Chapter 15 Monte carlo simulation and numerical integration

This chapter discusses simulation methods that are both important and useful in the solution of integration problems, and discusses the principles for the practical application of simulation in economics with a focus on integration problems. The simulation methods are generally straightforward for the investigator to implement, relying on an understanding of a few principles of simulation and the structure of the problem at hand. By contrast, deterministic methods require much larger problem-specific investments in numerical methods. Simulation methods can provide solutions for two related integration problems. One integration problem arises in model solution for agents whose expected utilities cannot be expressed as a closed function of state and decision variables. The other occurs, when the investigator combines sources of uncertainty about models to draw conclusions about policy. Markov chain Monte Carlo methods, which make use of samples that are neither independently nor identically distributed, have greatly expanded the scope of integration problems with convenient practical solutions.