Policy Brief The High Cost of Teacher Turnover Prepared for the National Commission on Teaching and America’s Future Thomas G. Carroll, President NATIONAL COMMISSION ON TEACHING AND AMERICA’S FUTURE Policy Brief The High Cost of Teacher Turnover Prepared for the National Commission on Teaching and America’s Future Thomas G. Carroll, Ph.D., President This policy brief is based upon research that was supported by Rockefeller Foundation Grants #2003WC097 & 2004WC129, along with Joyce Foundation Grant #29542 and Spencer Foundation Grant #200700049. Opinions in this paper reflect those of the National Commission on Teaching and America’s Future and do not necessarily reflect those of the granting agencies. Contents I. Introduction……………………………..…………………………………….………... 1 Teacher Turnover Is a Costly Problem Spiraling Out of Control…………………. 1 Box 1: Philadelphia, Pennsylvania: 70 Percent of New Teachers Drop Out in Six Years…………………………………………………………... 1 High-need Schools Pay the Highest Price………………………..……….............. 2 There’s a Hole in the Bucket………………………..…………………….............. 2 Chart 1: Teacher Attrition…………..…………………………………….............. 2 Is All Turnover Bad?................................................................................................ 3 II. The NCTAF Cost of Teacher Turn-over Study in Five Districts…………............... 3 Table 1: The Annual Cost of Teacher Turnover: A Five District Study…..……… 4 III. Solving the Problem of High Teacher Turnover…………………………………… 4 Step One: Measure Teacher Turnover and Its Costs………………........................ 4 Table 2: Cost of Teacher Turnover in Selected School Districts…………………. 5 Box 2: Fixing the Hole in the Bucket: Las Vegas, Nevada………………............... 6 Step Two: Invest in Well-Prepared Teachers and Comprehensive Induction Programs……………………………………… 6 Box 3: Improving Induction in Chicago……………………………………………7 Step Three: Transform School into Genuine Learning Organizations……………. 8 Box 4: The Benwood Initiative: Hamilton County, Tennessee…………………… 8 IV. Recommendations………………………………………………................…………. 9 Recommendations for Federal and State Action………………………….............. 9 Recommendation for School District Action………………………......................10 Appendix: Calculating the National Cost of Teacher Attrition…………………...11 I. Introduction Teacher Turnover Is a Costly Problem Spiraling Out of Control America’s schools are struggling with a Until we recognize that we have a retention problem growing teacher dropout problem that is costing the we will continue to engage in a costly annual recruit- nation over $7 billion a year. It is draining resources, ment and hiring cycle, pouring more and more teach- diminishing teaching quality, and undermining our ers into our nation’s classrooms only to lose them at ability to close the student achievement gap. a faster and faster rate. This will continue to drain The National Commission on Teaching and our public tax dollars, it will undermine teaching America’s Future (NCTAF) estimates that the national quality, and it will most certainly hinder our ability to cost of public school teacher turnover could be over close student achievement gaps. $7.3 billion a year1. This new estimate is significantly higher than the most recent estimate of $4.9 billion2 in It does not have to be this way. NCTAF has annual costs that was made in a report by the Alliance found that school leaders can reduce teacher turnover for Excellent Education in 2005, and takes into ac- and control their costs with coherent human resource count recent increases in the size of the teacher work- policies that begin with measuring teacher turnover force and the rate of teacher turnover. and understanding its consequences. Building on this NCTAF’s estimate, which is based on the knowledge, they should then focus on hiring well pre- cost generated by teachers who leave their school or pared teachers and giving them a strong start with district during a given year, does not include the dis- comprehensive induction programs. To sustain these trict’s cost for teachers who move from school to new teachers as they progress toward accomplished school within a district in search of a better position. teaching, their schools should be transformed into The estimate also does not include any federal or state genuine learning organizations. These policies will investments that are lost when a teacher leaves. If all achieve the greatest return on investment if they are of these costs were taken into account, the true cost to targeted at high-need schools. A discussion of these the nation would be far in excess of $7 billion. recommendations begins on page eight of this policy NCTAF’s findings are a clear indication that brief and additional information can be found at America’s teacher dropout problem is spiraling out of www.nctaf.org. control. Teacher attrition has grown by 50 percent over the past fifteen years. The national teacher turn- Philadelphia, Pennsylvania: 70 Percent of New over rate has risen to 16.8 percent. In urban schools it Teachers Drop Out in Six Years is over 20 percent, and, in some schools and districts, the teacher dropout rate is actually higher than the stu- In 1999, in the School District of Philadelphia, 919 dent dropout rate.3 new teachers began teaching and 12,000 students By allowing excessive teacher turnover to began ninth grade. Six years later, 58% of those continue unabated year after year, we have been dig- students had graduated from high school, but only ging a deep hole for ourselves. In 1994, former U. S. 30% of those new teachers were still teaching in Secretary of Education, Richard W. Riley, warned the Philadelphia. This means that the new teacher nation that we would need to hire two million teachers dropout rate (70%) over six years in Philadelphia within ten years to offset Baby Boom retirements. was higher than the student dropout rate (42%)5. Over the next decade we beat that goal by hiring ap- proximately 2.25 million teachers – but during that same decade we lost 2.7 million teachers, with over with over 2.1 million of them leaving before retire- ment.4 1. National Commission on Teaching and America’s Future, estimate based on NCTAF Teacher Turnover Cost Calculator applied to the Digest for Education Statistics data for all public school teachers in urban and non-urban public schools and districts (see Appendix to this Policy Brief). 2. Alliance for Excellent Education (2005). Teacher Attrition: A Costly Loss to the Nation and to the States, p. 1. 3. U.S. Department of Education, Teacher Attrition and Mobility: Results from the 2004-05 Teacher Follow-up Survey, pp. 7-9. 4. Useem, E., Offenberg, R., & Farley, E. (2007). Closing the Teacher Quality Gap in Philadelphia: New Hope and Old Hurdles. Philadel- phia, PA: Research for Action, p. 25; and Neild, R.C. & Balfanz, R. (2006). Unfulfilled Promise: The Dimensions and Characteristics of Philadelphia’s Dropout Crisis, 2000-2005, p. 18. 5. U.S. Department of Education, Teacher Attrition and Mobility: Results from the 2004-05 Teacher Follow-up Survey, pp. 7-9 1 High-need Schools Pay the Highest Price There's a Hole in the Bucket The consequences of high teacher turnover are Some would say that this is a simple recruit- particularly dire for our nation’s low-performing, high- ment problem – just find more teachers to replace those poverty schools. Many of these schools struggle to who leave. But focusing on the hiring of new teachers close the student achievement gap because they never won’t stem the costly exodus of teachers that is under- close the teaching quality gap – they are constantly cutting our nation’s ability to provide every child with rebuilding their staff. An inordinate amount of their quality teaching in a school organized for success. As capital – both human and financial – is consumed by this and previous NCTAF reports have noted, the con- the constant process of hiring and replacing beginning ventional wisdom that we can improve teaching quality teachers who leave before they have mastered the abil- by increasing the supply of new teachers is a misread- ity to create a successful learning culture for their stu- ing of the fundamental problem facing our schools to- dents. day. The problem is not finding enough teachers to do As a result of high turnover, high-need urban the job – the problem is keeping them in our schools. and rural schools are frequently staffed with inequitable The traditional children’s song “There’s a hole concentrations of under-prepared, inexperienced teach- in the bucket, Dear Liza, Dear Liza” is an apt metaphor ers who are left to labor on their own to meet the needs for the current state of affairs. In No Dream Denied of their students. This isolation has a crippling effect (2003), NCTAF reported that 287,370 teachers left on many new teachers who feel overwhelmed by the teaching during the 1999-2000 school year (220,582 challenges they face. They leave after several years of left for other pursuits and 66,788 retired). Recently working with a frustrating lack of support – perhaps released data from the 2003-04 Schools and Staffing they find a better school, but in too many cases they Survey show that this attrition is worsening. During abandon teaching altogether. And when they go, they the 2003-2004 school year, 332,700 left teaching leave a host of problems behind for the eager young (245,429 left for other pursuits, and 88,271 retired)6. teachers who take their place. Bright young teachers are leaving at an unsustainable rate. Teacher Attrition 350 300 ) 250 s d achers ousan 125000 e h T n t 100 (i 50 0 1987- 1993- 1999- 2003- 88 94 00 04 6. U.S. Department of Education, Teacher Attrition and Mobility: Results from the 2004-05 Teacher Follow-up Survey, pp. 7-9. 2 Is All Turnover Bad? achieve the maximum return on their investments in School leaders can’t manage what they don’t teachers. Many school leaders who are interested in measure. Because they have always relied on a steady controlling their turnover have asked for benchmarks or supply of new teachers, virtually no school district in turnover targets to shoot for. In setting benchmarks, the country has systems in place to track or control school leaders should look close to home. A school’s teacher turnover. Without these systems, they have no turnover target should be the turnover rate of the way to know how much money they are losing, schools with the highest performance in its district. whether they are losing good teachers or bad teachers, Similarly, a district’s turnover target should be the or which schools are suffering the greatest conse- turnover rate of the highest-performing districts in its quences of turnover. region. In various studies of teacher turnover, NCTAF has found more than a few school leaders who genu- II. The NCTAF Cost of Teacher Turn- inely believe that turnover is driven by forces that are over Study in Five Districts largely out of their control – and in the absence of ade- quate management information, they are probably Several previous studies have attempted to right. With accurate turnover and cost data, school estimate the cost of teacher turnover, but prior to leaders could better manage their human resources to this NCTAF study, only one has been based on ac- achieve a higher return on their teaching investments. tual cost data from specific districts7. Instead, the During our study of turnover we also have previous studies relied on turnover formulas derived found school leaders and members of the public who from industry to estimate turnover costs in educa- believe that high teacher turnover saves school districts tion. Those earlier estimates found that the nation money by lowering the average salary of teachers – was spending up to $4.9 billion dollars a year on high turnover schools and districts have more begin- teacher turnover.8 ning teachers who are concentrated at the lower end of But because the majority of these estimates the pay scale. This is a false economy. High turnover are not derived from a detailed analysis of actual schools incur significant costs associated with their school data, and because they do not provide school constant recruitment, hiring, training and separation of leaders with specific management tools they could teachers, and these costs are not being weighed against use to control costs, the findings of these previous possible salary savings; high turnover creates a con- studies have been downplayed by policymakers. stant drain on funding that offsets savings on low sala- To address this problem, the National ries for beginning teachers. And this is to say nothing Commission on Teaching and America’s Future of the lost teaching quality and diminished student (NCTAF) completed a pilot study of teacher turn- achievement in schools that are consistently staffed over and its costs in five school districts: Chicago with high concentrations of inexperienced beginners. Public Schools (Chicago, Illinois), Milwaukee Pub- As school leaders work to address this prob- lic Schools (Milwaukee, Wisconsin), Granville lem, it is important to remember that the goal is not to County Schools (Granville, North Carolina), Jemez achieve zero turnover. Teachers will continue to retire Valley Public Schools (New Mexico), and Santa or leave for personal reasons that can not be controlled Rosa Public Schools (New Mexico). This pilot by the district. Some beginning teachers may also find study was designed to develop tools for estimating that they are not well suited to teaching – they and their turnover costs, to test the feasibility of collecting districts may be better off if they leave teaching early. actual turnover cost data, and to determine the mag- Well-designed induction and peer mentoring programs nitude of these costs. that can help young teachers make this decision and are In both small and large districts, the study well worth the investment, especially if they are cou- found that when a teacher leaves, the costs of re- pled with coherent human resource management sys- cruiting, hiring, and training a replacement teacher tems that enable school leaders to know whether they are substantial. It is clear that thousands of dollars are losing effective or ineffective teachers. walk out the door each time a teacher leaves. The The goal is to ensure that teacher turnover is a cost per teacher leaver ranged from $4,366 in rural managed process and not a random series of events. Jemez Valley to $17,872 in Chicago. The total cost Each school district needs a comprehensive human of turnover in the Chicago Public Schools is over resource plan in place that enables school leaders to $86 million per year. 7. Shockley, R., Guglielmino, P., and Watlington, E. (2006). The Costs of Teacher Attrition. 8. Alliance for Excellent Education (2005). Teacher Attrition: A Costly Loss to the Nation and to the States, p. 1. 3 The NCTAF study also found that teacher III. Solving the Problem of High turnover is highest in high-minority, high-poverty, Teacher Turnover and low-performing schools. As a result, these at- risk schools spend a higher percentage of available One of the most important steps that school funding on teacher turnover than do high-performing, districts can take is to recognize that supply side so- low-minority, and low-poverty schools because they lutions focused on recruiting more teachers will not spend significantly more on teacher recruitment, hir- reduce the high cost of teacher turnover. School dis- ing, orientation, and separation. tricts must first recognize the importance of teacher For example, in Milwaukee low-performing retention and then develop a comprehensive and co- schools have double the teacher turnover of high- herent human resource strategy to reduce teacher performing schools. With an average school faculty turnover. of 55 and an average school cost of $8,300 per leaver, a typical low-performing school in MPS Step One: Measure Teacher Turnover and Its spends $67,000 more to deal with the consequences Costs of teacher turnover every year. If saved, this sum Education leaders need clear, current, accu- would allow the same school to provide additional rate data on teacher turnover and its costs, in formats support for new teachers or pay the salary of a read- that make it possible to analyze, manage, and control ing specialist. those costs. This is the first step toward reducing These costs do not include what may in fact turnover and making sound investments in teaching be the largest cost of teacher turnover: lost teaching quality. Prior to the NCTAF study, none of the five quality and effectiveness. Numerous studies have participating school districts tracked teacher turnover shown that teacher effectiveness improves with ex- or its costs. perience during the early years of a teacher’s career.9 The data collected by the districts for the New teachers struggle, but as they gain more knowl- study made it possible for them to analyze which edge and experience they are able to raise student teachers were leaving, from which schools, and how achievement. With the high rate of new teacher turn- much money was walking out the door each time a over, our education system is losing half of all teach- teacher left. Granville County Schools, for example, ers before they reach their peak effectiveness. Stu- learned that they were losing a high percentage of dents, especially those in at-risk schools, are too of- new teachers across all the schools in the district. ten left with a passing parade of inexperienced teach- Their teachers were leaving Granville and taking jobs ers who leave before they become accomplished edu- in surrounding districts. With a better handle on cators. teacher turnover data collected in this study, it is pos- The Annual Cost of Teacher Turnover: A Five District Study School District Number of Teachers Cost Per Teacher Leaver Chicago, Illinois 25,300 $17,872 Granville County, North Carolina 532 $ 9,875 Jemez Valley, New Mexico 41 $ 4,366 Milwaukee, Wisconsin 6,139 $15,325 Santa Rosa, New Mexico 58 unavailable Calculating the Cost of Teacher Leavers The cost of teacher leavers was calculated using district data on turnover and resources allo- cated to deal with turnover. Teachers who left the district altogether were considered leavers. In terms of costs, districts, along with a small subset of schools in each district, were asked to report time and money spent on activities associated with teacher leavers including: recruit- ment, hiring, administrative processing, professional development, and separation. 9. McCaffrey, Koretz, Lockwood, & Hamilton (2003); Rivkin, S. G., Hanushek, E. A., & Kain, J. F. (2005); Shkolnik, J., Hikawa, H., Suttorp, M., Lockwood, J.R., Stecher, B., & Bohrnstedt, G. (2002). 4 Even without a price tag on lost teaching quality and student opportunities to learn, the message is clear: high teacher turnover is draining school districts of precious dollars that could be used to improve teach- ing quality and student learning. Based on the pilot study, we have estimated the cost of teacher turnover in a number of selected school districts around the country. Cost of Teacher Turnover in Selected School Districts School District Annual Cost of Teacher Turnover Atlanta, Georgia $10,920,000 Baltimore, Maryland $19,013,750 Boston, Massachusetts $13,020,000 Cleveland, Ohio $12,538,750 Dallas, Texas $28,892,500 Detroit, Michigan $26,565,000 Denver, Colorado $14,988,750 Fairfax, Virginia $28,350,000 Hartford, Connecticut $4,462,500 Houston, Texas $35,043,750 Los Angeles, California $94,211,250 Louisville, Kentucky $18,208,750 Memphis, Tennessee $21,866,250 Miami, Florida $47,775,000 Nashville, Tennessee $14,393,750 New York City, New York $115,221,250 Oakland, California $12,005,000 Philadelphia, Pennsylvania $29,662,500 Pittsburgh, Pennsylvania $8,890,000 Prince Georges County, Maryland $23,292,500 Richmond, Virginia $6,072,500 San Francisco, California $11,865,000 Seattle, Washington $10,596,250 Washington, D.C. $16,598,750 You can calculate the cost for your own school district by using the NCTAF Teacher Turnover Cost Calculator at www.nctaf.org. 5 Fixing the Hole in the Bucket: Las Vegas, Nevada The Clark County School District in Las Vegas, Nevada is the fastest growing school district in the country. In the 2002-2003 school year, the Human Resources (HR) Department examined turnover data and recognized that the school district was losing a great deal of money because of high teacher turnover. Year after year the HR Department had worked hard to recruit and train high quality teachers only to see them leave the district. This cycle of recruitment, hiring, training, and turnover was both time-consuming and expensive. In examining the teacher data, the district found that twelve schools had especially high teacher turn- over – the average teacher tenure was 1.9 years, and the average experience of teachers in one of the 12 schools was only 1.3 years. The teacher dropout rate in the twelve schools was higher than the student drop- out rate. As a result, students were struggling on the state test and principals were struggling to create an effec- tive learning environment. Something needed to be done. In collaboration with the Clark County Education Association, the district piloted a multi-faceted program in 2004. First, the principals in the twelve at-risk schools were given a two month head start in the hiring process. The principals were able to access a large applicant pool and choose teachers who fit their school improvement plan. With this early hiring, the principals had time to fill vacancies and hold a summer urban studies program to prepare the newly-hired teachers. The pilot program also offered new teachers full- time mentoring and an advance of one column on the salary schedule. Of the first cohort, 91 percent of the teachers remained at their school after one year. An initial investment was needed to reduce teacher turnover and save money in Clark County. In this case, the initial investment came from a federal grant that paid for the summer urban studies program, the full-time mentoring, and the salary increase. The result of the initial investment was a drastic reduction in turnover and in the costs of recruiting, hiring, and training replacement teachers. And principals at the twelve schools had the opportunity to work with a stable group of teachers to transform their schools into genuine learning organizations. Three years after this initiative began, Clark County has sustained a retention rate of 85% to 95% in the twelve pilot schools, which are actually attracting teachers from higher-performing schools in the district. The program is now being expanded to 27 schools in the district. Step Two: Invest in Well-Prepared Teachers and prepared teachers reduced first year attrition by 50 Comprehensive Induction Programs percent.10 Well-prepared teachers possess strong con- Efforts to build a comprehensive, 21st cen- tent knowledge; they understand how students learn tury system of teacher induction continue to be sty- and demonstrate the teaching skills necessary to help mied by the fact that educators find it difficult to move all students meet high standards; they can use a variety beyond the factory-era mentality of the last century. of assessment strategies to diagnose student learning The culture of today’s schools continues to reinforce needs; and they can reflect on their practices to im- the practice of solo teaching in self-contained class- prove instruction in collaboration with their col- rooms. This mindset is compounded by a belief that leagues.11 Whether through traditional or alternative new teachers are interchangeable units who can easily preparation, teachers need to acquire the knowledge be replaced by the next cohort of beginners. As a re- and skills to be effective. It is not how new teachers sult, good teachers have little opportunity and few are prepared but how well they are prepared and sup- incentives to share their expertise with their col- ported in whatever preparation pathway they choose. leagues, and beginning teachers are left to fend for themselves without the collegial mentoring and coach- ing support they need to succeed. The first step to- ward breaking this mindset is to recognize the impor- tance of hiring and developing well-prepared teachers. In its 2003 report, No Dream Denied: A Pledge to America’s Children, NCTAF found that hiring well- 10. National Commission on Teaching and America’s Future (2003). No Dream Denied: A Pledge To America’s Children, p. 84 [www.nctaf.org]. 11. Ibid, p. 73 6 Once good teachers have been hired it is es- vestment in an intensive model of teacher induction sential to give them a strong start with the support they pays $1.66 for every $1 spent.”14 need to succeed. A recent national study of support Comprehensive induction programs are for new teachers found that comprehensive approaches based on four defining principles: (1) building and to teacher induction can reduce teacher turnover by deepening teacher knowledge; (2) integrating new more than 50 percent.12 In some well-designed pro- practitioners into a teaching community and school grams, such as those conducted by the New Teacher culture that support the continuous professional growth Center, or in Las Vegas and Chattanooga, the gains of all teachers; (3) supporting the constant develop- can be significantly greater. But in contrast to these ment of the teaching community in the school; and (4) comprehensive approaches, too many districts use a encouraging a professional dialogue that articulates the minimal approach to teacher induction that relies on an goals, values, and best practices of a community. untrained mentor or buddy who makes occasional vis- Comprehensive induction programs provide a its to new teachers. This buddy system has a negligi- package of support systems for a new teacher that in- ble effect, reducing new teacher turnover by just two cludes: (1) a mentor; (2) supportive communication percentage points.13 from the principal, other administrators, and depart- Because comprehensive induction programs ment chairs; (3) common planning or collaboration reduce teacher turnover and increase teacher effective- time with other teachers in the field; (4) reduced ness, they have been found to be very cost effective. preparations (course load) and help from a teacher’s In California, a recent study of a district induction pro- aide; and (5) participation in an external network of gram by the New Teacher Center found that “an in- teachers.”15 Improving Induction in Chicago NCTAF’s pilot study estimates that Chicago Public Schools (CPS) lost over $86 million in one year because of high teacher turnover. To reduce this loss, CPS has been developing a growing number of induc- tion programs for all new teachers in recent years. All first and second year teachers, for example, are re- quired to participate in the CPS GOLDEN Teachers Program, which assigns each new teacher a mentor and requires them to attend 15 workshops over the course of the school year. A 2007 report by the Consortium on Chicago School Research states that strong levels of support for new teachers greatly improved teachers' experiences and intentions to continue teaching in CPS.16 How- ever, more needs to be done. The report found that new teacher support was uneven across the district.17 About 20 percent of new teachers did not participate in CPS’ mandatory induction program. Seeking to address these issues, the Academy for Urban School Leadership (AUSL) in Chicago began offering a new approach to teacher preparation and support in 2001. Aspiring teachers learn to teach by apprenticing for a full year with a skilled teacher in classrooms similar to those in which they will eventu- ally teach. Annually, AUSL recruits 45 to 60 mid-career professionals and recent college graduates to par- ticipate in an intensive 12-month teacher preparation program. By 2006, 114 teachers had graduated from AUSL to teach in CPS. Ninety-five percent of them are still teaching, most staying in the high-need schools of their initial employment. To its credit, CPS has begun to upgrade the quality of new teacher induction and clearly recognizes that the loss of teachers is concentrated in low-performing and high-minority schools. CPS has committed to implement a comprehensive, two-year induction program run by the New Teacher Center. The program is being piloted in Englewood, an area with high teacher turnover and low student achievement. Prior to the induction program, Englewood had 140 vacancies in 20 schools. With comprehensive induction for new teachers, the number of vacancies in Englewood schools has fallen to single digits. 12. Ingersoll, R. & Smith, T. (2004). What are the effects of induction and mentoring on beginning teacher turnover?, p. 705. 13. Ibid, p. 705. 14. Villar, A. (2004). Measuring the Benefits and Costs of Mentor-Based Induction: A Value-Added Assessment of New Teacher Effective- ness Linked to Student Achievement, p. 36. 15. Fulton, K. (2005) Induction into Learning Communities, National Commission on Teaching and America’s Future, p. 8. 16. Kapadia. K, Coca V. & Easton J. (2007). Keeping New Teachers: A First Look at the Influences of Induction in Chicago Public Schools, Consortium on Chicago School Research, p. 2. 17. Ibid, p.2. 7