fessionals. We provide evidence that applicants just eligible for an exam school have peers that
score 0.17 to 0.36 standard deviations higher on eighth grade state tests and that are 6.4 to 9.5
percentage points less likely to be black or Hispanic. Surprisingly, however, the impact of exam
school eligibility on college enrollment and graduation is, if anything, negative. The results are
similar across gender, middle school type, and baseline state test scores.
Our results suggest that the typical applicant does not signicantly benefit from attending a
school with dramatically higher-achieving and more homogeneous peers. With that said, without
longer-term measures such as income, health, or life satisfaction, it is dicult to fully interpret our
results. To the extent that attending an exam school with higher-achieving peers increases social
capital in ways that are important for later outcomes that are independent of college enrollment,
graduation, or human capital, then there is reason to believe that our conclusions are premature
and the true impact of an elite exam school will only be understood with the passage of time. If, on
the other hand, college enrollment and graduation are \sucient statistics" for later life outcomes,
then our results show the impact of attending a school with elite peers for the marginal student is
likely to be small.
全文见内:
http://scholar.harvard.edu/files/fryer/files/dobbie_fryer_shs_07_2013.pdf