Topics in Identification, Limited Dependent Variables, Partial Observability, Experimentation, and Flexible Modeling

The first of two volumes in honor of the scholarship of professor Dale J. Poirier, this volume consists of 12 chapters on econometrics methods related to identification, limited dependent variables, partial observability, experimentation, and flexible modeling, including both Bayesian and classical contributions to theory and application. The volume begins with an interview with Poirier, then addresses macroeconomic nowcasting using Google probabilities; sentiment-based overlapping community discovery of Reddit's newsfeed users; a psychological model of violence and Israeli and Palestinian fatalities in the Second Intifada; Bayesian methodology for modeling local activation and global connectivity using data on magnetic resonance signals in the brain; robust estimation of ARMA (autoregressive moving average) models with near root cancellation; and the estimation of a stochastic volatility model. Others discuss a novel approach to the modeling of expectation formation and learning in models with time-varying parameters, particularly endogenous gain learning; an approach for checking the sensitivity of predictive modeling to prior hyperparameters; the estimation of a panel model and the use of a Stein-type shrinkage estimator; an out-of-sample Granger causality testing procedure; and the effect of compulsory schooling laws on educational attainment and labor market earnings. Essays were presented at a conference at the U. of California, Irvine, in June 2018, and contributors are data scientists, economists, and other researchers working in Europe, North America, Australia, China, and Saudi Arabia. -- Copyright 2019 * Portland, OR *