Econometric Model Performance In Forecasting And Policy Assessment PDF Download

The role of the REMI Policy Insight+ model in socioeconomic forecasting and economic impact analysis of transportation projects was assessed. The REMI PI+ model is consistent with the state of the practice in forecasting and impact analysis. REMI PI+, like its competitors, is vulnerable to the trends contained in the historical data it uses, especially recent trends. After the most recent periodic update in data, the performance of the REMI PI+ model improved, i.e., it produced long-term forecasts that were more credible. Zonal-level population and employment forecasts for direct input to the Indiana Statewide Travel Demand Model (ISTDM) can be achieved by applying disaggregation regression methods. IU's Center for Econometric Model Research (CEMR) model is also a sound forecasting model. Because of the knowledge of in-state economists, the CEMR-IBRC model could provide forecasts of the Indiana economy that reflect characteristics not known to out-of-state forecasters. The researchers also examined economic impact analysis models that are possible alternatives to REMI PI+. Acquiring a new economic impact analysis package does not seem necessary for INDOT, if REMI forecasts can be adjusted to (a) accommodate recent and reasonable expected trends in the Indiana economy, and (b) meet the geographic (TAZ) needs of the ISTDM. MCIBAS -- which is currently used by INDOT -- is a good hybrid system to use in the economic impacts analysis of transportation projects. IU's CEMR is capable of conducting economic impact analyses, with local knowledge of the Indiana economy, at a cost lower than REMI's. However, INDOT would have to decide whether these potential advantages justify changing the present relationship with REMI. In cases where the credibility of data, forecasts, and/or impact analyses needs to be verified, an INDOT version of an expert panel along the lines of Michigan's Transportation Technical Committee could be convened.