A system dynamics model of the dynamics of special education reimbursement policies, based upon work completed in Massachusetts, is presented. The model is used to analyze the causes of unstable growth in special education costs and to propose policy options for controlling such costs. Because of an elaborated ability to represent the behaviors of local school districts, the system dynamics technology was found to be ideally suited to this type of policy analysis. However, the existence of the elaborated feedback structure would make it extremely cumbersome to use the system dynamics model to project annual costs on a locality-by-locality basis, such projections being the principal strength of traditional education finance models. This trade-off between two modeling technologies suggests that analysts must carefully match their audience, purpose, and modeling technology to attain "best" analyses.
Analysis in the field of educational finance has been dominated by detailed, highly disaggregated, town-by-town analyses of the impact of funding policies for one or two years. This paper explores the question of what are the relative advantages and disadvantages of using an alternative, highly aggregate, feedback simulation technology to analyze educational finance policies. The case reported here is based upon an analysis of the policies used to fund special education in Massachusetts. The original analysis was undertaken by the Division of Special Education, Massachusetts State Department of Education [I].
Comparative analyses of the advantages and disadvantages of various simulation technologies are important for several reasons. First, as Roberts (1976) has noted, although the applications of simulation and other management science techniques have grown at the tremendous rate over the past twenty years, there has not been comparable growth in understanding concerning how such models are implemented