 1987-88 SASS and the 1990-91 SASS showed that the age distributions for these years were similar, though not identical. One factor that may change the age distribution over time is the aging of the baby boom generation. As this generation retires, there may be relatively fewer people in their forties and fifties who become newly hired teachers, thus pushing the average age of newly hired teachers lower. However, programs to encourage the rehiring of retirees may partially diminish this effect.

Using scenario 1 and teacher continuation rates from 1993-94 to 1994-95, the model projects that approximately 2.4 million newly hired public school teachers^* will be needed from 1998-99 to 2008-09. These newly hired teachers will be needed to replace teachers who retire or leave the profession for other reasons and to keep the pupil/teacher ratio constant as total enrollment increases.

The combination of three scenarios for total number of teachers and three teacher continuation rates produces a relatively wide range of estimates, from about 1.7 to 2.7 million newly hired teachers (table B).

Under scenario 1, using the most recent continuation rates (from 1993-94 to 1994-95), the model projects that 2.4 million newly hired teachers will be needed. Under the same scenario, but using continuation rates from 1990-91 to 1991-92, approximately 450,000 fewer newly hired teachers are predicted to be needed (19 percent lower than with the most recent continuation rates). If the 1987-88 to 1988-89 rates are used, approximately 350,000 fewer teachers will be needed (14 percent lower than with the most recent rates). These relatively large differences in the forecasts occur because of the cumulative impact of the differences in continuation rates when they are applied to the entire population of teachers over each year of the forecast period. The numbers of newly hired teachers needed are lower using the older sets of continuation rates because the older sets of continuation rates are generally higher.

Even for the same set of continuation rates, there is a considerable range in the estimates. Using the most recent set of continuation rates, for example, the forecast of 2.4 million newly hired teachers needed by 2008-09 under scenario 1 is 10 percent greater than the 2.2 million newly hired teachers projected under scenario 2, but 12 per- cent less than the 2.7 million teachers projected under scenario 3.

Table B.—Number of newly hired public school teachers needed for the 11 years from 1998-99 to 2008-09, by continuation rate used and teacher total assumption

Another way to compare results for the alternative scenarios is to look at projected age distributions. Since the estimated numbers of public school teachers at each age in 2008-09 look very much alike for each of the three scenarios and each of the three sets of continuation rates, this discussion concentrates on the results for scenario 1 and the most recent set of continuation rates.

The age distribution of FTE teachers is predicted to flatten over time, with a more equal distribution of teachers in each age group. Specifically, the proportion of teachers who are in their forties is expected to decrease over time, while other age groups, which had been underrepresented, are expected to increase. Yet, even in 2008-09, the model projects that a sizable number of the teachers who had been in their forties in 1993-94 will still be teaching. The model forecasts that there will be more public school teachers in their late fifties in 2008-09 than there were in 1993-94.

Under scenario 1, approximately 759,000 teachers will retire from 1998-99 to 2008-09. As there are fewer teachers each year in scenario 2 compared with scenario 1, there will be fewer teachers who will be retiring (745,000). Conversely, as there are more teachers in scenario 3, there will be more teachers who will be retiring (765,000). These numbers of retiring teachers are based on the most recent continuation rates, but the pattern is similar using the alternative rates.

Using the most recent set of continuation rates, scenario 1 projects that some 568,000 newly hired private school teachers will be needed from 1998-99 to 2008-09. The comparable number is somewhat lower under scenario 2 (524,000) and somewhat higher under scenario 3 (620,000). The range of projections using alternative continuation rates was small compared with the range for newly hired public school teachers: Under scenario 1, the projected numbers of newly hired private school teachers ranged from 2 percent lower (using the 1990-91 to 1991-92 rates) to 5 percent higher (using the 1987-88 to 1988-89 rates) than the number calculated using the most recent rates. A forecast of age distribution predicts that the numbers of both older and younger private school teachers will increase, while the number of teachers in their forties will fall.

Another source of national-level estimates of newly hired elementary and secondary school teachers is the U.S. Department of Labor's Bureau of Labor Statistics (BLS). BLS forecasts annual average job openings for elementary and secondary school teachers at approximately 400,000 per year from 1996 to 2006 (BLS 1998, table 1), for a total of approximately 4.5 million newly hired teachers over this 11-year period. The BLS total is significantly greater than the total of 3.3 million newly hired teachers projected for the same period under scenario 3, which yields this study's highest projections.

One reason for the larger BLS projections is that the BLS definition of teacher includes those working at all pre-primary institutions and training centers in addition to those working at traditional elementary and secondary schools. The broader BLS definition results in a greater overall number of teachers than the definition used in this study (3.8 million versus 3.0 million in 1996), and BLS inclusion of daycare staff may also contribute to lower continuation rates. A second reason for the larger BLS projections is that BLS forecasts greater growth in the number of teachers from 1996 to 2006 (21.1 percent from BLS versus 12.7 percent from scenario 3). Again, some of the growth projected by BLS would occur outside traditional elementary and secondary schools.

Despite differences in definitions and results, the projections in this report and the BLS projections both suggest a need for large numbers of newly hired teachers over the next decade.

If the pupil/teacher ratio remains constant, about 2 million newly hired public school teachers and about 500,000 newly hired private school teachers will be needed during the 11-year period from 1998-99 to 2008-09. Some of the alternative assumptions and scenarios result in higher forecasts, particularly scenario 3, which assumes some decline in the pupil/teacher ratio. Data from BLS also indicate a need for large numbers of newly hired teachers.

In the model used for this report's projections, the teacher continuation rate is a critical factor that can be influenced by supply and demand forces. These forces, in turn, are affected by both economic conditions and education policies. For example, a good economy tends to decrease continuation rates by creating greater opportunities for alternative employment. If faced with an aging teacher force and an inadequate supply, however, school districts or state education agencies could enact incentives to delay retirements, thus increasing continuation rates and reducing the demand for new hires, at least temporarily. Increases in salaries or other benefits could be used to help retain teachers who might otherwise leave the profession. Such policies could have a sizable impact on the number of newly hired teachers needed. Also, an economic downturn might make teaching positions more attractive because of their perceived stability.

Supply and demand forces also can influence the model's important, but less critical, assumption regarding the stable age distribution of the new teachers. Districts could enact policies to recruit older people into the teaching profession. The supply of qualified teachers available could be adjusted by changing teacher certification requirements to favor either new or less recent college graduates. These efforts would have an impact on the age distribution of newly hired teachers, which would later affect the teacher demand.