Association for Public Policy Analysis & Management

Do high school sport championships come at the cost of reduced student performance on statewide assessments? This question is examined using high school data on math, reading, and science scores for the state of New Mexico over a five-year panel. The identification strategy centers on the fact that statewide assessments are administered in the same testing windows year to year (once in the fall semester and once in the spring), which happen to correspond with statewide sports tournaments. Due to the rural nature of New Mexico, high schools that participate in the championship games of these tournaments often have student athletes who miss two to three days of instructional time that cannot be made up before the statewide assessments. Non-student athletes are also impacted if they choose to spectate these championship tournaments, class instruction is halted due to missing students attending the championship tournaments, or if school is altogether cancelled to allow all students to attend the championship tournament (this is a common practice at some rural New Mexico high schools when days built into the schedule for bad weather have not been used yet in the spring semester). This impact would then theoretically be largest for the smallest schools in the state, where student athletes comprise a larger percentage of the student body than at large high schools. Preliminary evidence from fixed effects models suggests that in semesters where a small high school (small defined as being in the bottom two quartiles of enrollment in the state) participates in the spring basketball championship game experience a 0.56 standard deviation decrease in the percentage of students performing proficient or above on statewide reading assessments (which are administered during a testing window that corresponds with spring basketball championship tournaments). There are inconclusive effects on statewide math assessments. When examining the effects of participating in a fall championship game, there are no effects on reading or math assessments that are administered in the spring. These findings suggest that school performance on spring statewide assessments can be negatively impacted by participating in spring basketball championship games through the mechanism of lost instructional time. Policy implications for these findings would suggest either moving the spring statewide assessments windows to not correspond with state basketball tournaments, moving the state basketball tournaments to a different time, or providing administrators a larger testing window to ensure that extra-curricular activities at specific schools do not interfere with statewide assessments.