A Comparison of Optimal Control Solutions
in a Labor Market Model
Ric D. Herbert *
The University of Newcastle, Faculty of Science and Information Technology,
Newcastle, NSW, 2308 Australia
Gareth D. Leeves
The University of Queensland, School of Economics, Brisbane, Qld, 4072,
Australia
Abstract
In this paper a variety of computational optimal control techniques are compared
using a nonlinear labor market model.
Key words: JEL Classification: J63, J64, J65, J68
1 Introduction
In this paper we compare a variety of computational optimal control tech-
niques to undertake the same problem. The problem we use for the compar-
ison is that of determining the optimal number of long-term unemployed to
enter Active Labour Market Programs (ALMPs) as a policy tool to reduce un-
employment. The model we use is nonlinear with the nonlinearities inherent
in the model due to a Nash bargaining rule between employers and work-
ers, and Cobb-Douglas technology in a job-matching function. The model is
complicated by distinguishing between various duration classes of short-term
unemployment. These features make the model difficult for the application of
optimal control techniques.
* Corresponding author.
Email addresses: [email protected] (Ric D. Herbert),
[email protected] (Gareth D. Leeves).
Preprint submitted to Elsevier Science
29 March 2004