Chapter 22 Panel data

摘要: Publisher Summary This chapter discusses the models that are static conditional on a latent variable. The panel aspect of data has been primarily used to control for Much work needs be done incorporate uncertainty and interesting dynamics. Exploiting martingale implications time additive utility seems fruitful. There is, however, potentially important distinction between averages cross-section averages. A average forecast errors over T periods should converge zero as T→ ∞. But an across N individuals surely need not N→ ∞; there is common component in those errors, due economy-wide innovations. same point applies when considering covariances with variables agent's information sets. If conditioning discrete, one can think averaging subsets errors; ∞, these but necessarily →