File(s) stored somewhere else
Please note: Linked content is NOT stored on Carnegie Mellon University and we can't guarantee its availability, quality, security or accept any liability.
Statistical Methods to Evaluate Geographically-Targeted Economic Development Programs
journal contributionposted on 01.01.2000, 00:00 by Daniele Bondonio
In recent years an increasing amount of efforts has been devoted to the evaluation of geographically-targeted economic development (GTED) programs. In the U.S. and in Great Britain, geographically-targeted business incentives (denominated Enterprise Zone programs) are an important policy instrument to revitalize local communities. Within the E.U., interest on the evaluation of GTED programs is fueled by the number of development programs cofunded by the European Regional Development Fund, the European Social Fund and the European Agricultural Guidance and Guarantee Fund. The surging interest for the evaluation of GTED programs is challenged by the difficulty to assess the causality link between the program intervention and the observed changes in the economic outcomes of interest. Evaluating GTED programs is a difficult task because it requires the evaluator to distinguish changes due to the program from changes due to the many factors independent from the program intervention. Such a task is particularly difficult also due to the lack of experimental data available to the evaluator. This paper illustrates the sources of the potential biases that can affect impact estimates of GTED programs, and develop a number of statistical methods that control for such sources. The proposed methods are then grouped and sorted out in a decision tree algorithm that provides guidance to select the most appropriate methodology for the analysis, based on the program characteristics and on the type of data available. An evaluation of the impact of the U.S. Enterprise Zones on local employment concludes the paper as an empirical application of the methods and the decision tree algorithm proposed. This application highlights how seriously distorted impact estimates can be when they are obtained using unsophisticated tools for the analysis. The methods proposed in the paper proved instead to be effective tools to avoid these distortions.