Working papers

2025

1. Location Characteristics of Conditional Selective Confidence Intervals via Polyhedral Methods

   Andreas Dzemski, Ryo Okui, and Wenjie Wang

   2025

   Accepted for publication at Statistical Science

   Abs arXiv

   We examine the location characteristics of a conditional selective confidence interval based on the polyhedral method. This interval is constructed from the distribution of a test statistic conditional upon the event of statistical significance. In the case of a one-sided test, the behavior of the interval varies depending on whether the parameter is highly significant or only marginally significant. When the parameter is highly significant, the interval is similar to the usual confidence interval derived without considering selection. However, when the parameter is only marginally significant, the interval falls into an extreme range and deviates greatly from the estimated value of the parameter. In contrast, an interval conditional on two-sided significance does not yield extreme results, although it may exclude the estimated parameter value.

2. Inference after multiple hypothesis testing

   Andreas Dzemski, Ryo Okui, and Wenjie Wang

   2025

   Revision requested by Biometrika

   Abs arXiv

   Significant treatment effects are often emphasized when interpreting and summarizing empirical findings in studies that estimate multiple, possibly many, treatment effects. Under this kind of selective reporting, conventional treatment effect estimates may be biased and their corresponding confidence intervals may undercover the true effect sizes. We propose new estimators and confidence intervals that provide valid inferences on the effect sizes of the significant effects after multiple hypothesis testing. Our methods are based on the principle of selective conditional inference and complement a wide range of tests, including step-up tests and bootstrap-based step-down tests. Our approach is scalable, allowing us to study an application with over 370 estimated effects. We justify our procedure for asymptotically normal treatment effect estimators. We provide two empirical examples that demonstrate bias correction and confidence interval adjustments for significant effects. The magnitude and direction of the bias correction depend on the correlation structure of the estimated effects and whether the interpretation of the significant effects depends on the (in)significance of other effects.

3. Long-Run Stock Return Distributions: Empirical Inference and Uncertainty

   Andreas Dzemski, Adam Farago, Erik Hjalmarsson, and 1 more author

   2025

   Revision requested by the Review of Asset Pricing Studies

   Abs Website

   We analyze empirical estimation of the distribution of total payoffs for stock investments over very long horizons, such as 30 years. Formal results for recently proposed bootstrap estimators are derived and alternative parametric methods are proposed. All estimators should be viewed as inconsistent for longer investment horizons. Valid confidence bands are derived and should be the focus when performing inference. Empirically, confidence bands around long-run distributions are very wide and point estimates must be interpreted with great caution. Consequently, it is difficult to distinguish long-run aggregate return distributions across countries; long-run U.S. returns are not significantly different from global returns.

Journal articles

2024

1. Confidence set for group membership

   Andreas Dzemski and Ryo Okui

   Quantitative Economics, 2024

   Abs arXiv Website

   Our confidence set quantifies the statistical uncertainty from data-driven cluster assignment in clustered panel models. It covers the true cluster memberships jointly for all units with pre-specified probability and is constructed by inverting many simultaneous unit-specific one-sided tests for group membership. We justify our approach under N,T to infinity asymptotics using tools from high-dimensional statistics, some of which we extend or develop in this paper. We provide an empirical application as well as Monte Carlo evidence that the confidence set has adequate coverage in finite samples.

2021

1. Convergence rate of estimators of clustered panel models with misclassification

   Andreas Dzemski and Ryo Okui

   Economics Letters, 2021

   Abs arXiv Website

   We study kmeans clustering estimation of panel data models with a latent group structure and N units and T time periods under long panel asymptotics. We show that the group-specific coefficients can be estimated at the parametric root NT rate even if error variances diverge as T approaches infinity and consequently some units are asymptotically misclassified. This limit case approximates empirically relevant settings and is not covered by existing asymptotic results.

2019

1. An empirical model of dyadic link formation in a network with unobserved heterogeneity

   Andreas Dzemski

   Review of Economics and Statistics, 2019

   Abs PDF Website

   I study a dyadic linking model in which agents form directed links that exhibit homophily and reciprocity. A fixed effect approach accounts for unobserved sources of degree heterogeneity. I consider specification testing and inference with respect to the homophily and reciprocity parameters. The specification test compares observed transitivity to predicted transitivity. All test statistics account for the presence of an incidental parameter by using formulas based on an asymptotic approximation. In an application to favor networks in Indian villages, the specification test detects that the dyadic linking model underestimates the true transitivity of the network.