Urban Economics
Credit Standards and Segregation
with Romain Ranciere (Paris School of Economics and International Monetary Fund)
working paperThis paper shows that the mortgage credit boom has significantly affected urban and school racial segregation from 1995 to 2007. We develop a model of urban segregation with credit constraints that shows that easier credit can either increase or decrease segregation, depending on the race of the marginal consumer who benefits from the expansion of credit. We then use school demographics from 1995 to 2007, matched to a national comprehensive dataset of mortgage originations, to show that higher leverage increases the segregation of African American and Hispanic students. Both segregation across schools and across school districts is higher when the leverage is higher. The estimates show that segregation could be substantially lower without the large increase in leverage.
Brokers and the Dynamics of Segregation
Update coming soon
Economics of Education
Estimating Perceptions of Discrimination:
Experimental Economics in Schoolsjoint with Lionel Page (Queensland Institute of Technology)
working paperWe put forward a unique experimental economics design to estimate people's perceptions of discrimination in a variety of contexts. We apply it in the context of teachers in the classroom. We estimate students' perceptions of teachers using monetary incentives: Students were given an endowment to bet on their own performance on a written verbal test, either when assessed non-anonymously by their teacher, in a random half of the classrooms, or when assessed anonymously by an external examiner. 1,200 British students in grade 8 classrooms across 29 schools took part. Female students invested more when assessed by a male teacher, and male students invested less when assessed by a female teacher. Interestingly, female students' perceptions are not in line with actual discrimination: Male teachers tended to be more lenient with male students, and tougher with female students. Ethnicity and socioeconomic status did not play a role.
Using Compulsory Mobility to Identify
Heterogeneous School Quality and Peer Effectsjoint with Francis Kramarz (CREST-INSEE, Ecole Polytechnique)
and Stephen Machin (University College London and London School of Economics).
working paper.This paper was previously circulated under the title What Makes a Test Score?, available
here.
A presentation for the U.S. National Academy of Science
here.
Educational production functions that estimate school quality and peer effects require sufficient and exogenous mobility of pupils across schools. We use a unique feature of English education that provides us with sufficient and exogenous pupil mobility across schools: pupils who attend schools that do not cater for grades 3 to 5 have to leave their school in grade 2. We show in an econometric model that focusing on compulsory movers reduces the endogenous mobility bias. We use this strategy to estimate a full educational production function where school quality and peer effects depend on student characteristics, i.e. there is a “match” between schools, peers, and pupils. Results suggest substantial heterogeneity in peer effects and school quality across subgroups of pupils, and questions the idea that there are bad or good schools regardless of students' characteristics. The model also incorporates long- and short-run effects of school quality, pupil background, and peer effects. Educational inputs have short-lived effects, but the pupil's background has effects of both the level and the progress of the pupil.
Programs for the paper are available
here.
Assessed by a Teacher Like Me:
Race, Gender, and Subjective Evaluations Paper.
CentrePiece featureRevise and Resubmit at Education Finance and PolicyThe underrepresentation of minority teachers and male teachers remains an issue in U.S. elementary education. Do teachers give better grades to children of their own race, ethnicity, or gender? A U.S. nationally representative longitudinal data set that includes both test scores and teacher assessments offers a unique opportunity to answer this question. I use student mobility and changes in classroom composition as credibly idiosyncratic variations. Results indicate that teachers give higher assessments to children of their own race conditional on test scores, child effects, and teacher effects; but not significantly higher assessments to children of their own gender. This effect seems to be driven largely by the lower assessments given by white teachers to black and hispanic children. Results are robust to various checks on stereotype threats, endogenous mobility, measurement error, and reverse causality.
Applied Econometrics
Predicting the Moments of Firms’ Earnings (work-in-progress)Current Coauthors
Francis Kramarz
Steve Machin
Lionel Page
Romain Ranciere
Birthe Larsen
Andrei Levchenko
Woo-Jin Chang
Steve Monahan
Rodney Ramcharan
