Derives sufficient statistics for feedback and heterogeneity in dynamic panel logit models, proves conditional likelihood identification is infeasible for Markov covariates, and proposes two assumptions to restore identification.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
econ.EM 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Sufficient Statistics for Markovian Feedback Processes and Unobserved Heterogeneity in Dynamic Panel Logit Models
Derives sufficient statistics for feedback and heterogeneity in dynamic panel logit models, proves conditional likelihood identification is infeasible for Markov covariates, and proposes two assumptions to restore identification.