>Research

Publications

« Inference on an Extended Roy Model, with an Application to Schooling Decisions in France », with Xavier d’Haultfœuille, Journal of Econometrics, 174(2) : 95-106 (2013).

Abstract : This paper considers the identification and estimation of an extension of Roy’s model (1951) of sectoral choice, which includes a non-pecuniary component in the selection equation and allows for uncertainty on potential earnings. We focus on the identification of the non-pecuniary component, which is key to disentangle the relative importance of monetary incentives versus preferences in the context of sorting across sectors. By making the most of the structure of the selection equation, we show that this component is point identified from the knowledge of the covariates effects on earnings, as soon as one covariate is continuous. Notably, and in contrast to most results on the identification of Roy models, this implies that identification can be achieved without any exclusion restriction nor large support condition on the covariates. As a byproduct, bounds are obtained on the distribution of the ex ante monetary returns. We also propose a three-stage semiparametric estimation procedure for this model, which yields root-n consistent and asymptotically normal estimators. Finally, we apply our results to the educational context, by providing new evidence from French data that non-pecuniary factors are a key determinant of higher education attendance decisions.

« Another Look at the Identification at Infinity of Sample Selection Models », with Xavier d’Haultfœuille, Econometric Theory, 29(1) : 213-224 (2013).

Abstract : It is often believed that without instruments, endogenous sample selection models are identified only if a covariate with a large support is available (see, e.g., Chamberlain, 1986, and Lewbel, 2007). We propose a new identification strategy mainly based on the condition that the selection variable becomes independent of the covariates for large values of the outcome. No large support on the covariates is required. Moreover, we prove that this condition is testable. We finally show that our strategy can be applied to the identification of generalized Roy models.

« Choosing the Field of Studies in Postsecondary Education : Do Expected Earnings Matter ? », with Magali Beffy and Denis Fougère, Review of Economics and Statistics, 94(1) : 334-347 (2012).

Abstract : This paper examines the determinants of the choice of the college major when the length of studies and future earnings are uncertain. We estimate a three-stage schooling decision model, focusing on the effect of expected earnings on major choice. We control for dynamic selection through the use of mixture distributions. Exploiting variations across the French business cycle in the relative returns to the majors, our results yield a very low, though significant, elasticity of major choice to expected earnings. This suggests that at least for the French university context, nonpecuniary factors are a key determinant of schooling choices.


« The Effect of Employment while in High School on Educational Attainment : a Conditional Difference-in-Differences Approach », with Franz Buscha, Lionel Page and Stefan Speckesser, Oxford Bulletin of Economics and Statistics, 74(3) : 380-396 (2012).

Abstract : Using American panel data from the National Education Longitudinal Study of 1988, this article investigates the effect of working during grade 12 on attainment. We employ, for the first time in the related literature, a semiparametric propensity score matching approach combined with difference-in-differences. We address selection on both observables and unobservables associated with part-time work decisions, without the need for instrumental variable. Once such factors are controlled for, little to no effects on reading and math scores are found. Overall, our results therefore suggest a negligible academic cost from part-time working by the end of high school.

Working papers and Work in progress

« College Attrition and the Dynamics of Information Revelation », with Peter Arcidiacono, Esteban Aucejo and Tyler Ransom. Slides(April 2013).

Abstract : This paper investigates the determinants of college attrition, in a setting where individuals have imperfect information about their schooling ability and labor market productivity. We estimate, using data from the National Longitudinal Survey of Youth 1997 (NLSY97), a dynamic structural model of schooling and work decisions, where students who graduated from high school (or obtained a GED) decide at each period whether to attend college, either in a two- or four-year institution, work part-time or full-time, or engage in home production. A key feature of our approach is to account for learning over time through college grades and wages, thus implying that individuals may leave or re-enter college as a result of the arrival of new information on their ability and productivity. We use our results to quantify the importance of informational frictions in explaining the observed school-to-work transitions, and to evaluate the value of information in this context.


« Overeducation and Skill Mismatch : a Dynamic Structural Approach », with Brian Clark and Clement Joubert. Companion paper providing reduced-form evidence on the dynamics of overeducation over the life cycle (October 2012).

Abstract : This paper investigates the determinants of educational mismatch and its effect on wages. We specify and estimate, using data from the National Longitudinal Survey of Youth 1979 (NLSY79), a dynamic structural model of schooling and career choice that incorporates search frictions, endogenous human capital accumulation, compensating wage differentials and unobserved heterogeneity in skills. Notably, this framework allows us to quantify the relative importance of different theories that have been used separately in the literature to explain why a substantial fraction of workers are more qualified than necessary to perform their job, despite low returns to surplus schooling. We argue that telling apart the role played by search frictions, versus other factors such as unobserved ability, non-pecuniary job benefits or career mobility prospects, is crucial to understand the welfare and policy implications of the overeducation phenomenon.


« Extremal Quantile Regression for Sample Selection Models, with an Application to Black-White Wage Gap », with Xavier d’Haultfœuille and Yichong Zhang.

Abstract : We consider the estimation of a semiparametric location-scale model subject to endogenous selection, without instruments nor large support regressor. Following D’Haultfoeuille and Maurel (2013), identification relies on independence between covariates and selection, when the potential outcome tends to infinity. In this context, we propose a simple estimator based on extremal quantile regressions. We establish asymptotic normality by extending the results of Chernozhukov (2005) to allow for selection. Finally, we apply our method to the estimation of the black-white wage gap among males from the NLSY79, revisiting the results of Neal and Johnson (1996).


« The Effect of Grade Retention on Schooling Performance », with Xavier d’Haultfœuille and Jean-Marc Robin.

Abstract : In this paper, we develop a new strategy to estimate the effects of early grade retention on schooling performance. In order to pin down the causal effect of grade retention, we assume a factor structure where selection into treatment is controlled for using a strategy at infinity. The key identifying condition relies on the fact that the best students face an arbitrarily small probability of repeating a grade, so that the selection issue vanishes once conditioning on sufficiently high initial test scores. No other restriction is imposed on the grade retention process. Under those assumptions, which allow for general form of essential heterogeneity, we show that the factor loadings, the distribution of the factors and the distribution of treatment effects are nonparametrically identified, and we propose to estimate them using a flexible parametric specification. Finally, we apply our estimation procedure to analyze the effects of grade retention between grades 1 and 3 in the context of French primary schools.


« Changes Across Cohorts in Wage Returns to Schooling and Early Work Experiences : Distinguishing Price and Composition Effects », with Jared Ashworth, V. Joseph Hotz, and Tyler Ransom.

Abstract : This paper investigates the evolution over the last three decades in the wage returns to schooling and early work experience. We isolate changes in skill prices from changes in composition by estimating, using data from the National Longitudinal Survey of Youth 1979 and 1997 cohorts, a dynamic model of schooling and work decisions. Importantly, this allows us to account for the endogenous nature of the changes in educational and accumulated work experience across the two cohorts. Our preliminary results point to the existence of significant composition effects.


« Estimation of Non-Stationary Job Search Models using CCP », with Peter Arcidiacono, Dave Bielen and Ekaterina Roshchina.


« Understanding College Financial Decision-Making : the Role of Time Preference, Risk Aversion, Skills and Preferences », with Christian Belzil and Modibo Sidibe.


« The Effect of Part-Time Work on Post-Secondary Educational Attainment : New Evidence from French Data », with Magali Beffy and Denis Fougère, IZA Discussion Paper No 5069.

Abstract : We provide new evidence on the effect of part-time work on post-secondary educational attainment. To do so, we use samples from the French Labor Force Surveys from 1992 to 2002. We use variations across departments in low-skilled youth unemployment rates and in their interactions with the father’s socio-economic status to identify the part-time work effect. Our results suggest a very large detrimental effect of holding a regular part-time job on graduation. We argue that this may be due to the lack of flexibility of the French university system, which does not offer much complementarity between schooling achievement and part-time work.