Chapter 12
Staffing the Nation's Schools with Skilled Teachers1

Richard J. Murnane

Harvard University

Students who are taught by effective teachers learn much more in school than students who are taught by ineffective teachers. This judgment by most parents is also the conclusion of many research studies.2 Devising strategies to provide all children with skilled teachers, both by improving the effectiveness of teachers already working in schools and by attracting a greater number of talented new teachers to the nation's classrooms, should be at the center of school reform efforts. These goals are related; providing opportunities for growth and learning on the job is important in attracting and retaining talented college graduates who seek not only good compensation but also rewarding work.

This chapter addresses the second objective, attracting a greater number of skilled teachers to the nation's classrooms. It considers the role of incentives in determining who trains to teach, who becomes a teacher, which teachers change districts, who leaves teaching, and who returns to teaching. It is through changes in incentives that public policies can influence which college graduates staff our nation's schools. Designing incentives to attract talented college graduates to teaching will be especially important over the next 15 years as an increase in retirement rates among the existing teaching force increases the opportunities for new hiring.

1  

The author would like to thank Dominic Brewer, Linda Darling-Hammond, Eric Hanushek, and Michael Podgursky for helpful comments on drafts of this paper.

2  

For a summary of the evidence, see Hanushek (1986).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 241
--> Chapter 12 Staffing the Nation's Schools with Skilled Teachers1 Richard J. Murnane Harvard University Students who are taught by effective teachers learn much more in school than students who are taught by ineffective teachers. This judgment by most parents is also the conclusion of many research studies.2 Devising strategies to provide all children with skilled teachers, both by improving the effectiveness of teachers already working in schools and by attracting a greater number of talented new teachers to the nation's classrooms, should be at the center of school reform efforts. These goals are related; providing opportunities for growth and learning on the job is important in attracting and retaining talented college graduates who seek not only good compensation but also rewarding work. This chapter addresses the second objective, attracting a greater number of skilled teachers to the nation's classrooms. It considers the role of incentives in determining who trains to teach, who becomes a teacher, which teachers change districts, who leaves teaching, and who returns to teaching. It is through changes in incentives that public policies can influence which college graduates staff our nation's schools. Designing incentives to attract talented college graduates to teaching will be especially important over the next 15 years as an increase in retirement rates among the existing teaching force increases the opportunities for new hiring. 1   The author would like to thank Dominic Brewer, Linda Darling-Hammond, Eric Hanushek, and Michael Podgursky for helpful comments on drafts of this paper. 2   For a summary of the evidence, see Hanushek (1986).

OCR for page 241
--> Measuring Teacher Quality The lack of reliable indicators of teacher quality has hindered researchers' attempts to answer the central policy question: How do alternative policies affect the supply of effective teachers? Most studies have focused on the role of incentives in the occupational decisions of all teachers, potential teachers, and former teachers and on how their decisions affect the total pool of college graduates willing to teach. These studies are useful in that they document that incentives do affect career decisions, but they do not directly address the central question of how policies affect the quality of the stock of teachers. One assumption used in many studies of teachers' career patterns is that academic talent is a good indicator of teaching effectiveness. The assumption rests on the results of studies showing that measures of teachers' academic talent are positively related to their students' test score gains. Some of these studies use teachers' scores on standardized tests as the measure of academic talent; others use indicators of the quality of the undergraduate college the teacher attended (Ehrenberg and Brewer, 1994, 1995; Ferguson, 1991; Hanushek, 1972; Summers and Wolfe, 1977; Winkler, 1975). The value of the assumption that college graduates' academic talent predicts their effectiveness as classroom teachers is that many databases that track college students' careers contain measures of academic talent. The assumption makes it possible to interpret the results of studies exploring the career decisions of academically talented college graduates as evidence of the factors affecting the career decisions of effective teachers. This assumption has intuitive appeal. Teaching is a complex job, and it makes sense that academically talented college graduates have an advantage in learning the many skills required to teach well. However, one should be cautious in equating academic talent with teaching effectiveness. The evidence supporting this assumption comes from databases in which academically talented teachers went through the same type of preservice training that other teachers did. There is no reason to believe that the preservice training received by the academically talented teachers in these studies was less intensive than that received by teachers with lower levels of academic talent. In fact, it might have been of higher quality since the academically talented teachers tended to attend colleges with more financial resources than the colleges attended by less academically able teachers. In recent years public and private policies have been introduced to attract academically talented college graduates to teaching, in part by reducing the amount of preservice training that participants must undergo before beginning classroom teaching. Examples include alternative certification (licensing) programs in a number of states and the highly publicized private program, Teach for America (TFA). These programs are attractive to academically talented college graduates who intend to teach for a few years and then pursue other, more lucra-

OCR for page 241
--> tive careers because they provide an opportunity to enter teaching without taking a large number of preservice teacher preparation courses. The question of whether alternative licensing programs improve the quality of teaching in the nation's schools is highly controversial. In a strongly worded critique of the TFA program, Darling-Hammond (1994) argues that it hurts children in urban schools by providing them with teachers who have little knowledge of the techniques needed to teach effectively. She cites the statement of one former TFA participant, a Yale graduate: I—perhaps like most TFAers—harbored dreams of liberating my students from public sector mediocrity and offering them as good an education as I had received… But I was not ready…. As bad as it was for me, it was worse for the students… Many of mine … took long steps on the path toward dropping out. … I was not a successful teacher and the loss to the students was real and large. (Schorr, 1993) At the center of Darling-Hammond's criticism is the belief that good preservice training is critical to effective teaching.3 I think she would agree that in the best of all worlds academically talented people who have participated in high-quality preservice programs are attracted to teaching. Later in this chapter strategies for accomplishing this are considered. However, no inferences about the value of programs that attract academically talented graduates to teaching by reducing training requirements can be made from studies in which academically talented teachers had preservice training that was at least as good as that of teachers with less academic talent. Studies examining the career decisions of academically talented college graduates are useful because, holding quality and quantity of training constant, academically talented teachers do seem to be more effective in the classroom. Learning about the factors that influence the career decisions of academically talented college graduates may provide ideas for policies to attract more of them to teaching. Nevertheless, these studies tell nothing about the efficacy of policies designed to attract talented college graduates to teaching by reducing training requirements. Factors Affecting Career Decisions The evidence on factors that affect college students' teaching career decisions can be organized according to the sequence of those decisions: whether to begin work as a teacher, whether to switch districts, how long to stay in teaching, and whether to return to teaching after a career interruption. 3   See Darling-Hammond (1990) for a summary of evidence supporting the importance of preservice training.

OCR for page 241
--> Obtaining a Teaching License To examine trends in the number of college graduates who obtain teaching licenses, data from individual states are needed. As job opportunities in teaching fell during the 1970s, the number of college students who were training to become teachers declined dramatically. In North Carolina the number of new teaching licenses granted by the state fell from 6,538 in 1975 to 2,830 in 1984, a period during which enrollments in North Carolina colleges and universities were quite stable (Murnane et al., 1991). Data from New York show a decline in the number of teaching licenses from 34,770 in 1974 to 16,002 in 1985 (Gilford and Tenenbaum, 1990). This decline is attributable to the decline in job opportunities in teaching, the fall in teachers' salaries relative to those of other occupations, and improved job opportunities for women and minorities in fields outside education.4 It is often observed that Scholastic Assesment Test (SAT) scores of college freshmen who intend to major in education are low, both in absolute terms and relative to the scores of freshmen planning to major in other fields; but two studies have found that the SAT scores of freshmen who plan to become teachers are poor predictors of the scores of those college students who actually complete education majors. Many freshmen with low scores drop out of college before completing degree programs, and more able students switch fields of study during their college years. Studies show that the academic skills of students who complete education majors are higher than those of students who announce as freshmen that they intend to major in education (Nelson, 1985; Hanushek and Pace, 1994), although probably not as high as those of college graduates not preparing to teach. 5 A recent study shows that college students are less likely to complete education majors in states that require candidates for teaching licenses to complete a relatively large number of education-related courses (Hanushek and Pace, 1994), which raise the cost of obtaining a teaching license, especially for college students who either plan to teach for a few years before moving to another occupation or want to obtain a teaching license as "insurance" in case opportunities in other fields prove unattractive. This finding points to a dilemma in designing licensing requirements for teaching. Schools have always relied heavily on teachers who intend to teach only for a few years. It is important to develop licensing requirements that do not discourage college students from considering teaching 4   Given the limited time series evidence, it is not possible to compare the relative importance of these complementary explanations. 5   See Ballou and Podgursky (1994), Hanushek and Pace (1994), and Nelson (1985). An exception to this pattern is a 1990 study reporting that newly qualified teachers had higher grade point averages than did all college graduates as a group (Gray et al., 1993). While provocative, it is difficult to interpret this pattern given the variation in grading practices among colleges and universities.

OCR for page 241
--> even if they do not plan to make it their life's work. At the same time, teachers need to acquire skills to teach effectively. As elaborated on below, basing licensure on the demonstration of teaching competence rather than on completion of course work is a strategy for reconciling these objectives. Among the factors affecting the number of college students who seek teaching licenses the role of standardized testing in licensing requirements has been the subject of a good deal of analysis. As part of attempts to raise entry standards over the past 20 years, several states have required that candidates for teaching licenses achieve scores above threshold values on certain standardized tests, primarily the NTE, formerly known as the National Teacher's Examination, developed by the Educational Testing Service. Until recently the NTE included a "Core Battery," which tested communication skills, general knowledge; and professional knowledge; 49 specialty-area tests that measured knowledge of specific academic subjects or fields; and the Pre-Professional Skills Tests, which measured basic reading, writing, and mathematics skills.6 As of 1994, 34 states used at least part of the NTE in their teacher licensing requirements. Other states use different standardized multiple-choice tests, such as the SAT or the California Achievement Test. Several studies have shown that requiring applicants for teaching licenses to score above a prespecified cutoff on the NTE reduces the number of college students who train to become teachers and the number of college graduates who obtain teaching licenses. The effect is particularly great on minority students, who tend to score lower on standardized tests such as the NTE (Murnane et al., 1991; Hanushek and Pace, 1994). The use of standardized multiple-choice tests in teacher licensing programs is controversial. Critics point to the low correlation between scores on these tests and measures of teaching effectiveness. One comprehensive review summarized the evidence as follows: The available evidence is none too good, but it indicates that teacher tests have little, if any, power to predict how well people perform as teachers, whether that performance is judged by ratings of college supervisory personnel, ratings by teachers, student ratings, or achievement gains made by students. (Haney et al., 1987, p. 199) Defenders of these tests argue that the lack of correlation between test scores and measures of teaching effectiveness is not the issue. They argue that the tests screen out applicants who lack the basic literacy skills necessary to serve as successful role models to students and to write grammatically correct and coherent prose. This argument has considerable merit, but the underlying issue is 6   See Educational Testing Service (1988). The number of NTE specialty-area tests was counted from a table entitled "NTE Programs: Table of User Qualifying Scores and Validity Study Status Information," provided by Carol Dwyer of the Educational Testing Service.

OCR for page 241
--> whether the use of multiple-choice tests in state licensing procedures increases the likelihood that our nation's children will be taught by effective teachers of varying backgrounds. One reason for doubt is that multiple-choice tests of "professional knowledge" do not reliably measure whether applicants possess the knowledge needed to teach effectively because the test items rarely provide the rich contextual information needed to respond thoughtfully to a problem situation. Nor do multiple-choice questions allow applicants to offer creative responses. These are critical limitations because the answer to almost all questions about how an effective teacher should respond in a particular classroom situation is "it depends."7 Given the lack of evidence relating scores on multiple-choice tests of knowledge to measures of teaching effectiveness, how has the use of these tests survived legal challenge? The courts have ruled that use of the NTE tests in licensing is justified because they measure "the content of the academic preparation of prospective teachers" (South Carolina, 1989, p. 241). In other words, it is argued that the tests measure particular types of knowledge and are not intended to measure teaching effectiveness. This is an exceedingly weak defense in view of the questionable validity of multiple-choice questions about professional knowledge and the demonstrable negative impact that the testing requirements have on the racial composition of the teaching force. A challenge in revising licensing requirements is to ensure that potential teachers do possess the skills necessary to communicate effectively with students and their parents and to provide incentives and opportunities for candidates deficient in basic skills to improve their skills. In response to this challenge, the Educational Testing Service recently made dramatic changes in the structure of the NTE. The new Praxis system has three components. Praxis I, which assesses reading, writing, and basic math skills, is designed to be taken by college students before entering a teacher education program. The idea is that students should be made aware of any deficits in their basic skills before they prepare to become teachers. Praxis II tests subject matter knowledge and is designed to be taken after college seniors have completed their course work. Unlike the old subject matter tests, Praxis II does not consist solely of multiple-choice items. It also includes constructed-response questions. Praxis III assesses teacher competence through interviews, classroom observations, and examination of documents; it is designed to be used with teachers in their second or third year of teaching as part of their licensing requirements (Forbes, 1995). As states are just beginning to make the transition from the old NTE to the new Praxis system, there is as yet no evidence on its effectiveness in keeping ineffective teachers out of the classroom or on the number of minorities becoming teachers. 7   See Wise and Darling-Hammond (1987, p. 22).

OCR for page 241
--> Entering Teaching In the late 1960s, one out of every three students who graduated from college in the United States taught in either a public or private school within five years of graduation. By the early 1980s, only one in 10 new graduates entered teaching (Murnane et al., 1991). This trend reflects a decline in teaching opportunities accompanying declines in student enrollment. Between 1971 and 1984 the number of students attending elementary and secondary schools fell from 46 million to 39 million. The decline in the number of college students preparing to teach and entering teaching is simply a response to declining demand, but changes in the composition of the pool of college graduates choosing to teach are disturbing. One striking change is the decline in minorities. In the late 1960s, approximately 60 percent of all black female college graduates entered teaching within five years of graduation; less than 40 percent of white female graduates made the same decision. During the 1970s, the percentages of black and white graduates who became teachers not only declined but converged. In the early years of the 1980s, black graduates were less likely to enter teaching than were white graduates (Murnane et al., 1991). In recent years the percentages of black and Hispanic teachers have increased slightly. In the 1987–1988 school year, 8 percent of the nation's public school teachers were black, and 2.6 percent were Hispanic (NCES, 1992). In the 1990–1991 school year the comparable figures are 8.3 and 3.4 percent, respectively (NCES, 1993). These percentages do not reflect the racial/ethnic composition of the public school student body, 16.1 percent of which was black and 11.1 percent Hispanic in the 1990–1991 school year (NCES, 1993). Another disturbing trend is the decline in the representation of the most academically able college graduates. In the late 1960s, college graduates with IQ scores of 130 were only slightly less likely to become teachers than graduates with IQ scores of 100. By 1980 a college graduate with an IQ score of 100 was more than four times as likely to become a teacher than was a graduate with a score of 130.8 In other words, throughout the 1970s the number of new entrants to teaching who were among the most academically able of all college graduates became smaller and smaller. In the 1990s the demand for new entrants to elementary and secondary school teaching has been rising for two reasons, both demographic. First, the teacher retirement rate is rising as the large group of teachers hired during the 1950s and 1960s, the baby boom years, reach retirement age. Second, school enrollments of the children of the last baby boom are rising moderately. There is little question that enough warm-bodied adults will be found to staff the schools, 8   See also Murnane et al. (1991). Hanushek and Pace (1994) show that this trend did not reverse itself over the 1980s.

OCR for page 241
--> but unless recent trends change, a declining proportion will be the nation's most academically able college students. Will salary increases attract skilled teachers in the years ahead? Although the evidence is scanty, one study showed that the occupational decisions of college graduates in England were extremely sensitive to relative salaries in teaching and other professions (Dolton, 1991). A second study showed that the higher the teaching salaries, the larger the pool of college graduates who enter teaching (Manski, 1987). But the study concluded that salary increases, by themselves, do not have a marked effect on the ability distribution of the set of college graduates who enter teaching. Higher salaries attract graduates with low SAT scores, the measure of ability in this dataset, as well as graduates with high SAT scores. Moreover, some school districts do not choose to hire the most academically able candidates. There are various reasons why many school districts do not value academic talent (Ballou and Podgursky, 1994). Explanations include patronage, the weight attached to skills other than teaching ability such as coaching skill, and the possibility that academically strong candidates lack thorough teacher preparation. The pattern complicates strategies to staff all of the nation's schools with skilled teachers, suggesting that incentives to attract skilled teachers may not be sufficient. It may also be necessary to ensure that the most effective applicants are hired or that ineffective teachers are removed from applicant pools. Changing Districts The evidence on factors influencing teachers' decisions to change school districts, although scanty, supports the notion that opportunities matter. Teachers who are "underpaid" (defined as receiving a salary lower than that predicted from a model in which teachers' salaries are a function of personal characteristics) have a relatively high probability of changing districts (Baugh and Stone, 1982). Male teachers are more likely to remain in a school district greater the number of administrative positions to which they can be promoted, and the higher the salaries in these administrative positions. Male teachers are also more likely to leave a school district the higher the salaries of administrators in neighboring districts. Administrative opportunities apparently do not affect the mobility decisions of female teachers (Brewer, 1994). How Long to Stay in Teaching Decisions about how long to stay in teaching are influenced by salaries and opportunity costs. A $2,000 difference in annual salary is associated with a difference in median employment duration of approximately one year for teachers in Michigan and approximately two years for North Carolina teachers. In both states, differentials are more likely to induce teachers to leave during the

OCR for page 241
--> first years on the job (Murnane et al., 1991). A study based on data from Indiana produced similar findings (Grissner and Kirby, 1992). Another study shows that the turnover rate of mathematics and science teachers in California is negatively related to teachers' salaries (Rumberger, 1987). Not only do teachers' salaries affect career decisions, so do opportunity costs. Murnane et al. (1991) used two measures of opportunity cost in studying the determinants of employment duration in teaching. The first was a teacher's score on the NTE, on the assumption that teachers who achieve high scores on the NTE are also likely to achieve relatively high scores on other tests, such as the law boards, that determine entry to relatively high paying occupations. The second measure of opportunity cost was a teacher's subject specialty, on the assumption that teachers with disciplinary specialties are likely to command higher salaries in fields outside education than are elementary school teachers who teach basic skills in many subject areas. Among secondary school teachers, those in such fields as chemistry and physics, which command relatively high salaries in business and industry, are likely to have the shortest teaching spells, especially since virtually all U.S. public school districts have a uniform salary scale that takes no account of variations in opportunity cost by field. In North Carolina 56 percent of white teachers with NTE scores at the 90th percentile remained in teaching for at least five years, compared with 71 percent with scores at the 10th percentile of the sample distribution (Murnane et al., 1991). In both North Carolina and Michigan, the median employment duration for secondary school teachers was two or more years shorter than that of elementary school teachers. Among secondary school teachers in both states, chemistry and physics teachers had the shortest teaching careers. The median length of time that chemistry and physics teachers in Michigan stayed in teaching was 2.2 years, compared to 4.0 years for social studies teachers, and 3.7 years for English teachers (Murnane et al., 1991). Other evidence concerning the role of salaries and opportunity costs in teachers' decisions to shift occupations comes from the first administration of the School and Staffing Survey, administered by the National Center for Education Statistics. White male teachers who left teaching within their first three years in the classroom to work full time in another industry earned $3,300 more, on average, in their new jobs, than they did in their last year teaching. The comparable numbers for nonwhite men and white women are increases of $6,200 and $300, respectively. The one exception to this pattern is nonwhite women, who earned $300 less in their first jobs in other industries than they did in their last year teaching. This helps explain why only 2.8 percent of nonwhite women left teaching in their first three years compared to 7.5 percent of white women, 9.0 percent of white men, and 18 percent of black men.9 9   These tabulations were conducted at my request by the National Data Resource Center, National Center for Education Statistics, in February 1991.

OCR for page 241
--> Whether to Return to Teaching One of the surprises from recent research on teacher supply is the increase over the past 20 years in the proportion of new hires who come from a ''reserve pool" of people who were neither teaching the previous year nor engaged in full-time study. In 1966, three out of 10 newly hired teachers came from the reserve pool; in 1991 the proportion was seven out of 10.10 It is therefore a mistake to think of newly minted college graduates as the only or principal source of teachers, an assumption behind many projections of teacher shortages in the years ahead. The reserve pool has been the principal source of new hires in recent years when demand has been low and relatively few college students have prepared to teach. What is not known is how large a source of supply the reserve pool will be in the years ahead when the demand for new hires will grow. Undoubtedly, the answer depends on the attractiveness of teaching salaries and working conditions relative to those in other occupations. It also likely will depend on the costs associated with licensing requirements of moving from other occupations into teaching. As yet there are no data on the sensitivity of mobility decisions to these factors. There is some information on the behavior of members of the reserve pool who left teaching and then returned to it. Approximately one in four teachers who leave the classroom return within five years. The teachers most likely to return are those with subject area specialties that provide limited opportunities for better-paying employment outside teaching (Beaudin, 1993; Murnane et al., 1991). This pattern supports the hypothesis that decisions to return to teaching are sensitive to opportunity costs. Policies for Staffing the Schools with Skilled Teachers Evidence on the career decisions of potential teachers, teachers, and former teachers show that they do respond to incentives. Salaries and opportunity costs influence who goes into teaching, who stays in teaching, and who returns to teaching after a career interruption. At the same time, some school districts do not hire candidates with the most promise for helping children acquire critical skills. Thus, the policymaking challenge is not only to design incentives to attract effective candidates and former teachers to the pool of those who want to teach but also to exclude from the pool those who lack the skills to teach. 10   See National Education Association (1992). The percentages of newly hired teachers drawn from the reserve pool in 1966 and 1991 were calculated by using information from Table 13.

OCR for page 241
--> Ill-Advised Policies The Mandatory Master's Degree Several states, including New York and California, require that teachers earn a master's degree within a specified period of time after initial hiring. Teachers who fail to do so lose their positions. This policy is ill advised for two reasons. First, the preponderance of evidence is that teachers with master's degrees are no more effective than teachers who do not hold these degrees.11 Second, the requirement raises the cost of choosing teaching as a career and may have the effect of dissuading potentially effective teachers from entering the profession. Some recent evidence indicates that "fifth-year" programs are successful in attracting academically talented college graduates to teacher preparation. These programs enable college graduates who did not prepare to teach to obtain teacher training and a master's degree in teaching. Graduates of the fifth-year program at the University of New Hampshire were more academically talented than graduates of the university's undergraduate education program, were more likely to enter teaching, and were more likely to remain in teaching. Graduates of the fifth-year program believed that their preparation for teaching was better than did graduates of the undergraduate program (Andrew, 1990). Master's degree programs can play a significant role in providing schools with academically talented, well-prepared teachers, but the evidence does not support a policy of requiring all teachers to complete a fifth-year program. Fifth-year programs probably attract talented college graduates in part because they want to distinguish themselves from graduates of undergraduate teacher education programs. Education schools certainly have incentives to make their programs effective in order to compete for students who can enter teaching without continuing in school for a fifth year. If a state mandates that all prospective teachers must complete a master's degree program, the incentives change markedly, and the education schools then have captive student bodies and, worse, face pressures to lower standards to produce enough graduates to staff the schools. Paying a Premium for Master's Degrees The uniform salary schedules of almost all public school districts in the country pay teachers with master's degrees premiums ranging from a few hun 11   For a summary of this evidence, see Hanushek (1986). One exception to this general pattern is a study by Ferguson (1991), which found that students in Texas schools taught by teachers with master's degrees had higher achievement scores than students taught by teachers without master's degrees. Ferguson made creative use of available data in conducting this study; however, the data have important limitations. First, it is not possible to match students to individual teachers. Second, it is not possible to measure the achievement gains of individual students taught by teachers with different characteristics.

OCR for page 241
--> dred to several thousand dollars. To obtain these premiums, more than half of the nation's teachers have earned master's degrees. In view of the lack of evidence of the superior effectiveness of teachers with master's degrees, this is a poor use of scarce resources. High quality teacher training is important in helping graduates learn to teach effectively. As Shulman (1987) observed, "Our question should not be, Is there really much one needs to know in order to teach? Rather, it should express our wonder at how the extensive knowledge of teaching can be learned at all during the brief period allotted to teacher preparation." The problem is that the automatic premium for the master's degree gives teachers an incentive to seek a degree program that makes as few demands on their time and energy as possible. Teacher education schools, to maximize enrollments, have an incentive to design undemanding programs. The salary premium gives both teachers and education schools the wrong incentives. Merit Pay Based on Supervisors' Evaluations Merit pay systems reward teachers with higher compensation for superior supervisor assessments of teaching quality. In principle, merit pay appears to be a good way to improve the performance of public schools. Does it not attract talented college graduates to the profession and give all teachers an incentive to teach as well as possible? Unfortunately, the results of extensive research are clear—merit pay based on supervisors' performance evaluations simply does not work. In the vast majority of cases where school districts have adopted merit pay plans for teachers, they have dropped them within five years. This is as true in districts without teachers' unions as in districts with unions. It was as true in the 1920s as it is in the 1990s. There is no example of a troubled district that has successfully used merit pay to improve its performance (Murnane and Cohen, 1986). A Combination of Policies with Promise The challenge is to attract academically able college students to the teaching profession, give aspiring and veteran teachers an incentive to undertake high-quality training, and adopt licensing requirements that discourage those who lack the skills necessary to teach effectively. A two-part strategy holds promise for making progress toward these goals. Performance-Based Teacher Licensing In most states a college graduate obtains a teaching license by documenting that he or she has completed a program of prescribed study at an accredited institution of higher education and by scoring above threshold levels on a series

OCR for page 241
--> of multiple-choice tests. These requirements are ill advised for several reasons. First, the prescribed courses are often of low quality. As a result, participants do not learn the critical skills needed to teach effectively, and some college students may decide not to teach because they do not want to take courses they perceive to be of little value. Second, scores on the multiple-choice tests keep many minorities from obtaining a teaching license and provide no assurance that graduates who do obtain teaching licenses possess the skills needed to teach effectively. Several changes would improve the effectiveness of licensing procedures in keeping incompetent teachers out of schools without discouraging talented college students from trying teaching. First, the multiple-choice tests should be replaced with constructed-response tests that assess whether candidates can perform certain tasks related to teaching, such as writing a lucid letter to parents advising them of a problem with their child. Aspiring teachers should be able to take these tests as early in their training as they desire so as to eliminate the uncertainty that deters many minorities from training to teach and provide the maximum possible time for remediation. Second, compulsory training requirements should be replaced with a system of performance assessments, under which candidates would obtain a long-term teaching license only after demonstrating that they can teach effectively.12 A system of performance-based assessments is much more costly than current licensing procedures but has the potential for improving the quality of teaching. A system of comprehensive, high-quality performance assessments should improve teacher training by focusing instruction on the critical skills needed to pass the assessments and giving aspiring teachers an incentive to seek the best training in the skills needed to pass the assessments. Ensuring that the skills needed to pass the assessment are closely related to the skills needed to help children learn is a serious challenge because there is little consensus, but there has been progress in designing performance-based assessment systems. Teachers' Salaries Attempts to improve the quality of the teaching profession cannot succeed unless talented college graduates want to teach. A salary schedule that attracts academically talented graduates to teaching and makes it worthwhile to acquire the skills needed to teach effectively is a necessary condition of improving the quality of teachers. The question is how to structure salaries to achieve these goals efficiently. The salary schedules in the vast majority of the nation's school districts reward longevity and advanced degrees. They do not pay premiums for subject specialties with shortages of teachers or provide incentives for teachers to invest 12   See Murnane et al. (1991, Chap. 7) for a detailed discussion of alternative designs for performance-based assessments.

OCR for page 241
--> in improving their skills, for example, through collaborative exploration of teaching strategies. Most salary schedules reward teachers for taking university-based graduate courses, but the reward is as large for undemanding courses quite unrelated to teachers' day-to-day work as it is for a demanding one that tries to help teachers change how they teach. Moreover, the best university-based courses may be no better than alternative forms of professional development that are not rewarded in most teachers' contracts. Over the past 20 years the funds available for increases in teachers' salaries have gone disproportionately to experienced teachers (Monk and Jacobson, 1985; Murnane et al., 1989). This "backloading" of pay increases reflects the increased power of experienced teachers, who became a majority in most districts, but as a result the salary structure becomes increasingly unattractive to college graduates interested in teaching for several years but not for an entire career. Across-the-board salary increases may not improve the average quality of the teaching force at all. Although higher salaries do attract more talented graduates to teaching, they induce teachers to stay on the job longer, reducing opportunities to upgrade the teaching stock by hiring talented new applicants. The lower probability of receiving a teaching offer induces candidates with the best alternatives outside teaching, often the more academically able graduates, to choose a different occupation.13 A number of changes in the structure of teachers' salary schedules would make it easier to recruit talented college graduates in every subject field and would encourage them to obtain the training needed to learn to teach effectively. Flexible salaries for teachers in fields with shortages of teachers . College graduates trained in specialties with the best salaries outside of teaching are the least likely to enter teaching, the most likely to leave teaching after only a very few years in the classroom, and the least likely to return to teaching after a career interruption. This evidence that career decisions are related to opportunity cost suggests that a policy of paying premiums to teachers in fields with shortages is an efficient strategy for attracting skilled teachers in all subject areas (Kershaw and McKean, 1962). Fortunately, teacher union opposition to this policy has weakened in recent years (AFT Task Force on the Future of Education, 1986). Large pay increase for passing the performance-based licensing exam. With performance-based licensing, teachers with provisional teaching licenses, or "interns," would take the licensing examination after one or two years of supervised teaching. The payment of a large salary increase for passing the performance-based licensing exam serves two related purposes. It encourages apprentice teachers to seek training that will best help them pass the exam, and it encourages them to stay in teaching for at least a few years after passing the test. 13   This paragraph is a paraphrasing of the conclusion to Ballou and Podgursky's (1995) paper.

OCR for page 241
--> This is important because the effectiveness of teachers tends to increase markedly during their first years in the classroom (Murnane and Phillips, 1981). Incentives for licensed teachers to improve their teaching skills . Giving all teachers an incentive to invest in improving their performance is a major challenge for the nation's education system. Merit pay based on supervisors' assessments of individual teachers' performances was thought to be the answer, but it is not. The problem remains. In January 1995, as part of a long-term project in which a talented group of educators sought to define what good teaching is and devise a set of strategies to measure teaching effectiveness, the National Board for Professional Teaching Standards (NBPTS) provided board certification to 81 middle school teachers who had demonstrated a remarkably high degree of teaching skill over a year of difficult, performance-based assessments. In the years ahead the NBPTS plans to expand its scope markedly and develop assessment programs that will provide experienced teachers of every subject and grade level the opportunity to apply for board certification. Although it is not yet known whether the benefits of board certification will be sufficient to justify the enormous amount of work needed to prepare for and complete the NBPTS assessment, initial reactions are promising. A number of states and school districts have agreed to pay the approximately $1,000 fee to go through the assessment process and to provide substantial salary bonuses to teachers who become board certified (NBPTS, 1994). The NBPTS program has several attractive characteristics. By defining good teaching in concrete terms and developing methods of measuring teaching success, it may stimulate improvements in teaching training. Second, it may provide experienced teachers with incentives to seek help to improve their teaching. Third, it may provide financial rewards and enhanced status that will help keep extremely effective teachers in the classroom. Learning from New Initiatives Research has provided a great deal of information about the factors that influence who staffs the nation's schools. A variety of studies confirm that incentives matter. Incentives influence who goes into teaching, where teachers teach, how long teachers stay in teaching, and whether teachers return to the classroom after a career interruption. To conclude that incentives matter is not the same as knowing what set of incentives will best contribute to staffing the nation's schools with skilled teachers. In fact, many policy initiatives designed to improve the quality of the teachers in our nation's schools have had negative results. Examples include merit pay for individual teachers and extra pay for earning a master's degree. Getting the incentives right is difficult. Responses to policy initiatives, even ones that in theory seem promising, cannot be predicted. This is especially the case for initiatives that increase risk. For example, performance-based licensing

OCR for page 241
--> increases the risk of preparing to teach. A person might find that he or she is unable to demonstrate satisfactory teaching performance and is barred from public school teaching after having spent several years preparing for it. Offering a sizable salary increase to teacher interns who pass a performance-based licensing exam should compensate for this risk, but as yet there is no evidence of the size of the salary increase that would be needed to make teaching attractive to talented college graduates under a performance-based licensing system. The decentralized nature of the nation's education system provides many opportunities to test the effectiveness of alternative policies. Many of the 50 states and 15,000 school districts in the United States are experimenting with policies to increase the quality of teachers. These initiatives represent natural experiments from which a great deal can be learned, but only rarely are policy initiatives accompanied by systematic evaluations (Hanushek, 1994). Careful research on the consequences of policy initiatives is critical to improved knowledge of the policies that will contribute to staffing all of the nation's schools with highly skilled teachers. References AFT Task Force on the Future of Education. 1986. The Revolution That is Overdue. Washington, D.C.: American Federation of Teachers. Andrew, M.D. 1990. "Differences between graduates of 4-year and 5-year teacher education programs." Journal of Teacher Education 41(2):45–51. Ballou, D., and M. Podgursky. 1994. Do public schools hire the best applicants? Working paper, Department of Economics, University of Massachusetts, Amherst. Ballou, D., and M. Podgursky. 1995. "Recruiting smarter teachers." Journal of Human Resources, 30(2):326–338. Baugh, W. H., and J. A. Stone. 1982. "Mobility and wage equilibrium in the educator labor market." Economics of Education Review 2(3):253–274. Beaudin, B. Q. 1993. "Teachers who interrupt their careers: characteristics of those who return to the classroom." Journal of Policy Analysis and Management 15(1):51–64. Brewer, D. J. 1994. "Career paths and quit decisions: evidence from teaching." Working paper, Rand Corporation, Santa Monica, Calif. Darling-Hammond, L. 1990. "Teaching and knowledge: policy issues posed by alternative certification for teachers." Peabody Journal of Education 67(3):123–154. Darling-Hammond, L. 1994. "Who will speak for the children? How 'Teach for America' hurts urban schools and students." Phi Delta Kappan 76(1):21–34. Dolton, P. 1991. "The economics of UK teacher supply: the graduate's decision." Economic Journal 199(4):91–104. Educational Testing Service. 1988. Guidelines for Proper Use of NTE Tests. Princeton, N.J.: Educational Testing Service. Ehrenberg, R. G., and D. J. Brewer. 1994. "Do school and teacher characteristics matter? Evidence from High School and Beyond." Economics of Education Review 13(1):1–17. Ehrenberg, R. G., and D. J. Brewer. 1995. "Did teachers' verbal ability and race matter in the 1960s? Coleman Revisited." Economics of Education Review 14(1):1–21. Ferguson, R. F. 1991. "Paying for public education: new evidence on how and why money matters." Harvard Journal on Legislation 28:465–498.

OCR for page 241
--> Forbes, E. 1995. "Statement about the Praxis series: professional assessments for beginning teachers." Princeton, N.J.: Educational Testing Service. Gilford, D. J., and E. Tenenbaum, eds. 1990. Precollege Science and Mathematics Teachers. Washington, D.C.: National Academy Press. Gray, L., and others. 1993. New Teachers in the Job Market: 1991 Update. Washington, D.C.: Office of Educational Research and Improvement, U.S. Department of Education. Grissner, D. W., and S. Nataraj Kirby. 1992. Patterns of Attrition Among Indiana Teachers, 1965–1987. Santa Monica, Calif.: Rand Corp. Haney, W., G. Madaus, and A. Kreitzer. 1987. "Charms talismanic: testing teachers for the improvement of American education." Review of Research in Education 14:169–238. Hanushek, E. A. 1972. Education and Race. Lexington, Mass.: Health-Lexington. Hanushek, E. A. 1986. "The economics of schooling." Journal of Economic Literature 24:1141–1177. Hanushek, E. A. 1994. Making Schools Work: Improving Performance and Controlling Costs. Washington, D.C.: The Brookings Institution. Hanushek, E. A., and R. R. Pace. 1994. "Understanding entry into the teaching profession." In Choices and Consequences: Contemporary Policy Issues in Education, R. G. Ehrenberg, ed. Ithaca, N.Y.: ILR Press. Kershaw, J. A., and R. N. McKean. 1962. Teacher Shortages and Salary Schedules. New York: McGraw-Hill. Manski, C. F. 1987. "Academic ability, earnings, and the decision to become a teacher: evidence from the national longitudinal study of the high school class of 1972." In Public Sector Payrolls, D. A. Wise, ed. Chicago: University of Chicago Press. Monk, D. H., and S. L. Jacobson. 1985. "The distribution of salary increments between veteran and novice teachers: evidence from New York state." Journal of Education Finance 11:157–175. Murnane, R. J., and D. K. Cohen. 1986. "Merit pay and the evaluation problem: why most merit pay plans fail and a few survive." Harvard Educational Review 56(1):379–388. Murnane, R. J., and B. Phillips. 1981. "Learning by doing, vintage, and selection: three pieces of the puzzle relating teaching experience and teaching performance." Economics of Education Review 1. Murnane, R. J., J. D. Singer, and J. B. Willet. 1989. "Changes in teacher salaries during the 1970s: the role of school district demographics." Economics of Education Review 6(4):379–388. Murnane, R. J., J. D. Singer, J. B. Willet, J. J. Kemple, and R. J. Olsen. 1991. Who will teach? policies that matter. Cambridge, Mass.: Harvard University Press. National Board for Professional Teaching Standards (NBPTS). 1994. "State Action Supporting National Board Certification." Detroit: NBPTS. National Center for Education Statistics (NCES). 1992. Schools and Staffing in the United States: A Statistical Profile, 1987–88. Washington, D.C.: U.S. Department of Education, Office of Educational Research and Improvement. National Center for Education Statistics (NCES). 1993. Schools and Staffing in the United States: A Statistical Profile, 1990–91. Washington, D.C.: U.S. Department of Education, Office of Educational Research and Improvement. National Education Association (NEA). 1992. Status of the American Public School Teacher: 1990–91. Washington, D.C.: NEA. Nelson, F. H. 1985. "New perspective on the teacher quality debate: empirical evidence from the National Longitudinal Study," Journal of Educational Research 78(3):133–140. Rumberger, R. W. 1987. "The impact of salary differentials on teacher shortages and turnover: the case of mathematics and science teachers." Economics of Education Review 6(4):389–399. Schorr, Jr. 1993. "Class action: what Clinton's National Service Program could learn from 'Teach for America.'" Phi Delta Kappan (Dec.):315–318. Shulman, L. S. 1987. "Knowledge and teaching: foundations of the new reform." Harvard Educational Review 57(1):1–22.

OCR for page 241
--> South Carolina. 1989. 445 F. Supp. 1094 (D.S.C. 1977, aff'd; 434 U.S. 1026 (1978), 1108). Summers, A. A., and B. L. Wolfe. 1977. "Do schools make a difference?" American Economic Review 67(4):639–652. Winkler, D. R. 1975. "Educational achievement and school peer group composition." Journal of Human Resources 10(3):189–204. Wise, A. E., and L. Darling-Hammond. 1987. Licensing Teachers: Design for a Teaching Profession. Santa Monica, Calif.: Rand Corp.