This paper considers a general class of stochastic dynamic choice models with discrete and continuous decision variables. This class contains a variety of models that are useful for modeling intertemporal household decisions under risk. Our examples are drawn from the field of development economics. We formalize this class as a dynamic programming problem, then propose a solution method that relies on value function iteration. Finally, in an example we show how our algorithm can be applied to solve and estimate a dynamic model with discrete and continuous controls.
value function iteration mixed continuous/discrete controls stochastic dynamic choice model