« Back to Results
Marriott Marquis, Del Mar
Hosted By:
Econometric Society
Identification and Estimation in Causal Models
Paper Session
Sunday, Jan. 5, 2020 1:00 PM - 3:00 PM (PDT)
- Chair: Victor Aguirregabiria, University of Toronto
(Non)Randomization: A Theory of Quasi-Experimental Evaluation of School Quality
Abstract
In centralized school admissions systems, rationing at oversubscribed schools often uses lotteries in addition to preferences. This partly random assignment is used by empirical researchers to identify the effect of entering a school on outcomes like test scores. This paper formally studies if the two most popular empirical research designs successfully extract a random assignment. For a class of data-generating mechanisms containing those used in practice, I show: One research design extracts a random assignment under a mechanism if and almost only if the mechanism is strategy-proof for schools. In contrast, the other research design does not necessarily extract a random assignment under any mechanism.Quantile Treatment Effects with Two-Sided Measurement Error
Abstract
This paper considers quantile effects when there is measurement error both in the outcome variable and in a continuous "treatment" variable while allowing for other covariates that are not measured with error. We identify these effects under two key conditions (i) the quantiles of the outcome and the quantile of the treatment are linear in the covariates and (ii) the measurement errors are "classical." In our framework, no instruments are required nor are repeated measures, and we do not need to impose distributional assumptions on the measurement errors. We develop two-step semiparametric estimators of the parameters of interest that are straightforward to implement in practice and show that the estimators are asymptotically normally distributed. We apply our method to study intergenerational income mobility. We find evidence that the income of sons and fathers are measured with error and that accounting for measurement error particularly appears to lower the income of low income sons of low income fathers; we find especially big differences relative to linear quantile regression of son's income on father's income and other covariates.Identification of Regression Models with a Misclassified and Endogenous Binary Regressor
Abstract
We study identification in nonparametric regression models with a misclassified and endogenous binary regressor when an instrument is correlated with misclassification error. We show that the regression function is nonparametrically identified if one binary instrument variable and one binary covariate that satisfy the following conditions are present. The instrumental variable (IV) corrects endogeneity; the IV must be correlated with the unobserved true underlying binary variable, must be uncorrelated with the error term in the outcome equation, and is allowed to be correlated with the misclassification error. The covariate corrects misclassification; this variable can be one of the regressors in the outcome equation, must be correlated with the unobserved true underlying binary variable, and must be uncorrelated with the misclassification error. We also propose a mixture-based framework for modeling unobserved heterogeneous treatment effects with a misclassified and endogenous binary regressor and show that treatment effects can be identified if the true treatment effect is related to an observed regressor and another observable variable.Identification and Estimation of Dynamic Structural Models with Unobserved Choices
Abstract
This paper develops identification and estimation methods for dynamic discrete choice structural models when agents‘ actions are unobserved by econometricians. We provide conditions under which choice probabilities and latent state transition rules are nonparametrically identified with a continuous state variable. Our identification results extend to models with serially correlated unobserved heterogeneity and to cases in which only discrete state variables are available. We apply our method to study moral hazard problems in US gubernatorial elections. We find that the probabilities of shirking increase as the governors approach the end of their terms.Identification of Firms' Beliefs in Structural Models of Market Competition
Abstract
This paper studies the joint identification of firms' beliefs and structural parameters in a general class of empirical games of market competition. In this framework, firms have incomplete information and their beliefs about the behavior of other firms are unrestricted -- nonparametric -- functions of a firm's information. Beliefs may be out of equilibrium. This framework includes --- as particular cases --- models of competition in prices or quantities, auction models, and static and dynamic discrete choice games. I present identification results under different scenarios on the data available to the researcher: only firms' choice data; data on consumers' demand; and data on firms' costs.JEL Classifications
- C2 - Single Equation Models; Single Variables
- C3 - Multiple or Simultaneous Equation Models; Multiple Variables