Research Report: A Coming Crisis in Teaching? Teacher Supply, Demand, and Shortages in the U.S.

Transcript

1 A Coming Crisis in Teaching? Teacher Supply, Demand, and Shortages in the U.S. Leib Sutcher, Linda Darling-Hammond, and Desiree Carver-Thomas SEPTEMBER 2016

2 Acknowledgments Acknowledgments The authors would like to thank a number of individuals for their valuable assistance in assembling The authors would like to thank a number of individuals for their valuable assistance in assembling and reviewing data that informed various aspects of our analysis: Sungti Hsu, Director of State and reviewing data that informed various aspects of our analysis: Sungti Hsu, Director of State Acknowledgments Affiliate and Partnership Support, American Association of Colleges for Teacher Education; and Affiliate and Partnership Support, American Association of Colleges for Teacher Education; and Allison Henderson, Vice President, and Carrie Murthy, Research Analyst, at Westat’s education Allison Henderson, Vice President, and Carrie Murthy, Research Analyst, at Westat’s education The authors would like to thank a number of individuals for their valuable assistance in assembling Acknowledgments group. We also thank the following LPI colleagues for their contributions to the research process: group. We also thank the following LPI colleagues for their contributions to the research process: and reviewing data that informed various aspects of our analysis: Sungti Hsu, Director of State Joseph Bishop, Tara Kini, Anne Podolsky, and Patrick Shields. In addition, we appreciate the insights Joseph Bishop, Tara Kini, Anne Podolsky, and Patrick Shields. In addition, we appreciate the insights Affiliate and Partnership Support, American Association of Colleges for Teacher Education; and The authors would like to thank Larry Rosenthal, Senior Lecturer of Public Policy at University of and feedback offered by our current and former colleagues: Roberta Furger, Roneeta Guha, Elise and feedback offered by our current and former colleagues: Roberta Furger, Roneeta Guha, Elise Allison Henderson, Vice President, and Carrie Murthy, Research Analyst, at Westat’s education California, Berkeley, for his insights and advice. We also thank the following LPI colleagues for their Levin-Guracar, and the entire Educator Quality Team. And we thank Laura Hayes and Lani Sinclair Levin-Guracar, and the entire Educator Quality Team. And we thank Laura Hayes and Lani Sinclair group. We also thank the following LPI colleagues for their contributions to the research process: valuable feedback: Leib Sutcher, Roneeta Guha, Tara Kini, and the entire Educator Quality team. We for their editing and design contributions to this project, and Lisa Gonzales for overseeing the for their editing and design contributions to this project, and Lisa Gonzales for overseeing the Joseph Bishop, Tara Kini, Anne Podolsky, and Patrick Shields. In addition, we appreciate the insights thank Bulletproof Services and Penelope Malish for their editing and design contributions to this editorial process. Without the generosity of time and spirit of all of the aforementioned, this work editorial process. Without the generosity of time and spirit of all of the aforementioned, this work and feedback offered by our current and former colleagues: Roberta Furger, Roneeta Guha, Elise project, and Lisa Gonzales for overseeing the editorial process. would not have been possible. would not have been possible. Levin-Guracar, and the entire Educator Quality Team. And we thank Laura Hayes and Lani Sinclair for their editing and design contributions to this project, and Lisa Gonzales for overseeing the Research in this area of work is funded in part by the S. D. Bechtel, Jr. Foundation. Core operating External Reviewers External Reviewers editorial process. Without the generosity of time and spirit of all of the aforementioned, this work support for the Learning Policy Institute is provided by the Ford Foundation, the William and Flora would not have been possible. Hewlett Foundation, and the Sandler Foundation. This report benefited from the insights and expertise of two external reviewers: Jason Grissom, This report benefited from the insights and expertise of two external reviewers: Jason Grissom, Associate Professor of Public Policy and Education, and Director of the Master of Public Policy Associate Professor of Public Policy and Education, and Director of the Master of Public Policy Program at Peabody College of Education and Human Development, Vanderbilt University; and Program at Peabody College of Education and Human Development, Vanderbilt University; and External Reviewers External Reviewers Richard Ingersoll, Board of Overseers Professor of Education and Sociology at the University of Richard Ingersoll, Board of Overseers Professor of Education and Sociology at the University of This report benefited from the insights and expertise of two external reviewers: Jason Grissom, This report benefited from the insights and expertise of two external reviewers: Li Feng, Brandon Pennsylvania. We thank them for the care and attention they gave the report. Any remaining Pennsylvania. We thank them for the care and attention they gave the report. Any remaining Associate Professor of Public Policy and Education, and Director of the Master of Public Policy Dee Roberts Excellence Assistant Professor at Texas State University; and Jesse Rothstein, Professor shortcomings are our own. shortcomings are our own. Program at Peabody College of Education and Human Development, Vanderbilt University; and of Public Policy and Economics and Director of the Institute for Research on Labor and Employment Richard Ingersoll, Board of Overseers Professor of Education and Sociology at the University of at University of California, Berkeley. We thank them for the care and attention they gave the report. Pennsylvania. We thank them for the care and attention they gave the report. Any remaining Any remaining shortcomings are our own. shortcomings are our own. The appropriate citation for this report is: The appropriate citation for this report is: A coming crisis in teaching? Teacher Sutcher, L., Darling-Hammond, L., & Carver-Thomas, D. (2016). A coming crisis in teaching? Teacher Sutcher, L., Darling-Hammond, L., & Carver-Thomas, D. (2016). . Palo Alto, CA: Learning Policy Institute. supply, demand, and shortages in the U.S. . Palo Alto, CA: Learning Policy Institute. supply, demand, and shortages in the U.S. The appropriate citation for this report is: https://learningpolicyinstitute.org/product/coming-crisis-teaching . This report can be found online at . This report can be found online at https://learningpolicyinstitute.org/product/coming-crisis-teaching Sutcher, L., Darling-Hammond, L., & Carver-Thomas, D. (2016). A coming crisis in teaching? Teacher And follow the conversation on Twitter at #SolvingTeacherShortages. And follow the conversation on Twitter at #SolvingTeacherShortages. supply, demand, and shortages in the U.S. . Palo Alto, CA: Learning Policy Institute. This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International This report can be found online at https://learningpolicyinstitute.org/product/coming-crisis-teaching . . License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc/4.0/ License. To view a copy of this license, visit http://creativecommons.org/licenses/by-nc/4.0/ . And follow the conversation on Twitter at #SolvingTeacherShortages. This work is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License. To view a copy of this license, visit . http://creativecommons.org/licenses/by-nc/4.0/ A COMING CRISIS IN TEACHING? LEARNING POLICY INSTITUTE | A COMING CRISIS IN TEACHING? B B LEARNING POLICY INSTITUTE | A COMING CRISIS IN TEACHING? LEARNING POLICY INSTITUTE | B

3 Table of Contents Executive Summary. ... 1 8 Teacher Supply, Demand, and Shortages in the U.S. ... I. Indicators of Shortages ... 10 10 Indicators of Shortages by Teaching Field ... Shortage Indicators by State ... 11 Indicators of Shortages of Teachers of Color ... 12 Shortage Indicators by Type of School and Student ... 13 National Indicators of Shortages ... 15 II. Teacher Demand ... 17 Demand Factors ... 17 Current Teacher Demand ... 21 III. Teacher Supply ... 23 Supply Factors ... 23 Current Teacher Supply ... 25 29 IV. The Teacher Labor Market: Predicting the Trends Ahead ... Supply Trends in Shortage Fields ... 30 Supply Trends in States ... 30 30 Availability of New Teachers ... ... 31 Salaries ... 32 Working Conditions ... 32 Attrition Rates ... Sources of Shortages ... 37 V. The Role of Teacher Attrition ... 38 The Costs of Attrition ... 40 Reasons for Attrition ... 42 VI. Policies to Meet Demand with a High-Quality Supply of Teachers ... 53 1. Creating Competitive, Equitable Compensation Packages ... 55 2. Increasing Supply in Shortage Fields and Areas ... 59 3. Improving Teacher Retention ... 63 4. Facilitating a National Labor Market for Teachers ... 68 VII. Conclusion ... 70 Appendix A: Methodology ... 71 79 Appendix B: State Indicators Influencing Supply and Demand ... C A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

4 Appendix C: Distribution of Uncertified and Inexperienced Teachers by State ... 82 84 Endnotes ... About the Authors ... 100 D A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

5 Figures and Tables Figures ... Figure 1: Projected Teacher Supply and Demand 2 Figure 2: Teacher Demand Factors Over Time ... 3 ... Figure 3: Projected Teacher Supply and Demand 15 Figure 4: Student Enrollment in Public Schools 1955–56 to 2024–25 ... 18 Figure 5: Public School Pupil-Teacher Ratios 1985–86 to 2024–25 ... 19 Figure 6: Components of Teacher Demand ... 20 Figure 7: Estimated New Teacher Hires Demanded ... 21 Figure 8: National Teacher Preparation Program Enrollments and Completers ... 26 Figure 9: Estimated Teacher Supply ... 27 Figure 10: Projected Teacher Supply and Demand ... 29 Figure 11: Teacher Reports of Administrative Support by State ... 33 Figure 12: Teacher Reports of Testing-Related Job Insecurity by State ... 34 Figure 13: Teacher Reports of Classroom Autonomy by State ... 35 36 Figure 14: Teachers Who Plan to Leave vs. Those Who Plan to Stay ... ... 38 Figure 15: Teacher Demand Factors Over Time 39 Figure 16: Attrition Rate Over Time ... Figure 17: Projected Supply and Demand with Varying Attrition Rates ... 40 Figure 18: Trends in Teacher Turnover ... 41 Figure 19: Distribution of Leavers by Voluntary Retirement Status ... 43 Figure 20: Teaching Experience of Public School Teachers, 1987–88, 2007–08, and 2011–12 ... 43 Figure 21: Teacher Turnover Rate by Primary Teaching Assignment ... 44 45 Figure 22: Leaver Rate, 1988–89 to 2012–13 ... Figure 23: Mover Rate, 45 1988–89 to 2012–13 ... Figure 24: Teacher Turnover by Teacher Characteristics ... 46 Figure 25: Turnover Rates by Region and District Type ... 48 Figure 26: Teacher Turnover by Source and State ... 48 Figure 27: Types of Reasons Given by Teachers for Leaving the Profession ... 49 Figure 28: Probability of Teacher Turnover in Relation to Administrative Support ... 52 57 Figure 29: What Would Bring Leavers Back? ... 72 Figure A1: Annual Change in Teacher Workforce Size Over Time ... Figure A2: Public School Experience-Based Continuation Rates, 2011–13 ... 73 74 Figure A3: Experience Distribution of New Teacher Hires, 2011–12 ... E A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

6 Tables Table 1: Reasons Teachers Listed as Important to Their Decision to Leave ... 50 Table A1: Workforce Growth ... 78 Table A2: Differences in Demand Estimates: Experience-Based vs. Age-Based Attrition Rates ... 78 79 ... Appendix B: State Indicators Influencing Supply and Demand Appendix C: Distribution of Uncertified and Inexperienced Teachers by State ... 82 F A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

7 Executive Summary Widespread media reports of local teacher shortages have become a hot topic in education since the summer of 2015. After years of teacher layoffs, districts began hiring again as the economy recovered from the Great Recession. Many were surprised to find they had serious difficulty finding qualified teachers for their positions, especially in fields like mathematics, science, special education, and bilingual education/English language development. A number of states greatly expanded emergency permits to allow hiring of untrained teachers to meet these demands—which is the classic definition of a shortage. To date, however, there has not yet been a detailed national analysis of the sources and extent of these shortages, or a prognosis for the future. This report details the outcomes of such a study, which analyzes evidence of teacher shortages, as well as national and regional trends in teacher supply and demand. Using several federal 1 databases, we examine the current context and model projections of future trends under several different assumptions about factors influencing supply and demand, including new entrants, re-entrants, projected hires, and attrition rates. We also investigate policy strategies that might 2 mitigate these effects based on research about effective approaches to recruitment and retention. We define shortages as the inability to staff vacancies at current wages with individuals qualified to teach in the fields needed. We find strong evidence of a current national teacher shortage that could worsen by 2017–18, if current trends continue. Combining estimates of supply and demand, our modeling reveals an estimated teacher shortage of approximately 64,000 teachers in the 2015–16 school year. By 2020, an estimated 300,000 new teachers will be needed per year, and by 2025, that number will increase to 316,000 annually. Unless major changes in teacher supply or a reduction in demand for additional teachers occur over the coming years, annual teacher shortages could increase to as much as 112,000 teachers by 2018, and remain close to that level thereafter. Based on the evidence available, the emerging teacher shortage is driven by four main factors: • A decline in teacher preparation enrollments, District efforts to return to pre-recession pupil-teacher ratios, • • Increasing student enrollment, and • High teacher attrition. The labor market should respond to the availability of jobs, so we can expect some increase in supply, but the extent of the increase and its distribution across subject fields and locations are likely to be uneven. 1 We analyzed the federal Schools and Staffing Surveys and Teacher Follow-Up Survey databases from 2012 and 2013, along with Baccalaureate and Beyond 2008:2012 databases, and the Higher Education Act Title II data from 2005 through 2014, as well as more recent data from the State of California. 2 A more complete review of the evidence regarding effective recruitment and retention is also available at Podolsky, A., Kini, T., Bishop, J., & Darling-Hammond, L. (2016). Solving the Teacher Shortage: How to Attract and Retain Excellent Educators. Palo Alto, CA: Learning Policy Institute. 1 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

8 Figure 1 Figure 1 Projected Teacher Supply and Demand 300,000 Estimated Demand 250,000 200,000 Estimated Supply 150,000 100,000 NUMBER OF TEACHERS 50,000 0 2021 2025 2005 2009 2013 2017 Projected Estimated Supply Estimated Supply Projected New Hires Demanded Actual New Hires Demanded Note: The supply line represents the midpoints of our upper- and lower-bound teacher supply estimates (see Figure 10 for full analysis). Source: U.S. Department of Education, multiple databases (see Appendix A). Trends in Demand Teacher demand is on the rise, as a function of changes in student enrollment, shifts in pupil- teacher ratios, and most significantly, high levels of teacher attrition. Based on the most recent data available, our modeling shows that teacher demand increased sharply after the Great Recession, leveling off at around 260,000 teacher hires annually by 2014. Projections show a large increase in 2017–18 and a projected plateau, bringing annual hires demanded to approximately 300,000 teachers a year. • After relatively flat student enrollment growth for the past decade, the National Center for Education Statistics (NCES) predicts the school-going population will increase by roughly 3 million students in the next decade. • Districts are looking to reinstate classes and programs that were cut or reduced during the Great Recession. It would require hiring an additional 145,000 teachers, on top of standard hiring needs, to reduce average pupil-teacher ratios from the current 16-to-1 to pre-recession ratios of 15.3-to-1. • High levels of attrition, estimated to be nearly 8% of the workforce annually, are responsible for the largest share of annual demand. The teaching workforce continues to be a leaky bucket, losing hundreds of thousands of teachers each year—the majority of them before retirement age. Changing attrition would reduce the projected shortages more than any other single factor. 2 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

9 FIGURE 13 Figure 2 Teacher Demand Factors Over Time 300,000 250,000 200,000 150,000 TEACHERS 100,000 50,000 NUMBER OF 0 -50,000 -100,000 1 201 2012 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2014 1989 2025 2013 2024 2015 2016 2017 2018 2019 2020 2021 2022 2023 Retirement Attrition Preretirement Attrition Workforce Change: Workforce Change: Pupil-Teacher Ratio Student Enrollment Note: Negative number of teachers represents a decrease in the total number of teachers. Source: U.S. Department of Education, multiple databases (see Appendix A). Trends in Supply Increased demand would not be an immediate reason for concern—if there were enough qualified teachers to enter the classroom, or if we could reduce the number of teachers leaving the classroom. Unfortunately, the supply of new teachers is atypically low and has been declining. The number of re-entrants (those who have stepped out of teaching) depends a great deal on whether policies make teaching an attractive and accessible possibility. Between 2009 and 2014, the most recent years of data available, teacher education • enrollments dropped from 691,000 to 451,000, a 35% reduction. This amounts to a decrease of almost 240,000 professionals on their way to the classroom in the year 2014, as compared to 2009. Although teacher re-entrants make up one-third to one-half of each year’s supply, • depending on aspects of the economy that make teaching more or less attractive, securing teachers even at the high end of this range will not be enough to overcome shortages. In theory, the pool of former teachers is large, but estimates suggest only around a third of teachers who exit the profession ever return. 3 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

10 Projections incorporating historical data on the teacher pipeline and estimates of re-entrants show a steady decline in teacher supply. According to our model, 2016 will have the lowest number of available teachers in 10 years—between 180,000 and 212,000 teachers. This projection varies depending on the percentage of newly prepared teachers who actually enter the profession and the number of former teachers who return to the classroom as re-entrants. The Significance of Attrition In times of shortages, it is most common to focus attention on how to get more teachers into the profession. However, it is equally important to focus on how to keep the teachers we have in the classroom. In fact, as we show in this report, reducing attrition by half could virtually eliminate shortages. Compared to high-achieving jurisdictions like Finland, Singapore, and Ontario, Canada—where only about 3% to 4% of teachers leave in a given year—U.S. attrition rates are quite high, hovering near 8% over the last decade, and are much higher for beginners and teachers in high-poverty schools and districts. If attrition rates were reduced to the levels of those nations, the United States would eliminate overall teacher shortages. In order to reduce attrition, we must know why people are leaving the profession, who is leaving the profession, where attrition is the greatest, and what factors are associated with different rates of attrition. • Why. Contrary to common belief, retirements generally constitute less than one-third of those who leave teaching in a given year. Of those who leave teaching voluntarily, most teachers list some type of dissatisfaction as very important or extremely important in their decision to leave the profession. Areas of dissatisfaction include concerns with the administration, ranging from lack of support to lack of input and control over teaching decisions; testing and accountability pressures; dissatisfaction with the teaching career; or unhappiness with various working conditions. The next largest category of reasons (43%) is associated with family or personal factors, including parental leave. Attrition varies by teacher subpopulations: Teachers with little preparation tend • Who. to leave at rates two to three times as high as those who have had a comprehensive preparation before they enter. Teachers in high-poverty, high-minority schools tend to have higher rates of attrition, as do teachers of color, who are disproportionately represented in these schools. • Where. Teacher attrition rates also vary considerably across the country. The South has a particularly high turnover rate (movers and leavers) compared to the Northeast, Midwest, and West. For most regions, teacher turnover is higher in cities than in suburban or rural districts. Administrative support is the factor most consistently associated • Associated Factors. with teachers’ decisions to stay in or leave a school. Our study found that teachers who find their administrators to be unsupportive are more than twice as likely to leave as those who feel well-supported. Many other factors that emerge from research on attrition are also associated with the quality of school leadership, including professional learning opportunities, instructional leadership, time for collaboration and planning, collegial relationships, and decision-making input. 4 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

11 Variations in Shortages Teacher shortages are not felt uniformly across all communities and classrooms, but instead affect some states, subject areas, and student populations more than others, based on differences in wages, working conditions, concentrations of teacher preparation institutions, as well as a wide range of policies that influence recruitment and retention. State-Level Shortages Currently reported shortages vary from state to state. California, for example, is experiencing severe shortages. The number of teachers on substandard credentials, an indicator that districts were unable to find fully-prepared teachers, increased by 63%, and comprised one-third of all credentials issued in 2015. Nearly half of all entering special education teachers were underprepared for the classroom. States like Arizona, Nevada, Oklahoma, and Washington also reported severe shortages. Meanwhile, higher-spending Massachusetts still reported an overall teacher surplus. Even there, however, schools are experiencing some shortages in special education and English learner education. Even though the teacher labor market might be balanced at the state level, subjects or regions within the state may be experiencing shortages. The factors influencing shortages—wages, working conditions, and attrition rates—vary substantially from state to state, as we show in Appendix B. These disparities, which are related to policy differences, create very different labor markets from one state to the next. For example, wages are generally highest and attrition rates are lowest in the Northeast, while salaries are lower and turnover rates are higher in the South and some parts of the West. Subject Area Shortages States across the country are currently experiencing subject area teacher shortages. In the 2015–16 school year, 48 states and the District of Columbia reported shortages in special education; 42 states plus DC did so in mathematics; and 40 states and DC reported teacher shortages in scienc e. In a 2014–15 educator supply and demand survey, all 10 special education subgroups were listed as severe shortage areas, comprising more than half of all severe shortage areas. Along with mathematics and science, this survey identified shortages in bilingual education/teachers of English learners. Equity Concerns Students in high-poverty and high-minority settings bear the brunt of teacher shortages. Considerable evidence shows that shortages historically have disproportionately impacted our most disadvantaged students and that those patterns persist today. Nationally, in 2013–14, on average, high-minority schools had four times as many uncertified teachers as low-minority schools. These inequities also exist between high-poverty and low-poverty schools. When there are not enough teachers to go around, the schools with the fewest resources and least desirable working conditions are the ones left with vacancies. 5 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

12 Policy Recommendations There are many policy decisions that can be made to relieve teacher shortages. These are generally aimed either at increasing the attractions to teaching or lowering the standards to become a teacher. Short-term solutions may temporarily curb the fear of empty classrooms, but, as we found, they can often exacerbate the problem over the long haul. For example, if teachers are hired without having been fully prepared, the much higher turnover rates that result are costly in terms of both dollars spent on the replacement process and decreases in student achievement in high-turnover schools. Long-term solutions focusing on recruitment and retention can ease the shortage, while also prioritizing student learning and a strong teacher workforce. Based on research we reviewed on what matters for recruiting and retaining teachers, policies should focus on: 1. Creating competitive, equitable compensation packages that allow teachers to make a reasonable living across all kinds of communities: • Leverage more competitive and equitable salaries so districts serving high-need students have a fair shot at recruiting well-qualified educators. including housing Create incentives that make living as a teacher more affordable, • supports, child care supports, and opportunities to teach or mentor after retirement to more effectively recruit and retain teachers. through 2. Enhancing the supply of qualified teachers for high-need fields and locations targeted training subsidies and high-retention pathways: • Offer forgivable loans and service scholarships to attract and retain teachers to high-need fields and locations. • Create career pathways and “Grow Your Own” programs to prepare committed individuals from urban and rural school districts. Establish teacher residency models for hard-to-staff districts to recruit and retain • talented and diverse candidates in high-need schools, while better preparing them for the challenges they will face. through improved 3. Improving teacher retention, especially in hard-to-staff schools, mentoring, induction, working conditions, and career development: • Develop strong, universally available mentoring and induction programs to increase retention and help slow the revolving door of beginning teacher turnover. including supportive working conditions, Create productive school environments, • administrative supports, and time for teachers’ collaborative planning and professional development—all of which help attract and keep teachers in schools. • to develop principals and district leaders who Strengthen principal training programs can create productive teaching and learning environments that have a major impact on a teacher’s decisions to stay or leave the classroom. 6 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

13 4. Developing a national teacher supply market that can facilitate getting and keeping teachers in the places they are needed over the course of their careers: • by removing unnecessary interstate barriers so Support teacher mobility states with teacher surpluses in particular fields can be connected to states with corresponding shortages. Conclusion The teacher shortage provides an opportunity for the United States to take a long-term approach, as was done in medicine more than half a century ago, to mitigating current shortages, while establishing a comprehensive and systematic set of strategies to build a strong teaching profession. At first, the price tag for these investments may seem substantial, but evidence suggests that these proposals would ultimately save far more in reduced costs for teacher turnover and student underachievement than they would cost. Preventing and solving teacher shortages so that all children receive high-quality instruction in every community every year is essential in a 21st century economy for the success of individuals as well as for society as a whole. 7 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

14 A Coming Crisis in Teaching? Teacher Supply, Demand, and Shortages in the U.S. As the 2015–16 school year got underway, headlines across the country broadcasted severe teacher shortages: 1 “Nevada needs teachers, and it’s shelling out $5 million to get them.” 2 “First marking period in Philly ends with many teacher shortages. ” 3 “[San Francisco] Principals say state teacher shortage now a crisis.” 4 “Why Oklahoma Is Racing to put Nearly 1,000 Uncertified Teachers in Its Classrooms.” These examples are but the tip of a large iceberg indicating current and future imbalances in the U.S. teaching force. In the five months between June 22 and November 22, 2015, more than 300 5 Two years earlier, there were only 21 newspaper articles appeared regarding teacher shortages. 6 such articles during the same time period. Many of the advertised shortfalls have been in mathematics and science. Special education is seeing the greatest shortages of all. A National Public Radio report indicated that 48 states and the District of Columbia have identified a shortage of teachers in special education or related services: Half of all schools and 90% of high-poverty schools are struggling to find qualified special education 7 As these statistics suggest, teacher shortages often have a disproportionate effect on the teachers. headline warned: “High-poverty schools often most disadvantaged students. One Washington Post 8 staffed by rotating cast of substitutes.” These shortfalls mark a dramatic change from the years of teacher layoffs that occurred during the economic recession of 2008 and the several years that followed. In those years, tens of thousands of pink slips were handed out each spring informing teachers they would not be needed the following 9 school year. State austerity measures resulted in eliminating support staff, reducing the number of 10 new teacher hires, and increasing class sizes. The recession left the public accustomed to a surplus of teachers, with policies aligned to this reality. However, as the economy improved and money began to come back into the system, districts have begun to hire again. Teacher demand has rapidly increased as schools begin to lower pupil-teacher ratios, and reinstate classes and programs that were reduced or eliminated in the Great Recession. This hiring increase comes at a time when teacher attrition is high, and teacher preparation program enrollments have fallen 35% nationwide in the last five years, a decrease of close to 11 240,000 teachers in total. Has the United States moved into an era of teacher shortages? If so, how large are the differences between supply and demand? Where and in what fields are they most severe? Will they persist? Most important, what can be done to prevent and mitigate the negative effects of such a teacher shortage? This report examines current indicators of a national teacher shortage, analyzes the severity and persistence of the labor market imbalance, discusses the impact on students and schools, and 8 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

15 proposes evidenced-based policies that could help create a sustainable supply of well-prepared teachers in the subjects and states where they are needed. Section I briefly reviews some of the current indicators of teacher shortages. Section II examines the factors that influence teacher demand, and then uses statistical techniques to model demand trends and project teacher demand into the future. Section III studies components of teacher supply, looking at current indicators and modeling supply trends. Section IV assesses supply and demand together to examine the current teacher labor market and implications for the U.S. education system, including factors influencing different states, regions, and types of communities. Section V highlights the role of teacher attrition in shortages, combining new evidence of trends with research on when and why teachers leave. Finally, section VI discusses policy opportunities to shift supply and demand to address both short-term teacher imbalances and long-term labor market stability in order to come to a better understanding of the current shortages, and provides evidence-based recommendations for future policy action. 9 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

16 I. Indicators of Shortages Not all commentary on current signs of teacher shortages has been unanimous. Some articles have 12 13 called the teacher shortage “a myth” or “overblown.” Dissent about teacher shortages can be rooted in different labor market realities that operate across locations and teaching fields, as well as different uses of the term “shortage.” is used to refer to an insufficient production of new teachers, Often, the term teacher shortage given the size of student enrollments and teacher retirements. In this narrow definition, a teacher shortage is measured only by teacher production in relation to these factors associated with teacher demand. However, a large body of research indicates that teacher staffing problems are driven by a myriad of factors, including not only production of new teachers in various fields, but also teacher turnover, changes in educational programs and pupil-teacher ratios, and the attractiveness of teaching generally and in specific locations. In this report, we use a theoretical framework of supply and demand that defines a teacher shortage as an inadequate quantity of qualified individuals willing to offer their services under prevailing wages and conditions. In other words, teacher shortages emerge in different fields and locations when there is an imbalance between the number of teachers demanded and the number of qualified teachers willing to offer their services to fill these demanded positions. This more robust definition allows for a more detailed diagnosis with clearer implications for policy analysis and recommendations. In this case, current data on the teacher labor market provide a compelling argument for teacher shortages, at least in some key teaching fields and locations. y employers have in filling One of the most widely used indicators of shortages is the difficult vacancies. When school districts are having trouble finding fully qualified applicants to fill empty positions, it is a sign of labor market imbalances in those fields or locations. Each year, the American Association for Employment in Education (AAEE) surveys a sample of higher As states are scrambling to find education institutions and districts across the country. In 2014–15, the survey found most special education teachers, many districts reporting as “a big challenge” not schools have no other choice but having enough candidates for open positions to hire unqualified teachers to fill (56%), and not finding candidates with the right credentials (53%). These proportions of these vacancies. districts were nearly double the rates from the 14 previous year. Indicators of Shortages by Teaching Field As reported by the media, districts, teacher education institutions, and states have all identified special education as the number one field with severe shortages. In 2015–16, 48 states and DC identified special education as a shortage area in their reports to the U.S. Department of Education. Most states identified special education as the most severe shortage they face, appearing on their 15 shortage reports for many grade levels and many subareas within special education. Similarly, in the 2014–15 AAEE educator supply and demand report, “considerable” shortages—the most severe rating—were reported in all 10 special education subareas, from dual certification to cognitive 10 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

17 disabilities. Of the 18 areas in which school districts reported “considerable” shortages, 10 of the 18 16 were in special education. As states are scrambling to find special education teachers, many schools have no other choice but to hire unqualified teachers to fill these vacancies. In California, for example, one study found that almost half (48%) of entering special education teachers in 2015 lack ed full preparation for 17 teaching. It is striking that the field that serves the most vulnerable students and, arguably, requires the most wide-ranging teacher knowledge—drawing on medical, psychological, and pedagogical fields—is increasingly populated by underprepared teachers. Mathematics and science are close behind special education as shortage areas. In the 2015–16 school year, 42 states and DC reported teacher shortages in mathematics, as did 40 states and DC 18 in science, a marginal increase from the previous year. With greater job opportunities offering stronger compensation in the broader labor market for individuals who have trained in mathematics and science fields, these subjects are up against a particularly difficult recruitment challenge. Along with special education, these fields have faced perennial staffing difficulties. Since 2000, over 19 One 10% of schools have reported serious difficulties filling mathematics and science vacancies. 2012 analysis suggested these problems are caused less by an underproduction of mathematics and 20 Since that time, demand has science teachers than by high levels of attrition for these teachers. rapidly increased, signaling a shift in the labor market. This shift could indicate that hiring for these already difficult-to-staff subject areas will only become more challenging. In 2014–15, the AAEE survey found that teacher preparation programs as well as school districts reported “considerable” or “some” shortage in mathematics and every science subject listed 21 (chemistry, biology, earth/physical science, physics). The fact that both school districts and oncerning, teacher preparation programs are reporting shortages in these fields is particularly c because this evidence suggests immediate help from the pipeline is not coming, and we may no longer have adequate production of mathematics and science teachers. In many states, bilingual education or teaching of English as a second language (ESL) is another 22 subject area with intense shortages. In all, 31 states and DC report shortages in these fields. The need for bilingual/ESL teachers clearly varies based on state demographics, so it makes sense that somewhat fewer states list shortages. However, the AAEE survey finds that the severity of the 23 World shortages is “considerable” for both teacher preparation programs and school districts. languages—including Chinese, Japanese, and classical languages—are also listed as showing considerable shortages. Shortage Indicators by State Each state experiences teacher supply and demand differently, because there are state-level labor markets created by different policies and contexts affecting teaching. These include funding levels and allocations, salary levels, teaching conditions, licensure and accreditation policies, concentration of preparation institutions, demographics of the teaching force, concentration or sparsity of the population, and topography, among others. 11 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

18 Regional, state, or local imbalances can lead to larger class sizes, cancellation of courses, or the assignment of underprepared or out-of-field teachers (those teaching subjects outside of their area of licensure or certification). National shortfalls can affect states that typically import many teachers from out of state more intensely than those that usually e xport teachers. States offering less competitive wages and working conditions will also typically feel shortages more acutely. Even when the teacher labor market is balanced overall, teaching fields or regions within a state may be prone to shortages. As just some examples of the variable landscape across states: California’s recent teacher shortages led to a tripling in the number of emergency and • temporary permits in the last three years. In 2014–15, when demand outstripped supply by more than 25%, fully 7,700, or just over a third of the credentials and permits issued that year, went to teachers who were not fully prepared for their teaching assignments. The greatest increase was in permits issued to individuals without preparation who were not even in an internship pathway that would support their training. Permits to underprepared teachers were most plentiful in special education, mathematics, and science, and in schools 24 serving concentrations of low-income and minority students. • Arizona’s teacher shortage is having an intense impact on the school system. In 2013–14, 62% of school districts had unfilled teaching positions three months into the school year. In the same school year, close to 1,000 teachers were on substitute credentials—a 29% increase from the previous year. With one of the highest turnover rates of any state and 24% of the workforce eligible to retire by June 2018, the future outlook points to 25 continued shortages. • Oklahoma’s teacher demand projections over the next five years are only slightly greater than its annual supply, or about 320 vacancies each year. However, with imbalances by field 26 and regional shortages in the southern half of the state, there has been a sharp increase in emergency credentials issued to underprepared teachers by the Education Department, 27 increasing from just 98 in 2010–11 to more than 900 by 2015–16. • By contrast, in high-spending Massachusetts, where there are many teacher preparing institutions, supply and demand projections show a current surplus. This varies by subject, though, with an expected surplus of more than 1,000 general education teachers (almost 2% of the teacher workforce) offset by shortages in special education (3% of the special education workforce) and English Language Learner (ELL) education (9% of the 28 ELL workforce). We discuss the factors influencing state and regional variations in supply and demand further in Section IV. Indicators of Shortages of Teachers of Color Shortages of teachers of color are different from other shortages in that there are not positions intended exclusively for teachers of color. Rather, there is a noticeable discrepancy between the proportion of teachers of color in the workforce and the proportion of students of color. Meanwhile, many districts are seeking to hire a greater number of teachers of color than are available in the 29 Despite the fact that students of color make up an increasing share of all students in the pool. 12 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

19 United States and are currently estimated to be in the majority, teachers of color comprised less 30 The number of teachers of color grew than 18% of the public school teacher workforce in 2012. from about 12% to nearly 18% between 1987 and 2012. At the same time, however, the share of 31 students of color increased by 75%: from 28% to 49% of students, leaving a very large differential. Although there have been increases in entry rates for teachers of color, successful minority teacher recruitment strategies are undone, in part, by high rates of teacher attrition, which we discuss in Section V of this report. While there are numerous theories touting the benefits of a racially diverse teacher workforce, research suggests benefits of two outcomes, in particular: (1) teachers of color enhance school experiences and academic outcomes for students of color, and (2) they are more 32 likely to seek out difficult-to-staff teaching positions in low-income communities of color. Both of these outcomes help combat acute teacher shortages and poor educational quality in the hardest- to-staff schools. Shortage Indicators by Type of School and Student Some of the largest variation in teacher shortages is not between states, but between schools, both within and across districts. Regardless of the state, students in high-poverty and high- minority schools typically feel the largest impact of teacher shortages. Historical patterns reveal a long-standing trend that has been a subject of many desegregation and school finance lawsuits: Students in high-poverty, high-minority schools are most likely to be taught by underprepared, inexperienced, and out-of-field teachers. These schools often experience difficulty hiring and high turnover on a regular basis, and they are the most severely affected when teacher shortages become 33 widespread. This happens, in part, because inequitable funding of schools leaves many low-wealth urban and rural communities with inadequate resources, so they must pay lower salaries and 34 typically have poorer working conditions. The last time shortages were common, in the late 1990s to the early 2000s, the inequitable distribution of teachers was widespread across the country. The practice of lowering teacher quality standards to fill vacancies, particularly in high-poverty, high-minority schools, was seen in California, Florida, Massachusetts, New York, South Carolina, and Texas, among other states. For instance, in California, there were over 42,000 teachers on emergency credentials, and students in high-minority schools were five times as likely to have a teacher who was underprepared as 35 In the 1990s, until the courts ordered a finance reform that students in low-minority schools. raised salaries, New York City hired nearly half of its new teachers without full preparation. In 2002, students in high-minority schools in South Carolina and Texas were four times as likely—and in Massachusetts, they were five times as likely—to have an underprepared teacher as students in 36 low-minority schools. The nationally representative Schools and Staffing Survey (SASS) reveals these patterns still existed 37 Across the country, teachers working on in 2011-12, even in a time of relative teacher surpluses. emergency credentials (the least qualified of the underprepared teachers), were three times more likely to serve in a high-poverty, high-minority school than in a low-poverty, low-minority school (4.0% vs. 1.4%). The combined categories of underprepared teachers were 61% more likely to be employed in a high-poverty, high-minority school than in low-poverty, low-minority schools (8.3% vs. 5.2%). These inequities were more intense in cities. Students in high-poverty, high-minority 13 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

20 schools in cities were nearly three times as likely to be taught by inexperienced teachers and by 38 individuals who were not “highly qualified” by the federal law’s definition. The Civil Rights Data Collection (CRDC), a universal survey of all public schools and districts, provides a more comprehensive overview of the distribution of teachers who lack certification. In 2013–14, on average, schools in the top quartile of minority student enrollment had four times as many uncertified teachers as schools in the bottom quartile of minority student enrollment. In Maryland, Louisiana, and Colorado, high-minority schools had, respectively, 12, eight, and five times as many uncertified teachers as their low-minority counterparts. In Colorado and DC, more than 20% of the teachers in high-minority schools were uncertified. Conversely, although there were still small equity gaps in states such as Idaho, Iowa, Kentucky, Nebraska, and North Carolina, both high-minority and low-minority schools employed fewer than 1% uncertified teachers. Nevada, South Dakota, and Vermont also had low levels of uncertified teachers (less than 1%) and no equity gap (see Appendix C). In an effort to hold states accountable for the inequitable distributions of teachers, the U.S. y plan to address differential Department of Education requires each state to create a state equit access to high-quality educators. The equity reports filed in 2015 reveal these same patterns across the nation, with unqualified, inexperienced, or out-of-field teachers disproportionately found in 39 high-poverty and high-minority schools. In Ohio, for example, high-poverty and high-minority schools had, respectively, 13 and 11 times as many unqualified teachers (teachers not licensed to teach that particular subject, student population, or grade level) as low-poverty and low-minority schools. In Arizona, high-poverty and high-minority schools had more than one in four of their teachers still in their first or second year of teaching, when they are typically much less 40 The equity reports filed effective than other teachers. This meant that the odds of these students being taught in 2015 reveal these by inexperienced teachers were about 70% same patterns across the greater than students in low-poverty and nation, with unqualified, low-minority schools. Although each state uses slightly different definitions to report inexperienced, or out-of-field equity comparisons of underprepared and teachers disproportionately inexperienced teachers, all but a handful of found in high-poverty and states have a higher percentage of teachers not fully certified, inexperienced, or out of high-minority schools. field in their high-poverty and high-minority schools than their low-poverty and low- 41 minority schools. These data can be interpreted as shortages in high-minority and high-poverty schools and/or as distributional problems within the contexts of districts and states. Either way, the extent to which underprepared teachers are hired in these schools is a constant factor that grows greater during times of broader shortages. 14 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

21 As we describe in the last section of this report, states and districts that have addressed these problems with forceful policies have reduced these disparities in the past. Without the appropriate policy response, however, in times of shortage, both the overall number of underprepared teachers and the inequities between schools tend to increase. National Indicators of Shortages National data on the indicators of teacher supply and demand point to the po tential for current and future shortages. The number of teachers entering the profession has significantly declined at the same time as more teachers are needed due to increasing student enrollments and distric ts’ efforts to replace the teachers and programs reduced during the Great Recession, between 2008 and 2012. Looking beyond simple indicators of supply and demand, statistical projections incorporating the most recent data on the teacher labor market suggest the current shortages will persist into the future if current trends continue (see Figure 3). Figure 1 Figure 3 Teacher Supply and Demand Projected 300,000 Estimated Demand 250,000 200,000 Estimated Supply 150,000 100,000 NUMBER OF TEACHERS 50,000 0 2021 2025 2005 2009 2013 2017 Estimated Supply Projected Estimated Supply Projected New Hires Demanded Actual New Hires Demanded Note: The supply line represents the midpoints of our upper- and lower-bound teacher supply estimates (see Figure 10 for full analysis). Source: U.S. Department of Education, multiple databases (see Appendix A). As we describe in this report, teacher supply and demand modeling based on data regarding new entrants, potential re-entrants, and projected new hires reveals an estimated current teacher shortage of approximately 64,000 teachers in the 2015–16 school year. Of roughly 260,000 new teachers demanded, the number of qualified entrants available was approximately 196,000. This shortage reflects the difference between the number of qualified new or re-entering teachers available and the number of teachers districts would ideally hire. As we have noted, the shortages 15 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

22 can be met by hiring individuals who are not fully prepared for their teaching assignments, hiring temporary substitutes (who may or may not be prepared) while a search continues, maintaining class sizes and pupil loads that are greater than desired, or canceling courses or programs for which teachers cannot be found. By 2020, an estimated 300,000 new teachers will be needed each year, and by 2025, that number will increase to 316,000. Unless major changes in teacher supply or a reduction in demand for additional teachers occurs over the coming years, annual teacher shortages could increase to as many as 112,000 teachers by 2018. Based on current data and these projections, it appears that teacher shortages are occurring—at least in some fields and locations—and may grow. To understand the full extent and seriousness of potential shortages, we must dig more deeply into the factors influencing teacher supply and demand, as well as the policies that can make a difference in the teacher labor market. Short-sighted solutions may temporarily curb fears of empty classrooms, but only long-term solutions focusing while also prioritizing student on sound recruitment and retention strategies can ease the shortage, learning and a sustainable future. 16 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

23 II. Teacher Demand The first component of the teacher labor market is teacher demand. Knowing how many teachers are needed is crucial to understanding labor market needs and thus ensuring that every community has access to a sufficient supply of well-qualified teachers. Demand in the teacher labor market can be thought of in two ways. Total teacher demand represents the total number of teachers required to educate the nation’s students. Next year, the total teacher demand will be roughly 3.1 million full-time-equivalent (FTE) public school teachers 42 in classrooms across the United States. Most of these positions are filled by teachers who remain 44 43 must be filled in the workforce from the year before. However, each year, about 250,000 positions due to some combination of the increase in student population, changing pupil-teacher ratios, and 45 teachers who leave the profession. This second way of considering teacher demand—the annual need for teachers to fill vacant positions—is what we focus on here. One of the biggest challenges in estimating demand is navigating the difference between Each year, about 250,000 ideal demand and actual demand. Ideal positions must be filled due demand requires defining the desired pupil- to some combination of the teacher ratio, geographic teacher distributions, and course requirements to determine the increase in student population, perfect number of teachers necessary each changing pupil-teacher ratios, year. The actual demand represents reality— and teachers who leave the need for teachers based on the number of teachers actually hired and employed. the profession. For example, in the Great Recession, actual demand for teachers dropped as budgets were cut, and schools could not afford to hire new teachers or even keep the teachers they already had. In this case, actual demand dropped, but ideal demand did not. In an ideal sense, schools would like, at a minimum, to be able to maintain the number of teachers and return to the class sizes and course offerings they had in place before the recession. The projections in this paper, while mindful of ideal demand, attempt to estimate actual demand using NCES projections of student enrollments, pupil-teacher ratios, and the teacher workforce. These estimates are just that—estimates. Using historical data and indications of future trends, these projections paint a potential picture of the workforce in the future. We first examine the factors that determine the quantity of teachers needed in the classroom. Next, we look at indicators that speak to current trends in demand. Finally, in order to further understand the teacher labor market and pressing policy needs, we use current data to model teacher demand 10 years into the future. Demand Factors Some elements of teacher demand, such as student enrollments, are predicted by trends not easily influenced by education policy. Others, such as pupil-teacher ratios and attrition rates, are very directly influenced by policy strategies. We discuss each of these in turn. 17 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

24 Student enrollment. Student enrollment directly influences teacher demand. An increase in the school-age population corresponds with an increase in the number of teachers needed in the education system, as long as pupil-teacher ratios remain constant. Future public school enrollment numbers can be estimated by looking at birth rates, public school attendance rates, and immigration 46 These indicate how many school-age children will enter school. Once the and migration patterns. number of students entering school is estimated, historical data can be used to model how many students will stay in school and for how long. Public school enrollment increased substantially (by 26%) from 1986 to 2007, growing from 39 million students to 49 million students. From 2007 to 2015, student enrollment has remained relatively flat, hovering around 49 million public school students (see Figure 4). NCES estimates a steady increase in public school student enrollment starting in 2016, growing from 50 million to 53 47 (see Figure 4). Although NCES does not specify what will drive this increase, the million by 2025 enrollment rate of 5- to 17-year-olds has changed less than three percentage points over the last 30 years, suggesting the majority of this growth, if the enrollment rates continue to remain relatively constant, will come from an increase of school-age children due to population growth (higher birth 48 rates and/or immigration), with teacher demand rising in response. Figure 4 Figure 2 Student Enrollment in Public Schools 1955–56 to 2024–25 50 million 40 million 30 million 20 million NUMBER OF STUDENTS 10 million 0 2020 1970 1980 1990 2000 2010 1960 Projected Actual Source: National Center for Education Statistics (2015). Digest of Education Statistics. Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. 49 The number Pupil-teacher ratios. Pupil-teacher ratio is another component of teacher demand. of additional teachers needed by a school district is dependent on the change in desired class 18 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

25 size, as well as any adjustments in how schools are staffed and how staff are used. To estimate pupil-teacher ratios, researchers first look at school budgets and economic conditions influencing a 50 district’s ability to hire more teachers. Second, state and local policies may influence pupil-teacher ratios by requiring certain class sizes. Current average public school pupil-teacher ratios are 16-to-1. Historically, pupil-teacher ratios have been slightly higher. In 1986, the average public school pupil-teacher ratio was just under 18-to-1. It remained over 17 until there was a large push in the late 1990s and early 2000s for 51 smaller class sizes and increased staffing for special education. By the 2008–09 school year, the national average pupil-teacher ratio was as low as 15.3-to-1. When the Great Recession hit and education budgets were slashed, average pupil-teacher ratios quickly increased to 16-to-1, and have remained there for some time (see Figure 5). These ratios vary across states, from a high of 24-to-1 52 (see Appendix B). in California to a low of 11-to-1 in Vermont FIGURE 3 Figure 5 Public School Pupil-T eacher Ratios 1985–86 to 2024–25 18:1 15:1 12:1 9:1 EACHER RATIO 6:1 PUPIL-T 3:1 0 2025 1995 2000 2005 2010 2015 2020 1985 1990 Projected Actual Source: National Center for Education Statistics (2015). Digest of Education Statistics. Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. Pupil-teacher ratios are incorporated into our analysis by using NCES pupil-teacher ratio projections. The NCES projections are produced by modeling both historical patterns and economic conditions, such as a teacher’s relative wage compared to that of other workers and state education budgets, to determine the relationship between these factors, and current and future pupil-teacher 53 The NCES projects that pupil-teacher ratios will remain at 16-to-1 through 2016 and then ratios. slowly dip downward to 15.3-to-1 by 2025, slightly below pre-recession levels (see Figure 5). The effects of such a decrease on demand are significant: Reducing the pupil-teacher ratio from 16-to-1 19 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

26 54 to 15.3-to-1 would require hiring an additional 145,000 teachers. Over a period of eight years, this reduction in the pupil-teacher ratio would increase annual demand by nearly 20,000 teachers a year. The final component of demand for new hires is attrition. The number of new teachers Attrition. 55 One form of demanded depends substantially on how many teachers leave the profession. 56 departure is retirement. This accounts for about one-third of teachers who leave. Preretirement attrition, due to school staffing decisions, life changes, or dissatisfactions with teaching, accounts for about two-thirds of all attrition and is estimated to have comprised around 66% of total new demand in the 2015–16 school year (see Figure 6). As we discuss in Section V, policies that change attrition rates can greatly influence demand and the extent of shortages. Based on the Schools and Staffing Surveys (SASS) from 2011–12 and the corresponding Preretirement attrition accounts Teacher Follow-Up Survey (TFS) from 2012–13, . . . for about two-thirds of the overall public school teacher attrition 57 all attrition. rate is calculated to be 7.68%. To put this in perspective, of the 3.1 million FTE teachers in 2011, 238,310 teachers did not teach the 58 following year. In 1989, the attrition rate was below 6%; however, this rate climbed to 8.4% in 2004 and has remained near 8% since. The difference between a 6% and 8% attrition rate might seem trivial, but in 2015–16 alone, this would cut demand by nearly 25%, eliminating the need to replace approximately 63,000 teachers. Each year, different factors contribute varying amounts to teacher demand. For example, in some years, student enrollment growth is responsible for a noticeable portion of increased demand, while in other years, demand is driven almost entirely by attrition. Figure 6 shows the estimated 59 In each of breakdown of the different components comprising demand in 2012, 2016, and 2020. these years, the largest portion of demand is driven by preretirement attrition. FIGURE 4 Figure 6 Components of Teacher Demand 2015 – 16 – 12 2011 2019 – 20 1% 1% 3% <1% 7% 6% 31% 31% 28% 59% 67% 66% Workforce Growth Due to Pupil-Teacher Ratio Reduction Retirement Attrition Workforce Growth Due to Student Enrollment Increases Preretirement Attrition Source: U.S. Department of Education, multiple databases (see Appendix A). 20 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

27 Current Teacher Demand Currently, all of these factors play a role in boosting teacher demand. Student enrollment is on the rise, and if economic conditions hold firm, class sizes are predicted to decrease as districts move toward pre-recession class The largest portion of demand is offerings and teacher workloads. Furthermore, driven by preretirement attrition. attrition remains at relatively high rates with little indication of any change to come in the near future. To model demand more precisely, we use the public school teacher projections from 2000 to 2025 60 We refine these estimates using several nationally representative data produced by the NCES. sources that examine teachers and schools over time: the Schools and Staffing Survey (SASS) Teacher File 2011–12, the SASS Teacher Follow-Up Survey 2012–13, and Common Core of Data 61 For a full description of the methodology, see the technical (CCD) years 1999–00 through 2012–13. notes in Appendix A. In Figure 7, the solid line represents observed demand, while the dashed line represents projected 62 The impact of the Great Recession is evident by the sharp decrease in demand between demand. 2008 and 2012 as a result of budget cuts and layoffs that occurred during these times of austerity. Teacher demand shows a sharp increase after 2012, leveling off at around 260,000 teacher hires by 2014. In 2017–18, there is a large projected increase, which brings annual demanded teacher hires to 300,000 a year. This model projects the largest demand increases are yet to come, eventually requiring over 300,000 public school teacher hires a year, the largest number of annual teacher hires in the last decade. FIGURE 5 Figure 7 Estimated New Teacher Hires Demanded 300,000 250,000 200,000 TEACHERS 150,000 100,000 NUMBER OF 50,000 0 2016 2025 2013 2022 2010 2004 2019 2007 Actual New Hires Demanded Projected New Hires Demanded Note: Data for teachers are expressed in full-time equivalents. Source: U.S. Department of Education, multiple databases (see Appendix A). A COMING CRISIS IN TEACHING? 21 | LEARNING POLICY INSTITUTE

28 This large increase in demand is not inherently In the years ahead, a sizable a cause for alarm. After a period of austerity, it is no surprise that schools are attempting demand is projected: New hires to return to pre-recession conditions. In fact, will fill positions that were lost the downturn in the teacher workforce during during the Great Recession, as the Great Recession was relatively modest compared to its considerable growth in the well as positions that are created years before. Between the late 1980s and 2008, to keep up with increased the teacher workforce grew at more than twice student enrollment and smaller the rate of student enrollment growth, driven not only by declines in pupil-teacher ratios, class sizes. but also by programmatic changes that called for more mathematics and science teachers when high school standards were raised, as 63 Other well as more special education and bilingual/ESL teachers when services were expanded. changes, like expansions of kindergartens from part- to full-day and increases in graduation rates 64 that kept more high school students in school, also influenced the need for teachers. In the years ahead, a sizable demand is projected: New hires will fill positions that were lost during the Great Recession, as well as positions that are created to keep up with increased student enrollment and smaller class sizes. The question is, as demand rapidly rises, can teacher supply expand to keep pace and provide all classrooms with well-qualified teachers? Alternatively, can we reduce attrition sufficiently to make the supply of teachers more adequate? 22 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

29 III. Teacher Supply The second component of the teacher labor market is teacher supply. States and districts need accurate information about production and availability of future teachers to assess and plan for their workforce needs. To evaluate the current condition of teacher supply, we first identify indicators of teacher preparation program enrollments and completions. Next, we take a more complex methodological approach to model teacher supply over the last decade and to project supply into the next decade. To look further at labor market trends, we disaggregate national supply data by subject area. These analyses reveal a significant reduction in teacher supply over the last five years and shortages continuing over the next decade, if current teacher preparation enrollment trends continue. Given these sharp decreases, it is critical to understand the policy levers surrounding recruitment and retention that could support a high-quality, sustainable teacher workforce. Like teacher demand, there are several different ways to think about teacher supply. One way to analyze supply-side dynamics is total teacher supply. Total activated teacher supply could be conceptualized as the entire teacher workforce, for example, the 3.1 million FTE public school 65 The majority of these teachers remain in the teachers in classrooms across the United States. workforce the following year; however, a substantial number of positions vacate and must be filled. Examining the pool of teachers who will potentially be available to fill empty classrooms is a second way of looking at teacher supply and is most relevant to labor force analysis. This teacher supply can be thought of as the number of potential teacher entrants who were not teaching in the previous year, who are either new entrants to the profession or re-entrants who have stepped out from teaching for a period of time. At the local or state level, individuals who may be recruited from other schools or states are also considered a source of supply. The teacher supply that has the potential to meet the number of newly demanded employees is adjusted by two main 66 influences: new entrants and re-entrants. When fully credentialed teachers Individuals who hold valid teaching cannot be found, sometimes credentials, whether or not they are currently in the workforce, are thought of as the supply teachers are hired who are not pool. Some of these individuals may enter or fully prepared for the subjects re-enter teaching under the right conditions. they teach. Understanding these conditions is important to policymakers, particularly in times when shortages emerge. Supply Factors 67 New entrants. In most years, new entrants comprise over half of the annual teacher supply. The potential supply of new entrants is often estimated as the number of recent fully credentialed teachers who have not yet taught. When fully credentialed teachers cannot be found, sometimes teachers are hired who are not fully prepared for the subjects they teach. State laws generally indicate that underprepared teachers cannot be hired when fully credentialed teachers are available. Thus, when underprepared teachers are hired, it typically indicates evidence of a shortage. 23 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

30 Estimating the number of new entrants into the workforce in a given year can be tricky. Universal data collected by the federal government on teacher preparation programs can be used to help deduce the number of candidates who enroll in and complete teacher preparation programs. However, these data do not reveal how many and how soon recent completers end up teaching in the classroom. Longitudinal studies conducted by institutions of higher education that follow college graduates in their educational and occupational pursuits can provide estimates of when and how many college graduates enter teaching. These sample sizes are relatively small and non- representative of all teachers. The federal government periodically conducts some surveys, like the Baccalaureate and Beyond survey, that follow college graduates into their careers, which can provide some information as well. A combination of these types of data—individuals who have completed teacher preparation programs and individuals who have accepted teaching jobs—is necessary to properly investigate patterns of entry into the teacher labor market. Past estimates in the literature generally find between 70% and 90% of newly minted teachers end up in the classroom in the year following their teacher preparation, with higher rates of 68 entry from postbaccalaureate programs than from undergraduate programs. Historically, new 69 teachers often came straight from undergraduate teacher preparation programs. Today, it is 70 increasingly common to obtain a teacher credential through a postgraduate program. One survey in 2011 found that about one-third (34%) of respondents had entered the profession through a 71 postbaccalaureate program. While some education graduates select occupations other than teaching, others enter teaching after a delay. The nature of this delay from preparation to practice adds delayed entrants as a sub- category of teachers that must be accounted for when examining new entrants. Estimates of completer-to-teacher entrant rates are tied to labor market conditions. For instance, the comparative wage a teacher is paid relative to other jobs in the economy can affect the decision 72 of a newly qualified teacher to enter the classroom. Furthermore, the relative availability of teaching positions alters the likelihood that a newly qualified teacher who wants to enter the profession finds a job. In times of high demand, when there are more job openings, the percentage 73 of hires coming straight from a teacher preparation program increases. Some evidence suggests delayed entry also varies by subject area. High rates of delayed entry are evident in some fields such as physical education and elementary education, but a recent study using national data found that almost all newly prepared mathematics and science teachers entered 74 teaching within a year of graduating. We used survey responses from the 2008:2012 Baccalaureate and Beyond (B&B), a longitudinal study that follows college graduates after they receive their BA degree, to determine the percentage of students who were prepared to teach but did not end up in the classroom in a four-year period. The data provide an estimate that 75% of newly minted teachers ended up teaching within four 75 years. It is worth noting that the B&B sample includes only recent BA graduates; therefore, this estimate does incorporate many teachers coming from graduate-level pathways or nontraditional routes, which typically have higher entry rates. The four-year average has the benefit of accounting for many delayed entrants. However, the years sampled were during the Great Recession in precisely the years that many beginning teachers were being laid off, and fewer new teachers were able to get teaching jobs than would normally be the 24 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

31 case. Thus, we think of this as a lower-bound estimate of entry rates into the profession. For this reason, our analysis used a lower bound of 75% and an upper bound of 90% to model the flow from teacher preparation to the classroom. Re-entrants. Re-entrants make up the other component of teacher supply. A proportion of teachers 76 each year leave the occupation for personal and professional reasons. These exits create a reserve 77 pool of teachers who have a teaching credential and experience but are not currently teaching. Some of these individuals come back to teaching at a later date as re-entrants. The size and willingness of candidates in the reserve pool to re-enter is difficult to estimate, but existing data allow us to examine the share of annual hires who come from this pool. In recent years, research 78 has found between 37% and 49% of new hires entering the workforce are re-entrants. Proportions of re-entrants as a share of total hires likely vary with labor market and economic conditions as 79 alternate employment opportunities fluctuate. For example, the 2011–12 estimate of 49% is likely elevated because after the Great Recession, as the workforce began to expand, schools rehired many teachers who had been laid off just a few years prior as re-entrants. In 2003–04 and 2007–08, the 80 proportion of new hires that were re-entrants was lower, at 41% and 37%, respectively. Research suggests that many re-entrants left with the intention of returning, including teachers 81 who stepped out for child-rearing or other personal reasons. Estimates suggest that just under a third (between 28% and 32.3%) of teachers who leave teaching come back to the classroom 82 On average, teachers who are female and who have more teaching experience within five years. 83 (controlling for age) are more likely to re-enter teaching, as are those who receive higher salaries. Most teachers who choose to re-enter teaching do so after just one or two years out of the classroom. The more time a teacher spends away from the classroom, the less likely he or she is to 84 return to the profession. Understanding that the proportion and rate Estimates suggest that just of re-entrants can vary, in our analysis we under a third (between 28% estimate an upper and lower bound for the rate at which teachers who leave return to the and 32.3%) of teachers who classroom within five years of leaving. The leave teaching come back to the 85 upper-bound rate is 32.3% and the lower 86 classroom within five years. bound is 28%, which are based on the range of estimates found in high-quality studies on teacher re-entrance. Current Teacher Supply Currently, key indicators point to a significant decrease in the supply of teachers. Enrollments in teacher preparation programs across the country have decreased steadily in recent years. Between 2009 and 2014, the most recent years of data available, there was a 35% reduction in undergraduate and postbaccalaureate teacher preparation enrollments, which amounts to a decrease of almost 240,000 fewer professionals working their way toward the classroom in 2014 as compared to 2009. Another way of looking at the future teacher supply is by observing the number of prospective candidates who attend a teacher preparation program and complete the requirements for a 87 credential. The number of completers decreased by over 23% from 2009 to 2014. Together, these decreases indicate significantly reduced teacher supply (see Figure 8). 25 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

32 FIGURE 6 Figure 8 Teacher Preparation Program Enrollments and Completers National 690,667 700,000 600,000 451,155 500,000 TEACHERS 400,000 300,000 233,282 178,891 200,000 NUMBER OF PROSPECTIVE 100,000 0 2014 2012 2010 2009 201 2013 1 Completers Enrollments Source: LPI analysis of the Title II Data Collection, 2004 2014, U.S. Department of Education. – A common hypothesis for the dramatic downturn in supply is that the highly publicized teacher layoffs during the Great Recession left a mark on the public psyche, including individuals who might have been considering a teaching career. In addition to the fact that there were few jobs available, budget cuts resulted in frozen salaries and deteriorated teaching conditions as resource limitations led to increased class sizes, and fewer materials and instructional supports. One sign of the cumulative impact of these factors is that only 5% of the students in a recent survey of those taking the ACT college entrance exam were interested in pursuing a career in education, a 88 decrease of 29% between 2010 and 2014. In an annual national survey of college freshmen, only 4.2% of students indicated their probable field of study would be education. This is fewer than half the share who expressed interest Only 5% of the students in a in 2007, when 9.2% of students intended to recent survey of those taking the major in education, and the lowest proportion of students considering teaching in the last ACT college entrance exam were 89 45 years. interested in pursuing a career These simple indicators—enrollments and in education, a decrease of 29% completers—reveal important information between 2010 and 2014. about the current teacher labor market. Modeling teacher supply using these data 26 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

33 in tandem with longitudinal analysis of college graduates and teacher re-entry allows a more powerful analysis. Figure 9 estimates new teacher entrants over the past decade using universal data on teacher preparation programs collected by the U.S. Department of Education under Title II of the Higher 90 Education Act —the most recent and complete national data on teacher preparation. We use the average ratio of completers to enrollments (.303) to project current enrollments into future completers. We then obtain a range of new teacher estimates by multiplying the number of completers by a lower and upper bound (75% and 90%) for the rate at which completers are expected to end up in the classroom. We estimate total teacher supply using the average annual attrition rate reported by the National 91 the upper (32.3%) and lower (28%) bounds of re-entry Center for Education Statistics (7.88%), 92 rates, and estimates of the timing of return for these re-entrants by Grissom and Reininger (2012). To estimate supply into the future, we exponentially smooth new teacher entrants through 2025 following the same methods as described above to produce total teacher supply. Figure 9 shows a steady decline in the teacher supply starting in 2009. According to our model, we estimate 2016 to have the lowest number of available new entrants in over 10 years, at around 93 The upper and lower bounds in brackets indicate where total supply 113,000 first-time teachers. will lie based on variation in re-entrants and the entry rate. In 2015–16, the overall teacher supply (new entrants and re-entrants) is estimated to be between 180,000 and 212,000 teachers, while the demand is approximately 260,000. FIGURE 7 Figure 9 Estimated Teacher Supply 280,000 240,000 200,000 160,000 120,000 80,000 TEACHER ENTRANTS 40,000 0 1 201 2021 2019 2017 2015 2013 2009 2007 2005 New Entrants Projected New Entrants Estimated Total Supply Projected Estimated Total Supply Entry Rate V ariation Note: Error bars represent high and low projections for total teacher supply to account for varying economic conditions that affect the entry and re-entry rates. Source: LPI analysis of the Title II Data Collection, 2004–2014, U.S. Department of Education; N2Derived Data File, 2012, from the Baccalaureate and Beyond (B&B), National Center for Education Statistics. 27 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

34 Depending on economic conditions and Candidates do not necessarily changes in the desirability of teaching, teacher supply will vary. Assuming that new entrants choose the fields and subject and re-entrants remain at 2016 levels, this areas in which there are large model estimates that supply will decrease numbers of vacancies, nor do through 2017 and remain near that low level through 2021. Constant supply is unlikely they choose to teach in the given the reactive nature of supply and hard-to-staff locations where the demand; however, projecting current trends shortages are most pronounced. into the future illuminates the extent of the labor market gap if nothing is done to alter current trends. As one indicator of current trends at the state level, where more recent data are available, there was a small (roughly 4%) uptick in California teacher preparation program enrollments and completers 94 in 2015 as the labor market shifted. Although promising, the increases were far too small to meet the sharply increased demand, and there was continued enrollment decline in the shortage fields that had been experiencing large shortages even during the downturn (mathematics, science, and special education). Thus, even as they respond to the market, candidates do not necessarily choose the fields and subject areas in which there are large numbers of vacancies, nor do they choose to teach in the hard-to-staff locations where the shortages are most pronounced. In fact, candidates are often less likely to go to these higher-need schools when there are more vacancies in well-heeled districts that tend to have easier working conditions and better salaries. Policymakers must ensure not only that there are enough teachers to meet demand, but also that there is an adequate supply of teachers for the fields and locations where they are needed most. 28 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

35 IV. The Teacher Labor Market: Predicting the Trends Ahead So far, we have looked at supply and demand separately. For each component of the labor market, we have examined the contributing factors and the current outlook, and we have modeled estimates of current and future trends. Looking at these components separately allowed us to investigate each in a degree of detail necessary to later understand how supply and demand interact together in a policy context. Now, combining supply and demand together allows us to investigate the teacher labor market as a whole. Figure 10 shows the size of the predicted teacher shortage. In the early 2000s, demand fell roughly within the brackets of total teacher supply, suggesting relative balance in the teacher labor market. In 2010, 2011, and 2012, supply was greater than demand. This finding matches with reality—during the Great Recession, there was a teacher “surplus”: Teachers were being laid off, and it was difficult to find work. In 2013, however, as the economy recovered, demand rose steeply, while supply continued to remain low and declined further in 2014 and 2015. During this period, the teacher labor market moved into a shortage condition. Currently, there are not enough qualified teachers to meet the demand. The shortage in the 2015–16 school year is estimated to be between 47,000 and 80,000 teachers. If supply trends were to persist at these current lows, in 2018, as demand increases again, supply could be around 112,000 teachers short of demand. We can expect some increase in the number of individuals entering teaching in response to greater demand. Nonetheless, even if supply increases to pre-recession levels of 260,000 teachers a year, demand would still be outstripping supply by approximately 40,000 teachers. Furthermore, the perennial areas of acute shortages (special education, mathematics, and science) thus far show little sign of response to labor market demand. FIGURE 8 Figure 10 Teacher Supply and Demand Projected 300,000 250,000 200,000 TEACHERS 150,000 100,000 NUMBER OF 50,000 0 2005 2009 2013 2017 2021 2025 Total Supply Projected Estimated Estimated Total Supply Total Supply Estimated Projected New Hires Demanded New Hires Demanded Actual New Hire Demanded Source: U.S. Department of Education, multiple databases (see Appendix A). 29 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

36 Figure 10 suggests that the United States is in the midst of a teacher shortage that, if current trends continue, may worsen further. The effects are likely to be felt in some subject areas, states, and communities more than others, as we describe below. Supply Trends in Shortage fields Looking at the Title II teacher education completer data from the last two years available (2011–12 and 2012–13), there was a decrease in new teacher entrants in almost every subject 95 96 More recent area, including those with perennial shortages, such as mathematics and science. data from California continued to show ongoing declines in mathematics and science credentials through 2014–15, even when supply was beginning to improve in other areas in response to greater 97 These trends will further exacerbate hiring difficulties in the mathematics and science demand. fields. Constant shortfalls exist in these areas because average non-teacher wages for individuals 98 with mathematics and science degrees are so high relative to teaching. No two states are the same; thus, a national teacher shortage will be felt and experienced by each state differently. Examining the distribution of teacher preparation programs, starting salaries, and working conditions, as we do below, helps us understand the variability in teacher labor markets by state, and how those differences could have an impact on teacher supply and demand. Supply Trends in States Availability of New Teachers States vary in the number and size of preparation programs they sponsor, and their yield in relation to state hiring demands. According to the Title II data collection, in 2014 there were 2,085 teacher preparation programs across the nation. Texas, California, and New York have the largest number of providers, 199, 143, and 139, respectively, and in 2013–14 prepared around 50,000 teachers, a quarter (27%) of the newly prepared teachers in the country. By contrast, Wyoming has only one 99 teacher preparation program, and in 2013-14 prepared just 256 teachers. However, the number of teachers prepared as a percentage of a state’s total teacher workforce can tell a different story. For example, California prepares a large number of teachers (the third most teachers of any state), but this is a smaller proportion of its total workforce than in most other states. On average, each year, states prepare 6% of their teacher workforces. Yet Alaska, California, Connecticut, Florida, Nevada, and Wyoming all prepare less than 4% of their teacher workforces. Conversely, Arizona, New York, Utah, and Washington, DC, prepare far greater numbers of teachers in relation to the size of their workforces (between 9% and 12%.) This variation creates states that are net importers of teachers and o thers that are net exporters. For example, in 2013–14, states, on average, awarded 34% of their initial teaching credentials 100 Some of these teachers to candidates who received their teacher preparation in another state. represent an interstate trade—for example, a teacher who prepared in Connecticut teaching in Massachusetts, balanced by a teacher who prepared in Massachusetts teaching in Connecticut—but others represent net imports. Wyoming, Alaska, and North Dakota—net importers—issued 72%, 75%, and 100%, respectively, of their initial teaching credentials to out-of-state prepared candidates 30 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

37 in 2013–14. On the other hand, in the same year, New Jersey and New York—net exporters—issued 101 effectively no initial teaching credentials to individuals who prepared out of state. Teachers do move around the country; one analysis found about one-quarter of all applicants 102 for teaching positions across the country were from out of state. At the same time, states that prepare a great many teachers, relative to their needs, have at least two advantages. First, many 103 and some will have job offers teachers prefer to teach near where they grew up or went to school, from the districts where they student taught. In addition, teachers earn state-specific credentials while they are preparing, and candidates sometimes encounter a variety of barriers in trying to transfer credentials to other states, creating steep transaction costs for moving. States that prepare fewer teachers must attract out-of-state candidates to fill their vacancies, which can be even more difficult in times of shortage. Thus, the concentration of teacher preparation providers and number of candidates completing these programs is likely one reason for the variation in labor markets by state. Salaries State differences in starting teacher salaries can also contribute to the variability in teacher labor markets. Salaries can affect the attractiveness of teaching jobs in ways that impact both recruitment 104 and retention. According to data from the National Education Association, the average starting 105 but the range was very wide. In 2013, the teacher salary in the United States in 2013 was $36,141, District of Columbia had the highest starting salary at $51,539, and Montana had the lowest starting salary at $27,274, unadjusted for cost of living differentials (see Appendix B). Within states, school district funding often varies, with strong consequences for salary levels. Great inequalities in salaries across districts typically cause shortages in those that are not able to offer 106 Some of this variation can be explained by cost-of-living differences, but a competitive wage. even in higher-paying states, such as California, there can be wide variations in wages that are not associated with cost-of-living differentials, leaving many teachers struggling with the higher cost 107 of living and lower purchasing power. These variations in salaries are another reason why states experience teacher shortages differently. The competitiveness of teachers’ wages to those of non-teaching occupations requiring Great inequalities in salaries similar levels of education can be just as across districts typically cause important as teacher salaries themselves, as shortages in those that are not these are an influential factor in teachers’ 108 Wage decisions to enter the profession. able to offer a competitive wage. competitiveness also varies from state to state. After controlling for age, education level, hours worked per week, and weeks worked per year, teachers in Wyoming earned 94% of what non-teachers in the state earned in 2012. The ratio in Alaska and Iowa was 85%. Conversely, teachers in Arizona and Virginia earned only 62% and 63%, 109 respectively, of what this group of non-teachers earned (see Appendix B). 31 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

38 Working Conditions Working conditions that teachers report as important to their decisions to stay, leave, or return to the classroom vary by state. One working condition important to teachers is the average number of students they teach: their class sizes or pupil loads. Although a higher number than average pupil-teacher ratios, the two are correlated, and widely varying across the country. In 2014, when the national average pupil-teacher ratio was 16-to-1, California’s ratio was 24-to-1, while Vermont’s 110 was 11-to-1 (see Appendix B). Other conditions found to be related to teachers’ decisions to stay in or leave a school—such as competent and supportive leadership, a school’s testing and accountability environment, and teacher autonomy in making key classroom decisions—also vary substantially. For instance, according to data from the 2011–12 SASS, more than twice as many teachers in Arkansas strongly agreed that their administration was supportive as did teachers in the District of Columbia (58% vs. 24%) (see Figure 11). More than 10 times as many teachers in Indiana and More than 10 times as many Florida strongly agreed their job security was impacted by the performance of their students teachers in Indiana and Florida or school on state or local tests as in Vermont strongly agreed their job and North Dakota (25–26% vs. 2%) (see Figure security was impacted by the 12). Whereas roughly 88% of teachers in Montana, North Dakota, and Vermont agreed performance of their students they have decision-making autonomy in the or school on state or local tests classroom, fewer than 60% felt this way in 111 as in Vermont and North Dakota Delaware, Florida, and Maryland (see Figure 13). All of these differences are associated with (25–26% vs. 2%). differences in teachers’ plans to leave teaching across these states. Attrition Rates It should be no surprise, then, that teachers’ estimates of how long they will stay in the profession also differ by state. Teachers in Washington, DC, are more than five times as likely to report they plan to leave the classroom as soon as possible as are teachers in South Dakota. Teachers are most likely to say they plan to leave teaching in Arizona, Nevada, and Washington, DC, and least likely to 112 report plans to leave in Illinois, Rhode Island, and South Dakota (see Figure 14). Indeed, federal data show that close to one in four teachers moves schools or leaves the profession annually in Arizona and Washington, DC, more Teachers in Washington, DC, are than three times the rate in Rhode Island, for more than five times as likely 113 example (see Appendix B). Not surprisingly, to report they plan to leave the states with more teachers who plan to leave teaching tend to be the same states with a classroom as soon as possible smaller percentage of teachers planning to as are teachers in South Dakota. teach as long as possible or until they are eligible for retirement benefits. 32 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

39 FIGURE 9 Figure 11 Teacher Reports of Administrative Support by State Percent of teachers who strongly agree that "the school administration's behavior toward the staff is supportive and encouraging" Distr ict of Columbia Connecticut Iowa Ne w Hampshire yland Mar Wisconsin Minnesota Illinois Michigan Rhode Island or k w Y Ne ania Pennsylv Massachusetts Ver mont Indiana ware Dela Ar iz ona Texas ota th Dak Nor ado Color Ohio Maine aska Nebr vada Ne Ne xico w Me Virginia Calif or nia Montana y w Jerse Ne Kentucky Alaska Mississippi Washington waii Ha Wyoming South Dak ota ida Flor West Virginia th Carolina Nor i Missour Oklahoma Louisiana South Carolina Georgia Kansas Oregon Utah Tennessee Alabama Idaho Ar kansas 70 60 80 50 40 30 20 PERCENTAGE WHO STRONGLY AGREE Note: Bars represent 95% confidence interval. Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 33 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

40 FIGURE 10 Figure 12 Teacher Reports of Testing-Related Job Insecurity by State Percent of teachers who strongly agree that they "worry about the security of my job because of the performance of my students or my school on state and/or local tests” Ver mont Nor ota th Dak Alaska Montana ota South Dak Oregon Maine Minnesota Ne w Hampshire West Virginia Massachusetts Nebr aska Kansas Iowa South Carolina Ar kansas Calif or nia waii Ha Virginia Washington Alabama Nor th Carolina Connecticut Wyoming Illinois Utah Kentucky Georgia i Missour Texas Oklahoma Pennsylv ania Mississippi vada Ne y w Jerse Ne Ar iz ona Ohio Idaho yland Mar ado Color ware Dela Wisconsin or k w Y Ne Michigan w Me Ne xico ict of Columbia Distr Louisiana Tennessee Rhode Island ida Flor Indiana 40 30 20 10 0 PERCENTAGE WHO STRONGLY AGREE Note: Bars represent 95% confidence interval. Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 34 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

41 h actual contr ou ha ve IN Y OUR CLASSR Ho w m uc OOM at this sc ol do y hool o ver FIGURE 11 Figure 13 the f hing? our planning and teac wing areas of y ollo Teacher Reports of Classroom Autonomy by State Percent of teachers who strongly agree that they have control in their classroom in the lass materials, content and skills to be taught, xtbooks and c (te following areas of planning and teaching: textbooks and class materials, content and teac hniques, e valuating students, discipline, home work) hing tec skills to be taught, teaching techniques, evaluating students, discipline, and homework Dela ware Flor ida yland Mar Louisiana Rhode Island Virginia Georgia Texas Ne vada Nor th Carolina Tennessee West Virginia South Carolina Kentucky Alabama Ar iz ona or nia Calif w Jerse Ne y Connecticut w Me xico Ne Mississippi aska Nebr Michigan Color ado Pennsylv ania Indiana ict of Columbia Distr Alaska Ar kansas Utah Massachusetts Ohio Idaho Wyoming Washington or k w Y Ne i Missour Maine waii Ha Illinois Wisconsin w Hampshire Ne Oregon Minnesota Kansas Iowa Oklahoma ota South Dak Ver mont th Dak Nor ota Montana 60 100 90 80 70 50 40 PERCENTAGE WHO STRONGLY AGREE P ercentage That Strongly Agree Composite Scale Note: Bars represent 95% confidence interval. Teacher autonomy in the classroom is measured using a Cronbach Alpha-generated construct of classroom control derived from the six components listed above. eacher Date File val. Source: 2011 Note: Bars Represent 95% Confidence Inter 12 SASS T Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 35 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

42 or when a better job comes ar ound FIGURE 12 vs. Figure 14 le or P er cent of teac her hing as long as ab s who plan to remain in teac Teachers Who Plan to Leave vs. Those Who Plan to Stay until the y are eligible f or retirement benefits Percent of teachers who plan to either leave teaching immediately or when a better job comes around vs. percent of teachers who plan to remain in teaching as long as able or until they are eligible for retirement benefits Rhode Island Plan to Stay Plan to Lea ve ota South Dak Illinois Massachusetts Wyoming Ohio th Dak ota Nor Connecticut Iowa aska Nebr Kentucky w Jerse y Ne or nia Calif Pennsylv ania Alaska Ver mont Ar kansas i Missour or k w Y Ne Georgia Oregon Minnesota w Hampshire Ne Utah Alabama Wisconsin Michigan Montana Louisiana ware Dela West Virginia Mississippi xico w Me Ne Maine Washington Kansas Tennessee Oklahoma Indiana waii Ha ado Color South Carolina Idaho th Carolina Nor ida Flor Virginia Texas yland Mar vada Ne Ar iz ona ict of Columbia Distr 30 90 0 10 100 40 20 70 60 50 80 PERCENT OF TEACHERS Percent of T eachers Note: Bars represent 95% confidence interval. eacher Date File 12 SASS T val. Source: 2011 Note: Bars Represent 95% Confidence Inter Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 36 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

43 Sources of Shortages In order to respond effectively, it is important to understand what factors are driving these shortages and what can be done to shift teacher supply and demand to bring the teacher labor market to equilibrium. Based on the evidence available, the emerging teacher shortage appears to be driven by four main factors: Decline in teacher preparation enrollments, 1. 2. An effort to return to pre-recession course offerings and class sizes, 3. Increasing student enrollment, and 4. High teacher attrition. We identify the first three factors because these have shown the largest changes at the Reducing attrition would actually same time teacher shortages have increased. make a greater difference in For example, between 2009 and 2014, there balancing supply and demand was a 35% decrease in teacher preparation program enrollment. Meanwhile, the teacher than any other intervention. workforce has been increasing, although it is still 145,000 teachers short of pre-recession pupil-teacher ratios, and student enrollment has begun to increase and is projected to grow by 3 million students. While these factors have been changing, a more constant factor—and by far the largest component of annual demand—is the high rate of teacher attrition in U.S. schools. As we discuss below, this rate is much higher than in many other countries that offer more competitive wages for teachers and more uniformly supportive working conditions. Reducing attrition would actually make a greater difference in balancing supply and demand than any other intervention. Given the large impact that attrition has on teacher demand and on the adequacy of supply, we examine its characteristics and effects before turning to a broader discussion of policy drivers and potential solutions. 37 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

44 V. The Role of Teacher Attrition Using workforce estimates, student enrollment, The most important driving factor and attrition data from the SASS 1988–89 to 2011–12, Figure 15 shows the relative of teacher shortages is high contribution of different factors to teacher teacher attrition. 114 demand. Estimates vary from actual demand for the years 1989 to 2012 because the model does not perfectly predict demand, especially in unusual circumstances that stray from averages, like the Great Recession. Figure 15 highlights two main points: First, although the number of teachers who leave the profession has not dramatically changed in recent years, it constitutes the lion’s share of demand, representing anywhere from two-thirds to nearly 100% of the demand for teachers in any given year. Thus, the most important driving factor of teacher shortages is high teacher attrition. Second, while workforce growth due to student enrollment increases and reductions in pupil-teacher ratios will play a bigger role in demand from 2018 to 2025 than they have in recent years, attrition still swamps these variables as a driver of teacher demand. FIGURE 13 Figure 15 Teacher Demand Factors Over Time 300,000 250,000 200,000 150,000 TEACHERS 100,000 50,000 NUMBER OF 0 -50,000 -100,000 1 201 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 1989 1990 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Retirement Attrition Attrition Preretirement Workforce Change: Workforce Change: Pupil-Teacher Ratio Student Enrollment Note: Negative number of teachers represents a decrease in the total number of teachers. Source: U.S. Department of Education, multiple databases (see Appendix A). 38 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

45 Attrition has not always been such a dominant factor in demand. In the late 1980s and early 1990s, attrition was less than 6%, and demand was, consequently, lower, as fewer teachers had to be replaced each year (see Figures 15 and 16). The higher attrition rates of recent years have had a very large impact on demand. Between 1989 and 2005, attrition rates increased by 50%, and they have stayed high since then. Consequently, compared to 25 years ago, attrition is now responsible for a larger number of teachers demanded each year. FIGURE 14 Figure 16 Attrition Rate Over Time 8.4% 8.0% 8 7.7% 7.4% 6.6% 6 5.6% 4 2 ANNUAL ATTRITION RATE (PERCENT) 0 2013 1989 2009 1995 2001 2005 Source: National Center for Education Statistics (2015). Digest of Education Statistics. Mobility of public elementary and secondary teachers, by selected teacher and school characteristics: Selected years, 1987–88 through 2012–13. Washington, DC: U.S. Department of Education. In times of shortages, policymakers often focus attention, understandably, on how to get more teachers into the profession. However, it is equally important to focus on how to keep effective teachers in the workforce. In fact, as we show below, reducing attrition could virtually eliminate overall shortages, with the exception of distributional imbalances across fields and locations that may require specific incentives. Compared to high-achieving nations like Finland, Singapore, and Ontario, Canada—where only 115 about 3% to 4% of teachers leave in a given year —U.S. attrition rates are quite high: hovering near 8% over the last decade, and much higher for new teachers and teachers in high-poverty schools 116 and districts. In many states and districts, relatively little attention is paid to supporting teachers after they have joined the profession. 39 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

46 The Costs of Attrition Figure 17 displays the teacher labor market under different assumptions of teacher attrition. Relatively small differences in the attrition rate have huge implications for the teacher labor market. The attrition rate during the projection period (black dashed line) is 8%. If the annual attrition rate could be reduced from the current projection of 8% to 6%—slightly higher than the U.S. teacher attrition rate in 1989—hiring needs would decrease by at least 60,000 teachers annually. This modest decrease in the attrition rate would reduce the demand for teachers in each 117 year to about 200,000 teachers instead of 260,000. 118 If the attrition rate could be reduced to 4%, closer to where it is in some other countries, U.S. hiring needs would be reduced by roughly 130,000 teachers annually, cutting annual demand by nearly half. This large reduction in demand would not only largely eliminate current teacher shortages, but also allow for increased selectivity, boosting the quality of our nation’s teachers. FIGURE 15 Figure 17 Projected Supply and Demand with Varying Attrition Rates 300,000 250,000 200,000 TEACHERS 150,000 100,000 NUMBER OF 50,000 0 2021 2017 2013 2009 2025 2005 Projected Estimated Total Supply Estimated Total Supply Projected New Hires (8% Attrition) Actual New Hires Demanded Projected New Hires (4% Attrition) Projected New Hires (6% Attrition) Source: U.S. Department of Education, multiple databases (see Appendix A). Beyond those who leave the profession, teachers who move from one school to another are an equally important component of teacher turnover. We use the term “turnover” to denote the rate at which teachers leave a school, whether to teach elsewhere (movers) or to leave the profession entirely (leavers). At the school level, teachers who leave for a different school have the same impact as teachers who leave the profession—a vacancy that must be filled, along with both fiscal and academic costs associated with the turnover. Nationally, on average, close to 16% of teachers leave the school at which they teach each year (see Figure 18). These rates are higher in some schools than others, based on conditions in the school, which we discuss further below. 40 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

47 FIGURE 16 Figure 18 Trends in Teacher Turnover 18 16 14 12 10 8 6 4 TURNOVER RATE (PERCENT) 2 0 2013 1992 1995 2001 2005 2009 1989 Movers Leavers Source: National Center for Education Statistics (2015). Digest of Education Statistics. Mobility of public elementary and secondary teachers, by selected teacher and school characteristics: Selected years, 1987–88 through 2012–13. Washington, DC: U.S. Department of Education. Hard-to-staff schools with high turnover rates typically end up with a disproportionate number of relatively inexperienced teachers, which can both create greater churn, if they leave rapidly as many beginning teachers do, and undermine student achievement as a function of both teacher 119 inexperience and overall instability. High teacher turnover rates have been found to negatively 120 affect the achievement of all students in a school, not just students in a new teacher’s classroom. Research shows that stability, coupled with shared planning and collaboration, helps teachers to 121 and that teachers improve more rapidly in supportive and collegial improve their effectiveness, 122 working environments. High teacher turnover undermines these benefits, which arise through shared knowledge and collaboration among colleagues. As a principal in a poorly resourced, high-turnover school in California explained: (H)aving that many new teachers on the staff at any given time meant that there was less of a knowledge base. It meant that it was harder for families to be connected to the school because, you know, their child might get a new teacher every year. It meant there was less cohesion on the staff. It meant that every year, we had to re-cover ground in professional development that had already been covered and try to catch people up to sort of where the 123 school was heading. 41 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

48 Such schools must continually pour money into recruitment efforts and professional support for new teachers, many of them untrained, without reaping dividends from these investments. Other teachers, including the few who could serve as mentors, are stretched thin and often feel overburdened by the needs of their colleagues as well as their students. Scarce resources are squandered trying to reteach the basics each year to teachers who come in with few tools and leave 124 before they become skilled. Each time a teacher leaves a district, it not only increases demand but also imposes replacement costs on districts. A decade ago, replacement costs for teachers were estimated to range from around $4,400 in a small rural district to nearly $18,000 in a large urban district for every teacher 125 who leaves—a national price tag of over $7 billion a year. With inflation, those costs would be more than $8 billion today. A comprehensive approach to reducing attrition would effectively both lessen the demand for teacher hiring and save money that could be better spent on mentoring and other evidence-based approaches to supporting teacher development. Attrition is one of the most important aspects of demand to focus on, both because it is such Each time a teacher leaves a a large component of demand and because district, it not only increases it is policy malleable. To understand the demand but also imposes policy actions that could reduce attrition, we must understand why people are leaving replacement costs on districts. the profession, who is leaving the profession, where attrition is the greatest, and what factors are associated with it. Reasons for Attrition Rhetoric used in discussions about teacher demand often suggests that high levels of teacher attrition are driven by an aging teaching force, but current data suggest otherwise. Contrary to common belief, only a third of departing teachers in 2012 listed retirement as a very or extremely 126 important reason for leaving. Only 12.6% of teachers who left the teaching workforce said the 127 most important factor for their departure was retirement. High attrition rates in the United States are driven much more by teachers leaving for other reasons. Even during a period of substantial layoffs and incentives for early retirement, most teachers left voluntarily for reasons other than retirement in 2011–12 (see Figure 19). This rate varies year to 128 For example, in 2008–09, 69.3%, or seven out of year with the economy and retirement incentives. every 10 leavers, left voluntarily for reasons other than retirement. 129 National estimates New teachers leave at greater rates than others in the preretirement period. have suggested that new teachers leave at rates of somewhere between 19% and 30% over their 130 first five years of teaching. Because more new teachers have been hired over the last decade, this high attrition is having a greater and greater impact on the teacher labor market and on students’ 131 experiences in school (see Figure 20). 42 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

49 FIGURE 17 Figure 19 Distribution of Leavers by Voluntary Retirement Status 1% 9% 37% 53% Involuntary Retirement Leavers Involuntary Preretirement Leavers Voluntary Preretirement Leavers Voluntary Retirement Leavers Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. FIGURE 18 Figure 20 Teaching Experience of Public School Teachers, 1987–88, 2007–08, and 2011–12 Source: Ingersoll, R. M., Merrill, L., and Stuckey, D. (2014). Seven Trends: The Transformation of the Teaching Force. Philadelphia, PA: Consortium for Policy Research in Education. 43 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

50 Which Teachers Leave? The literature suggests that turnover rates and reasons for leaving vary for different teacher populations. Mathematics and science teachers, for example, move schools and leave teaching at higher rates than humanities teachers and general elementary teachers (see Figure 21). Special education teachers and teachers of English language learners leave and move at even higher rates. FIGURE 19 Figure 21 Teacher Turnover Rate by Primary Teaching Assignment 6.4 4.9 General Elementary 7.3 5.3 Humanities 7.2 6.1 Math/Science 5.6 8.6 Special Education Teaching English 6.9 12.0 Learners 10 8 16 12 14 0 18 20 2 4 6 TURNOVER RATE (PERCENT) Leavers Movers Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. Teachers of color and teachers working in high-poverty, high-minority schools also tend to have higher turnover rates. Historically, teachers of color and white teachers have left the workforce at similar rates (see Figure 22), but teachers of color have moved schools at noticeably greater rates for most of the last two decades (see Figure 23). Some of these teacher attrition differences can be explained by the association between teacher characteristics and school working conditions. More than three-quarters of all teachers of color 132 These schools, which are often teach in the quartile of schools with the most students of color. under-resourced and plagued by poor working conditions, typically experience greater turnover. For example, the turnover rate in Title I schools is nearly 50% greater than that of non-Title I schools. Furthermore, across content areas, teachers in Title I schools have less experience and 133 shorter tenures. 44 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

51 FIGURE 20 Figure 22 Leaver Rate 1988–89 to 2012–13 12 10 8 6 PERCENT 4 2 0 1995 1985 1990 2015 2000 2005 2010 Teachers o f C olor White T eachers Source: Ingersoll, R. M., and May, H. (2011). The minority teacher shortage: Fact or fable? Kappan Magazine, 93(1), 62–65; LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. FIGURE 21 Figure 23 Mover Rate 1988–89 to 2012–13 12 10 8 6 PERCENT 4 2 0 1985 2000 1995 1990 2015 2010 2005 White Teachers Teachers of Color Source: Ingersoll, R. M., and May, H. (2011). The minority teacher shortage: Fact or fable? Kappan Magazine, 93(1), 62–65; LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. | A COMING CRISIS IN TEACHING? 45 LEARNING POLICY INSTITUTE

52 The disparities are even greater for schools with different racial compositions, as shown in Figure 24. In schools in the quartile with the most students of color, the turnover rate is 64% greater than in the quartile of schools with the fewest students of color. Turnover rates are higher in these schools for teachers across content areas, years of experience, and certification type. Annual turnover rates are especially high in these schools for inexperienced teachers (24%) and for those who have entered through alternative certification pathways (20%). Alternatively certified teachers were 2½ times more likely to leave high-minority schools than they were to leave schools with few students of color. FIGURE 22 Figure 24 Teacher Turnover by Teacher Characteristics 17% All T eachers*** 10% H umanities 16% Teachers 12% Math and Science 19% Teachers*** 10% Special E ducation 20% Teachers* * 11% 21% ELL T eachers 22% Regular 16% Certification*** 10% Alternative 20% Certification*** 8% 4+ Yea rs 16% Experience*** 9% 3 o r F ewer 24% Yea erience rs Exp 18% 5 10 15 20 25 30 0 TURNOVER RATE (PERCENT) Top Quartile, Students of Color Bottom Quartile, Students of Color Note: Asterisks indicate significant differences in means between quartiles: *p <.10; **p < .05; ***p < .01. In the top quartile of schools, 55% of students are students of color. In the bottom quartile, 9% of students are students of color. Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics; Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 46 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

53 Our analysis of the Schools and Staffing Surveys found that teachers who enter the profession through an alternative certification program have higher rates of annual turnover (17% versus 13% for regular pathway teachers). Among full-time teachers, after controlling for school and teacher characteristics, alternatively certified teachers are 20% more likely to leave their schools than 134 teachers who entered teaching with standard certification. More than half of all alternatively certified teachers teach in schools serving primarily students of color, and they account for 21% of teachers in these schools. In contrast, alternatively certified teachers account for less than 9% of teachers in schools with predominantly white students. We find 135 similar patterns in Title I schools. Unfortunately, many of the teachers in More than half of all alternatively hard-to-staff fields receive less pedagogical preparation because they are encouraged to certified teachers teach in enter before they have completed training, schools serving primarily as districts seek to meet their pressing hiring needs. For example, mathematics and science students of color, and they teachers are more likely to be certified via account for 21% of teachers in an alternative pathway (21% of the total) these schools. than those teaching other subjects (less than 136 14%), and they have had less pedagogical 137 training than other teachers on average. Similarly, more than twice as many teachers of color are certified via an alternative pathway (25%) as white teachers (12%). Where is turnover greatest? Teacher turnover rates also vary considerably across the country. At over 16% annually, the South has a particularly high turnover rate compared to the Northeast, Midwest, and West census regions, where turnover rates average about 13% or less. Southern and Midwestern cities have the highest rates of teacher turnover, followed by Southern suburbs, towns, and rural areas (see Figure 25). The higher-spending Northeast averages the lowest turnover rates 138 In the west, turnover in cities and suburbs is among the lowest, while in across all district types. towns and rural areas, it is among the highest. For most regions, turnover is higher in cities than in 139 any other district type. The variation in annual turnover is even greater between states, ranging from under 9% in Utah to 24% in Arizona (see Figure 26). Retirement represents less than a third of all turnover in every state except for two: Oregon and New Jersey. By contrast, in Delaware, less than 6% of turnover can be attributed to retirement. Finally, the highest rates of teachers leaving the profession for reasons other than retirement are found in North Dakota (10.7%), Arizona (8.7%), and Indiana (8.5%). 47 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

54 FIGURE 23 Figure 25 Turnover Rates by Region and District Type 24 20 16 12 8 4 AVERAGE TURNOVER RATE (PERCENT) 0 West West West West South South South South Midwest Midwest Midwest Midwest Northeast Northeast Northeast Northeast RURAL SUBURB TOWN CITY Note: Interval bars represent upper and lower bounds of a 95% confidence interval; District types are based on 2000 Census population and geography information. For more information, see Documentation to the NCES Common Core of Data Public Elementary/Secondary School Universe Survey: School Year 2009–10 (NCES 2008–332). Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. FIGURE 24 Figure 26 Teacher Turnover by Source and State 25 20 15 10 ER RATE (PERCENT) 5 TURNOV 0 IA IL ID IN NJ TX MI LA FL AL UT CT WI AZ KY CO CA GA PA KS OH AR TN AK VA SD OR SC DE OK MA NV MT NY ND NC MS NE WV MO ME MN MD WA NM Movers tirement Leavers Pre-re Retirement Leavers New England Note: States with fewer than 25 teachers surveyed were excluded (DC, HI, and WY). Three small New England states with similar data patterns were combined (NH, RI, and VT). Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. | A COMING CRISIS IN TEACHING? 48 LEARNING POLICY INSTITUTE

55 Studies find that attrition is often due to family reasons, lack of Why Do Teachers Leave? 140 administrative support, low salaries, and poor working conditions. More recently, with the rise of accountability systems in education, many teachers have reported that high-stakes testing 141 Figure 27 shows the importance of policies have encouraged them to leave the profession early. different types of factors on a teacher’s decision to leave. Most teachers who voluntarily leave the classroom list some area of job dissatisfaction as very important or extremely important in their decision to leave the profession. As shown in Table 1, these areas of dissatisfaction can range from physical conditions—such as class sizes, facilities, and classroom resources—to unhappiness with administrative practices—such as lack of support, classroom autonomy, or input to decisions—to policy issues, such as the effects of testing and accountability. FIGURE 25 Figure 27 Types of Reasons Given by Teachers for Leaving the Profession 55% Dissatisfaction Family/Personal 43% Reasons 31% Retirement 31% To Pursue Another Job 18% Financial Reasons 60 0 10 20 30 40 50 PERCENT Note: These five factors combine survey questions into common categories (see Table 1). Percentages do not add to 100 because teachers can select multiple reasons. Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 49 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

56 Within the above categories, the top five reasons teachers identified as important or very important in their decision to leave the classroom, other than retirement, were child care or pregnancy (37%), pursuit of another career (28%), dissatisfaction with recent school accountability measures (25%), dissatisfaction with the administration (21%), and dissatisfaction with teaching as a career (21%) 142 (see Table 1). Table 1 Reasons Teachers Listed as Important to Their Decision to Leave % Rated Very or Extremely Important Survey Questions in Each Category Reason for Exit Dissatisfaction (55%) Dissatisfied because of assessments 25% and accountability measures Dissatisfied because not enough 17% support to prepare students for assessments Dissatisfied with compensation tied to 8% student performance Dissatisfied with the administration 21% Too many intrusions on teaching time 18% Discipline issues were an issue 17% at school 14% Not enough autonomy in the classroom Dissatisfied with lack of influence over 13% school policies and practices 21% Dissatisfied with teaching as a career Not enough opportunities for leadership 9% or professional advancement Dissatisfied with job description 12% or assignment 10% Dissatisfied with large class sizes Dissatisfied with working conditions 9% (facilities, classroom resources, school safety) Family/Personal Wanted to take a job more 11% conveniently located Reasons (43%) Other personal life reasons (e.g., pregnancy/child care, health, caring 37% for family...) Decided to pursue another career 28% To Pursue Another Job (31%) Taking courses to improve career 13% opportunities within the field of education Taking courses to improve career 5% opportunities outside the field of education Decided to retire or receive Retirement (31%) 31% retirement benefits Financial Reasons (18%) Wanted or needed a higher salary 13% Needed better benefits 8% 7% Concerned about job security 50 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

57 As we noted earlier, aside from the personal reasons that individuals step out of teaching, dissatisfactions with the job are a predominant driver of attrition. Researchers have identified a number of workplace conditions associated with teacher attrition, including instructional leadership, school culture, collegial relationships, time for collaboration and planning, teachers’ decision-making power, experiences with professional development, facilities, lack of parental 143 support or involvement, and resources. Poor workplace conditions are, unfortunately, common in schools with disadvantaged student 144 Although rates populations and explain much of the high attrition rates those schools experience. of attrition are higher in high-poverty schools, in studies that control for workplace conditions, the explanatory power of student demographic characteristics is reduced or eliminated. Turnover has been found to be largely a product of the lower salaries and poorer working conditions at these 145 schools, rather than a function of the students’ characteristics. Among these working conditions in Poor workplace conditions recent years has been the threat of firing, displacement, and school closures associated are, unfortunately, common in with accountability schemes under No schools with disadvantaged Child Left Behind. Increased accountability measures and the resulting decrease in student populations and explain classroom autonomy, including the limited much of the high attrition rates ability for teachers to select texts, content, those schools experience. and assessments, are heavily associated with 146 minority teacher turnover. Accountability pressures particularly impact teachers of color because there are more teachers of color in schools serving higher-needs students, which are the schools that have struggled most to meet 147 test-based accountability requirements. Unfortunately, the recent rise of accountability measures undermines efforts to increase teacher diversity, especially in high-need schools. A comprehensive research review of attrition in high-poverty schools finds that the most significant workplace conditions associated with teacher attrition are teachers’ perceptions of their principal, 148 collegial relationships, and school culture. Similarly, in our own analyses, we found that the single most predictive workplace condition was whether teachers reported lacking administrative support. When teachers strongly disagree that their administrator encourages and acknowledges staff, communicates a clear vision, and generally runs a school well, turnover rates for movers and leavers jump to nearly one in four, more than double the rate of those who feel their administrators 149 are supportive (see Figure 28). In the quartile of schools with the most students of color, teachers are almost twice as likely to report severe dissatisfaction with their administration as teachers in schools with the fewest 150 students of color. This may be in part because the same factors that produce teacher shortages 151 also produce shortages of administrators in high-need schools. 51 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

58 FIGURE 26 Figure 28 Probability of Teacher Turnover in Relation to Administrative Support Teachers' reports about the extent to which their “school administration is supportive” 30 25 20 15 PREDICTED TURNOVER RATE (PERCENT) 10 Somewhat Agree Strongly Agree Strongly Disagree Somewhat Disagree TEACHER REPORTS Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. In sum, more than half of teachers who leave the profession list dissatisfaction with some part of their job as “very” or “extremely important” in their decision. Research indicates that turnover is driven less by student or teacher characteristics than by teaching circumstances. Among working conditions, administrative support is especially central, along with other factors that are often a function of the administrator’s approach: school culture and collegial relationships, time for collaboration, and decision-making input. Reducing attrition from the current 8% to about 4%—the norm in high-achieving nations like Finland, Singapore, and Ontario, Canada, and a few states—could virtually eliminate overall teacher shortages. Because most attrition is voluntary, preretirement attrition policy can have a big impact on reducing the number of people who leave the classroom and on increasing the number of those who have left but who later return. In the next section, we discuss policy responses that research suggests can help resolve shortages and create a stable, long-term, high-quality teaching force. 52 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

59 VI. Policies to Meet Demand with a High-Quality Supply of Teachers While policymakers often focus on how to recruit more teachers when there are shortages, it is important to recognize that policies which decrease teacher attrition are at least equally important. Cutting teacher attrition in half—and hence reducing teacher demand—would make our current supply largely adequate and would allow greater focus on the quality of teachers. This is essentially what high-achieving nations do. As we noted earlier, high-achieving countries like Finland, Singapore, and Ontario, Canada, have rates of attrition that are typically less than half the rate the United States has recently been experiencing each year. These countries manage to attract, prepare, and distribute well-trained teachers to all students by increasing incentives rather than lowering standards. These countries and others that rarely experience teacher shortages have made substantial investments in teacher training and distribution in the last two decades, including: • Salaries that are competitive with other professions, such as engineering, and equitable across schools (often with additional incentives for hard-to-staff locations); High-quality teacher education, usually at the graduate level and largely at government • expense, including extensive practice teaching in a clinical school connected to the university; Mentoring for beginners in their first year of teaching from expert teachers, coupled with a • reduced teaching load and shared planning time; • Collegial work settings offering ongoing professional learning embedded in 10–20 hours a week of planning and professional development time; and • Opportunities for expert, veteran teachers to be engaged in leading curriculum 152 development, professional development, and mentoring/coaching for their peers. Unfortunately, unlike other industrialized nations that are high-achieving, the United States lacks a systematic approach to recruiting, preparing, and retaining teachers, or for using the skills of accomplished teachers to help improve schools. With unequal resources across states and districts, and few governmental supports for preparation or mentoring, teachers in the United States enter: • With dramatically different levels of training—with those least prepared typically teaching the most educationally vulnerable children; • Earning salaries typically below those of other occupations—with those teaching the neediest students often earning the least; Working under radically different teaching conditions—with those in the most affluent • communities benefiting from small classes and a cornucopia of materials, equipment, specialists, and supports, while those in the poorest communities teach classes of 30 or more without adequate books and supplies; and • With little time for collaboration and uneven access to on-the-job mentoring or 153 professional learning to help improve their skills. 53 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

60 The United States needs a systemic approach if it is ever to build a strong teaching profession, and recruit and retain teachers where they are most needed. There are a number of states and districts 154 that have undertaken successful approaches that could be emulated. Ultimately, a national 155 coupled with a rethinking of the teaching career so teacher supply policy is critically needed, that teachers can become highly effective, have strong reasons to stay in the career, use their skills where they are needed most, and enhance the expertise of the profession as a whole. No single policy can solve the nation’s The United States needs a emerging teaching shortage. What is needed is a comprehensive set of strategies at the systemic approach if it is ever federal, state, and local levels that are focused to build a strong teaching on increasing the number of well-prepared profession, and recruit and entrants to the field of teaching, directing them to the fields and locations where they retain teachers where they are are needed, and plugging the leaky bucket of most needed. teacher attrition, which has high costs for both district budgets and student achievement. Without policy interventions, it is likely that even if more new candidates—heartened by reports of greater hiring—consider teaching, they will fail to choose the fields in which there are shortages or go to the high-po verty communities where they are more sorely needed. Furthermore, a status quo approach will not leverage better 156 preparation that supports student achievement or stem turnover where it is currently high. To ensure that all children have access to effective teachers, we need a national teacher supply policy based on well-developed federal-state partnerships, like those the nation has pursued in the field of medicine. Intelligent, targeted subsidies for preparation, coupled with stronger supports at entry, incentives for competitive salaries, and productive working conditions, could go a long way toward ensuring that all students have access to a strong cadre of teachers prepared to support their learning. Based on what we know about what matters and what works, policies should focus on: that allow teachers to make a • Creating competitive, equitable compensation packages reasonable living across all kinds of communities; Enhancing the supply of qualified teachers targeted to high-need fields and locations • through training subsidies and high-retention pathways; Improving teacher retention, especially in hard-to-staff schools, through more • effective mentoring, induction, working conditions, and career development; and • that can facilitate getting and keeping Developing a national teacher supply market teachers in the places they are needed over the course of their careers. We treat each of these areas, with associated recommendations, below. 54 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

61 1. Creating Competitive, Equitable Compensation Packages Even if teachers may be more motivated by altruism than some other workers, teaching must compete with other occupations for talented college and university graduates. Since the early 1990s, teacher salaries have been declining in relation to other professional salaries. Even after adjusting for the shorter work year in teaching, beginning teachers nationally earn about 20% less than 157 individuals with college degrees in other fields, a wage gap that widens to 30% by mid-career. This gap is even wider in a number of states, especially in the South and West, as we show in Appendix B. By contrast, in high-achieving nations, teachers’ salaries are generally comparable to those of other 158 college graduates. A study by the Center for American Progress (CAP) in 2014 found that, increasingly, a teacher’s salary in much of the United States is too low to support a middle-class lifestyle. In 11 states, more than 20% of teachers work second jobs during the school year to supplement their incomes, as well as the additional work they take on in the summertime. The CAP study found that experienced teachers with 10 years of teaching experience made less than unskilled workers in a number of 159 states. In 30 states, mid-career teachers who head families of four or more qualify for three or more public benefit programs, such as subsidized children’s health insurance or free or reduced- price school meals. Compounding the problem of low wages in the teaching profession overall are great inequities in teacher salaries among districts within the same labor market, leaving some high-need, under- resourced districts at a strong disadvantage in hiring. An analysis of the Schools and Staffing Survey found that the best-paid teachers in low-poverty schools earned 35% more than their counterparts 160 Teachers in more advantaged communities also usually experience much in high-poverty schools. better working conditions, including smaller class sizes and more control over decision-making in 161 Given these salary differentials among districts, high-poverty districts consistently their schools. struggle to attract and retain effective teachers, who can often take a less-demanding, higher- 162 paying job in another district down the road. The Importance of Compensation Research finds that teachers are more likely to choose to enter the occupation when teacher salaries 163 Starting salaries and salary growth have are competitive in comparison to other occupations. an impact on recruiting and retaining individuals both to the profession and to particular states 164 or districts. Salaries also influence teacher attrition. Recent data from the National Center for Education Statistics show a 10 percentage point attrition gap between beginning teachers whose first-year 165 salary was $40,000 or more as compared to those earning less. Both beginning and veteran teachers are more likely to quit when they work in districts with lower wages and when their salaries are low relative to alternative wage opportunities, especially in high-demand fields like 166 mathematics and science. th grade teacher who had recently transferred to a district 15 miles away and Alishia Morris, a 4 across the border in Arkansas after six years of teaching in Oklahoma, explained how compensation and teaching resources influenced her decision: 55 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

62 “It wasn’t the school’s fault. If it was, it wouldn’t have been so difficult for me to leave. It’s just that Arkansas has more resources—they just make teaching easier,” Morris said. On top of a salary increase of $8,000 to $9,000 over the $33,500 she made at Westville [in Oklahoma], Morris now has reading and math facilitators to help her with her students, and Siloam Springs teachers [in Arkansas] get a $500 annual allowance for 167 classroom materials. States and districts that have addressed these compensation issues have eliminated shortages and, especially when they have also put in place policies to improve preparation and licensing, have improved quality (see box). Investing in Compensation and Quality Together Two of the most comprehensive examples, Connecticut and North Carolina, made omnibus investments in teaching during the 1980s and ʼ90s that solved perennial teacher shortages and created a strong supply of well-qualified teachers. Both states coupled major statewide increases in teacher salaries with improved pay equity across districts. A local control state, Connecticut provided funding incentives to help districts raise beginning salaries to a statewide minimum, offering more aid to lower-wealth districts, which made it possible for urban and poor rural districts to compete for qualified teachers. North Carolina increased its statewide salary schedule to reach the national average. In both cases, teacher shortages rapidly turned to surpluses. At the same time, both states strengthened teacher education and raised licensing standards. In this way, supply and quality were improved simultaneously. Connecticut required a content major and more extensive pedagogical training for literacy, special education, and teaching new English learners. North Carolina also boosted requirements and mandated that all public teacher education programs secure national accreditation. Both raised entry and licensing standards. Connecticut ended the practice of granting emergency credentials. Furthermore, because state teacher salary assistance could be spent only for fully certified teachers, districts had greater incentives to recruit those who had met the high new standards, and individuals had greater incentives to meet them. Both states instituted scholarships and forgivable loans targeted to individuals preparing to teach in high- demand fields. North Carolina launched a mentoring program for new teachers that greatly increased their access to early career support. Connecticut provided trained mentors for all beginning teachers and student teachers as part of its staged licensing process. Both also invested in extensive professional development for teachers and principals, aiming to create more productive school environments where strong instruction could flourish. North Carolina also adopted a groundbreaking performance-based salary increase—12% of base salary—for teachers who achieved National Board Certification—a process of demonstrating accomplished 168 teaching that is associated with greater teacher effectiveness. 169 both states experienced steep gains in As documented by the National Assessment Governing Board, achievement on the National Assessment of Educational Progress, with Connecticut becoming the top-ranked state in the nation after a decade of these investments, and North Carolina posting the largest gains and the sharpest reduction in the achievement gap of any state during that same period of time—becoming the first 170 Southern state to achieve above national norms in reading and mathematics. A comprehensive study of North Carolina’s teaching workforce also found that student achievement gains were significantly greater for students whose teachers had the strong academic and teaching preparation the state’s strategic policy investments have tried to leverage, along with those whose teachers were National 171 Board Certified. A study of Connecticut’s achievement gains found that in districts with sharply improved achievement, educators cited the high quality of teachers and administrators as a critical reason for their gains. They also noted that “when there is a teaching opening in a Connecticut elementary school, there are 172 often several hundred applicants.” Most notably, both states held to the course of teacher improvement over a sustained period—more than 15 years in each case. They demonstrate what state policy in support of good teaching can accomplish. 56 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

63 Another way to mitigate shortages is by attracting former teachers back to the classroom. A number of factors that past teachers say would make them consider returning to the classroom are related to compensation in one way or another. According to the National Center for Education Statistics (NCES), 53% of teachers who left the profession said they would consider returning to the classroom. When asked what would bring them back to teaching, aside from the availability of jobs, leavers’ responses included salary increases and the ability to maintain retirement benefits (both at 68%), as well as student loan forgiveness and housing incentives (about 25% each). Some former teachers who said they would consider returning to the profession listed the availability of part-time teaching positions and having child care options (41% and 30%, respectively), strategies that would allow them to balance home and work life (see Figure 29). FIGURE 27 Figure 29 What Would Bring Leavers Back? Factors rated by former teachers as important in a decision to return Ability to maintain teaching retirement benefits An increase in salary Smaller class sizes or smaller student load Easier & less costly renewal of certification State certification reciprocity HIGHEST-RATED FACTORS Availability of part-time teaching positions Availability of suitable child care options Forgiveness of student loans Housing incentives 40 50 60 70 80 0 10 20 30 PERCENT The percent of leavers who rated the factor as extremely or very important in their decision to return Note: Leavers included in this table are the 53% of teachers who left and said they would consider returning to the teaching workforce. Source: LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. A COMING CRISIS IN TEACHING? | 57 LEARNING POLICY INSTITUTE

64 To address these issues, we recommend that states and districts consider policies addressing compensation both directly and indirectly, as described below: Leverage more competitive and equitable salaries As many states and nations have discovered, a balanced teacher labor market during a time of high demand requires salaries that are competitive with other fields and that are equitable across districts, so that those serving high-need students have a fair shot at recruiting well-qualified educators. Given the inequalities in funding across most states, it is critical to leverage more resources to the districts that, because of this, are frequently hard to staff. States can accomplish this by equalizing funding across districts, as California and Massachusetts have done with weighted student funding formulas that direct resources to districts in relation to the different kinds of students they serve (e.g., those in poverty, English learners, foster care students). States can provide districts incentives to raise salaries to a competitive minimum level, as Connecticut did, subsidizing poor districts more extensively. They can increase statewide salary schedules, as North Carolina did. Different strategies will be appropriate in distinctive state contexts. Whatever the tactic, enabling teachers to earn a competitive wage across districts is associated with stronger recruitment and retention, and with a more robust and highly qualified supply of teachers. Create incentives that make living as a teacher more affordable In addition to salary hikes, states and districts can consider multiple strategies for making teaching more financially viable. Other forms of compensation are incentives that support teachers’ ability to stay in or re-enter the profession, such as: • mortgage guarantees, down payment assistance, or other housing supports, in exchange for service commitments; child care supports; and • • opportunities to continue teaching and mentoring after retirement. About 25% of teachers nationally point to housing incentives as an important factor in their potential decision to return to teaching, and some districts are developing housing solutions as part of their recruitment and retention initiatives. Among many others, for example, San Francisco recently passed a measure to provide stabilized housing for 500 teachers by 2020, and is building housing units on district-owned land. Texas provides eligible teachers with low, fixed-rate home 173 The federal government could help by expanding loans and grants for down payment assistance. its efforts through the U.S. Department of Housing and Urban Development (HUD) to discount housing prices for public servants in revitalization areas, and by creating more broad-based initiatives to support housing for teachers. Another 30% of prospective re-entrants point to child care supports as a potential enticement to return to teaching. Two-thirds of those who have left note that the ability to maintain retirement benefits if they returned could encourage them to re-enter the profession. All of these are potential incentives that districts or states could put in place. 58 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

65 2. Increasing Supply in Shortage Fields and Areas As policymakers address shortages, there is often a great deal of pressure to try to get warm bodies into classrooms quickly, even if that means skipping teacher preparation. This strategy, however, creates a leaky bucket that ultimately feeds greater shortages, as unprepared entrants into teaching leave at much higher rates, and they can create churn in schools that is costly both for student achievement and for district pocketbooks. science, special It is clear that in the fields with nationwide shortages—such as mathematics, education, and bilingual/ESL education—the pipelines into teaching must be expanded. Given the strong effects of preparation and mentoring on candidates’ effectiveness and rates of retention in teaching, it is important that strategies for increasing supply do so by strengthening incentives to enter along with supports to succeed. This can be accomplished by drawing in part on the federal experience with medical manpower programs. Since 1944, the federal government has subsidized medical training and facilities to meet the needs of underserved populations, to fill shortages in particular fields, and to increase diversity in the medical profession. Beginning with the Medical Manpower Act during the 1950s, the Health Professions Education Assistance Act of 1963, and continuing ever since, the federal government has invested in the preparation and distribution of physicians for high-need fields and locations, and has collected data to monitor and plan for medical manpower needs. This consistent commitment has contributed significantly to America’s world-renowned system of medical training and care. Making Preparation Affordable A growing body of evidence indicates that attrition is unusually high for those who lack preparation 174 for teaching. One recent study found that teachers who received little pedagogical training were more than twice as likely to leave teaching after their first year than teachers who had received a comprehensive preparation, including observing others teaching, student teaching a full semester, receiving feedback, and taking five or more courses in teaching methods, in addition to receiving 175 training in learning theory and selecting instructional materials. Another study found that 30% of uncertified Teachers who received little entrants left the profession within five years, 176 As we compared to 15% of certified entrants. pedagogical training were more described earlier, our analyses of the Schools than twice as likely to leave and Staffing database found that alternatively certified teachers (who typically enter without teaching after their first year than student teaching, and take courses on nights teachers who had received a and weekends) were much less likely to stay in comprehensive preparation. teaching. Of note, they also had a much shorter tenure in schools serving concentrations of low-income and minority students than teachers who were fully prepared—averaging 177 about 5½ years in those schools in comparison to nine years for regularly certified teachers. 59 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

66 Other national studies, plus local studies in New York, North Carolina, and Texas, have found that individuals who enter teaching before they have had the opportunity to prepare are less effective 178 when they are teaching while still in training, and leave teaching at much higher rates than those 179 who are fully prepared when they enter. A key issue, however, is how candidates can afford adequate preparation—especially when they may have had to go into debt to prepare to enter a profession that earns less than other alternatives. Many prospective teachers are saddled with college debt incurred while undergoing their teacher 180 training. Research shows that the more debt college students incur, the less likely they are to choose to work in a lower-wage profession, like teaching. One study of students at a highly selective undergraduate institution found that incurring debt increased the odds that students chose “substantially higher-salary jobs” and “reduce[d] the probability that students [chose] low-paid 181 ‘public interest’ jobs.” The influence of debt on job choice was “most notable on the propensity to work in the education industry.” To make teaching affordable, some states and the federal government have provided forgivable loans and service scholarships that subsidize preparation, just as the Health Professions Education Assistance Act has long done for doctors. These subsidies are paid back with a number of years of service in the profession. A recent review of research on these strategies in both medicine and teaching found that loan forgiveness or service scholarship programs covering a significant portion of tuition and/or living costs are effective in recruiting candidates into the profession and into high-need locations and 182 fields. Among the things that federal and state governments can do are the following: Offer forgivable loans and service scholarships First, as it does in medicine, the federal government should maintain a substantial, sustained program of service scholarships that cover training costs in high-quality preparation programs at the undergraduate or graduate level for those who will teach in a high-need field or location for at least four years. (After three years, candidates are much more likely to remain in the profession and to make a difference for student achievement. State governments can augment such an approach ) with programs targeted to specific local needs. As noted earlier, research on such programs in medicine and teaching has found that they are effective in recruiting and retaining candidates in the profession, and into high-need locations and 183 fields, if they cover a large enough share of candidates’ training costs. Whereas other high-achieving countries underwrite all of the costs of high-quality teacher preparation and often offer additional wages to those who go to high-need areas, in the United States, there are few supports for teacher education or distribution. The major supports that were enacted in the 1960s and ʼ70s to underwrite preparation for teachers to go to high-need fields and high-need schools (under the National Defense and Education Act) ended in the 1980s, and have not been fully reinstated since then. While some federal grants are currently available, they are not designed to serve as an adequate incentive to candidates. Service scholarships (as opposed to post hoc forgivable loans), which have the benefit of being able to target candidates who might not otherwise enter teacher preparation, can be used proactively to y are needed. recruit specific kinds of candidates to the fields and locations where the 60 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

67 Perhaps the best-known model of such an approach—subsequently copied in other states—is the . In operation for more than 25 years, the program North Carolina Teaching Fellows Program selected highly able high school students and paid all college costs, including an enhanced and 184 fully funded teacher education program, in return for several years of teaching. The program has recruited nearly 11,000 candidates into teaching, representing approximately 10% of all teachers credentialed each year in North Carolina. Among these have been a larger than usual number of males, minority candidates, and mathematics and science teachers. A recent study of the program found that the Teaching Fellows are on average more effective than their peers in supporting student achievement. In addition, they are much more likely to stay in teaching, with 75% still in teaching by the fifth year, as compared to only about 60% of other prepared entrants 185 and 40% of alternative route entrants. Many of the remaining Fellows were still in education as administrators. Similar programs were later started in South Carolina (the South Carolina Teaching Fellows ) Governor’s Teaching Fellowships ). In South Carolina’s program, over 82% and California (the of 1,502 graduates between 2000 and 2011 either satisfied the service scholarship commitment 186 or are currently teaching to fulfill the commitment in a South Carolina public school district. California’s program, which established $20,000 service scholarships to high-ability college graduates who would prepare to teach in underperforming schools, recruited candidates entering one-year graduate-level teacher education programs. An evaluation found this program provided a supply of high-ability, well-trained candidates to high-need schools in a short time period, encouraging many who would not otherwise have gone to such schools to do so, with high 187 retention rates. The financial incentives offered by service scholarships like these enhance the attractions to teaching and support stronger preparation by eliminating student debt payments, while improving candidates’ preparedness. States can launch similar programs. As in medicine, the federal government should also play a major role, for at least two reasons. First, a federal approach can influence the flow of talent across areas of the country, allowing candidates trained in exporting states (those that prepare more teachers than they need) to pay off their loans or scholarships by working in states that prepare too few. Second, the budgetary implications are extremely modest for the federal government relative to the states. A relatively small federal outlay could go a long way—and ultimately save the nation sizable sums by reducing attrition. For example, a program costing $800 million a year could support 40,000 new teachers with $20,000 loans or scholarships, addressing most of the nation’s teacher shortfall over the course of a few years. Create career pathways and “Grow Your High school career pathway Own” programs programs, such as cadet To attack the systemic nature of teacher programs and teaching shortages in urban and isolated rural schools, it is important also to develop pipelines of academies, are ways to interest candidates who are committed specifically high school students in teaching to those communities. Since many young and start them on the path to teachers have a strong preference to teach close to home, states and the federal the profession. government can help urban and rural schools 61 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

68 enhance the pipeline of potential teachers from these neighborhoods by developing high school 188 career pathways and “Grow Your Own” programs. High school career pathway programs, such as cadet programs and teaching academies, are ways to 189 interest high school students in teaching and start them on the path to the profession. Teacher 191 190 and North Carolina Cadet programs in South Carolina offer college-level courses in education to high school students, along with extracurricular activities focused on teaching and tutoring, which result in many later entering teacher preparation. In some states, high school career academies have focused on teaching as a profession, combining college preparatory academic study with education-related courses and real-world experiences 192 for students. The Education Academy at Skyline High School in the Oakland Unified School District, in California, for example, now has graduates serving as teachers and administrators in 193 Hamline University’s partnership with Mounds View Public Schools outside of Saint that district. Paul, Minnesota, allows students to earn credits toward both a teaching credential and high school 93 Targeted expansions graduation requirements during their junior and senior years in high school. of federal Perkins Career Technical Education Act funds could support the growth of these kinds of programs. Grow Your Own programs recruit community members into teaching and support them as they complete their bachelor’s degree and teaching credential. One model is the California Teacher Pathway program, which recruits young people interested in becoming educators; supports them through the process of earning their associate’s degree, bachelor’s degree, and teaching credential; and helps them gain stable employment in after-school programs, while they are studying to gain 194 experience working with youths and to support them through their studies. Other models support paraprofessionals and teachers’ aides in earning a teaching credential. Such programs have proven highly successful for recruiting diverse individuals who are rooted in communities and already have a great deal of experience working with students, often in special education and bilingual settings where there are significant shortages, and retaining them in 195 California’s Paraprofessional Teacher Training Program, for example, trained more teaching. th than 2,200 graduates, two-thirds of whom were people of color and bilingual. As of its 13 year of 196 operation, 92% remained California school employees. In addition to specifically targeted funding sources, federal Workforce Innovation and Opportunity Act funds can be used to support these types of comprehensive teacher training programs designed to address local workforce needs in times of teacher shortages. Establish teacher residency models in hard-to-staff districts New approaches to training and recruitment are needed if we are to solve shortages in communities and fields that have long-standing challenges with the interrelated problems of adequate preparation and adequate supply. When underprepared recruits leave the profession quickly, students can face a revolving door of teachers. Under these circumstances, everyone loses. Student achievement is undermined by high rates of teacher turnover and teachers who are inadequately prepared for the challenges they face. Schools suffer from continual churn, undermining long-term improvement efforts. Districts pay the costs of both students’ underachievement and teachers’ high attrition. 62 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

69 The teacher residency model is a new and When underprepared recruits important strategy that solves many of these problems. Urban and rural residency programs, leave the profession quickly, sponsored by individual districts or consortia students can face a revolving of districts, place candidates who plan to teach door of teachers. in shortage fields and who want to commit to high-need urban or rural schools into paid year-long apprenticeships with expert mentor teachers, while they complete tightly linked credential coursework and earn a master’s degree from partnering universities. When they become teachers, these recruits also receive two years of mentoring. In exchange for this high-quality preparation—which is directly focused on becoming an excellent teacher in the specific community where they are hired—candidates pledge to spend three to five years in the district’s schools. Most stay on for the long haul, becoming leaders in the schools rather than dropouts from the profession. Research has shown residencies to be effective at recruiting and retaining talented and diverse 197 candidates in high-need schools, and better preparing them for the challenges they will face. These initiatives are much like the federally funded urban medical training models that have created programs specifically designed to prepare doctors and other health professionals for urban 198 communities, and have established community health centers to support clinical preparation. Such programs solve several problems simultaneously—creating a pipeline of committed teachers who are well prepared to engage in best practice for children in high-need schools, while creating demonstration sites that serve as models for teaching and teacher education. To significantly expand the reach of these programs, the Teacher Quality Partnership grants, which fund residencies, could be expanded to their authorized level of $300 million. AmeriCorps funds and TEACH grants or other scholarship programs can also support these candidates. 3. Improving Teacher Retention In addition to incentives for entering teaching, improving teacher retention is a critical goal. High attrition and turnover are a major part of the problem, especially for beginning teachers and in hard-to-staff schools. A great unfinished task in American education is to create conditions for better support of new teachers, as well as the conditions under which experienced teachers will want to stay in the profession and in schools serving high-need students. The Importance of High-Quality Mentoring and Induction for Beginners Expert mentoring in the first years of teaching enhances the retention effects of strong initial preparation. Early induction opportunities offering mentoring and other teaching supports are an important part of keeping beginning teachers in the profession and building their competence and self-efficacy, which in turn enhances retention. The more teachers know about how to do their 199 jobs well, the more they experience a sense of self-efficacy and derive satisfaction from teaching. They are, in turn, supported by having had useful coursework and clinical learning experiences both before entry and while on the job. 63 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

70 A number of studies have found that well-designed mentoring programs improve retention rates for 200 Key to success new teachers, as well as their attitudes, feelings of efficacy, and instructional skills. is having a mentor teacher in the same subject area, common planning time with teachers in the 201 same subject, and regularly scheduled collaboration with other teachers. Beginning teachers’ practice is enhanced further when their mentors also receive formal training, and are released from some of their own classroom duties to provide one-to-one observation and coaching in the classroom, so they can demonstrate effective methods and help new teachers solve immediate 202 problems of practice. The amount and kind of induction support matters. In addition to mentoring, Well-designed mentoring collaboration time for planning instruction programs improve retention rates with other teachers is key and can, if coupled with a beginning teacher seminar, reduce for new teachers, as well as their turnover by 34%. In the Cadillac version, attitudes, feelings of efficacy, and if a teacher receives basic mentoring, plus instructional skills. collaboration, a strong teacher network, and extra resources (a reduced number of preparation periods and a teacher’s aide), first-year turnover (leaving the school or the 203 profession) is cut by more than half, from 41% to 18%. Researchers have found that beginning teachers who participate in induction are more able to keep students on task, develop workable lesson plans, use effective questioning practices, adjust classroom activities to meet students’ interests, maintain a positive classroom atmosphere, and 204 At least one study has found that students of demonstrate successful classroom management. 205 beginning teachers who participated in induction showed stronger achievement gains. Despite the fact that most states have some kind of requirement for beginning teacher induction, few offer the suite of supports that are most effective for keeping teachers in the profession. As of 2011–12, only 33% of beginning teachers had access to mentoring, common planning, supportive communication with their principal, and seminars, and only 2.5% had access to the most 206 comprehensive set of supports described above. Furthermore, the funding for these programs has been cut in many states in recent years as a function of shrinking budgets. The proportions of teachers receiving induction services of all kinds dropped between 2008 and 2012. For example, the share of beginners working with a mentor and supported by periodic conversations with the principal dropped from 75% to only 59% in that 207 Consequently, beginning teachers’ access to high-quality mentoring and induction period of time. is much more variable than it was a few years ago. Develop strong, universally available induction programs With new teachers leaving at high rates, especially in urban and poor rural areas, the revolving door cannot be slowed until the needs for beginning teacher support are addressed. Other high-achieving countries invest heavily in structured induction for beginning teachers. They fund schools to provide released time for expert mentors, and they fund other learning opportunities for beginners, such as seminars in key areas (e.g., classroom management, working with special education 64 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

71 students, and engaging with parents), visits to other teachers’ classrooms, and joint planning time. As noted above, such strategies are also found effective in reducing beginning teacher turnover in the United States, but very few teachers receive the complement of supports that are most effective. Federal or state matching grant programs aimed at these features could ensure support for every new teacher in the nation through investments in district mentoring programs. Based on the funding model historically used in California’s Beginning Teacher Support and Assessment Program, for example, an allocation of $4,000 for each beginning teacher, matched by local 208 a district funds, could fund mentoring for every novice. At 125,000 new teachers each year, federal investment of $500 million per year could ensure that each novice is coached by a trained, accomplished mentor with expertise in the relevant teaching field. Alternatively, federal leverage could be applied by offering matching grants to states to develop or expand induction programs that provide the key elements known to make the most difference in teacher retention: mentoring by a trained mentor in the same teaching field, collaborative planning time, a reduced teaching load, and additional learning opportunities on key issues. How Teaching Conditions Matter Surveys of teachers have long shown that teaching conditions play a major role in their decisions to move schools or leave the profession. As Alishia Morris noted of her decision to move schools, teaching conditions such as the availability of resources for materials and instructional specialists factored into the decision, along with salaries. The relatively poor teaching conditions in many low-wealth schools are a major reason why teachers in high-poverty schools are more than twice as 209 likely to leave due to dissatisfaction as those in low-poverty schools. Beyond resources, teachers’ plans to stay in teaching and their reasons for actually having left are strongly associated with how they feel about administrative support, collegial opportunities, and teacher input into decision-making. When these conditions are present, retaining teachers is much easier. As a 20-year public school veteran in Minneapolis noted, For the past decade, I’ve worked at a school where 97% of the children qualify for free and reduced-price lunch. I stay because the school climate is good for children and teachers alike. I stay because my principal is wonderful, supports us, does what’s best for children, and because I trust her. I stay because my colleagues are gifted teachers and good company, 210 and because I continually learn from them. A poll by the Public Agenda Foundation found that almost 80% of teachers said they would choose to teach in a school where administrators supported them, as opposed to only about 20% who said 211 they would teach at one with significantly higher salaries. Some policies have emphasized monetary bonuses or stipends to attract teachers to high-need schools. Under some circumstances, such bonuses have helped make a dent in schools’ hiring 212 However, they are rarely enough, by themselves, to solve shortage problems. For example, needs. one study of efforts to recruit high-performing teachers to struggling schools found that, among 1,500 such teachers in the Talent Transfer Initiative, only 22% were willing to apply to transfer to high-need schools for a two-year bonus of $20,000, and fewer than 5% ultimately did so. Although 65 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

72 the targeted teachers filled most of the 81 vacancies, attrition rates of these teachers soared to 40% 213 after the bonuses were paid out and the money disappeared. As another review found: [S]chool districts have tried offering additional pay for high-needs schools without much positive result, even when substantial bonuses are awarded. In 2004, Palm Beach, Florida, eliminated its $7,500 high-needs school stipend after few teachers took the offer. Dallas’s offer of $6,000 to accomplished teachers to move to challenging schools also failed to generate much interest. ... A decade ago, South Carolina set out to recruit “teacher specialists” to work in the state’s weakest schools. Despite the offer of an $18,000 bonus, the state attracted only 20% of the 500 teachers it needed in the first year of the program, 214 and only 40% after three years. Although money can help, teachers are primarily attracted by principals who are good instructional leaders, by like-minded colleagues who are committed to the same goals, by having the teaching conditions and instructional materials they need readily available, and by having learning supports that enable them to be effective. As one National Board Certified teacher noted in a discussion of what would attract him to a high-needs school: I would move [to a low-performing school], but I would want to see social services for parents and children, accomplished leadership, adequate resources and facilities, and flexibility, freedom, and time. ... One of the single greatest factors in school success is principal leadership. Effective administrators are magnets for accomplished teachers. In addition, it is amazing to me that attention is being paid to teaching quality in hard-to- staff schools when little is done to address the sometimes appalling conditions in which teachers are forced to work and students are forced to learn. ... Finally, as an accomplished teacher, my greatest fear is being assigned to a hard-to-staff school, and not being given the time and the flexibility to make the changes that I believe are necessary to bring about 85 student achievement. Create productive school environments To create school environments that attract and retain great teachers, it is critical to provide both the tangible resources that enable teachers to teach—instructional materials, texts, computers, reasonable class sizes, and instructional specialists—and the context that allows them to feel supported and become effective. To keep high-quality teachers in high-poverty communities, schools need to offer working conditions that support teacher and student success. One policy strategy is to allocate state and/or district funds specifically to improve teaching conditions in hard-to-staff schools: smaller classes and pupil loads, administrative supports for necessary materials and supplies, and time for teacher planning and professional development—all of which help attract and keep teachers in schools. Montgomery County, Maryland, pursued a policy that reduced class sizes, and expanded teacher collaboration time and training to ensure stronger supports for teachers in its schools serving concentrations of low-income and minority students. These moves enhanced recruitment and retention, strengthened the teaching force, and increased the levels of achievement in those 215 schools. California implemented a similar approach in its Teachers as a Priority Program, which sent resources to high-need schools to recruit and retain fully certified teachers through improving 66 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

73 working conditions, adding mentors, reducing class sizes, and providing hiring or retention bonuses. The state also implemented $10,000 bonuses for National Board Certified teachers, increased to $20,000 for such teachers who taught in low-performing schools—a strategy that led to 216 much more equitable distribution of these accomplished teachers than common in other states. As part of a broader career ladder initiative, federal and/or state matching grants to districts can provide incentives for the design of innovative approaches to attract and keep accomplished teachers in priority low-income schools, including through compensation for accomplishment and for additional responsibilities, such as mentoring and coaching. Finally, as recent data and research have highlighted, teachers have been discouraged from 217 staying in high-need schools by accountability pressures tied to negative labels and sanctions. Of the teachers who voluntarily left the profession after the 2011–12 school year for reasons other than retirement, approximately 25% reported that dissatisfaction with the influence of school assessment and accountability measures on their teaching or curriculum was extremely or very important in their decision to leave teaching; 17% reported that dissatisfaction with in their decision to very important or support preparing students for assessments was extremely 218 leave teaching. With the advent of the Every Student Succeeds Act (ESSA), which has replaced No Child Left Behind, it will be important for the federal government, states, and districts to consider how to create productive approaches to accountability that enable schools to improve without creating disincentives for staff to work in the most challenging environments. State and district efforts to focus on identifying areas for improvement with productive resources to enable that improvement would better support the recruitment and retention of capable teachers in high-need schools. Strengthen principal training programs Teachers are clear that their decisions to stay in the profession rest substantially on the capacity of administrators to create a productive teaching and learning environment in which they can be effective and continue to develop their skills. As we described earlier, teacher attrition rates increase sharply when teachers feel they do not have administrative support. However, school leadership 219 influences teacher turnover in ways that go beyond administrative support of teachers, since principals influence just about every major area of school functioning, from hiring, evaluation, and professional development to the organization of time and resources to opportunities for teacher input and creativity in the classroom, all of which affect the context of teaching. Efforts to improve the knowledge and skills of school leaders would go a long way toward improving teacher recruitment and retention, as well as teacher effectiveness. This is particularly true in disadvantaged schools. Multiple studies of teacher attrition in high-poverty schools have found that teachers’ perceptions of their school’s leader is a dominant factor in their decision to remain in the 220 Teachers working in schools with large proportions of low-income and minority students school. tend to rate their principals as less effective, and the impact of this rating on their decisions to leave 221 is larger in such schools. A synthesis of six studies analyzing teacher turnover in high-poverty schools found that effective school leaders were: 67 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

74 • Effective school managers (ensuring that teachers have the necessary resources, communication channels, sensible budgets). • Effective instructional leaders (strategically hiring teachers and staff, providing regular and fair teacher evaluations, helping their teachers to continually improve). • Inclusive decision-makers (listening to teachers’ ideas and engaging them in change, 222 providing teacher autonomy within their classrooms as appropriate). Teachers in high-poverty schools identify an inclusive environment characterized by respect and trust among colleagues, formal structures that promote collaboration, and the presence of a shared 223 mission among teachers as important to them. Creating a cadre of principals who know how to create such environments is critically important, and it is an area where the United States lags behind other leading nations. There are no major federally funded initiatives, and few state initiatives to support the high-quality recruitment and training of school principals. North Carolina’s Principal Fellows Program, which supports preparation that includes a full-time internship under the mentorship of an e xpert principal, is an exception, as is Delaware’s mentoring program for beginning principals, provided through its state 224 leadership academy. State and federal policies should support efforts to recruit promising candidates for school leadership positions, something that has become increasingly important as the challenges of the 225 job often discourage strong candidates from entering the field. State and federal agencies could offer grant funding and technical assistance for creating and expanding high-quality principal training programs that focus on these abilities, as well as competitive service scholarship programs for principal preparation to attract exemplary candidates to the field. States can fund leadership training through state leadership academies. Finally, states and districts should consider means to recruit the most successful principals to the settings where the needs for their skills are the greatest. This could take the form of compensation, but also the provision of school resources needed to do the job well: wraparound social supports for students; overstaffing that allows for mentoring, reasonable pupil loads, and collaborative time; opportunities to recruit teachers; and plentiful professional development resources to support learning and success. 4. Facilitating a National Labor Market for Teachers Finally, there is a need to enable teachers to stay in the profession when they move across the country. In addition to the one-quarter of teachers who end up teaching in a state where they did 226 not prepare, there are many others who drop out of teaching when they move across state lines. This is frequently because of the challenges of securing a different state license; maintaining pension benefits, which vary from state to state; or transferring their years of seniority into the salary schedule, so that they do not fall back in compensation. Of teachers who have left and would consider returning to teaching, more than 40% cited state certification reciprocity as an important factor, and nearly 70% cited the ability to keep teaching retirement benefits. 68 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

75 Support teacher mobility Federal policy can help create the foundation of a national labor market for teachers, including the removal of unnecessary interstate barriers to teacher mobility. Because teacher supply and demand vary regionally, the country can benefit only if states with teacher surpluses in particular fields can be connected to states with corresponding shortages. States could work together, potentially with federal incentives like those used to create the interstate highway system, to accomplish three goals: 1. Support common licensing exams that evaluate teaching performance and interstate agreements about content and pedagogical coursework that would facilitate more complete license reciprocity. 2. Create a system of pension portability across the states. 3. Provide labor market data and analyses for federal, state, and local planning. Several groups already are working on these agendas in ways that could be leveraged toward genuine changes. For example, National Board Certification—a rigorous demonstration of teaching accomplishment—is already accepted for license reciprocity in some states. The Interstate New Teacher Assessment and Support Consortium (INTASC), sponsored by the Council of Chief State School Officers, has developed common licensing standards adopted or adapted in most states that could provide the basis for agreements on coursework and assessments. A new national performance assessment for beginning teachers, modeled on the National Board portfolio, has been developed based on these standards and has been piloted or adopted for use by programs in more 227 than 30 states. These elements could be the basis for an eventual national system. The organization of State Higher Education Executive Officers, along with the Education Commission of the States, has examined how to achieve teacher pension portability, and TIAA- CREF has developed such analyses as well. A public/private partnership to stimulate the next steps in these plans could be extremely productive. In addition, the long-standing federal role of keeping statistics and managing research is well suited to the job of creating a database and analytic agenda for monitoring teacher supply and demand. Such a system, which would inform other policies, could document and project shortage areas and fields; determine priorities for federal, state, and local recruitment incentives; and support plans for institutional investments where they are needed. 69 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

76 VII. Conclusion Current data and projections reveal an emerging teacher shortage in the United States that will, if trends continue, grow worse before it improves and exacerbate perennial shortages in areas such as mathematics, science, and special education. These shortages also worsen the inequitable distribution of qualified teachers to schools serving concentrations of low-income students and students of color. Attention to the current status of the teaching force could provide an opportunity for the United States to take a long-term approach to establishing a comprehensive and systematic set of strategies to build a strong, stable profession, as was done in medicine more than half a century ago. This would include: Creating competitive and equitable compensation packages that make teaching an • affordable choice for candidates across communities. Enhancing the supply of qualified teachers targeted to high-need fields and locations • through targeted training subsidies and high-retention pathways. • Improving teacher retention, especially in hard-to-staff schools, through improved mentoring, induction, working conditions, and career development. • Developing a national teacher supply market, with license reciprocity and portable pensions that can facilitate getting and keeping teachers in the places they are needed over the course of their careers. Although these proposals are associated with costs, it is likely that they would ultimately save far more than they would cost. The savings would include the more than $8 billion now wasted annually on replacement costs because of high teacher turnover, plus much of the cost of grade retention, summer schools, and remedial programs required because too many children are poorly taught—not to mention the broader social costs of dropouts, unemployment, and incarceration that are associated with a failure to ensure that high-quality teaching can occur in our highest-need communities. In the competition for educational investment, the evidence strongly points to the importance of teacher quality for educational improvement. Preventing and solving teacher shortages so that all children receive competent, continuous instruction in every community every year is, in a 21st century economy, essential for the success of individuals as well as for society as a whole. 70 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

77 Appendix A: Methodology Data The analyses in this report rely on a number of databases. To analyze and project demand, we first examine teacher turnover in the workforce using the Schools and Staffing Survey (SASS) Teacher File 2003–04, 2007–08, and 2011–12, and the SASS Teacher Follow–Up Survey 2004–05, 2008–09, and 2012–13. Both of these are nationally representative data sources that monitor teachers and schools over time. The Common Core of Data (CCD), years 1999–00 through 2012–13, a universal database on teachers and students in the United States, provides accurate teacher and student counts. Finally, public school teacher projections 2000 to 2025 published in the Digest of Education Statistics allow for our model to estimate workforce trends a decade into the future. To examine teacher production and the supply side of the labor market, this analysis uses universal data on teacher preparation programs collected by the U.S. Department of Education under Title II 228 —the most recent and complete national data on teacher preparation. of the Higher Education Act To further investigate an individual’s journey from teacher preparation to the classroom, we use the 2008:2012 Baccalaureate and Beyond (B&B), a longitudinal dataset that follows recent baccalaureates from 2008 until 2012, four years after their graduation, with a special focus on careers in education. This collection of data is used in tandem with the modeling techniques below to provide the estimates and projections in this paper. Modeling Demand The demand model aims to predict the number of teachers districts want to hire in each school year. As explained in the main text, teacher demand is shifted by three factors: student enrollment, pupil-teacher ratio, and the attrition rate. Due to the structure of data available, in the context of this methodology, demand is thought of as having two main sources: (1) growth or decline in the workforce due to student enrollment and pupil-teacher ratios, and (2) vacancies created due to attrition. Together, these two factors make up demand for teacher hires in a given year. Our overall methodological approach comprises four steps: 1. Estimate the increase or decrease in the necessary number of public school teachers to educate all students between 2011–12 and 2012–13, using the NCES workforce projections. 2. Estimate the number of teachers who left the classroom between 2011–12 and 2012–13— vacancies created due to attrition. 3. Combine the change in total teachers demanded in step 1 with the number of teachers who left the classroom and need to be replaced in step 2 to calculate the demanded hires in 2012–13. 4. Repeat for years 2014 to 2025, assuming trends affecting public school enrollment (e.g., birthrates, immigration, and grade progression rates), teacher wages and education revenue, teacher experience-based continuation rates, the experience distribution for newly hired teachers, and general economic and political conditions remain similar during the projected period. 71 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

78 Each of these steps has its own more complex methodological strategy to produce responsible and accurate projections. As explained above, the number of teacher hires needed due to growth or decline in the workforce is driven by both student enrollments and changes in the pupil-teacher ratio, both of which are 229 directly incorporated in the NCES teacher projections. Calculating the year-to-year difference in projected public school teachers reveals the estimated growth or decline in the teacher workforce. For example, the teacher workforce growth between 2011–12 and 2012–13 is estimated to be 230 5,839 teachers. Change in workforce estimates can be found for every year in Table A1 below, also shown in Figure A1. These data reveal severe shrinking of the workforce during the recession, followed by a slow increase in the workforce during recovery. Most notably, there is a predicted increase of 47,000 teachers in the 2017–18 school year. This increase persists through the end of the projection period, signaling a continued increase in demand. It is worth noting these estimates account for only student enrollment, pupil-teacher ratio, and the subsidiary factors that are used to calculate those figures, such as birthrates, migration, immigration, and current school budget—all assumed to continue with similar trends in the projected period. Some of these intermediate projections can be quite difficult to produce, making these estimates just that—estimates. They do not account for exogenous shocks that could take place in the future, such as a sudden recession or a shift in education policy that funds smaller class sizes. In all, these projections are assuming the continuation of conditions in the current time period. FIGURE A-1 Figure A1 Annual Change in Teacher Workforce Size Over Time 60,000 40,000 20,000 0 -20,000 -40,000 -60,000 -80,000 -100,000 2025 2019 2022 2021 2024 2023 2020 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Note: Negative number of teachers represents a decrease in the total number of teachers. Source: National Center for Education Statistics. (2015). Digest of Education Statistics. Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. 72 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

79 The next step is to calculate the number of hires needed due to attrition. In order to better estimate the number of teachers who left in a given year, it is important to differentiate teachers by shared characteristics, because a first-year teacher will not leave at the same rate as a teacher in her or his th year. This paper assumes that teachers with similar levels of experience will exit the teaching 10 workforce at similar rates. Other analyses disaggregate attrition by age; however, because it is 231 and some evidence suggests that more and more common to enter teaching at a variety of ages, teacher attrition behavior at the start of the career is associated with years of experience, there are arguments for both methods. Early in one’s career, it makes sense that experience may be a driving factor in differential attrition rates; however, when retirement rolls around, age might be a bigger factor than experience. Results using both methods are displayed here for comparison purposes. The projections turn out to be very similar: Age-based attrition estimates are on average lower by 3,306 teachers (see Table A2 in this Appendix). To calculate demand due to attrition, the first step is to apply the experience distribution observed 232 Because there are in the SASS 2011–12 TFS to the total number of teachers in 2011 per the CCD. small sample sizes for teachers with more than 40 years of experience, these teachers are clumped into one category of teachers with over 40 years of experience. This decision yields a sample size of at least 34 for all experience groups under 40. Next, for each experience group of teachers, an experience-specific continuation rate is applied. For example, 92.45% of first-year teachers continue on to the next year, 92.5% of second-year teachers continue on to the next year, and so on. When projecting into the future, these rates were exponentially smoothed to appropriately weight observed data from the 2004–05 TFS and 2008–09 TFS into the projections. In general, teacher-leaving rates increase noticeably after about 25 years of experience. Figure A2 displays the continuation rates for each experience level. FIGURE A-2 Figure A2 –13 Public School Experience-Based Continuation Rates, 2011 100 80 60 PERCENT 40 20 0 >40 5 29 27 25 23 21 19 17 15 13 11 3 1 9 7 39 37 35 33 31 YEARS OF EXPERIENCE Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics; Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 73 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

80 To determine how many teachers left the classroom, each experience group is aggregated to calculate how many teachers continued from 2011–12 to 2012–13 school years. This number is then subtracted from the total number of teachers in 2011–12 to obtain the number of teachers who left teaching. By combining the new teachers demanded due to workforce growth with the new teachers demanded due to teachers leaving the workforce, we derive estimated new hires for 2012–13. The final step is to repeat this process for years 2014–25. In this step, because another school year has passed, teachers must advance by one year in the experience distribution. Each teacher is moved up one experience category, with teachers with 40 years of experience being moved to the over 40 years of experience category. By shifting these teachers’ experience by one year, there are no new teachers. It would be easy to fill this empty value if all new hires were first-year teachers; however, many new hires are actually re-entrants and have prior teaching experience. In order to appropriately account for the different experience levels of new hires, using the SASS 2011–12 public school teacher data file, the experience distribution of new teachers is applied to the estimated new teacher hires from the prior year. Figure A3 displays the experience distribution of new teacher hires. As expected, most new hires are first-year teachers, but a significant portion are re-entrants. FIGURE A-3 Figure A3 – 12 Teacher Hires, 201 Experience Distribution of New 1 50 40 30 PERCENT 20 10 0 >40 39 37 35 33 31 29 27 25 23 21 19 17 15 13 11 9 5 3 7 1 YEARS OF EXPERIENCE Source: LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. Once the new hires have been distributed across the experience groupings, steps 1 through 3 are repeated for each year. Experience-based continuation rates are exponentially smoothed into the 74 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

81 future to appropriately weight each year of observed data. The most recent data are weighed most heavily, with exponentially diminishing value given to the older observed data from the SASS 2008–09 TFS and SASS 2004–05 TFS. This process yields projected teacher demand estimates to 2024–25. The main limitation of the aforementioned methodology, which is used in most forecasting, is that the model heavily relies on one year of data—data collected from the SASS 2011–12. Although using exponential smoothing for continuation rates helps incorporate older data, most of the data and assumptions are derived from the 2011–12 school year. In summary, the demand estimates of new teacher hires are projected a decade into the future using the data described above. Demanded new hires are calculated by combining the number of new teachers in response to changes in the workforce and the number of new teachers necessary to replace teachers who left. To model these estimates into the future, remaining teachers gain an additional year of experience, and new hires are distributed into the corresponding experience category according to the SASS 2012–12 experience distribution of new hires. Additionally, experience-based continuation rates are exponentially smoothed to incorporate more base data into the projections, while at the same time weighting older observations. Disaggregating demand—analysis used in Figures 2, 6, and 15 As described above, teacher demand in a given year is driven by two factors: additional teachers due to marginal increases (or decreases) in the size of the teacher workforce, and additional teachers to replace those who left teaching. To disaggregate these estimates into smaller components, workforce growth was separated into student enrollment-driven workforce growth and pupil- teacher ratio-driven workforce growth. Teacher demand due to student enrollment growth can be calculated by dividing the increase in student enrollment by the current teacher-pupil ratio. The difference between the number of teachers necessary under the current pupil-teacher ratio and the number of teachers necessary under the NCES projected pupil-teacher ratio represents the increase in teachers needed due to changes in the pupil-teacher ratio. To break down attrition into retirement and non-retirement attrition, this analysis uses data from the SASS 2004–05, 2008–09, and 2012–13 TFS that find on average 32% of teachers who leave teaching do so to retire. A new hire is defined as a teacher who is teaching in a public school in the current year but was not teaching in a public school the previous year. Additionally, a private school teacher moving to the public sector is a “new hire.” Modeling Supply The goal of the supply model is to estimate the number of qualified teachers available to enter the classroom in a given year. The analysis can be broken down into two questions: (1) how many new entrants (i.e., first-time teachers) are available to be hired? and (2) how many re-entrants, teachers who had left teaching but are now returning to the classroom, are in the labor market? In order to estimate new entrants, this report leverages the fact that there is a lag New Entrants. between candidates’ initial enrollment in and subsequent completion of a preparation program. Decreased enrollment in teacher preparation does not bring down the number of completers in that year but does so over the subsequent years, depending on how long it takes candidates to finish 233 The steadily decreasing enrollment numbers indicate that completers will their requirements. continue to decrease at least in the several years after the most recent national data available for 75 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

82 234 enrollments—2013–14. The ratio of current enrollments to future completers is .303. In other words, on average 30% of teacher preparation enrollments two years ago are completers in the current year. This estimation is not the graduation rate; the other 70% do not drop out; more often than not, they complete in other years. This is to be expected given that some teacher preparation programs take one year, some two years, and others up to four years. The extrapolation from 235 where 2015 and 2016 enrollments to completers provides data on completers from 2005 to 2016, are projected using these lagged ratios. One of the biggest challenges of modeling new entrants is determining how many of these program completers end up in the classroom the following year, how many are delayed entrants, and how many never end up teaching at all. We used survey responses from the 2008:2012 Baccalaureate and Beyond (B&B), a longitudinal study that follows college graduates four years after they receive their BA degree to determine the percentage of students who were prepared to teach but did not end up 236 in the classroom. This longitudinal analysis found that 25.01% of students who received a BA degree in 2008, and had prepared for teaching or did so within the next three years, did not end up 237 teaching by 2012. The four-year average rate allows for more completers to enter the classroom than the one-year rate that is typically calculated. This larger time window controls for many 238 delayed entrants. It is important to note that the years sampled were during the Great Recession during precisely the years that many beginning teachers were being laid off, and fewer new teachers were able to get teaching jobs than would normally be the case. Thus, in order to control for economic variation in the entry rate, we use the B&B estimate of 75% as a lower bound and 90% as an upper bound. This allows us to more accurately represent the range of entry rates found in the literature and be responsive to shifts in economic conditions. The next step is to model re-entrants. Re-entrants are often a forgotten segment of Re-entrants. supply that turns out to be a fairly significant piece of the labor market. We model teacher re-entry 239 These authors by applying adapted re-entrant rates by year estimated by Grissom and Reininger. use the National Longitudinal Surveys of Youth 1979 to follow over 200 teachers who leave and 240 re-enter teaching. They find that in all, 38% of teachers who leave the profession return at a later date. Because our forecast period is not long enough to accurately account for teachers re-entering over a 10-year period, like much of the literature, we use a five-year re-entry rate. We know from Grissom and Reininger’s time to re-entry estimates that 85% of re-entry happens in the first five years. Therefore, if 38% of all leavers re-enter at some point, 85% of the 38% of leavers enter within the first five years, which yields a five-year re-entry rate of 32.3%. Next, to adapt the Grissom and Reininger’s estimates of a leaver’s time to re-entry, we apply the existing proportions of re-entry to the new re-entry rate of 32.3%. Consequently, of the 32.3% who return to teaching within five years, 55.9% of them return after just one year, 24.1% after two years, 8.7% after three years, 9.2% after 241 four years, and 2.9% after five years. This paper uses these re-entry rates in combination with the Common Core of Data (CCD), a 242 universal national dataset on teachers and students, National Center for Education Statistics 243 First, we estimates of total attrition per year, and Title II data to model re-entrants over time. 244 This apply the average attrition rate between 2000–12 of 7.88% to the total teacher supply. calculation yields the number of teachers who left teaching each year. Next, because the research 245 we calculate an upper indicates the re-entry rate is dependent on current economic conditions, 76 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

83 246 and the lower bound and lower bound for the re-entry rate. The upper bound rate is 32.3%, 247 based on the range of estimates found in high-quality studies on five-year teacher is 28%, 248 This step produces the number of teachers who will return in each year. Because re-entrance. some teachers do return after five years, and we are using a five-year re-entry rate, we can assume our estimate for teacher supply in some years is a slight underestimate. Next, these re-entering teachers are distributed by year based on Grissom and Reininger’s (2012) estimates. For example, 55.9% of the projected re-entrants from teachers who left in 2008 are modeled as re-entrants in 2010, 24.1% are modeled as re-entrants in 2011, and so on for each 249 additional year. This is repeated for each year’s attrition-based re-entrant projections. Due to data limitations, there is no reliable way to determine the exact number of re-entrants in a given year over time; however, using literature from the field, we can use upper and lower bounds to approximate the number of re-entrants to produce more reliable teacher supply estimates. Finally, to project total supply into the future, this analysis exponentially Projecting Supply. smooths total new entrant estimates and then combines those estimates with the re-entrant estimates to project total teacher supply. Assuming supply holds constant, meaning new entrants and re-entrants remain constant at 2016 levels, these projections illustrate future supply. Constant supply is unlikely given the reactive nature of supply and demand; however, projecting current trends into the future illuminates the extent of the labor market gap if nothing is done to alter current trends. 77 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

84 Table A2 Table A1 Workforce Growth Differences in Demand Estimates: Experience-Based vs. Age-Based Attrition Rates Experience-Based Age-Based Demand Year Workforce Change Year Estimates Demand Estimates 42,273 2005 236,407 2004 236,407 52,078 2006 2005 2007 23,388 2006 33,604 2008 2007 2009 22,159 247,964 247,964 2008 2010 -12,482 2009 2010 -110,577 2011 2011 4,168 2012 172,754 172,754 2012 2013 5,839 245,930 249,632 2013 4,663 2014 259,531 276,063 2014 8,236 2015 269,391 262,032 2015 7,000 2016 259,777 267,465 2016 2017 9,000 2017 268,314 260,459 2018 47,000 299,813 2018 306,743 2019 39,000 299,987 2019 295,433 40,000 2020 299,959 2020 304,200 2021 38,000 2021 302,092 304,314 2022 40,000 2022 307,996 306,366 41,000 2023 311,622 2023 311,860 319,114 2024 319,510 46,000 2024 313,758 2025 316,013 2025 37,000 Note: Data for teachers are expressed in full-time equivalents (FTE). Note: Negative number of teachers represents Bolded values represent actual teacher hires demand measured using a decrease in the total number of teachers. the Schools and Staffing Survey 2004, 2008, and 2012. Source: National Center for Education Source: LPI analysis of the Public School Teacher File, 2004, 2008, Statistics. (2015). Digest of Education and 2012, from the Schools and Staffing Survey, National Center for Statistics. Public and private elementary and Education Statistics; Teacher Follow-Up Survey (TFS), 2005, 2009, secondary teachers, enrollment, pupil-teacher and 2013, from the Schools and Staffing Survey, National Center ratios, and new teacher hires: Selected years, for Education Statistics; State Universe Survey, 2004-2014, from fall 1955 through fall 2024. Washington, DC: the Common Core of Data, National Center of Education Statistics; U.S. Department of Education. National Center for Education Statistics. (2015). Digest of Education Statistics. Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. 78 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

85 79 5 13 Highest 2.42 2.92 2.25 2.75 2.82 2.00 2.73 2.17 3.64 1.91 3.25 1.50 3.17 3.58 3.67 3.67 3.42 3.33 3.42 3.73 4 Rating Teaching Attractiveness 3 12 2 9.1% 9.8% 9.8% 9.3% 5.9% 8.9% 17.9% 17.6% 12.7% 12.8% 12.4% 10.7% 10.5% 11.0% 15.3% 11.5% 14.1% 15.1% 14.8% 28.6% (2014) 1 Lowest... Quintile % of Teachers Inexperienced 11 Not % of Teacher Qualifications 2.10% 2.08% 0.65% 0.66% 4.33% 1.29% 0.87% 0.60% 0.88% 0.01% 0.55% 1.49% 1.10% 1.45% 1.18% 4.20% 3.58% 5.04% 17.84% 11.33% (2014) Teachers Certified 10 13 13 14 14 17.5 17.5 12.6 22.8 19.8 11.9 15.3 15.8 15.3 15.9 16.2 24.3 15.8 15.2 14.2 16.6 Pupil- Ratio (2014) Teacher 9 61% 81% 71% 66% 83% 58% 81% 79% 72% 76% 77% 54% 76% 78% 73% 71% 81% 77% 83% 71% (2012) Control in Their Who Feel They Have Classroom % of Teachers 8 † % of Staff 27.9% 37.9% 37.2% 39.7% 39.3% 39.2% 39.6% 39.2% 40.8% 36.8% 43.4% 35.9% 43.3% 38.5% 43.2% 36.1% 38.0% 38.1% 38.1% (2012) Teachers Who Report Cooperation 7 Working Conditions A COMING CRISIS IN TEACHING? | † 6% 7% 8% 9% 9% 8% 9% 3% 7% 21% 11% 11% 25% 15% 15% 20% 15% 26% 15% % of (2012) Worried Testing Security Teachers about Job Because of 6 54% 47% 49% 55% 51% 55% 52% 57% 48% 46% 37% 24% 45% 44% 45% 58% 57% 50% 46% 39% (2012) by Their Who Feel Supported % of Teachers Administrator 5 LEARNING POLICY INSTITUTE and 9.6% 17.3% 12.7% 10.3% 21.4% 20.5% 10.6% 10.9% 15.8% 15.1% 14.1% 13.2% 15.4% 14.6% 23.1% 13.7% 13.8% 23.6% 13.4% 16.8% (2013) Teacher (Movers Turnover Leavers) 4 † † † † † † 7.0% 9.9% 9.3% 5.5% 5.3% 4.6% 6.6% 4.6% 6.1% 6.3% 6.8% 8.2% 14.8% 18.8% (2013) Attrition Teacher (Leavers) 3 Teacher Turnover † 7.6% 7.0% 7.7% 7.0% 7.9% % of 2.9% 9.3% 5.4% 4.2% 5.3% 4.4% 4.1% 4.2% 4.7% 6.0% 8.9% 8.4% 11.9% 14.8% (2012) to Leave Planning Teachers Possible as Soon as 81 75 71 68 77 73 72 75 68 71 70 68 75 73 74 85 71 85 70 62 (Teachers Wage 2 (2012) Ratio Competitiveness to Non-teachers) 1 Compensation (2013) Salary $37,166 $42,924 $32,126 $32,691 $39,338 $31,835 $41,027 $35,166 $31,159 $35,166 $31,874 $33,664 $44,166 $34,696 $33,386 $33,226 $38,655 $36,198 Average Starting $41,259* $51,539* State Maine Louisiana Hawaii Kentucky Georgia Idaho Florida Delaware Illinois California Connecticut Colorado Kansas District of Columbia Alabama Alaska Indiana Arizona Arkansas Iowa cases, sample sizes are relatively small. We do not report data for states where the samples are too small to meet NCES guidelines for reporting. 2013, or 2014. Some states may have recently experienced changes in policies or conditions that would change the statistic reported if it were collected today. In addition, in some representing the most recent data available for analysis. Interpretations of the data should keep in mind that, depending on the specific statistic, these sources are from 2012, attractiveness rating,” indicating how supportive it appears to be of teacher recruitment and retention. The data are drawn from national data sources (listed in the footnotes), This table highlights a number of key factors that reflect and influence teacher supply and attrition, and signal whether states are likely to have an adequate supply of qualified teachers to fill their classrooms. Based on these data—which treat compensation, teacher turnover, working conditions, and qualifications—each state is assigned a “teaching Appendix B: State Indicators Influencing Supply and Demand

86 80 — 2.18 2.42 2.75 2.75 2.18 2.00 2.50 2.58 2.50 2.67 2.27 2.18 3.40 4.00 3.08 3.17 3.82 3.00 3.09 3.33 3.92 3.00 3.33 4.09 3.82 3.27 3.50 3.45 3.55 3.42 3.55 13% 7.5% 7.9% 7.0% 9.4% 9.8% 6.9% 6.8% 8.8% 12.1% 12.7% 12.9% 10.5% 12.6% 11.6% 10.0% 15.1% 14.3% 11.4% 10.4% 11.5% 10.4% 13.2% 15.7% 13.8% 11.3% 11.5% 14.4% 13.2% 11.2% 14.1% 13.8% 1.9% 7.68% 2.39% 2.27% 0.19% 2.09% 1.07% 0.49% 0.94% 0.26% 0.48% 1.72% 0.86% 0.69% 0.54% 1.53% 0.90% 3.16% 1.70% 0.40% 0.19% 3.45% 0.33% 1.55% 3.10% 0.53% 1.42% 0.26% 0.20% 1.48% 3.52% 1.14% 23 14 12 12.3 22.2 19.3 12.6 15.1 10.6 14.8 15.5 14.1 16.1 15.6 15.1 15.4 13.6 15.3 11.8 13.8 14.5 20.6 13.8 14.5 15.4 16.3 16.2 14.1 18.1 15.3 13.2 13.7 82% 79% 77% 59% 69% 78% 76% 86% 71% 83% 69% 63% 67% 78% 75% 80% 88% 78% 86% 82% 76% 80% 89% 88% 65% 80% 69% 76% 68% 73% 73% 82% 38% 37.7% 37.9% 37.6% 39.5% 42.6% 40.8% 40.9% 40.7% 26.9% 36.8% 31.6% 43.9% 43.3% 44.0% 44.2% 36.5% 34.5% 44.2% 30.9% 34.1% 30.4% 38.2% 46.2% 38.3% 34.7% 36.0% 38.0% 34.1% 38.0% 38.2% 38.0% A COMING CRISIS IN TEACHING? | † 9% 7% 6% 8% 5% 2% 6% 6% 9% 8% 2% 9% 7% 6% 15% 12% 15% 18% 22% 23% 12% 11% 14% 11% 15% 13% 13% 16% 14% 15% 20% 41% 51% 48% 41% 45% 42% 44% 51% 55% 56% 44% 46% 56% 45% 51% 53% 56% 53% 48% 47% 53% 45% 51% 48% 46% 44% 53% 47% 48% 40% 49% 48% LEARNING POLICY INSTITUTE 7.4% 9.9% 9.3% 9.7% 9.7% 9.2% 8.5% 8.7% 17.3% 17.4% 17.9% 17.4% 12.7% 12.5% 12.9% 10.5% 11.9% 20.7% 19.8% 14.2% 13.4% 11.9% 16.2% 16.0% 10.4% 13.2% 14.0% 11.1% 14.6% 14.6% 18.7% 23.2% † † † † † † † † † † † † † † † † 7.7% 7.2% 3.0% 5.9% 5.6% 4.1% 4.5% 5.5% 8.0% 8.2% 8.8% 10.5% 10.2% 10.0% 13.9% 14.9% † † † 7.8% 7.1% 7.8% 7.7% 7.2% 2.8% 9.2% 3.7% 3.5% 5.6% 6.2% 6.6% 5.8% 5.2% 6.5% 5.3% 5.6% 6.8% 4.0% 4.4% 4.0% 8.9% 4.2% 5.7% 4.4% 10.9% 10.7% 10.2% 11.5% 76 94 75 69 71 78 68 73 69 66 71 75 72 68 74 75 78 75 67 80 63 77 70 69 77 67 82 81 73 76 78 74‡ $27,274 $37,848 $32,306 $29,851 $32,533 $32,019 $39,196 $40,600 $35,901 $30,064 $34,505 $31,184 $35,541 $33,546 $43,269 $43,235 $31,606 $41,901 $34,098 $36,141 $30,844 $30,778 $33,081 $33,549 $35,358 $31,960 $33,096 $38,091 $36,335 $34,280 $43,839 $48,631 Massachusetts Wisconsin Wyoming United States Maryland Minnesota South Dakota Michigan South Carolina Tennessee Texas Mississippi Missouri Montana Rhode Island Utah Vermont Oklahoma Pennsylvania Oregon Virginia Ohio Washington West Virginia Nebraska North Carolina North Dakota Nevada New Hampshire New Mexico New York New Jersey

87 81 A COMING CRISIS IN TEACHING? | . Table 5. Princeton, NJ: Educational Testing Service. pp. 15. LEARNING POLICY INSTITUTE Public elementary and secondary teachers, enrollment, and pupil-teacher ratios, by state or jurisdiction: Selected years, fall 2000 through fall 2013. Washington Quintile 1st Quintile 5th Quintile 4th Quintile 3rd Quintile Digest of education statistics. 2nd Quintile Does not meet reporting standards KEY Mind the gap: 20 years of progress and retrenchment in school funding and achievement gaps † Color Teaching attractiveness ratings are calculated by assigning point values for each indicator according to the quintile: 5 points for quintile 5, 4 points for quintile 4, and so on. Teacher turnover indicators, teacher qualification indicators, and pupil- An inexperienced teacher is defined as a teacher in his or her first or second year of teaching. LPI Analysis of Civil Rights Data Collection, Public-Use Data File 2013-14, National Center for Education Statistics. The Office of Civil Rights defines certified teachers as those who have “met all applicable state teacher certification requirements for a standard certificate” for a beginning teacher or one who has completed the state-required probationary National Center for Education Statistics. (2015). Note: % of Teachers Planning to Leave as Soon as Possible, Teacher Attrition, the least desirable response. Teachers Inexperienced are reverse coded such that the 1st quintile is always Teacher Turnover, Pupil-Teacher Ratio, % of Teachers Not Certified, and % of NEA Collective Bargaining/Member Advocacy’s Teacher Salary Database, based on affiliate reporting as of December 2013; see www.nea.org/home/2012-2013-average-starting-teacher-salary.html. Percent of teachers who strongly agree that there is a great deal of cooperative effort among the staff members. Data are from the 2011-12 school year. LPI analysis of Public School Teacher File, 2012, from the Schools and Staffing Survey, Teacher autonomy in the classroom is measured using a Cronbach Alpha-generated construct of classroom control derived from six components: Control over textbooks and materials, content and skills to be taught, teaching techniques, Percent of teachers who strongly agree that their school administration’s behavior toward the staff is supportive and encouraging. Data are from the 2011-12 school year. LPI analysis of Public School Teacher File, 2012, from the Schools and Percent of teachers who plan to leave as soon as possible or until a more desirable job opportunity. Data are from the 2011-12 school year. LPI analysis of Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center LPI analysis of Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. Interpret with caution—all estimates other than Texas and Ohio have a coefficient of variation (CV) between The competitiveness wage index is calculated by dividing the predicted annual wage of elementary and secondary teachers by the predicted wage of non-teachers working in the same state with master’s degrees at both age 25 and 45. Baker, LPI analysis of Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics; Schools and Staffing Survey. (2013). Teacher status file 2012-13. Washington, DC: National Center for Percent of teachers who strongly agree that they worry about the security of their job because of the performance of their students or school on state and/or local tests. Data are from the 2011-12 school year. LPI analysis of Public School DC: U.S. Department of Education. https://nces.ed.gov/programs/digest/ (accessed 5/18/16). Public-Use Data File 2013-14, National Center for Education Statistics. period. “A teacher who is working toward certification by way of alternative routes, or a teacher with an emergency, temporary, or provisional credential is not considered to have met state requirements.” LPI Analysis of Civil Rights Data Collection, teacher ratio are reverse coded such that the 1st quintile is always the least desirable response. Each state’s point total was then divided by the number of available indicators to generate an average teaching attractiveness score for each state. This rating represents the average quintile rank for each state. Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. Interpret estimates from Alaska, Connecticut, North Dakota, and Vermont with caution—each estimate’s coefficient of variation (CV) is between Education Statistics. B., Farrie, D., & Sciarra, D.G. (2016). 30% and 45%. 12 7 5 13 † Reporting standards not met. Either there are too few cases for a reliable estimate or the coefficient of variation (CV) is greater than 50%. 6 * NEA salary data are from 2011-12. 9 4 1 11 3 8 ‡ Average of the 51 states (including Washington, DC) calculations, rather than a separate average for the United States as a whole. 2 10 for Education Statistics.; Interpret estimates from District of Columbia, New Hampshire, and New Mexico with caution—each estimate’s coefficient of variation (CV) is between 30% and 35%. 30% and 50%. evaluating students, discipline, and homework. The Cronbach Alpha value was equal to 0.76. LPI analysis of Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. Staffing Survey, National Center for Education Statistics. National Center for Education Statistics.

88 82 1 5 Highest 4 3 3 4 4 3 3 2 3 4 — 3 3 3 3 2 3 2 2 3 3 2.7 2.3 2.3 2.3 2.8 3.7 3.3 3.2 3.3 3.5 4.7 3.3 3 2.2 2.5 2.5 2.5 2.3 2.3 2.2 2.7 1.8 1.3 3.7 3.7 4.2 3.8 3.5 3.7 3.2 3.5 4.5 3.7 2 Teacher Equity Rating 1 Lowest... Quintile 1.4 1.7 1.9 2.68 1.8 1.1 0.93 1.76 1.35 1.09 1.56 1.67 1.25 1.47 1.82 1.49 1.63 0.91 1.28 3.08 1.72 1.61 1.06 1.99 2.71 2.32 2.38 2.38 2.72 2.09 2.22 1.45 1.27 1.74 1.79 0.86 1.86 1.64 1.28 0.91 1.08 1.26 1.13 1.61 1.18 1.29 1.98 1.54 1.72 1.27 1.07 3.75 (2014) Ratio of the % Minority Schools in High-Minority to Low- Inexperienced Teachers 9.9% 7.02% 7.59% 9.94% 9.48% 9.27% 5.43% 4.62% 8.80% 4.85% 8.99% 8.19% 7.26% 9.56% 9.10% 9.41% 9.69% 9.06% 9.35% 9.91% 12.02% 6.53% 10.72% 10.45% 3.11% 8.77% 6.78% 8.49% 8.36% 11.29% 11.51% 8.82% 10.75% 13.31% 8.64% 8.71% 8.76% 11.13% 13.17% 13.26% 12.59% 12.91% 12.39% 12.42% 29.20% 20.33% 10.90% 10.74% 10.29% 13.72% 15.58% 18.09% (2014) Schools* Percent of Teachers in Low-Minority Inexperienced A COMING CRISIS IN TEACHING? | 7.43% 9.47% 9.63% 16.6% 9.66% 12.44% 9.35% 12.40% 12.55% 11.71% 15.46% 15.04% 8.46% 16.31% 16.62% 16.72% 13.90% 11.78% 21.30% 16.86% 16.78% 16.82% 14.39% 18.80% 18.53% 17.47% 17.27% 17.00% 22.48% 12.76% 12.37% 19.70% 12.04% 22.89% 19.47% 10.80% 21.67% 15.02% 25.54% 15.18% 15.45% 16.61% 15.82% 25.61% 14.64% 14.60% 25.39% 14.22% 36.67% 13.88% 16.82% 18.88% (2014) Teachers in High-Minority Schools* Percent of Inexperienced 1.7 4.1 4.0 2.73 2.48 0.3 1.6 0.79 10.4 0.74 1.45 0.92 0.79 0.13 6.44 4.31 4.72 6.41 5.49 4.36 N/A 7.87 2.64 2.74 2.41 2.56 2.59 0.86 12.98 1.21 1.11 1.98 0.55 1.37 1.48 1.28 1.82 1.14 1.36 1.23 4.57 6.12 6.38 4.61 5.91 10.72 3.29 23.08 88.99 11.97 13.97 13.54 (2014) LEARNING POLICY INSTITUTE Ratio of the % Minority Schools High-Minority to Low- Uncertified Teachers in 0.9% 2.845% 2.355% 0.172% 0.294% 0.297% 0.036% 0.170% 0.319% 0.099% 0.638% 0.676% 1.079% 1.440% 0.308% 0.776% 0.784% 0.527% 0.245% 0.960% 1.277% 4.642% 0.806% 0.246% 0.785% 0.452% 0.378% 0.859% 0.446% 0.620% 0.000% 0.711% 0.685% 0.649% 0.501% 0.335% 0.472% 0.522% 0.082% 1.383% 1.384% 1.867% 1.346% 1.343% 1.492% 3.026% 4.544% 4.086% 3.964% 3.964% 8.408% (2014) 20.686% Certified in Quintile 1st Quintile 5th Quintile 4th Quintile 3rd Quintile 2nd Quintile Low-Minority Schools* Percent of Teachers Not KEY Undefined (not able to divide by zero) 3.6% 7.043% 2.285% 2.088% 2.748% 2.897% 0.811% 0.855% 0.497% 1.898% 1.789% 1.823% 0.722% 1.971% 0.169% 3.207% 4.138% 1.003% 3.661% 4.009% 3.664% 4.180% 7.225% 2.070% 2.067% 2.355% 2.051% 2.383% 2.175% 1.569% 0.184% 0.884% 0.040% 0.538% 0.713% 1.538% 4.938% 1.820% 1.179% 1.222% 1.116% 3.671% 5.281% 5.030% 5.422% 5.995% 3.336% 4.045% 4.423% 22.884% 20.964% 11.749% High-Minority Not Certified in Schools* (2014) Percent of Teachers N/A Color State reverse coded such that the 1st quintile is always the least desirable response. Minority Schools, and Percent of Inexperienced Teachers in Low-Minority Schools are Note: Percent of Teachers Not Certified in High-Minority Schools, Percent of Teachers Not Certified in Low-Minority Schools, Percent of Inexperienced Teachers in High- North Carolina North Dakota Ohio Rhode Island South Carolina Pennsylvania Oregon Oklahoma New York New Mexico New Jersey South Dakota Tennessee Utah Texas Vermont United States Wyoming West Virginia Wisconsin Washington Virginia Teacher Equity ratings are calculated by assigning point values for each indicator according to the quintile: 5 points for quintile 5, 4 points for quintile 4, and so on. (Note: Percent of Teachers Not Certified in High-Minority Schools, Percent of Arizona Alaska Alabama District of Columbia Arkansas California Colorado Delaware Connecticut New Hampshire Nevada Nebraska Iowa Kansas Hawaii Georgia Florida Montana Missouri Mississippi Minnesota Michigan Maryland Maine Massachusetts Kentucky Idaho Indiana Illinois Louisiana 1 Teachers Not Certified in Low-Minority Schools, Percent of Inexperienced Teachers in High-Minority Schools, and Percent of Inexperienced Teachers in Low-Minority Schools are reverse coded such that the 1st quintile is always the least desirable response.) Each state’s point total was then divided by the number of indicators to generate an average Teacher Equity score for each state. This rating represents the average quintile rank for each state. Source: LPI Analysis of the Civil Rights Data Collection, Public-Use Data File 2013-14, National Center for Education Statistics. period. “A teacher who is working toward certification by way of alternative routes, or a teacher with an emergency, temporary, or provisional credential is not considered to have met state requirements.” Note: The Office of Civil Rights defines certified teachers as those who have “met all applicable state teacher certification requirements for a standard certificate” for a beginning teacher or one who has completed the state-required probationary * “High-minority schools” are schools in the top quartile of minority enrollment in each state. “Low-minority schools” are those in the bottom quartile of minority enrollment in each state. Drawing on data from the Office of Civil Rights, this table identifies the extent to which uncertified or inexperienced teachers are hired within states and the extent to which they assigned uncertified or inexperienced teachers. are disproportionately assigned to students of color. Each state is assigned a “teacher equity rating,” indicating the extent to which students, in particular students of color, are Appendix C: Distribution of Uncertified and Inexperienced Teachers by State

89 83 3 4 3 4 3 3 2 3 4 — 2.7 2.3 2.3 2.3 2.8 3.3 3.7 3.2 3.3 3.5 4.7 3.3 1.4 1.7 1.9 2.68 0.93 1.76 1.35 1.09 1.56 1.67 1.47 1.25 1.82 1.49 1.63 0.91 1.28 3.08 1.72 1.06 1.61 1.99 9.9% 7.02% 7.59% 9.94% 9.48% 9.27% 5.43% 4.62% 8.80% 4.85% 8.99% 8.19% 12.02% 10.72% 10.45% 11.29% 11.51% 10.75% 13.31% 11.13% 13.17% 13.26% A COMING CRISIS IN TEACHING? | 7.43% 9.47% 9.63% 16.6% 12.44% 12.40% 12.55% 11.71% 15.46% 15.04% 16.31% 16.62% 16.72% 13.90% 11.78% 21.30% 16.86% 16.78% 16.82% 14.39% 18.80% 18.53% 1.7 4.1 4.0 2.73 2.48 0.79 10.4 0.74 1.45 0.92 0.13 0.79 6.44 4.31 4.72 6.41 5.49 4.36 12.98 10.72 23.08 88.99 LEARNING POLICY INSTITUTE 0.9% 2.845% 2.355% 0.172% 0.294% 0.297% 0.036% 0.170% 0.319% 0.099% 0.638% 0.676% 1.079% 1.440% 0.308% 0.776% 0.784% 0.527% 0.245% 0.960% 1.277% 4.642% Quintile 1st Quintile 4th Quintile 5th Quintile 3rd Quintile 2nd Quintile KEY Undefined (not able to divide by zero) 3.6% 7.043% 2.088% 2.285% 2.748% 2.897% 0.811% 0.855% 0.497% 1.898% 1.789% 1.823% 0.722% 1.971% 0.169% 3.207% 4.138% 1.003% 4.009% 3.661% 3.664% 4.180% N/A Color North Dakota Ohio North Carolina reverse coded such that the 1st quintile is always the least desirable response. Note: Percent of Teachers Not Certified in High-Minority Schools, Percent of Teachers Minority Schools, and Percent of Inexperienced Teachers in Low-Minority Schools are Not Certified in Low-Minority Schools, Percent of Inexperienced Teachers in High- Rhode Island South Carolina Pennsylvania Oregon Oklahoma New York New Mexico New Jersey South Dakota Tennessee Texas Utah Vermont United States Wyoming West Virginia Wisconsin Washington Virginia Teacher Equity ratings are calculated by assigning point values for each indicator according to the quintile: 5 points for quintile 5, 4 points for quintile 4, and so on. (Note: Percent of Teachers Not Certified in High-Minority Schools, Percent of 1 response.) Each state’s point total was then divided by the number of indicators to generate an average Teacher Equity score for each state. This rating represents the average quintile rank for each state. Teachers Not Certified in Low-Minority Schools, Percent of Inexperienced Teachers in High-Minority Schools, and Percent of Inexperienced Teachers in Low-Minority Schools are reverse coded such that the 1st quintile is always the least desirable period. “A teacher who is working toward certification by way of alternative routes, or a teacher with an emergency, temporary, or provisional credential is not considered to have met state requirements.” Note: The Office of Civil Rights defines certified teachers as those who have “met all applicable state teacher certification requirements for a standard certificate” for a beginning teacher or one who has completed the state-required probationary Source: LPI Analysis of the Civil Rights Data Collection, Public-Use Data File 2013-14, National Center for Education Statistics. * “High-minority schools” are schools in the top quartile of minority enrollment in each state. “Low-minority schools” are those in the bottom quartile of minority enrollment in each state.

90 Endnotes 1. Whitaker, I. (2015, November 27). Nevada needs teachers, and it’s shelling out $5 million to get them. The . http:lasvegassun.comnews2015nov27nevada-needs-teachers-and-its-shelling-out-5-milli. Las Vegas Sun Graham, K. A. (2015). First marking period in Philly ends with many teacher shortages. The Inquirer, Philly. 2. com . http://articles.philly.com/2015-11-20/news/68416094_1_vacancies-arlene-kempin-school-year. Markovich, M. (2015, November 30). Principals say state teachers shortage now a crisis. SF Gate . http:// 3. www.sfgate.com/local/komo/article/Principals-say-state-teacher-shortage-now-a-crisis-6666518.php. 4. Nix, N. (2015, December 1). Why Oklahoma is racing to put nearly 1,000 uncertified teachers in its classrooms. The Seventy Four. https://www.the74million.org/article/why-oklahoma-is-racing-to-put -nearly-1000-uncertified-teachers-in-its-classrooms. 5. We conducted a search of LexisNexis Academic (a newspaper archive database) for all articles in the United States containing “teacher shortage” between June 22 and November 22, 2015. The search was conducted May 12, 2016, and 332 articles were retrieved. We conducted a search of LexisNexis Academic (a newspaper archive database) for all articles in the 6. United States containing “teacher shortage” between June 22 and November 22, 2012. The search was conducted May 12, 2016, and 24 articles were retrieved. 7. Hale, L. (2015). Behind the shortage of special ed teachers: Long hours, crushing paperwork. NPR . http://www.npr.org/sections/ed/2015/11/09/436588372/behind-the-shortage-of-special-ed-teachers -long-hours-crushing-paperwork. 8. Brown, E. (2015, December 4). High-poverty schools often staffed by rotating cast of substitutes. The Washington Post . https://www.washingtonpost.com/local/education/how-can-students-learn- without-teachers-high-poverty-schools-often-staffed-by-rotating-cast-of-substitutes/2015/12/04/ be41579a-92c6-11e5-b5e4-279b4501e8a6_story.html. 9. In California alone at least 10,000 pink slips were issued according to the California Teachers Association in press releases from April 7, 2008, through March 15, 2012, http://www.cta.org/ About-CTA/News-Room/Press-Releases/2008/04/20080407-1.aspx ; http://www.cta.org/About- CTA/ News-Room/Press-Releases/2009/03/20090313-1.aspx ; http://www.cta.org/en/About-CTA/ News- Room/Press-Releases/2010/03/20100311_1.aspx ; http://www.cta.org/About-CTA/News-Room/ ; Press-Releases/2011/03/20110315_1.aspx http://www.cta.org/About-CTA/News-Room/Press- Releases/2012/03/20120315_1.aspx (all accessed 10/6/15). 10. Hussar, W. J., & Bailey, T. M. (2014). Projections of education statistics to 2022 . Washington, DC: National Center for Education Statistics. https://nces.ed.gov/pubs2014/2014051.pdf. 11. LPI analysis of the Title II Data Collection, 2009–2014, U.S. Department of Education. https://title2. ed.gov/Public/Home.aspx (accessed 9/3/15). 12. Gantert, T. (2015, January 19). Looming “teacher shortage” appears largely mythical. Capcon . http://www. michigancapitolconfidential.com/21831. 13. Barshay, J. (2016, April 25). Cries about national teacher shortages might be overblown. The Hechinger Report . http://hechingerreport.org/national_teacher_shortages_overblown/. 14. American Association for Employment in Education (AAEE) surveyed 144 school districts across the country. Teacher shortage areas 15. U.S. Department of Education Office of Postsecondary Education. (2015). nationwide listings 1990 – 1991 through 2015 – 16 . Washington, DC: U.S. Department of Education. 16. American Association for Employment in Education. (2015). Educator supply and demand report 2014 – 15 . Slippery Rock, PA: American Association for Employment in Education. 17. Darling-Hammond, L., Furger, R., Shields, P. M., & Sutcher, L. (2016). Addressing California’s emerging teacher shortage: An analysis of sources and solutions . Palo Alto, CA: Learning Policy Institute. 18. U.S. Department of Education Office of Postsecondary Education. (2015). Teacher shortage areas . Washington, DC: U.S. Department of Education. 16 – – 1991 through 2015 nationwide listings 1990 84 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

91 19. Ingersoll, R. M., & Perda, D. (2010). Is the supply of mathematics and science teachers sufficient? , 47 (3), 563–594. American Educational Research Journal 20. Ingersoll, R. M., & May, H. (2010). The magnitude, destinations, and determinants of mathematics and science teacher turnover. Educational Evaluation and Policy Analysis , 34 (4), 1–64. 21. American Association for Employment in Education. (2015). Educator supply and demand report 2014 – 15 . Slippery Rock, PA: American Association for Employment in Education. 22. U.S. Department of Education Office of Postsecondary Education. (2015). Teacher shortage areas 16 nationwide listings 1990 – 1991 through 2015 – . Washington, DC: U.S. Department of Education. 23. American Association for Employment in Education. (2015). – 15 . Educator supply and demand report 2014 Slippery Rock, PA: American Association for Employment in Education. 24. Darling-Hammond, L., Furger, R., Shields, P. M., & Sutcher, L. (2016). Addressing California’s emerging teacher shortage: An analysis of sources and solutions . Palo Alto, CA: Learning Policy Institute. 25. Educator Recruitment & Retention Task Force. (2015). Educator retention and recruitment report . Phoenix, AZ: Arizona Department of Education. 26. Berg-Jacobson, A., & Levin, J. (2015). Oklahoma study of educator supply and demand trends and projections . Washington, DC: American Institutes for Research. 27. Nix, N. (2015, December 1). Why Oklahoma is racing to put nearly 1,000 uncertified teachers in its classrooms. The Seventy Four . https://www.the74million.org/article/why-oklahoma-is-racing-to -put-nearly-1000-uncertified-teachers-in-its-classrooms. 28. Levin, J., Berg-Jacobson, A., Atchison, D., Lee, K., & Vontsolos, E. (2015). Massachusetts study of teacher . Washington, DC: American Institutes for Research. supply and demand trends and projections Kappan 29. See, for example, Ingersoll, R. M., & May, H. (2011). The minority teacher shortage: Fact or fable? Magazine 93 (1), 62–65.; Simon, N., Johnson, S.M., & Reinhorn, S.K. (2015). The challenge of recruiting , and hiring teachers of color: Lessons from six high-performing, high-poverty, urban schools . Working Paper, The Project on the Next Generation of Teachers, Harvard Graduate School of Education. http://robobees. seas.harvard.edu/files/gse-projectngt/files/the_challenge_of_recruiting_and_hiring_teachers_of_color_ diversity_july_2015.pdf 30. LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics.; National Center for Education Statistics. (2015). Digest of education statistics. Enrollment and percentage distribution of enrollment in public elementary and secondary schools, by race/ethnicity and region: Selected years, fall 1995 through fall 2024. Washington, DC: U.S. Department of Education. http://nces.ed.gov/programs/digest/d15/tables/dt15_203.50.asp, (accessed 6/27/16). 31. Ingersoll, R. M., & May, H. (2011). Recruitment, retention and the minority teacher shortage . Philadelphia, PA: Consortium for Policy Research in Education.; Achinstein, B., Ogawa, R. T., Sexton, D., & Freitas, C. (2010). Retaining teachers of color: A pressing problem and a potential strategy for “hard-to-staff” schools. Review of Educational Research (1), 71–107. 80 , 32. Villegas, A. M., & Irvine, J. J. (2010). Diversifying the teaching force: An examination of major arguments. Urban Review , 42 (3), 175–192. 33. Darling-Hammond, L. (2010). The flat world and education: How America’s commitment to equity will determine our future . New York, NY: Teachers College Press. 34. Adamson, F., & Darling-Hammond, L. (2012). Funding disparities and the inequitable distribution of teachers: Evaluating sources and solutions. Education Policy Analysis Archives , 20 (7), 1–46.; Darling- Hammond, L. (2010). The flat world and education: How America’s commitment to equity will determine our . New York, NY: Teachers College Press. future The flat world and education: How America’s commitment to equity will 35. Darling-Hammond, L. (2010). determine our future . New York, NY: Teachers College Press.; Shields, P. M., Humphrey, D. C., Wechsler, Teaching and M. E., Riehl, L. M., Tiffany-Morales, J., Woodworth, K., Young, V. M., & Price, T. (2001). California’s future: The status of the teaching profession 2001 . Santa Cruz, CA: The Center for the Future of Teaching and Learning. 85 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

92 36. Darling-Hammond, L. (2010). The flat world and education: How America’s commitment to equity will . New York, NY: Teachers College Press. p. 40–43. determine our future 37. LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 38. “Highly qualified” is the designation under the No Child Left Behind Act indicating that a teacher holds a bachelor’s degree from a four-year institution, is fully certified or licensed in the state he or she is a teacher, and demonstrates competency in each core academic subject area in which he or she teaches. Teacher inexperience is defined as a teacher in his or her first, second, or third year of teaching. LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 39. U.S. Department of Education. (2015). Equitable Access to Excellent Educators. http://www2.ed.gov/ programs/titleiparta/resources.html (accessed 12/9/15). 40. Kini, T., & Podolsky, A. (2016). Palo Alto, CA: Does teaching experience increase teacher effectiveness? Learning Policy Institute. 41. U.S. Department of Education. (2015). Equitable Access to Excellent Educators. http://www2.ed.gov/ programs/titleiparta/resources.html (accessed 12/9/15). 42. National Center for Education Statistics. (2014). Digest of education statistics . Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. https://nces.ed.gov/ programs/digest/ (accessed 5/18/16). 43. LPI analysis of the Public School Teacher File, 2004, 2008, and 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 44. Averaged public school teacher hires the last four years (2012–13 to 2015–16) to calculate the typical number of current teacher hires. National Center for Education Statistics. (2015). Digest of education . Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and statistics new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. https://nces.ed.gov/programs/digest/ (accessed 5/18/16). 45. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. p. 36. 46. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. p. 39–40. 47. Hussar, W. J., & Bailey, T. M. (2014). Projections of education statistics to 2022 . Washington, DC: U.S. Department of Education. https://nces.ed.gov/pubs2014/2014051.pdf. These estimates are based on historical data on grade progressions and enrollment data as well as U.S. Census Bureau’s school-age population estimates to project total public school enrollment to 2025. 48. National Center for Education Statistics. (n.d.). https://nces.ed.gov/ Digest of education statistics: 2014. programs/digest/d14/ (accessed, 6/15/16). 49. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. 50. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. p. 38. 51. Loeb, S. & Beteille, T. (2009). Teacher quality and teacher labor markets. In G. Sykes, B. Schneider & D. N. (p. 596–612). Plank (Eds.), Handbook of Education Policy Research . Public elementary and Digest of education statistics 52. National Center for Education Statistics. (2015). secondary teachers, enrollment, and pupil-teacher ratios, by state or jurisdiction: Selected years, fall 2000 through fall 2013. Washington, DC: U.S. Department of Education. https://nces.ed.gov/programs/digest/ (accessed 3/7/16). 86 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

93 53. For an extended explanation of the methodology, see: Hussar, W. J., & Bailey, T. M. (2014). Projections . Washington, DC: U.S. Department of Education. https://nces.ed.gov/ of education statistics to 2022 pubs2014/2014051.pdf. 54. Calculated using 2015 data of total teachers and total student in public schools. Data retrieved from: Digest of education statistics . Public and private National Center for Education Statistics. (2015). elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. https://nces.ed.gov/ programs/digest/ (accessed 4/14/16). 55. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. p. 36. 56. According to data from the Schools and Staffing Survey TFS 2004, 2008, and 2012, retirements on average account for 32% of leavers who report they are currently retired. 57. LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 58. Calculated by taking the total number of FTE teachers in 2011–12 and multiplying it by the 2011–12 attrition rate. 59. For methodological details see Appendix A. 60. These NCES teacher workforce projections were modeled based on pupil-teacher ratios predicted by current economic conditions and future student enrollment estimates. The NCES teacher projections are extended into the future by looking at total student enrollment and the pupil-teacher ratio projected over time. For more, see: Hussar, W. J., & Bailey, T. M. (2014). Projections of education statistics to 2022 . Washington, DC: U.S. Department of Education. https://nces.ed.gov/pubs2014/2014051.pdf. 61. To find out more about each of these data sources for the SASS, see: https://nces.ed.gov/surveys/sass/; for CCD see https://nces.ed.gov/ccd/. 62. The actual number of new hires demanded is calculated using the SASS teacher data file. A new hire urrent year, but was not teaching in a is defined as a teacher who is teaching in a public school in the c public school the previous year. Note: A private school teacher moving to the public sector is a “new hire.” Total teacher counts were converted to full time equivalents (FTE) teacher counts by multiplying total new hires by the ratio of total FTE teachers/total teachers in each 2003 (.9379), 2007 (.9399), and 2011 (.9167). For example, if in 2011 you had 100 total teachers, it is estimated that would be equivalent to 91.67 FTE teachers. 63. Ingersoll, R., Merrill, L., & Stuckey, D. (2014). Seven trends: the transformation of the teaching force, updated April 2014 . CPRE Report (#RR-80). Philadelphia, PA: Consortium for Policy Research in Education, University of Pennsylvania. 64. Kena, G., Hussar W., McFarland J., de Brey C., Musu-Gillette, L., Wang, X., Zhang, J., Rathbun, A., The Wilkinson- Flicker, S., Diliberti M., Barmer, A., Bullock Mann, F., and Dunlop Velez, E. (2016). condition of education 2016 . Washington, DC: U.S. Department of Education, p. 70–71; Chapman, C., Laird, Trends in high school dropout and completion rates in the United States: J., and KewalRamani, A. (2010). . Washington, DC: U.S. Department of Education, p. 10. 1972 – 2008 65. National Center for Education Statistics. (2014). Digest of education statistics . Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. https://nces.ed.gov/ programs/digest/ (accessed 5/18/16). 66. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. 67. LPI analysis of the Public School Teacher File, 2004, 2008, and 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 68. Darling-Hammond, L. (2000). Solving the dilemmas of teacher supply, demand, and standards: How we can ensure a competent, caring, and qualified teacher for every child . Kutztown, PA: National Commission on 87 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

94 Teaching and America’s Future. http://nctaf.org/wp-content/uploads/2012/01/supply-demand-standards. pdf, p. 37-38. 69. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. 70. Some states (CT and NY, for example) have required that teachers achieve a master’s degree in order to receive professional certification; in California, essentially all teaching credentials are issued in postbaccalaureate teacher education programs. In addition, federal incentives for alternative certification have created postbaccalaureate pathways in many states. See Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply policy for education: The right way to meet the “highly qualified teacher” challenge. Education Policy Analysis Archives , 11 (33), 1–55. ; U.S. Department of Education Office of Postsecondary Education. (2013). Preparing and credentialing America’s teachers: The secretary’s ninth . Washington, DC: U.S. Department of Education. https://title2.ed.gov/Public/ report on teacher quality TitleIIReport13.pdf. Profile of teachers in the U.S. 2011 . Washington, DC: National Center for 71. Feistritzer, C. E. (2011). Education Information. 72. Loeb, S. & Beteille, T. (2009). Teacher quality and teacher labor markets. In G. Sykes, B. Schneider & D. N. Plank (Eds.), Handbook of Education Policy Research (p. 596–612). S ee also, Dolton, P., Tremayne, A., & Chung, T. P. (2003). The economic cycle and teacher supply . Paris: OCED. 73. Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply policy for education: The 11 (33), right way to meet the “highly qualified teacher” challenge. Education Policy Analysis Archives , 1–55. 74. Ingersoll, R. M. (2011). Do we produce enough mathematics and science teachers? , Kappan Magazine 92 (6), 37–41, p. 40. 75. It is possible some of these teachers end up teaching private scho ol, but the percentage would be very small. 76. Shields, P. M., Esch, C. E., Humphrey, D. C., & Young, V. M. (1999). The supply of and demand for teachers in California public schools . Santa Cruz, CA: Center for the Future of Teaching and Learning. http://images. pearsonassessments.com/images/NES_Publications/2000_05Shields_428_1.pdf. 77. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. p. 30. 78. Recent estimates from the 2011–12 SASS survey find that approximately 49% of teachers entering the teaching workforce are re-entrants. Estimates from 2008–09 SASS survey find that approximately 37% of teachers entering the teaching workforce are re-entrants. A re-entrant is defined as a new hire (someone who was not teaching in public schools the previous year) who has prior experience teaching in public schools. LPI analysis of the Public School Teacher File, 2004, 2008, and 2012, from the Schools and Staffing Survey, National Center for Education Statistics. We note that labor market conditions in 2011–12 were unusual in that year, because many districts were laying off teachers, especially beginners, and hiring very few new teachers. Assessing teacher supply and demand. 79. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Santa Monica, CA: RAND Corporation. p. 32. See also, Dolton, P., Tremayne, A., & Chung, T. P. (2003). The economic cycle and teacher supply . Paris: OCED. 80. LPI analysis of the Public School Teacher File, 2004, 2008, and 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 81. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior of exiting teachers. Education Finance and Policy , 7 (4), 425–454. 82. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior 7 Education Finance and Policy , (4), 425–454. p. 428–429; Murnane, R., Singer, J. D., of exiting teachers. . Cambridge, MA: Harvard Who will teach? Policies that matter Willett, J. B., Kemple, J., & Olsen, R. (2009). University Press. 88 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

95 83. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior of exiting teachers. , 7 (4), 425–454. Education Finance and Policy 84. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior , 7 of exiting teachers. See also , Education Finance and Policy Haggstrom, G. W., Darling- (4), 425–454. Hammond, L., & Grissmer, D. W. (1988). Assessing teacher supply and demand. Santa Monica, CA: RAND Corporation. 85. Grissom and Reininger (2012) found that 38% of teachers who leave the classroom return at some point. However, using their time to re-entry estimates by year, we know within the first five years, 85% of teachers return; therefore, 32.3% of all teachers who left, returned within five years. See: Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior of exiting teachers. Education Finance and Policy , 7 (4), 425–454. 86. Murnane, R., Singer, J. D., Willett, J. B., Kemple, J., & Olsen, R. (2009). Who will teach? Policies that matter . Cambridge, MA: Harvard University Press. 87. LPI analysis of the Title II Data Collection, 2009–2014, U.S. Department of Education. https://title2. ed.gov/Public/Home.aspx (accessed 9/3/15). 88. ACT. (2014). The condition of future educators 2014 . Iowa City, IA: ACT. http://www.act.org/newsroom/ data/2014/states/pdf/FutureEducators.pdf. 89. Cooperative Institutional Research Program, Higher Education Research Institute, U. of California at Los Angeles accessible through: Backgrounds and Beliefs of College Freshmen. (2016). The Chronicle of Higher Education . http://chronicle.com/interactives/freshmen-survey. 90. For more information on the Title II data, see https://title2.ed.gov/Public/Home.aspx 91. This average was calculated using public school attrition estimates available from years 2000–2013. National Center for Education Statistics. (2015). . Mobility of public Digest of education statistics elementary and secondary teachers, by selected teacher and school characteristics: Selected years, 1987–88 through 2012–13. Washington, DC: U.S. Department of Education. https://nces.ed.gov/programs/ digest/ (accessed 11/10/15). 92. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior of Education Finance and Policy , 7 (4), 425–454. exiting teachers. 93. This estimate is calculated by finding the mid-point between the upper and lower bound of new entrants. 94. Darling-Hammond, L., Furger, R., Shields, P. M., & Sutcher, L. (2016). Addressing California’s emerging teacher shortage : An analysis of sources and solutions . Palo Alto, CA: Learning Policy Institute. 95. These calculations are produced using Title II data on completers by subject area. New teacher entrants are calculated by taking completers and applying an 82.5% entry rate (the midpoint between the upper and lower bound estimates of 75% and 90%). However, it is worth noting that there are several issues with the Title II data that produced these subject estimates. The main issues stem from the fact that the subject categories are not mutually exclusive. For example, a student can be counted in the “Secondary education” category in addition to the “Science” category. Furthermore, each state and the Institute of Higher Education categorize their students differently. Some double count and some treat them as mutually exclusive. We have no way to know if the data are underestimating or overestimating the true completers. However, since states started to require subject specific credentials for secondary teachers, we can assume this is less of an issue in recent years. 96. Hale, L. (2015). Behind the shortage of special ed teachers: Long hours, crushing paperwork. NPR . http://www.npr.org/sections/ed/2015/11/09/436588372/behind-the-shortage-of-special-ed-teachers- long-hours-crushing-paperwork; Ingersoll, R. M., & May, H. (2010). The magnitude, destinations, and determinants of mathematics and science teacher turnover. Educational Evaluation and Policy Analysis , 34 (4), 1–64., p. 2. 97. Darling-Hammond, L., Furger, R., Shields, P. M., & Sutcher, L. (2016). Addressing California’s emerging teacher shortage : An analysis of sources and solutions . Palo Alto, CA: Learning Policy Institute. 98. Murnane, R. J., Singer, J. D., & Willett, J. B. (1989). The influences of salaries and “opportunity costs” on , teachers’ career choices: Evidence from North Carolina. Harvard Educational Review (3), 325–347. 59 89 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

96 99. LPI analysis of the Title II Data Collection, 2014, U.S. Department of Education. https://title2.ed.gov/ Public/Home.aspx (accessed 9/3/15). 100. LPI analysis of the Title II Data Collection, 2014, U.S. Department of Education. https://title2.ed.gov/ Public/Home.aspx (accessed 9/3/15). 101. New Jersey and New York issued 0% of their initial teaching credentials to out-of-state prepared individuals. With rounding, 0% could mean there were some initial credentials issued to out-of-state prepared candidates, but this would at most represent less than half a percent of the new initial credentials. Note: This analysis excludes New Mexico, which did not report these data. LPI analysis of the Title II Data Collection, 2014, U.S. Department of Education. https://title2.ed.gov/Public/Home.aspx (accessed 9/3/15). 102. Ingersoll, R. M., & Perda, D. (2010). Is the supply of mathematics and science teachers sufficient? American Educational Research Journal , 47 (3), 563–594. 103. Boyd, D., Lankford, H., Loeb, S., & Wyckoff, J. (2005). The draw of home: How teachers’ preferences for proximity disadvantage urban schools. Journal of Policy Analysis and Management , 24 (1), 113–132. 104. Adamson, F., & Darling-Hammond, L. (2012). Funding disparities and the inequitable distribution of (7), 1–46. teachers: Evaluating sources and solutions. Education Policy Analysis Archives , 20 105. National Education Association. (n.d.). 2012-2013 Average starting teacher salaries by state. http://www. nea.org/home/2012-2013-average-starting-teacher-salary.html (accessed 3/14/16). Speaking of salaries: What it will take to get qualified, effective 106. Adamson, F., & Darling-Hammond, L. (2011). teachers in all communities . Washington, DC: Center for American Progress. 107. One analysis of California teacher salaries found a 2 to 1 ratio of wages for similarly experienced and educated teachers across the state, which actually grew to a 3 to 1 differential when cost of living differentials were taken into account. These differences were closely associated with teacher shortages in Speaking of salaries: What it will take to low-wage districts. Adamson, F., & Darling-Hammond, L. (2011). get qualified, effective teachers in all communities . Washington, DC: Center for American Progress. 108. Loeb, S. & Beteille, T. (2009). Teacher quality and teacher labor markets. In G. Sykes, B. Schneider & D. N. Handbook of Education Policy Research (p. 596–612). Plank (Eds.), 109. It is important to remember there are two moving parts in these ratios—teacher wages and non-teacher wages. So a high-wage competiveness ratio could represent high teacher wages and average non-teacher wages, or average teacher wages and low non-teacher wages, or any combinations in between. 110. National Center for Education Statistics. (2015). Digest of education statistics . Public elementary and secondary teachers, enrollment, and pupil-teacher ratios, by state or jurisdiction: Selected years, fall 2000 through fall 2013. Washington, DC: U.S. Department of Education. https://nces.ed.gov/programs/digest/ (accessed 3/7/16). 111. Teacher autonomy in the classroom is measured using a Cronbach Alpha generated construct of classroom control derived from six components: Control over textbooks and materials, content and skills to be taught, teaching techniques, evaluating students, discipline, and homework. The Cronbach Alpha value was equal to 0.76. 112. 14.8% of teachers in Washington, DC, 11.9% of teachers in Arizona, and 11.5% of teachers in Nevada reported they planned to leave as soon as possible or when a more desirable job opportunity comes along; 1.9% of teachers in Rhode Island, 2.8% of teachers in South Dakota, and 2.9% of teachers in Illinois reported they planned to leave as soon as possible or when a more desirable job opportunity comes along. LPI analyses of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 113. 23.6% of teachers in Arizona, 23.3% of teachers in New Mexico, and 23.1% of teachers in Washington, DC, did not stay in their current school; 7.4% of Rhode Island teachers did not remain teaching in their current school. LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 114. For methodological details, see Appendix A. 90 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

97 115. Darling-Hammond, L., & Rothman, R. (2011). Teacher and leader effectiveness in high-performing education systems . Washington, DC: Alliance for Excellent Education; Stanford, CA: SCOPE. http://eric.ed.gov/ ERICWebPortal/recordDetail?accno=ED517673. 116. Loeb, S., Darling-Hammond, L., & Luczak, J. (2005). How teaching conditions predict teacher turnover. Peabody Journal of Education , 80 (3), 44–70. 117. National Center for Education Statistics. (2015). Digest of education statistics. Public and private elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. Washington, DC: U.S. Department of Education. https://nces.ed.gov/ programs/digest/ (accessed 12/4/15). 118. Darling-Hammond, L., & Rothman, R. (2011). Teacher and Leader effectiveness in high-performing education . Washington, DC: Alliance for Excellent Education; Stanford, CA: SCOPE. http://eric.ed.gov/ systems ERICWebPortal/recordDetail?accno=ED517673. Does teaching experience increase teacher effectiveness? Palo Alto, CA: 119. Kini, T., & Podolsky, A. (2016). Learning Policy Institute.; Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How teacher turnover harms student achievement. American Educational Research Journal , 50 (1), 4–36. 120. Ronfeldt, M., Loeb, S., & Wyckoff, J. (2013). How teacher turnover harms student achievement. American Educational Research Journal , 50 (1), 4–36. 121. Jackson, C. K., & Bruegmann, E. (2009). Teaching students and teaching each other: The importance of (4), 85–108. 1 peer learning for teachers. American Economic Journal: Applied Economics , 122. Kraft, M. A., & Papay, J. P. (2014). Can professional environments in schools promote teacher Educational Evaluation and development? Explaining heterogeneity in returns to teaching experience. Policy Analysis 36 (4), 476–500. , 123. Carroll, S., Reichardt, R., & Guarino, C. M. (2000). The distribution of teachers among California’s school districts and schools. Santa Monica, CA: RAND Corporation. http://www.dtic.mil/dtic/tr/fulltext/u2/ a384997.pdf. 124. Darling-Hammond, L. (2003). Keeping good teachers: Why it matters, what leaders can do. Educational Leadership , 60 (8), 6–13. 125. Carroll, T. G. (2007). Policy brief: The high cost of teacher turnover . Washington, DC: National Commission on Teaching and America’s Future. http://nctaf.org/wp-content/ uploads/NCTAF CostofTeacherTurnoverpolicybrief.pdf\ nhttp://scholar.google.com/scholar?hl=en&btnG=Search &q=intitle:Policy+Brief+The+High+Cost+of+Teacher+Turnover#0. 126. LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 127. LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 128. For example, in 2011–12, 90.3% of leavers were voluntary, while in 2008–09, 94.6% of leavers were voluntary. LPI analysis of the Teacher Follow-Up Survey (TFS), 2009 and 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 129. National Center for Education Statistics. (2015). Digest of education statistics. Mobility of public elementary and secondary teachers, by selected teacher and school characteristics: Selected years, 1987–88 through 2012–13. Washington, DC: U.S. Department of Education. https://nces.ed.gov/programs/ digest/ (accessed 4/15/16). See also , Ingersoll, R. M. (2003). Is there really a teacher shortage . Seattle, WA: Center for the Study of Teaching and Policy. 130. Studies have produced a range of estimates for beginning teacher attrition, all of which have shortcomings. For example, one recent estimate using national longitudinal data put the attrition rate around 17%, finding 83% of beginning teachers still teaching at the end of their fifth year, including some who had left and re-entered. (See Gray, L., Taie, S., & O’Rear, I. (2015). Public school teacher attrition and – 2008 Beginning Teacher mobility in the first five years: Results from the first through fifth waves of the 2007 . Washington, DC: U.S. Department of Education. http://search.ebscohost.com/login.asp Longitudinal Study x?direct=true&db=eric&AN=ED556348&site=ehost-live.) However, the analysis omitted the large number 91 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

98 of individuals who did not respond to the survey at various points during these years without adjusting for nonresponse bias. In general, survey evidence finds that those who do not respond to such surveys are more likely to have left their position than to have continued teaching. For that reason, the 17% figure likely underestimates attrition by an unknown margin. Our own imputations to adjust these data based on the characteristics of nonrespondents suggest that the attrition rate is likely at least 19%. Older estimates of attrition using national cross-sectional data suggested about a 30% attrition rate at the end of five years. See Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply policy Education Policy Analysis for education: The right way to meet the “highly qualified teacher” challenge. , 11 (33), 1–55. Archives 131. Ingersoll, R. M. (2012). Beginning teacher induction: What the data tell us. Kappan , 93 (8), 47–52. 132. See also Carver-Thomas, D. (2016). National trends in teacher attrition: An analysis of 2011 – 12—2012 – 13 stayers, movers, and leavers. Palo Alto, CA: Learning Policy Institute. (forthcoming). 133. Carver-Thomas, D. (2016). National trends in teacher attrition: An analysis of 2011 – 12—2012 – 13 stayers, movers, and leavers. Palo Alto, CA: Learning Policy Institute. (forthcoming). 134. Holding other factors constant: school size, school level (i.e., elementary or secondary), district type (i.e., suburb, rural, town, city), percentage of students eligible for free/reduced-price lunch, percentage students of color, city x percent eligible for FRPL, average class size, teacher gender, teacher age (i.e., younger than 30, older than 50), years of experience, teacher race, main teaching assignment (i.e., mathematics/science, humanities, special education, elementary, arts, foreign languages, physical education, career technical education, miscellaneous), school working conditions (i.e., administrative support, salary satisfaction, student behavior, parental support, school resources, interference on teaching time, collegiality, job security, classroom control, school influence, beginning district salary, highest district salary). – 13 stayers, – – 135. Carver-Thomas, D. (2016). National trends in teacher attrition: An analysis of 2011 2012 12 movers, and leavers. Palo Alto, CA: Learning Policy Institute. (forthcoming). National trends in teacher attrition: An analysis of 2011 – 12 – 2012 – 13 stayers, 136. Carver-Thomas, D. (2016). Palo Alto, CA: Learning Policy Institute. (forthcoming). movers, and leavers. What are the effects of teacher education and preparation on 137. Ingersoll, R. M., Merrill, L., & May, H. (2014). beginning teacher retention? Philadelphia, PA: Consortium for Policy Research in Education. 138. In cities and suburbs, the difference between the northeast and west is not statistically significant. 139. Only in the west, the city attrition rate mirrors that in all other district types (except for western suburbs, which have lower attrition rates). 140. Shields, P. M., Esch, C. E., Humphrey, D. C., & Young, V. M. (1999). The supply of and demand for teachers in California public schools . Santa Cruz, CA: Center for the Future of Teaching and Learning. http://images. pearsonassessments.com/images/NES_Publications/2000_05Shields_428_1.pdf. 141. LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 142. LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 143. For a comprehensive review, see Simon, N. S., & Johnson, S. M. (2015). Teacher turnover in high-poverty schools: What we know and can do. Teachers College Record , 117 (March), 1–36. http://isites.harvard.edu/ fs/docs/icb.topic1231814.files/Teacher Turnover in High-Poverty Schools.pdf. 144. Ingersoll, R. M., & May, H. (2011). Recruitment, retention and the minority teacher shortage . Philadelphia, PA: Consortium for Policy Research in Education.; Achinstein, B., Ogawa, R. T., Sexton, D., & Freitas, C. (2010). Retaining teachers of color: A pressing problem and a potential strategy for “hard-to-staff” (1), 71–107. schools. Review of Educational Research , 80 145. Loeb, S., Darling-Hammond, L., & Luczak, J. (2005). How teaching conditions predict teacher turnover. (3), 44–70. , Peabody Journal of Education 80 . Philadelphia, Recruitment, retention and the minority teacher shortage 146. Ingersoll, R. M., & May, H. (2011). PA: Consortium for Policy Research in Education. p. 45. 92 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

99 147. See also Carver-Thomas, D. (2016). National trends in teacher attrition: An analysis of 2011 12 – 2012 – 13 – Palo Alto, CA: Learning Policy Institute. (forthcoming). stayers, movers, and leavers. 148. Simon, N. S., & Johnson, S. M. (2015). Teacher turnover in high-poverty schools: What we know and can Teachers College Record , 117 (March), 1–36. http://isites.harvard.edu/fs/docs/icb.topic1231814.files/ do. Teacher Turnover in High-Poverty Schools.pdf. 2012 – 12 – 149. See also Carver-Thomas, D. (2016). – 13 National trends in teacher attrition: An analysis of 2011 stayers, movers, and leavers. Palo Alto, CA: Learning Policy Institute. (forthcoming). 150. See also Carver-Thomas, D. (2016). National trends in teacher attrition: An analysis of 2011 – 12 – 2012 – 13 stayers, movers, and leavers. Palo Alto, CA: Learning Policy Institute. (forthcoming). 151. Darling-Hammond, L., LaPointe, M., Meyerson, D., Orr, M. T., & Cohen, C. (2007). Preparing school leaders for a changing world: Lessons from exemplary leadership development programs . Stanford, CA: Stanford University, Stanford Educational Leadership Institute. http://www.wallacefoundation.org/knowledge- center/Documents/Preparing-School-Leaders.pdf. 152. Darling-Hammond, L., Burns, D, Campbell, C., Goodwin, A. L., Hammerness, K., Low, E.L., McIntyre, A., Sato, M., & Zeichner, K. (2017, in press). Empowered educators: Teaching quality around the world . San Francisco, CA: Jossey-Bass. 153. Darling-Hammond, L. (2010). The flat world and education: How America’s commitment to equity will determine our future . New York, NY: Teachers College Press. 154. See, for example, Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply policy for education: The right way to meet the “highly qualified teacher” challenge. Education Policy Analysis , Archives 11 (33), 1–55. 155. For a fuller treatment of the design of a national teacher supply p olicy, see Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply policy for education: The right way to meet the “highly qualified teacher” challenge. 11 (33), 1–55. Education Policy Analysis Archives , 156. For a comprehensive review of influences on teacher recruitment and retention, see Podolsky, A., Kini, Solving the teacher shortage: How to attract and retain T., Bishop, J., & Darling-Hammond, L. (2016). excellent educators . Palo Alto, CA: Learning Policy Institute. (forthcoming). Is school funding fair? A national report card 157. Baker, B., Sciarra, D. G., & Farrie, D. (2015). . Newark, NJ: Education Law Center. http://www.schoolfundingfairness.org/. 158. Darling-Hammond, L., Burns, D, Campbell, C., Goodwin, A. L., Hammerness, K., Low, E.L., McIntyre, A., Sato, M., & Zeichner, K. (2017, in press). Empowered educators: Teaching quality around the world . San Francisco, CA: Jossey-Bass. 159. Boser, U., & Straus, C. (2014). Mid- and late-career teachers struggle with paltry incomes . Washington, DC: Center for American Progress. 160. Adamson, F., & Darling-Hammond, L. (2011). Speaking of salaries: What it will take to get qualified, effective teachers in all communities . Washington, DC: Center for American Progress. p. 7. Speaking of salaries: What it will take to get qualified, effective 161. Adamson, F., & Darling-Hammond, L. (2011). . Washington, DC: Center for American Progress. p. 7.; citing Henke, R. R., teachers in all communities America’s teachers: Profile of a profession, 1993 Choy, S. P., Chen, X., Geis, S., & Alt, M. N. (1997). 94 . – Washington, DC: U.S. Department of Education. papers3://publication/uuid/32604DB9-A4A7-4F02-A21C- 6120E0BCE8DF.; Loeb, S., Darling-Hammond, L., & Luczak, J. (2005). How teaching conditions predict teacher turnover. Peabody Journal of Education , 80 (3), 44–70. 162. Adamson, F., & Darling-Hammond, L. (2011). Speaking of salaries: What it will take to get qualified, effective teachers in all communities . Washington, DC: Center for American Progress. 163. Loeb, S. & Beteille, T. (2009). Teacher quality and teacher labor markets. In G. Sykes, B. Schneider & D. N. Plank (Eds.), Handbook of Education Policy Research (p. 596–612). 164. Adamson, F., & Darling-Hammond, L. (2012). Funding disparities and the inequitable distribution of , Education Policy Analysis Archives (7), 1–46. teachers: Evaluating sources and solutions. 20 93 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

100 165. Gray, L., Taie, S., & O’Rear, I. (2015). Public school teacher attrition and mobility in the first five years: Results – . U.S. Department from the first through fifth waves of the 2007 2008 Beginning Teacher Longitudinal Study of Education. http://search.ebscohost.com/login.aspx?direct=true&db=eric&AN=ED556348&site=eh ost-live. 166. Grissom, J. A., Viano, S. L., & Selin, J. L. (2015). Understanding employee turnover in the public sector: Insights from research on teacher mobility. Public Administration Review, 76.; Loeb, S., Darling-Hammond, Peabody Journal of Education , L., & Luczak, J. (2005). How teaching conditions predict teacher turnover. (3), 44–70.; 80 Stockard, J. & Lehman, M. B. (2004). Influences on the satisfaction and retention of 1st-year Educational Administration Quarterly, teachers: The importance of effective school management. 40(5), 742 – 771. Adamson, F. & Darling-Hammond, L. (2011). Speaking of salaries: What it will take to get qualified, effective teachers in all communities . Washington, DC: Center for American Progress. NewsOK . 167. Eger, A. (2015, September 20). Teachers struggle with low pay, working conditions. 168. Darling-Hammond, L. (2010). The flat world and education: How America’s commitment to equity will determine our future . New York, NY: Teachers College Press. 169. National Education Goals Panel. (1998). The national education goals report: Building a nation of learners, 1998 . Washington, DC: U.S. Government Printing Office.; Baron, J. B. (1999). Exploring high and improving reading achievement in Connecticut. Lessons from the states . Washington, DC: National Education Goals Panel. 170. Darling-Hammond, L. (2010). The flat world and education: How America’s commitment to equity will determine our future . New York, NY: Teachers College Press.; Adamson, F., & Darling-Hammond, L. (2011). Speaking of salaries: What it will take to get qualified, effective teachers in all communities . Washington, DC: Center for American Progress. 171. Clotfelter, C. T., Ladd, H. F., & Vigdor, J. L. (2010). Teacher credentials and student achievement in high school. (3), 655–681. The Journal of Human Resources 45 , 172. Baron, J. B. (1999). Exploring high and improving reading achievement in Connecticut. Lessons from the states . Washington, DC: National Education Goals Panel. p. 28. 173. Texas State Affordable Housing Corporation. (n.d.). Teacher home loans. http://www.tsahc.org/ homebuyers-renters/teacher-home-loans (accessed 7/28/16). 174. Marinell, W. H., Coca, V. M., Arum, R., Goldstein, J., Kemple, J., Pallas, A., Bristol, T., Buckley, C., Scallon, Who stays and who leaves? Findings from a three-part study of teacher turnover A., & Tanner, B. (2013). in NYC middle schools . New York, NY: Research Alliance for New York Schools.; Ingersoll, R., Merrill, What are the effects of teacher education preparation on beginning teacher attrition? L., & May, H. (2014). Philadelphia, PA: Consortium for Policy Research in Education, University of Pennsylvania. 175. Ingersoll, R., Merrill, L., & May, H. (2014). What Are the Effects of Teacher Education Preparation on Beginning Teacher Attrition? Philadelphia, PA: Consortium for Policy Research in Education, University of Pennsylvania. 176. Gray, L., Taie, S., & O’Rear, I. (2015). Public school teacher attrition and mobility in the first five years . Washington, DC: U.S. Department of Education. http://nces.ed.gov/pubs2015/2015337.pdf. 177. Carver-Thomas, D. (2016). National trends in teacher attrition: An analysis of 2011 – 12 – 2012 – 13 stayers, movers, and leavers. Palo Alto, CA: Learning Policy Institute. (forthcoming). 178. Darling-Hammond, L., Holtzman, D. J., Gatlin, S. J., & Heilig, J. V. (2005). Does teacher preparation matter? Evidence about teacher certification, Teach for America, and teacher effectiveness. Education Policy Analysis Archives , 13 (42), 51.; Glazerman, S., Isenberg, E., Dolfin, S., Bleeker, M., Johnson, A., Grider, M., & Jacobus, M. (2010). Impacts of comprehensive teacher induction . Washington, DC: U.S. Department of Education.; Boyd, D., Grossman, P., Lankford, H., Loeb, S., & Wyckoff, J. (2006). How changes in entry requirements alter the teacher workforce and affect student achievement. Education Finance 1 , (2), 176–216.; Clotfelter, C. T., Ladd, H. F., & Vigdor, J. L. (2007). Teacher credentials and and Policy , Economics of Education Review student achievement: Longitudinal analysis with student fixed effects. 26 (6), 673–682. 94 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

101 179. Darling-Hammond, L., Holtzman, D. J., Gatlin, S. J., & Heilig, J. V. (2005). Does teacher preparation Education matter? Evidence about teacher certification, Teach for America, and teacher effectiveness. Policy Analysis Archives , 13 (42), 51.; , D., Grossman, P., Lankford, H., Loeb, S., & Wyckoff, J. (2006). How changes in entry requirements alter the teacher workforce and affect student achievement. Education Finance and Policy , 1 (2), 176–216. 180. Staklis, S., & Henke, R. (2013). Who considers teaching and who teaches? Washington, DC: U.S. Department of Education. 181. Rothstein, J., & Rouse, C. E. (2011). Constrained after college: Student loans and early-career occupational choices. Journal of Public Economics , 95 (1–2), 149–163. 182. Podolsky, A., & Kini, T. (2016). How effective are loan forgiveness and service scholarships for recruiting teachers? Palo Alto, CA: Learning Policy Institute. 183. Podolsky, A., & Kini, T. (2016). How effective are loan forgiveness and service scholarships for recruiting teachers? Palo Alto, CA: Learning Policy Institute. 184. In 2011, the North Carolina legislature voted to eliminate funding for the NC Teaching Fellows Program. Until new appropriations are made, the last class of funded fellows graduated in the spring of 2015. 185. Henry, G. T., Bastian, K. C., & Smith, A. A. (2012). Scholarships to recruit the “best and brightest” into teaching: Who is recruited, where do they teach, how effective are they, and how long do they stay? Educational Researcher , 41 (3), 83–92. 186. Center for Educator Recruitment, Retention, and Advancement (CERRA), South Carolina. (March 2016). Teaching fellows program, data from the 2000 – 2011 cohorts . Rock Hill, SC. 187. Steele, J. L., Murnane, R. J., & Willett, J. B. (2010). Policy brief: Do financial incentives draw promising teachers to low-performing schools? Assessing the impact of the California Governor’s Teaching Fellowship. Stanford, CA: PACE, School of Education, Stanford University. http://www.ncbi.nlm.nih.gov/ pubmed/24578399. 188. Reininger, M. (2012). Hometown disadvantage? It depends on where you’re from: Teachers’ location preferences and the implications for staffing schools. Educational Evaluation and Policy Analysis 34 (2), , 127–145.; Boyd, D., Lankford, H., Loeb, S., & Wyckoff, J. (2005). The draw of home: How teachers’ Journal of Policy Analysis and Management , 24 (1), preferences for proximity disadvantage urban schools. 113–132.; Darling-Hammond, L., & Sykes, G. (2003). Wanted: A national teacher supply policy for education: The right way to meet the “highly qualified teacher” challenge. Education Policy Analysis Archives , 11 (33), 1–55. 189. Warford L. J. (2006). College and career transitions initiative: Responding to a quiet crisis. In L. J. Warford (Ed.) Pathways to student success: Case studies from the College and Careers Transitions Initiative. p. 8. Phoenix, AZ: League for Innovation in the Community College; cited in Bragg, D. D. (2007). Teacher pipelines: Career pathways extending from high school to community college to university. Community College Review 35 (1), 10–29. , 190. Teacher Cadets South Carolina. (n.d.) . The teacher cadet program. http://teachercadets.com/overview. aspx (accessed 12/1/15); What Works SC Clearing House. (2013). The Teacher Cadet Program. The Riley Institute at Furman University: Center for Education Policy and Leadership. http://riley.furman.edu/sites/ default/files/docs/WWSCClearinghouse_tcadet2_4.8.pdf (accessed 12/1/15). 191. Teacher Cadets South Carolina. (n.d.) . The Teacher Cadet Program. http://teachercadets.com/overview. aspx (accessed 12/1/15) 192. Warford L. J. (2006). College and career transitions initiative: Responding to a quiet crisis. In L. J. Warford (Ed.) Pathways to student success: Case studies from the College and Careers Transitions Initiative. p. 8. Phoenix, AZ: League for Innovation in the Community College; cited in Bragg, D. D. (2007). Teacher pipelines: Career pathways extending from high school to community college to university. Community 35 College Review , (1), 10–29. 193. Yee, G. (2014, March 21). Dispatches from OUSD: How do you find great leaders for our schools. Oakland . http://linkedlearning.org/zoomfolio/ dispatches-from-ousd-how-do-we-find-great-leaders-for-our- Local schools-community-voices/. 95 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

102 194. California Teacher Pathway. (n.d.). http://californiateacherpathway.org/ (accessed 10/28/16). 195. See, e.g., Clewell, B. C., & Villegas, A. M. (2001). Absence unexcused: Ending teacher shortages in high- . Evaluating the Pathways to Teaching Careers Program . Washington, DC: Urban Institution. need areas 196. California Commission on Teacher Credentialing. (2008). California school paraprofessional teacher training program . Sacramento, CA: Commission on Teacher Credentialing. 197. Darling-Hammond, L., Guha, R., & Hyler, M. (2016). Teacher residencies: An innovative model for preparing teachers. Palo Alto, CA: Learning Policy Institute. (forthcoming). 198. Title 42, chapter 6A, subchapter V of the U.S. Code details the many components of this system. 199. Darling-Hammond, L., Chung, R., & Frelow, F. (2002). Variation in teacher preparation: How well do Journal of Teacher Education different pathways prepare teachers to teach? 53 (4), 286–302. , 200. Ingersoll, R. M., & Strong, M. (2011). The impact of induction and mentoring programs for beginning Review of Educational Research , 81 (2), 201–233.; Headden, S. teachers: A critical review of the research. (2014). Beginners in the classroom . Stanford, CA: Carnegie Foundation for the Advancement of Teaching.; Henke, R. R., Chen, X., Geis, S., & Knepper, P. (2000). Progress through the teacher pipeline: 1992 – 93 College graduates and elementary / secondary school teaching as of 1997 . Washington, DC: U.S. Department of Education. ; California Commission on Teacher Credentialing. (2015). Report on New Teacher Induction . Sacramento, CA: California Commission on Teacher Credentialing. 201. Ingersoll, R. M., & Strong, M. (2011). The impact of induction and mentoring programs for beginning teachers: A critical review of the research. Review of Educational Research , 81 (2), 201–233.; Headden, S. (2014). . Stanford, CA: Carnegie Foundation for the Advancement of Beginners in the classroom Teaching.; Henke, R. R., Chen, X., Geis, S., & Knepper, P. (2000). Progress Through the Teacher Pipeline: 1992 – 93 College Graduates and Elementary / Secondary School Teaching as of 1997 . Washington, DC: U.S. Department of Education. ; California Commission on Teacher Credentialing. (2015). Report on New Teacher Induction . Sacramento, CA: California Commission on Teacher Credentialing. See also, The impact of mentoring on teacher retention : What the research Ingersoll, R., & Kralik, J. M. (2004). ECS Research Review . A . Denver, CO: Education Commission.; Chang, M., & Brown, R. S. (1992). says two-year evaluation of the peer support pilot project: 1990 – . Toronto, Ontario, Canada: Toronto 1992 Board of Education, Research Department.; Odell, S. J., & Ferraro, D. P. (1992). Teacher mentoring and teacher retention. , 43 (3), 200–204.; Spuhler, L., & Zetler, A. (1995). Montana Journal of Teacher Education beginning teacher support program. Final Report . Helena, MT: Montana State Board of Education.; Fuller, Beginning teacher retention rates for TxBESS and non-TxBESS teachers , Unpublished white paper. E. (2003). Austin, TX: State Board for Educator Certification. 202. Bartell, C. A. (1995). Shaping teacher induction policy in California. Teacher Education Quarterly , (Fall), 27–43.; Olebe, M. (2001). A decade of policy support for California’s new teachers: The beginning teacher support and assessment program. Teacher Education Quarterly , 28 (1), 71–84.; Smith, T. M., & Ingersoll, R. M. (2004). What are the effects of induction and mentoring on beginning teacher turnover? American Educational Research Journal , (3), 681–714.; Wang, J., Odell, S. J., & Schwille, S. A. (2008). Effects of 41 Journal of Teacher teacher induction on beginning teachers’ teaching: A critical review of the literature. Education 59 (2), 132–152. , 203. Ingersoll, R. M., & Smith, T. M. (2004). Do teacher induction and mentoring matter? NASSP Bulletin , 88 (638), 28–40. 204. Ingersoll, R. M., & Strong, M. (2011). The impact of induction and mentoring programs for beginning (2), 201–233. Review of Educational Research , 81 teachers: A critical review of the research. 205. Glazerman, S., Isenberg, E., Dolfin, S., Bleeker, M., Johnson, A., Grider, M., & Jacobus, M. (2010). Impacts of comprehensive teacher induction . Washington, DC: U.S. Department of Education. On finding significant differences in student achievement for teachers who had participated in two years of induction versus See also, the control group. Ingersoll, R. M., & Strong, M. (2011). The Impact of induction and mentoring , Review of Educational Research programs for beginning teachers: A critical review of the research. (2), 201–233. 81 206. LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 96 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

103 207. LPI analysis of the Public School Teacher File, 2012, from the Schools and Staffing Survey, National Center for Education Statistics. 208. About 250,000 teachers are hired each year, but typically only 40–60% of them are new to teaching. The others are experienced teachers changing schools or returning teachers who are re-entering the labor force. 209. Ingersoll, R. M. (2001). Teacher turnover and teacher shortages: An organizational analysis. American Educational Research Journal 38 (3), 499–534.; Loeb, S., Darling-Hammond, L., & Luczak, J. (2005). How , , 80 (3), 44–70. teaching conditions predict teacher turnover. Peabody Journal of Education , Star Tribune 210. Ragatz, K. (March 26, 2014). Counterpoint: Why do teachers leave the toughest schools? http://www.startribune.com/counterpoint-why-do-teachers-leave-the-toughest-schools/252548431/ (accessed 12/1/15). 211. Rochkind, J., Ott, A., Immerwahr, J., Doble, J., & Johnson, J. (2007). Lessons learned: New teachers talk . Washington, DC: National Comprehensive Center for about their jobs, challenges and long-range plans Teacher Quality and New York, NY: Public Agenda. http://www.publicagenda.org/files/lessons_learned_3. pdf (accessed 11/10/15). 212. Clotfelter, C. T., Glennie, E., Ladd, H., & Vigdor, J. (2008). Would higher salaries keep teachers in high- poverty schools? Evidence from a policy intervention in North Carolina. , Journal of Public Economics 92 (5-6), 1352–1370. 213. Glazerman, S., Protik, A., Teh, B., Bruch, J., & Max, J. (2013). Transfer incentives for high- performing teachers: Final results from a multisite randomized experiment . Washington, DC: U.S. Department of Education. Children of poverty deserve great teachers: One union’s commitment to changing the status 214. Berry, B. (2009). quo . Carrboro, NC: Center for Teaching Quality and Washington, DC: National Education Association. http://www.nea.org/assets/docs/HE/children_of_poverty_09.pdf (accessed 11/4/15). 215. Elmore, R., Thomas, D., Clayton, T.C. (2006). Differentiated treatment in Montgomery County public schools. Cambridge, MA: Public Education Leadership Project, Harvard University. http://pelp.fas.harvard. edu/files/hbs-test/files/pel028p2.pdf 216. Humphrey, D. C., Koppich, J. E., & Hough, H. J. (2005). Sharing the wealth: National Board Certified Teachers and the students who need them most. Education Policy Analysis Archives , 13 (18), 1–50. 217. Clotfelter, C. T., Ladd, H. F., Vigdor, J. L., & Diaz, R. A. (2004). Do school accountability systems make it more difficult for low-performing schools to attract and retain high-quality teachers? Journal of Policy , (2), 251–271.; Tye, B. B., & O’Brien, L. (2002). Why are experienced teachers Analysis and Management 23 th leaving the profession? Phi Delta Kappan , 84 (1), 24–32.; DeVise, D. (November 5 , 1999). A+ plan prompts teacher exodus in Broward County. Miami Herald What might school accountability do? .; Figlio, D. (2001). NBER Working Paper Series. 218. LPI analysis of the Teacher Follow-Up Survey (TFS), 2013, from the Schools and Staffing Survey, National Center for Education Statistics. 219. Grissom, J. A. (2011). Can good principals keep teachers in disadvantaged schools? Linking principal effectiveness to teacher satisfaction and turnover in hard-to-staff environments. Teachers College Record , 113 (11), 2552–2585. See also, Boyd, D., Grossman, P., Ing, M., Lankford, H., Loeb, S., & Wyckoff, J. (2010). The influence of school administrators on teacher retention decisions. American Educational Research Journal , 48 (2), 303–333.; Horng, E. L., Kalogrides, D., & Loeb, S. (2010). Principal preferences and the uneven distribution of principals across schools. Educational Evaluation and Policy Analysis , 32 (2), 205–229. 220. Goodpaster, K. P. S., Adedokun, O. A., & Weaver, G. C. (2012). Teachers’ perceptions of rural STEM teaching: Implications for rural teacher retention. Rural Educator , 33 , 9–22. http://search.proquest.com/ docview/1312420898?accountid=12339\nhttp://mcgill.on.worldcat.org/atoztitles/link?sid=ProQ:&issn =0273446X&volume=33&issue=3&title=Rural+Educator&spage=9&date=2012-01-01&atitle=Teacher s’+Perceptions+of+Rural+STEM+Teaching:+Implic; Boyd, D., Grossman, P., Ing, M., Lankford, H., Loeb, S., & Wyckoff, J. (2010). The influence of school administrators on teacher retention decisions. American Educational Research Journal (2), 303–333.; Marinell, W. H., Coca, V. M., Arum, R., Goldstein, J., Kemple, 48 , 97 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

104 J., Pallas, A., Bristol, T., Buckley, C., Scallon, A., & Tanner, B. (2013). Who stays and who leaves? Findings from a three-part study of teacher turnover in NYC middle schools . New York, NY: Research Alliance for New York Schools. 221. Grissom, J. A. (2011). Can good principals keep teachers in disadvantaged schools? Linking principal effectiveness to teacher satisfaction and turnover in hard-to-staff environments. Teachers College Record , 113 (11), 2552–2585.; Horng, E. L., Kalogrides, D., & Loeb, S. (2010). Principal preferences and the uneven distribution of principals across schools. Educational Evaluation and Policy Analysis , 32 (2), 205–229. 222. Simon, N. S., & Johnson, S. M. (2015). Teacher turnover in high-poverty schools: What we know and can do. , Teachers College Record 117 , 1–36. http://isites.harvard.edu/fs/docs/icb.topic1231814.files/Teacher Turnover in High-Poverty Schools.pdf. 223. Simon, N. S., & Johnson, S. M. (2015). Teacher turnover in high-poverty schools: What we know and can do. Teachers College Record , 117 , 1–36. http://isites.harvard.edu/fs/docs/icb.topic1231814.files/Teacher Turnover in High-Poverty Schools.pdf. 224. Darling-Hammond, L., LaPointe, M., Meyerson, D., Orr, M. T., & Cohen, C. (2007). Preparing school leaders for a changing world: Lessons from exemplary leadership development programs . Stanford, CA: Stanford University, Stanford Educational Leadership Institute. http://www.wallacefoundation.org/knowledge- center/Documents/Preparing-School-Leaders.pdf. 225. Whitaker, K. S. (2003). Principal role changes and influence on principal recruitment and selection: An , 41 (1), 37–54.; Darling-Hammond, L., & international perspective. Journal of Educational Administration . Stanford, CA: Stanford University, Institute for Orphanos, S. (2007). Leadership development in California Research on Education Policy & Practice. https://edpolicy.stanford.edu/sites/default/files/publications/ leadership-development-california.pdf.; Tyre, P. (2015). Why do more than half of principals quit after five The Hechinger Report , http://hechingerreport.org/why-do-more-than-half-of-principals-quit-after- years? five-years/.; Cooper, B. S., & Conley, S. eds. (2011). Keeping and improving today’s school leaders: Retaining and sustaining the best . Lanham, MD: Rowan & Littlefield Education. 226. Ingersoll, R. M., & Perda, D. (2010). Is the supply of mathematics and science teachers sufficient? American Educational Research Journal , 47 (3), 563–594. 227. Darling-Hammond, L. (2010). Evaluating teacher effectiveness: How teacher performance assessments can measure and improve teaching. Washington DC: Center for American Progress. 228. For more information on the Title II data, see https://title2.ed.gov/Public/Home.aspx Projections of education statistics to 2022 229. Hussar, W. J., & Bailey, T. M. (2014). . Washington, DC: US Department of Education Digest of education statistics . Public and private 230. National Center for Education Statistics. (2015). elementary and secondary teachers, enrollment, pupil-teacher ratios, and new teacher hires: Selected years, fall 1955 through fall 2024. https://nces.ed.gov/programs/digest/ (accessed 12/9/15). 231. Ingersoll, R. M., Merrill, L., & Stuckey, D. (2014). Seven trends: The transformation of the teaching force . Philadelphia, PA: Consortium for Policy Research in Education. http://www.cpre.org/7trends. 232. Note: Total teacher counts are in FTE, but the attrition rate/continuation rates were calculated using the SASS which has weights calibrated to total teachers. Because FTE still represents a combination of full- and part-time teachers, this should not have an effect on the estimates (other than the figures that represent FTE counts rather than total teacher counts). 233. Esch, C. E., Chang-Ross, C. M., Guha, R., Humphrey, D. C., Shields, P. M., Tiffany-Morales, J. D., Wechsler, M. E., & Woodworth, K. R. (2005). The status of the teaching profession 2005 . Santa Cruz, CA: The Center for the Future of Teaching and Learning. http://files.eric.ed.gov/fulltext/ED491141.pdf. 234. Using the Title II data from academic years 2009–2014, the ratio was evaluated by dividing the number for the 50 states and District of Columbia t-2 by the number of enrollments in year t of completers in year over a three-year period. These estimated completer rates are relatively stable across time as well as over varying lags. 235. Title II data go as far back as the academic year of 1999–00, but the early years of data collection were incomplete. For this reason, this report uses only completer totals as far back as 2004–05. Using this 98 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

105 10-year range, there were only three missing values: Montana in 2004 and 2005, and Delaware in 2008. These were imputed using the median completers for these states in every other year of available data for Montana (788) and Delaware (739). 236. For more information of the B&B 2008:2012, see https://nces.ed.gov/surveys/b&b/ 237. By looking at respondents’ experiences with teaching as of 2012, we can see 24.5% of the B&B sample prepared to teach. At the same time, 6.13% of the entire sample prepared to teach but never ended up in the classroom as of 2012. Therefore, 25.01% (6.13 divided by 24.5) of teachers who prepared to teach, prepared but had not entered the classroom as of 2012. 238. This benefit is a double-edged sword: it allows us to largely account for delayed entrants in the teacher supply but doesn’t properly indicate which specific years these delayed entrants surface in the teacher labor force. Delayed entrants are not distributed over each year, which is partially captured in the four- year average completer to teacher rate. Therefore, each year’s estimate of supply is a slight overestimate because future delayed entrants are captured in the year directly after they prepare rather than later in time. However, if in year 1, we overestimate supply because we are counting delayed entrants who haven’t yet entered, in year 2 we underestimate supply because the delayed entrants that might have entered in year 2 were already accounted for in year 1. It is difficult to know whether these over- and under-estimates each year cancel each other out or if there is bias in one direction or another. Although there may be some biases year to year, we believe that including imperfect measurements of delayed entrants paints a more reliable long-term picture than not accounting for them at all. 239. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior of 7 (4), 425–454. exiting teachers. Education Finance and Policy , 240. The authors note that the sample of 200 teacher career paths is slightly younger than the true age distribution of teachers’ potential adding upward bias to their estimates. 241. Grissom and Reininger’s (2012) original estimates are as follow: Of the 38% who return to teaching, 47.5% of them return just after one year, 20.5% after two years, 7.4% after three, 7.8% after four years, and 2.5% after five years. To adjust for the fact we are now using a five-year re-entrant rate, we scale the yearly percentages by (100/85). The new adjusted percentages yield the same results as the original estimates because the proportion remains the same. 242. For more information on the CCD, see: https://nces.ed.gov/ccd/ 243. Re-entrants are able to be modeled only to 2013–14 because the CCD last year of total teacher estimates is 2013–14. As previously mentioned, the Title II data has missing values for years prior to 2004. Using the range between 2000–12, there were only 22 missing values from nine states. These missing values were imputed using the median completers for these states in every other year of available data for Alabama (1,270), Idaho (1,235), Iowa (2,649), Nebraska (1,722), Wisconsin (4,239), Montana (788), North Dakota (668), Washington (2,857), and Delaware (739). 244. This average was calculated using public school attrition estimates available from years 2000 to 2013. National Center for Education Statistics. (2015). Digest of education statistics. Mobility of public elementary and secondary teachers, by selected teacher and school characteristics: Selected years, 1987–88 through 2012–13. Washington, DC: U.S. Department of Education. https://nces.ed.gov/programs/ digest/ (accessed 11/10/15). Assessing teacher supply and demand. 245. Haggstrom, G. W., Darling-Hammond, L., & Grissmer, D. W. (1988). Santa Monica, CA: RAND Corporation. p. 32. See also, Dolton, P., Tremayne, A., & Chung, T. P. (2003). The economic cycle and teacher supply . Paris: OCED. 246. Grissom, J. A., & Reininger, M. (2012). Who comes back? A longitudinal analysis of the reentry behavior of , (4), 425–454. 7 exiting teachers. Education Finance and Policy 247. Murnane, R., Singer, J. D., Willett, J. B., Kemple, J., & Olsen, R. (2009). Who will teach? Policies that matter . Cambridge, MA: Harvard University Press. 248. It is worth noting both these re-entry rates are calculated using data from the ‘90s and early 2000s. Because labor market conditions are constantly changing, we must acknowledge the possibility that shifts in economic conditions could alter the re-entry figures. 249. This is done only for the first five years. Beyond five years, the additional teachers added is smaller than the margin of error on our estimates. 99 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

106 About the Authors Leib Sutcher is a Research Associate on LPI’s Educator Quality Team. His current work focuses on teacher labor market issues, including recruitment and retention, the inequitable distribution of unqualified and inexperienced educators, and teacher shortages. He is a co-author of the Institute’s California teacher supply and demand report. is President of the Learning Policy Institute and Charles E. Ducommun Linda Darling-Hammond Professor of Education Emeritus at Stanford University. She has conducted extensive research on issues of educator supply, demand, and quality. Among her award-winning publications in this area are What Matters Most: Teaching for America’s Future; Teaching as the Learning Profession; Powerful Teacher Education ; and Preparing Teachers for a Changing World: What Teachers Should Learn and Be . Able to Do is a Research and Policy Associate on LPI’s Educator Quality Team. She is Desiree Carver-Thomas the lead author of a forthcoming study on teacher attrition. Previously, she taught in New York City public schools for five years, and consulted on strategies for diverting recidivism and implementing a full-service community schools initiative. Carver-Thomas recently received her M.P.P. degree from the Goldman School of Public Policy at UC Berkeley. 100 A COMING CRISIS IN TEACHING? | LEARNING POLICY INSTITUTE

107 1530 Page Mill Road, Suite 200 Palo Alto, CA 94304 p: 650.332.9797 1301 Connecticut Avenue, Suite 500 Washington, DC 20036 p: 202.830.0079 www.learningpolicyinstitute.org The Learning Policy Institute conducts and communicates independent, high-quality research to improve education policy and practice. Working with policymakers, researchers, educators, community groups, and others, the Institute seeks to advance evidence-based policies that support empowering and equitable learning for each and every child. Nonprofit and nonpartisan, the Institute connects policymakers and stakeholders at the local, state, and federal levels with the evidence, ideas, and actions needed to strengthen the education system from preschool through college and career readiness.

Related documents