• Tidak ada hasil yang ditemukan

(Farley-Ripple et al., 2012; Grissom and Bartanen, 2019a). Replacement effects may vary in these cases. To the extent that some kinds of transitions are anticipated and planned for more than others, disruptive effects may vary as well.

Building on existing research, our study makes three main contributions to our under- standing of the effects of principal turnover. First, we employ an identification strategy that more plausibly isolates the impact of principal turnover from the circumstances that may have led to the transition. Second, we move beyond a binary turnover measure to examine whether specific types of principal transitions are more or less harmful to school performance. Finally, we investigate a range of potential mechanisms—both compositional and disruptive—through which the effects of principal turnover may operate.

Education Statistics. We use information about student enrollment size, proportion of Black students, proportion of Hispanic students, and the proportion of students eligible for free- or reduced-price lunch, which serves as a measure of school poverty.

III.2.1 Measuring Principal Turnover

Both datasets contain job classification and location information that allows identification of principals and what school they worked in each year. From these longitudinal administrative data files, we create a binary principal turnover variable, which takes a value of 1 if a principal in a school in yeartis not the principal in the school in yeart+ 1, and zero otherwise. Figure III.1 shows the distribution of the number of principal transitions experienced by a school across the study period. In both states, almost all schools changed principals at least once and many had multiple transitions, a complication that we address in the modeling section.

We also examine the effects of different types of principal turnover: transfers, exits, promotions, and demotions. Transfers are principals in year t who move to a principal position in a different school in year t+ 1. In Missouri, roughly half of transfers are to a different district, compared to only 10% in Tennessee.4 Exits are principals in yeartwho are not working in the state’s K–12 public education system in year t+ 1. These exits include principals who retire, principals who are fired by the district (or whose contracts are not renewed), and principals who chose to leave public education (e.g., to work in a private school or a different industry); we cannot differentiate among these exit types. Promotions are principals in year t who move “upward” to a central office position in year t+ 1, such as superintendent or principal supervisor. Demotions are moves “downward” from a principal position in year t to a school-based, non-principal position in year t+ 1. The majority of demotions in both states are principals moving to assistant principal positions.

Table III.1 shows yearly principal turnover rates in each state for the analysis period.

Principal turnover rates are slightly higher in Missouri, on average, at 20% each year, com-

4Missouri has a much larger number of school districts and schools, despite having a smaller population than Tennessee.

Figure III.1: Number of Principal Transitions Experienced Within Schools

Notes: Number of principal transitions refers to the number of times that a given school changed principals during the study period (shown in the plot header).

pared to 18% in Tennessee. The breakdown of turnover into types is also fairly similar between the states. Among the four types of principal turnover we categorize (transfers, exits, promotions, and demotions), exits are the most common, followed by transfers, which are slightly more frequent in Missouri. While overall rates of position changes are similar between Missouri and Tennessee, demotions are relatively more common in Tennessee, with more principal promotions to central office in Missouri.

III.2.2 Outcome Measures

Our main outcome of interest is student achievement in math and reading. Specifically, we examine school-average achievement scores on statewide exams in grades 3 to 8 and end-of- course exams for high school students. We do not have student-level scores for Missouri, so we convert school mean scale scores to school-level standard deviations (standardized within year). We construct a parallel outcome in Tennessee by aggregating student-level scale scores to the school level. Thus, our achievement outcomes reflect school performance relative to the distribution of schools in each state.

In addition to achievement outcomes, we estimate the effect of principal turnover on teacher turnover. Specifically, we operationalize teacher turnover in two ways: the proportion of teachers in the school in yeart−1 who are still in the school in year t and the number of teachers who are new to the new school in yeart. These two measures are highly correlated.

We provide both to complement traditional analysis of turnover rates with an estimate (number of new-to-school teachers) that is concrete and helpful in interpreting magnitude.

We also draw on measures of teacher quality and school climate to probe the potential mechanisms connecting principal turnover to school performance. These measures are only available in the Tennessee data. Specifically, we construct a yearly measure of a school’s average teacher quality by estimating teacher value-added (VA) in math and reading using the drift-adjusted approach proposed by Chetty et al. (2014).5 We average the teacher VA

5The estimation steps are as follows. First, we residualize student test scores (separately by subject) on a vector of prior-year test scores, student characteristics (race/ethnicity, gender, FRPL eligibility, gifted

Table III.1: Principal Turnover by Year in Missouri and Tennessee

All Turnover Transfers Exits Promotions Demotions Missouri

2001 0.198 0.060 0.072 0.042 0.024

2002 0.197 0.054 0.077 0.040 0.026

2003 0.203 0.060 0.076 0.037 0.033

2004 0.189 0.046 0.081 0.042 0.022

2005 0.196 0.055 0.071 0.041 0.029

2006 0.207 0.058 0.074 0.051 0.026

2007 0.202 0.056 0.071 0.045 0.033

2008 0.225 0.061 0.101 0.042 0.023

2009 0.187 0.047 0.081 0.035 0.028

2010 0.202 0.055 0.087 0.039 0.031

2011 0.187 0.051 0.076 0.042 0.021

2012 0.223 0.063 0.090 0.042 0.030

2013 0.217 0.058 0.092 0.045 0.024

2014 0.176 0.047 0.064 0.039 0.027

2015 0.214 0.061 0.075 0.049 0.030

All 0.202 0.056 0.079 0.039 0.025

Tennessee

2007 0.165 0.041 0.065 0.032 0.030

2008 0.191 0.044 0.081 0.031 0.037

2009 0.174 0.042 0.073 0.029 0.030

2010 0.159 0.041 0.066 0.018 0.037

2011 0.185 0.044 0.078 0.034 0.030

2012 0.196 0.048 0.083 0.035 0.033

2013 0.178 0.040 0.076 0.031 0.033

2014 0.185 0.038 0.076 0.035 0.040

2015 0.179 0.039 0.078 0.032 0.033

All 0.179 0.042 0.075 0.031 0.033

Notes: Transfers are principals who move to a different school in the next year but remain a principal. Exits are principals who leave the state education system. Promotions are principals who move to a central office position in the next year. Demotions are principals who move to non-principal, school-based position, such as assistant principal or teacher.

estimates for each school-by-year observation and subsequently standarize them.

Our measure of school climate comes from teachers’ responses on a yearly statewide survey in Tennessee, which was first administered in 2011–12. Part of the survey includes items that assess teachers’ perceptions of school climate. Examples of items include, “The staff at this school like being here; I would describe us as a satisfied group”, and “There is an atmosphere of trust and mutual respect within this school.” Using factor analysis, we reduce the full set of responses to a single teacher-by-year score, then average these scores within each school-by-year cell. Details on the factor analysis, including the full set of items, is available in Appendix III.9. Importantly, the survey window runs from early March to the middle of April, which is roughly one month prior to the last day of school. This timing means that many teachers likely complete the survey prior to knowing whether their principal will remain in the school for the next year.