Embedded Files
belieffit is a Stata command for estimating posterior mean log beliefs and correcting parametrically for bias from posterior variance using fixed-point iteration. Developed for Pay Beliefs and the Amenity-Pay Tradeoff
ebayes_shrink for Stata
ebayes_shrink for Stata
ebayes_shrink is a Stata command for applying Empirical Bayes shrinkage to one or more already-estimated fixed effects using per-observation variance-covariance estimates. Developed for Pay Beliefs and the Amenity-Pay Tradeoff
R package to test for violations of monotonicity or LATE assumptions in subsets of data, using machine learning and sample splitting. Allows clustering, weights, fixed effects and various testable conditions. Developed for Testing the Monotonicity Assumption in Instrumental Variable Models, with Tymon Słoczyński and Martin Huber
Stata package for estimating bunching using the polynomial bunching estimator, consistent with theory. Tests for consistency of shapes of the distributions to the left and right of cutoff, estimates analytic standard errors and allows for various adjustments to account for distortions above the cutoff. Developed for A better polynomial bunching estimator.
Flexible Stata package for estimating marginal treatment effects using parametric and semiparametric models and various estimation methods. Accepts fixed effects, estimates treatment effect parameters and weights and provides a flexible plotting suite to graph output. See also Stata Journal article. Install from ssc or development version from Github.
Stata package for testing instrument validity in settings with binarized treatment. The package estimates all first stage thresholds and uses moment inequality tests to test (necessary, not sufficient) conditions for instrument validity. Download from Github. See also paper with Martin Huber.
Stata package for plotting specification curves (Simonsohn et al. 2015), useful for reporting robustness and specification checks. Development version on Github.
Google Sites
Report abuse