EXPLOITING SEQUENTIAL LEARNING TO ESTIMATEESTABLISHMENT-LEVEL PRODUCTIVITY DYNAMICS ANDDECISION RULES

This paper develops a sequential learning estimator of production functions and productivitydynamics for unbalanced establishment panels. Extending an idea from the literatureon dynamic industry models, establishments are uncertain about their own idiosyncraticproductivities and update productivity beliefs using information revealed by their productionexperience. The estimator relies on the structure of this iterative learning process andthereby avoids placing any restriction on establishment strategic behavior. Consequently,the estimator is suitable for comparative studies of the behavioral sources of technologicalchange across all types of industry. Estimation of productivity dynamics and of behavioraldecision rules are separated into recursive stages. Using sequential learning estimates of productivitybeliefs from the first stage, decision rules for exit, investment, and innovation effortcan be estimated in a second stage. A test application with four Chilean industries confirmsthat the estimator produces plausible estimates with small standard errors. Decision ruleestimates show that productivity beliefs affect investment and exit hazards in the expecteddirection....