Using fuzzy rule based economic models to measure the performance of state funded agencies

US state governments have been seeking analytical methods to show that taxpayer dollars devoted to economic development are utilized effectively. In New York, efforts to quantify the increase in value added as a result of technology development have involved the use of multiple regression to analyze client surveys. In this paper, we describe work with one group of agencies to develop a fuzzy rule-based system. Initial results show that fuzzy models can outperform multiple regression when used on even terms and have the potential to substantially outperform regression models when used to their full advantage.