You are here
We evaluate the feasibility of estimating test-score growth with a gap year in testing data, informing the scenario when state testing resumes after the 2020 COVID-19 induced test stoppage. Our research design is to simulate a gap year in testing using pre-COVID-19 data—when a true test gap did not occur—which facilitates comparisons of district- and school-level growth measures that are estimated with and without a gap year. We find that growth estimates based on the full data and gap-year data are generally similar, and our results highlight an advantage of using comprehensive growth models with rich controls for student and schooling circumstances. With the caveat that there is looming uncertainty about which students will be tested in public schools when testing resumes, and how they will be tested, our findings establish the potential for estimating useful growth measures with a gap year in testing.
Citation: Ishtiaque Fazlul, Cory Koedel, Eric Parsons, Cheng Qian (2021). Estimating Test-Score Growth with a Gap Year in the Data. CALDER Working Paper No. 248-0121
The decision about how and when to open schools to in-person instruction has been a key question for policymakers throughout the COVID-19 pandemic. The instructional modality of schools has implications not only for the health and safety of students and staff, but also student learning and the degree to which parents can engage in job activities. We consider the role of instructional modality (inperson, hybrid, or remote instruction) in disease spread among the wider community. Using a variety of regression modeling strategies to address unobserved heterogeneity, we find that simple correlations show in-person modalities are correlated with increased COVID cases, but accounting for both pre-existing cases and a richer set of covariates brings estimates close to zero on average. In Ordinary Least Squares (OLS) specifications, in-person modality options are not associated with increased spread of COVID at low levels of pre-existing COVID cases, but cases do increase at moderate to high pre-existing COVID rates. A bounding exercise suggests that the OLS findings for in-person modality are likely to represent an upper bound on the true relationship. These findings are robust to the inclusion of county and district fixed effects in terms of the insignificance of the findings, but the models with fixed effects are also somewhat imprecisely estimated.
Working Paper No. 247-1220 was originally released in December 2020 and has since been updated to Working Paper No. 247-1220-2, released in January 2021.
Citation: Dan Goldhaber, Scott A. Imberman, Katharine Strunk, Bryant Hopkins, Nate Brown, Erica Harbatkin, Tara Kilbride (2020). To What Extent Does In-Person Schooling Contribute to the Spread of COVID-19? Evidence from Michigan and Washington. CALDER Working Paper No. 247-1220-2
How much do teachers value compensation that is deferred until retirement? This question is important because the vast majority of public school teachers are covered by defined benefit (DB) pension plans that “backload” a large share of compensation to retirement relative to the compensation structure in the private sector. There is little evidence, other than Fitzpatrick (2015), however, about whether DB pensions are consistent with teacher preferences for current and deferred compensation. This study examines a unique setting in Washington State, where teachers enroll in a hybrid, DB-Defined Contribution (DC) pension system, and have choices over their DC contribute rate. These choices reveal preferences about the value teachers place on current versus retirement compensation. We find that teachers choose to contribute an average of 8.18 percent, significantly more than the minimum required contribution of 5 percent. This suggests that teachers value retirement compensation significantly more than previously estimated by Fitzpatrick. Potential explanations for the difference in findings from prior evidence are discussed, including estimation strategies, differences in state settings, overall plan generosity, and the potential for teachers to view DB and DC pensions as different products.
Citation: Dan Goldhaber, Kristian Holden (2020). How Much do Teachers Value Deferred Compensation? Evidence from Defined Contribution Rate Choices. CALDER Working Paper No. 242-0920
Prior work has documented a substantial penalty associated with taking the Partnership for Assessment of Readiness for College and Careers (PARCC) online relative to on paper (Backes & Cowan, 2019). However, this penalty does not necessarily make online tests less useful. For example, it could be the case that computer literacy skills are correlated with students’ future ability to navigate high school coursework, and thus more predictive of later outcomes. Using a statewide implementation of PARCC in Massachusetts, we test the relative predictive validity of online and paper tests. We are unable to detect a difference between the two and in most cases can rule out even modest differences. Finally, we estimate mode effects for the new Massachusetts statewide assessment. In contrast to the first years of PARCC implementation, we find very small mode effects, showing that it is possible to implement online assessments at scale without large online penalties.
Citation: Benjamin Backes, James Cowan (2020). Is Online a Better Baseline? Comparing the Predictive Validity of Computer- and Paper-Based Tests. CALDER Working Paper No. 241-0820
Many states enhanced benefits in teacher retirement plans during the 1990s. This paper examines the school staffing effects of one such enhancement in a major urban school district with mostly high poverty schools. Pension rule changes in 1999 for St. Louis public school teachers resulted in very large increases in pension wealth for active teachers, as well as a powerful increase in “push” incentives for earlier retirement. Simple descriptive statistics on retirement patterns before and after the enhancements suggest much earlier retirement resulted. Shorter teaching spells imply a steady state with more teaching vacancies and a larger share of novice teachers in classrooms. To better understand the long run effects of these changes and alternatives policies, the authors estimate a structural model of teacher retirement. Simulations of retirement behavior for a representative senior teacher point to shorter completed teaching spells and earlier retirement age as a result of the enhancements. By contrast, moving from the post-1999 to a DC- type plan would extend the teaching career of a representative senior teacher by roughly three years. Simulations of voluntary DC conversation plans suggest that many senior teachers would enroll, thereby reducing workforce turnover, and overall pension costs.
Citation: Shawn Ni, Michael Podgursky, Xiqian Wang (2020). Teacher Pension Enhancements and Staffing in an Urban School District. CALDER Working Paper No. 240-0620
The clinical teaching experience is one of the most important components of teacher preparation. Prior observational research has found that more effective mentors and schools with better professional climates are associated with better preparation for teacher candidates. We test these findings using an experimental assignment of teacher candidates to placement sites in two states. Candidates who were randomly assigned to higher quality placement sites experienced larger improvements in performance over the course of the clinical experience, as evaluated by university instructors. The findings suggest that improving clinical placement procedures can improve the teaching quality of candidates.
Citation: Dan Goldhaber, Matt Ronfeldt, James Cowan, Trevor Gratz, Emanuele Bardelli, Matt Truwit, Hannah Mullman (2020). Room for Improvement? Mentor Teachers and the Evolution of Teacher Preservice Clinical Evaluations. CALDER Working Paper No. 239-0620
Defined benefit (DB) pension systems determine the size of pension payments using an employee’s “final average salary”. Thus, employees enrolled in DB pension systems face an incentive to “salary spike” – strategically increase late career pensionable compensation – to increase their retirement income. This is an important issue given that public pension systems face increasing scrutiny due to ongoing concerns about their fiscal sustainability. This paper develops an empirical method to quantify the prevalence of salary spiking by identifying cases where end-of-career compensation deviates from expected levels of compensation. We apply this method to the teacher pension systems in Illinois and examine how salary spiking changed in response to policy reform. The results suggest that salary spiking is very common, with about half of late career employees observed as having pensionable compensation that exceeds expectations. Policies designed to dissuade salary spiking by internalizing its costs across districts appear to reduce the prevalence of salary spiking, but there may be unintended consequences for individuals who are not actually spiking, as such discouraging the assignment of supplementary responsibilities to late career employees.
Citation: Dan Goldhaber, Cyrus Grout, Kristian Holden (2020). A Method for Identifying Salary Spiking: An Assessment of Pensionable Compensation and Reform in Illinois. CALDER Working Paper No. 238-0620
There is growing interest in using measures of teacher applicant quality to improve hiring decisions, but the statistical properties of such measures are poorly understood. We present evidence on structured ratings solicited from teacher applicants’ references. We find that the reference ratings capture only one underlying dimension of applicant quality, which may indicate a need to broaden the range of questions posed to professional references. Point estimates of inter-rater reliability range between 0.23 and 0.31 and are significantly lower for novice applicants. It is difficult to judge whether these levels of reliability are high or low in the current context given so little evidence on comparable applicant assessment tools.
Citation: Dan Goldhaber, Cyrus Grout, Malcolm Wolff, Patricia Martinkova (2020). Evidence on the Dimensionality and Reliability of Professional References’ Ratings of Teacher Applicants. CALDER Working Paper No. 237-0620
We use a novel database of student teaching placements in Washington State to investigate teachers’ transitions from student teaching classrooms to first job classrooms and the implications for student achievement. We find that first-year teachers are more effective when they are teaching in the same grade, in the same school level, or in a classroom with student demographics similar to their student teaching classroom. We also document that only 27% of first-year teachers are teaching the same grade they student taught, and that first-year teachers tend to begin their careers in higher-poverty classrooms than their student teaching placements. This suggests that better aligning student teacher placements with first-year teacher hiring could be a policy lever for improving early-career teacher effectiveness.
Citation: John Krieg, Dan Goldhaber, Roddy Theobald (2020). Disconnected Development? The Importance of Specific Human Capital in the Transition from Student Teaching to the Classroom. CALDER Working Paper No. 236-0520
Testing students and using test information to hold schools and, in some cases, teachers accountable for student achievement has arguably been the primary national strategy for school improvement over the past decade and a half. Tests are also intended to be used as a diagnostic tool to identify individual student needs, so that students can be set on a trajectory for long-term academic success. We use panel data from three states – North Carolina, Massachusetts and Washington State – to investigate how accurate early measures of achievement are in predicting later high school outcomes. We contribute to the literature in three distinct ways. First, the long panels we employ allow us to quantify the accuracy of models predicting how early (3rd and 4th grade) measures of student background and achievement predict several later schooling outcomes: 8th grade test achievement, high school course-taking, and high school graduation. Second, we test the extent to which predictions based on distinct segments of student data (e.g. grades 3 to 8, then 8 to 12) sacrifice forecast accuracy; this is of particular policy relevance for states or localities that do not yet have long administrative data panels. Finally, we test the degree to which the use of parameter estimates from models predicting schooling outcomes derived from one state diminish the accuracy of predicting outcomes in other states.
Citation: Dan Goldhaber, Malcolm Wolff, Timothy Daly (2020). Assessing the Accuracy of Elementary School Test Scores as Predictors of Students’ High School Outcomes. CALDER Working Paper No. 235-0520
The Community Eligibility Provision (CEP) is a policy change to the federally-administered National School Lunch Program that allows schools serving low-income populations to classify all students as eligible for free meals, regardless of individual circumstances. This has implications for the use of free and reduced-price meal (FRM) data to proxy for student disadvantage in education research and policy applications, which is a common practice. We document empirically how the CEP has affected the value of FRM eligibility as a proxy for student disadvantage. At the individual student level, we show that there is essentially no effect of the CEP. However, the CEP does meaningfully change the information conveyed by the share of FRM-eligible students in a school. It is this latter measure that is most relevant for policy uses of FRM data.
Note: Portions of this paper were previously circulated under the title “Using Free Meal and Direct Certification Data to Proxy for Student Disadvantage in the Era of the Community Eligibility Provision.” We have since split the original paper into two parts. This is the first part.
WP 234-0320 was originally released in March 2020. This is the third update, WP 234-0320-3, which was released in September 2020.
Citation: Cory Koedel, Eric Parsons (2020). The Effect of the Community Eligibility Provision on the Ability of Free and Reduced-Price Meal Data to Identify Disadvantaged Students. CALDER Working Paper No. 234-0320-3
This study examines the effects of internal migration driven by severe natural disasters on host communities, and the mechanisms behind these effects, using the large influx of migrants into Florida public schools after Hurricane Maria. I find adverse effects of the influx in the first year on existing student test scores, disciplinary problems, and student mobility among high-performing students in middle and high school that also persist in the second year. I also find evidence that compensatory resource allocation within schools is an important factor driving the adverse effects of large, unexpected migrant flows on incumbent students in the short-run.
WP 233-0320 was originally released in March 2020. This updated version, WP 233-0320-2, was released in January 2021.
Citation: Umut Özek (2020). Examining the Educational Spillover Effects of Severe Natural Disasters: The Case of Hurricane Maria . CALDER Working Paper No. 233-0320-2
This study adds to the currently limited evidence base on the efficacy of interventions targeting non-college-ready high school students by examining the impact of Kentucky’s Targeted Interventions (TI) program. We focus on interventions that students received under TI in the senior year of high school based on their 11th grade ACT test scores. Using difference-in-regression discontinuity and difference-in-difference designs with seven cohorts of 11th grade students, we find that, for an average per-student cost of about $600, TI significantly reduces the likelihood that students enroll in remedial course in both 2- and 4-year postsecondary institutions by 5–10 percentage points in math and 3–4 percentage points in English. These effects are similar among students who are eligible for free-or reduced-price lunch, Black and Hispanic students, students with remediation needs in multiple subjects, and students in lower-performing schools. Evidence also shows that TI increases the likelihood that students enroll in and pass college math before the end of the first year by four percentage points in 4-year universities. However, little evidence exists for TI affecting credit accumulation or persistence.
Citation: Zeyu Xu, Benjamin Backes, Amanda Oliveira, Dan Goldhaber (2020). Targeted Interventions in High School: Preparing Students for College. CALDER Working Paper No. 232-0220
We use data on the teacher preparation experiences and workforce outcomes of more than 1,300 graduates of special education teacher education programs in Washington to provide a descriptive portrait of special education teacher preparation, workforce entry, and early career retention. We find high rates of workforce entry for special education candidates (over 80%), but we document considerably lower rates of entry into special education classrooms for candidates who hold a dual endorsement in special education and another subject. We also find that special education teachers who are dual endorsed and begin their careers teaching in special education classrooms are less likely stay in these classrooms. Both sets of findings are supported by an instrumental variable analysis that exploits passing score cutoffs on required licensure tests to provide plausibly causal evidence that obtaining a dual endorsement significantly reduces the likelihood that special education candidates teach in special education classrooms.
Citation: Roddy Theobald, Dan Goldhaber, Natsumi Naito, Marcy Stein (2020). The Special Education Teacher Pipeline: Teacher Preparation, Workforce Entry, and Retention. CALDER Working Paper No. 231-0220
Using detailed administrative data for public schools, we document racial and ethnic segregation at the classroom level in North Carolina, a state that has experienced a sharp increase in Hispanic enrollment. We decompose classroom-level segregation in counties into within-school and between-school components. We find that the within-school component accounted for a sizable share of total segregation in middle schools and high schools. Recognizing its importance could temper the praise for school assignment policies that reduce racial disparities between schools but allow large disparities within them. More generally, we observe between the two components a complementary relationship, with one component tending to be large when the other one is small. Comparing the degree of segregation for the state’s two largest racial/ethnic minority groups, we find that White/Hispanic segregation was more severe than White/Black segregation, particularly within schools. Finally, we examine enrollment patterns by course and show that school segregation brings with it differences by race and ethnicity in the courses that students take, with White students more likely to be enrolled in advanced classes.
WP 230-0220 was originally released in February 2020. This is the third update, WP 230-0220-3, released in January 2021.
Citation: Charles Clotfelter, Helen Ladd, Calen R. Clifton, Mavzuna R. Turaeva (2020). School Segregation at the Classroom Level in a Southern ‘New Destination’ State. CALDER Working Paper No. 230-0220-3
Teacher turnover has adverse consequences for student achievement and imposes large financial costs for schools. Some have argued that high-stakes testing may lower teachers’ satisfaction with their jobs and could be a major contributor to teacher attrition. In this paper, we exploit changes in the tested grades and subjects in Georgia to study the effects of eliminating high-stakes testing on teacher turnover and the distribution of teachers across grades and schools. To measure the effect of testing pressures on teacher mobility choices we use a "difference-in-differences" approach,comparing changes in mobility over time in grades/subjects that discontinue testing vis-à-vis grades/subjects that are always tested. Our results show that eliminating testing did not have an impact on the likelihood of leaving teaching, changing schools within a district, or moving between districts. We only uncover small negative effects on the likelihood of grade switching. However,we do find relevant positive effects on retention of beginning teachers in the profession. In particular, the average probability of exit for teachers with 0-4 years of experience fell from 14 to13 percentage points for teachers in grades 1 and 2 and from 14 to 11 percentage points in grades 6 and 7.
Citation: Dillon Fuchsman, Tim Sass, Gema Zamarro (2020). Testing, Teacher Turnover and the Distribution of Teachers Across Grades and Schools. CALDER Working Paper No. 229-0220
While the majority of students with disabilities (SWDs) receive instruction from general education teachers, little empirical work has investigated the ways in which these students have equitable access to high-quality teachers. We explore the differences in teacher quality experienced by SWDs and general education (GEN) students and how that access varies with school-level disadvantage by estimating SWD teacher quality gaps in the Los Angeles Unified School District. We examine several different indicators of teacher effectiveness (hiring scores, teacher experience, teachers’ ratings on their observation-based performance evaluations, and value-added measures) for general education teachers who instruct both SWDs and general education (GEN) students. We find that SWDs are significantly more likely to have lower math VAM teachers than their GEN peers, and these gaps do not vary by school-level disadvantage. We find no differences on the other indicators of teacher effectiveness.
Citation: Ijun Lai, W. Jesse Wood, Scott A. Imberman, Nathan Jones, Katharine Strunk (2020). Teacher Quality Gaps by Disability and Socioeconomic Status: Evidence from Los Angeles. CALDER Working Paper No. 228-0220
We use administrative microdata from six states to study academic mobility in K-12 education, by which we mean the extent to which students’ ranks in the distribution of academic performance change during their schooling careers. We find that there is substantial heterogeneity across districts in the academic mobility of students. The heterogeneity across districts is largest in terms of absolute mobility—i.e., in some districts, students throughout the distribution, including initially low performers, gain on other students in the statewide distribution, and vice-versa—whereas there is less cross-district variation in relative mobility. The most prominent correlates of high-mobility districts include value-added to achievement and the socioeconomic status of the student population.
Citation: Wes Austin, David Figlio, Dan Goldhaber, Eric Hanushek, Tara Kilbride, Cory Koedel, Jaeseok Sean Lee, Jin Lou, Umut Özek, Eric Parsons, Steven Rivkin, Tim Sass, Katharine Strunk (2020). Where are Initially Low-performing Students the Most Likely to Succeed? A Multi-state Analysis of Academic Mobility (Preliminary Draft). CALDER Working Paper No. 227-0220
We describe the postsecondary transitions of students taking CTE courses in high school using administrative data on one cohort of high school graduates in Washington State. Conditional on observable characteristics, CTE concentrators—high school graduates who complete at least four CTE credits—are about 4 percentage points less likely to enroll in college than other high school graduates. However, CTE students are significantly more likely to enroll in and complete vocational programs, especially in certificate programs in applied STEM and public safety fields. Among students not enrolled in college, CTE students are also more likely to obtain full-time employment—and to work more intensively—within the first three years following high school graduation. Although the improvements in employment outcomes do not offset reductions in college enrollment, the higher completion rates of vocational credentials among CTE concentrators indicate some important positive outcomes for this population.
Citation: James Cowan, Dan Goldhaber, Harry Holzer, Natsumi Naito, Zeyu Xu (2019). Career and Technical Education in High School and Postsecondary Pathways in Washington State. CALDER Working Paper No. 224-1119
We evaluate the predictive validity of the Massachusetts Candidate Assessment of Performance (CAP), a practice-based assessment of teaching skills that is typically taken during a candidate’s student teaching placement and is a requirement for teacher preparation program completion in Massachusetts. We find that candidates’ performance on the CAP predicts their in-service summative performance evaluations in their first 2 years in the teaching workforce and provides a signal of these ratings beyond what is already captured by the state’s traditional licensure tests, but is not predictive of their value added to student test scores. These findings suggest that the CAP captures aspects of candidates’ skills and competencies that are better reflected in their future performance evaluations than by their impacts on student performance.
Working paper 223-1019 was orginally published in October 2019. This is an updated version, published in February 2021.
Citation: Bingjie Chen, James Cowan, Dan Goldhaber, Roddy Theobald (2019). From the Clinical Experience to the Classroom: Assessing the Predictive Validity of the Massachusetts Candidate Assessment of Performance . CALDER Working Paper No. 223-1019-2