STOCHASTIC GROWTH WITH HETEROGENEOUS AGENTS
This paper is a simulation analysis of the stochastic growth model with heterogeneous agents. The environment is one of two agents, with a common technology for production, individual labor endowment shocks, constant relative risk aversion utility functions, and limited borrowing/lending opportunities with no-Ponzi game constraints on government behavior.The analysis makes use of parameterized expectations to solve the model, with a neural network approximation for the expectations of each agent. The genetic algorithm is used to solve for the parameterized expectations coefficients, which satisfy pre-set borrowing/lending limits.This paper analyzes the following questions: is there a higher capital stock when there are less restrictive limits on lending/borrowing between agents, than when there are more restrictive limits? Secondly, are the dynamics considerably different, in the case of identical utility functions but different endowment shocks for the agents, than in the case of different utility functions with identical endowment shocks? Finally, what happens when a government is introduced, with preset expenditures, which can borrow, and tax labor, capital, and consumption of each agent? How do particular tax schemes affect overall capital accumulation, borrowing/lending between agents, and overall welfare?
Year of publication: |
2000-07-05
|
---|---|
Authors: | McNelis, Paul |
Institutions: | Society for Computational Economics - SCE |
Saved in:
Saved in favorites
Similar items by person
-
Inflation Targeting and Q Volatility in Small Open Economies
McNelis, Paul, (2004)
-
The money-age distribution: Empirical facts and economic modelling
Heer, Burkhard, (2006)
-
Forecasting inflation with thick models and neural networks
McNelis, Paul, (2004)
- More ...