Job Market Paper

  • Locally Robust Policy Learning: Inequality, Inequality of Opportunity and Intergenerational Mobility
    • Abstract: Policy makers need to decide whether to treat or not to treat different individuals. The optimal choice depends on the welfare function that the policy maker has in mind. I study a general setting for policy learning with general semiparametric Social Welfare Functions (SWFs), possibly defined by semiparametric U-statistics, which accomodate a wide range of distributional preferences. I use locally robust/orthogonal scores to provide strong statistical guarantees for the estimated policy rules even in observational settings where the propensity score is unkown. This work expands previous results in Athey and Wager (2021). Three main applications of the general theory motivate the paper: (i) Inequality aware SWFs, (ii) Inequality of Opportunity aware SWFs and (iii) Intergenerational Mobility SWFs. I use the Panel Study of Income Dynamics (PSID) to asses the effect of attending preschool on adult earnings and estimate optimal policy rules based on parental years of education and parental income. Cite.

Books

  • Causal Inference and Machine Learning: A Locally Robust Approach: with Juan Carlos Escanciano, in progress (Taylor & Francis Group)

Working Papers

  • Debiased Machine Learning U-statistics: with Juan Carlos Escanciano, R&R at Review of Economic Studies (resubmitted), previously distributed as Machine Learning Inference on Inequality of Opportunity.
    • Abstract: We propose a method to debias estimators based on U-statistics with Machine Learning (ML) first-steps. Standard plug-in estimators often suffer from regularization and model-selection biases, producing invalid inferences. We show that Debiased Machine Learning (DML) estimators can be constructed within a U-statistics framework to correct these biases while preserving desirable statistical properties. The approach delivers simple, robust estimators with provable asymptotic normality and good finite-sample performance. We apply our method to three problems: inference on Inequality of Opportunity (IOp) using the Gini coefficient of ML-predicted incomes given circumstances, inference on predictive accuracy via the Area Under the Curve (AUC), and inference on linear models with ML-based sample-selection corrections. Using European survey data, we present the first debiased estimates of income IOp. In our empirical application, commonly employed ML-based plug-in estimators systematically underestimate IOp, while our debiased estimators are robust across ML methods. Cite.
  • Educational Inequality of Opportunity and Mobility in Europe
    • Abstract: Educational attainment generates labor market returns, societal gains and has intrinsic value for individuals. We study Inequality of Opportunity (IOp) and intergenerational mobility in the distribution of educational attainment. We propose to use debiased IOp estimators based on the Gini coefficient and the Mean Logarithmic Deviation (MLD) which are robust to machine learning biases. We also measure the effect of each circumstance on IOp, we provide tests to compare IOp in two populations and to test joint significance of a group of circumstances. We find that circumstances explain between 38% and 74% of total educational inequality in European countries. Mother’s education is the most important circumstance in most countries. There is high intergenerational persistence and there is evidence of an educational Great Gatsby curve. We also construct IOp aware educational Great Gatsby curves and find that high income IOp countries are also high educational IOp and less mobile countries. Cite.