Does attending a STEM high school improve student performance? Evidence from New York City

•We investigate whether STEM high schools in New York City promote performance in science and mathematics.•Comparisons of means indicate an advantage to attending a STEM school, but more thorough analysis reduces or eliminates this advantage.•Females and males in STEMs do better than their counterparts in Non-STEMs, but the gender gap is larger in these schools.•Black-white and Hispanic-white gaps are smaller in STEM relative to Non-STEM schools, but the Asian-white gap is larger in STEMs.

We investigate the role of specialized science, technology, engineering, and mathematics (STEM) high schools in New York City (NYC) in promoting performance in science and mathematics and in closing the gender and race gaps in STEM subjects. Using administrative data covering several recent cohorts of public school students and a rich variety of high schools including over 30 STEMs, we estimate the effect of attending a STEM high school on a variety of student outcomes, including test taking and performance on specialized science and mathematics examinations. While comparisons of means indicate an advantage to attending a STEM school, more thorough analysis conditioning on a rich set of covariates, including previous grade test performance, reduces or eliminates this advantage. Females and males in STEMs do better than their counterparts in Non-STEMs, but the gender gap is also larger in these schools. We also find that the black-white and Hispanic-white gaps are smaller in STEM relative to Non-STEM schools across almost all outcomes, but the Asian-white gap, in contrast, is larger in STEMs relative to Non-STEMs.