Abstract:This paper studies estimation of a panel data model with latent structures where individuals can be classified into different groups where slope parameters are homogeneous within the same group but heterogeneous across groups. To identify the unknown group structure of vector parameters, we design an algorithm called Panel-CARDS which is a systematic extension of the CARDS procedure proposed by Ke, Fan, and Wu (2015) in a cross section framework. The extension addresses the problem of comparing vector coefficients in a panel model for homogeneity and introduces a new concept of controlled classification of multidimensional quantities called the segmentation net. We show that the Panel-CARDS method identifies group structure asymptotically and consistently estimates model parameters at the same time. External information on the minimum number of elements within each group is not required but can be used to improve the accuracy of classification and estimation in finite samples. Simulations evaluate performance and corroborate the asymptotic theory in several practical design settings. Two empirical economic applications are considered: one explores the effect of income on democracy by using cross-country data over the period 1961-2000; the other examines the effect of minimum wage legislation on unemployment in 50 states of the United States over the period 1988-2014. Both applications reveal the presence of latent groupings in these panel data.

Uniform Inference in Linear Panel Data Models with Two-Dimensional Heterogeneity?

Linear panel regressions with two-way unobserved heterogeneity

Bi-integrative analysis of two-dimensional heterogeneous panel data model

Inference for Heterogeneous Effects using Low-Rank Estimation of Factor Slopes

Panel data models with nonadditive unobserved heterogeneity: Estimation and inference

Homogeneity and Sparsity Analysis for High-Dimensional Panel Data Models

Low-rank Panel Quantile Regression: Estimation and Inference

Homogeneity Pursuit in Panel Data Models: Theory and Applications

Determining Individual or Time Effects in Panel Data Models

Nonstationary Panel Models with Latent Group Structures and Cross-Section Dependence

Nonparametric dynamic panel data models: Kernel estimation and specification testing

Feasible Generalized Least Squares for Panel Data with Cross-sectional and Serial Correlations

Inference In Partially Identified Panel Data Models With Interactive Fixed Effects

Determination of Different Types of Fixed Effects in Three-Dimensional Panels*

Grouped Heterogeneity in Linear Panel Data Models with Heterogeneous Error Variances

TESTING HOMOGENEITY IN PANEL DATA MODELS WITH INTERACTIVE FIXED EFFECTS

Specification Test for Panel Data Models with Interactive Fixed Effects

Robust Nearly-Efficient Estimation of Large Panels with Factor Structures

Panel Data Models with Time-Varying Latent Group Structures

Heteroscedastic stratified two-way EC models of single equations and SUR systems

Discretizing Unobserved Heterogeneity