This paper provides a unified framework for partial identification of counterfactual parameters in a general class of discrete outcome models allowing for endogenous regressors and multidimensional latent variables, all without parametric distributional assumptions. Our main theoretical result...

This paper proposes a new methodology for estimating teacher value-added. Rather than imposing a normality assumption on unobserved teacher quality (as in the standard empirical Bayes approach), our nonparametric estimator permits the underlying distribution to be estimated directly and in a...

The instrumental variable (IV) estimator in a cross-sectional or panel regression model is often taken to provide valid causal inference from contemporaneous correlations. In this exercise we point out that the IV estimator, like the OLS estimator, cannot be used effectively for causal inference...

The existing measures of housing affordability are essentially short-run indicators that compare current income with property prices. Taking into consideration that a housing purchase is a long-horizon decision and the property price reflects the discounted present value of future mortgage...

Statistical models of unobserved heterogeneity are typically formalised as mixtures of simple parametric models and interest naturally focuses on testing for homogeneity versus general mixture alternatives. Many tests of this type can be interpreted as C (a) tests, as in Neyman (1959), and shown...

The link between mental stress and cancer is still a belief, not a well established scientific fact. Scientists have relied largely on opinions of cancer stricken patients to establish a link between stress and cancer. Such opinion surveys tend to produce contradictory statistical inferences....

This paper proposes to incorporate product customization in the Maskin and Riley (1984) nonlinear pricing model in order to capture major features of mobile service data. In particular, consumers are characterized by a two-dimensional type. One dimension is observed by the provider and...

Auction data often contain information on only the most competitive bids as opposed to all bids. The usual measurement error approaches to unobserved heterogeneity are inapplicable due to dependence among order statistics. We bridge this gap by providing a set of positive results. First, we show...