Essays on the organization of science and education

• Main Author: Li, Danielle
• Other Authors: David H. Autor and Pierre Azoulay.
• Format: Others
• Language: English
• Published: Massachusetts Institute of Technology 2012
• Subjects: Economics.
• Online Access: http://hdl.handle.net/1721.1/72836

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Economics, 2012. === Cataloged from PDF version of thesis. === Includes bibliographical references (p. 217-226). === This dissertation consists of four chapters exploring how organizations inform and distort the implementation of public policy in two empirical settings. Chapters 1 and 2 study the non-market allocation of research funding to scientists while Chapters 3 and 4 examine the market for schools and school leaders. Experts are likely to have more information regarding the potential of projects in their area, but are also more likely to be biased. Chapter 1 develops a theoretical and statistical framework for understanding and separately identifying the effects of bias and information on expert evaluation and applies it in the context of peer review at the National Institutes of Health (NIH). I use exogenous variation in review committee composition to examine how relationships between reviewers and applicants, as measured by citations, affect the allocation and efficiency of grant funding. I show that, due to bias, each additional related reviewer increases the chances that an applicant is funded by 2.9 percent. Reviewers, however, are also more informed about the quality of proposals from related applicants: the correlation between scores and quality is approximately 30 percent higher for related applicants. On net, the presence of related reviewers improves the quality of research that the NIH supports by two to three percent, implying that reductions in conflicts of interest may come at the direct cost of reducing the quality of funding decisions. In Chapter 2, I examine how women are treated in grant review at the US National Institutes of Health (NIH). Analyzing funded R01 grants, I show that women receive a half-percentile worse score than men for research that produces the same number of publications and citations. Allowing reviewers to observe applicant gender reduces the number of women who are funded by approximately 3 percent. Analysis of study sections shows that the presence of women attenuates bias, suggesting that diversity in study sections can improve peer review. Chapter 3 considers the effect of labor market for school leaders. School accountability may affect the career risks that school leaders face without providing commensurate changes in pay. Since effective school leaders likely have significant scope in choosing where to work, these uncompensated risks may limit the ability of low-performing schools to attract and retain effective leaders. This paper analyzes the effect of No Child Left Behind (NCLB) on principal mobility and the distribution of high-performing principals across low- and high-performing schools. I show that NCLB decreases average principal quality at schools serving disadvantaged students by inducing more able principals to move to schools less likely to face NCLB sanctions. Finally, Chapter 4 explores the viability of voucher base school market reforms by estimating the demand elasticity for private schooling using variation from sibling discounts at Catholic schools. Because families differ in their number and spacing of children, this variation allows us to isolate within-neighborhood variation in tuition prices. We find that a standard deviation decrease in tuition prices increases the probability that a family will send its children to private school by one half percentage point, which translates into an elasticity of Catholic school attendance with respect to tuition costs of -0.19. Our subgroup results suggest that a voucher program would disproportionately induce into private schools those who, along observable dimensions, are unlike those who currently attend private school. === by Danielle Li. === Ph.D.