[3] Brian A. Jacob's research focuses on the labor markets of teachers, school accountability and choice, and housing and criminal justice (among other topics).
Together with Lefgren, Jacob finds that raising teachers' in-service training has no significant effect on students' achievement in reading or math, raising questions on whether small-scale staff development investments are an effective strategy in high-poverty schools.
With Steven Levitt, Jacob uses unexpected test score fluctuations to assess teacher cheating in Chicago public schools and estimates that cheating by teachers or school administrators on standardized tests occurs in at least 4-5% of elementary school classrooms, with the frequency of cheating responding strongly to minor changes in incentives, which constitutes a key pitfall of high-powered incentive systems in education.
[15] Finally, in research with Lars Lefgren, Jacob has analyzed remedial education, parents' preferences regarding teaching, and the impact of grade retention.
[20] Finally, along with Enrico Moretti, Jacob and Lefgren use weather shocks to study the dynamics of criminal behavior, finding - unlike previous research - that criminal behavior displays natural mean reversion, thus casting doubt on the long-run effects of temporary crime prevention efforts.
[21] Other findings of Jacob's research include the lack of significant impact of mandatory high school graduation exams on 12th-grade students' achievement in math or reading, though they increase the likelihood of the students with the lowest ability dropping out of school.
[23] Examining why nearly 3 out of 5 college students are women, Jacob finds that nearly 90% of the gender gap in higher education can be accounted for by gender differences in non-cognitive skills (e.g. the ability to pay attention in class, to cooperate, to organize and to seek help) and college premia.