Content area
Abstract
Using the method of ordinary least squares (OLS), regressions were run on 22 revenue sources and total general fund revenue for the City of Dover, Delaware. Various specifications were attempted for each revenue source. To forecast each source, the regression with the lowest mean absolute percent error (MAPE) that had coefficients with logical signs was chosen. Revenue forecasts were then created for fiscal years 2007, 2008, and 2009. A tracking system was also created so city officials can easily see if the city is on pace to collect their budgeted revenue.