Abstract
Purpose. The forecasting of school enrollment is vital to the fiscal planning of California school districts. Most public school district allocation formulas usually include a per pupil funding component. Current, popular enrollment forecasting methods base predictions on prior year's enrollment from the preceding grade. Although these methods may be reliable for most grades, they often are unreliable for kindergarten enrollment prediction, because accurate counts of pre-school enrollment are usually unavailable. Methodology. This study examined 8 regional, 9 local economic factors and kindergarten enrollments from 36 San Diego County school districts for the years 1980 through 1990. Factor analysis was used to reduce each of the economic variable sets to its principal components. Factor scores were used as independent variables of regression analysis. Predictions were made for kindergarten enrollment, for the current year as well as for lag periods of 1, 2, and 3 years. Predictions from two other kindergarten enrollment forecasting methods currently being used in California were also examined. The predictions for the economic forecasting method and each of the other two forecasting methods for each of the 36 school districts for each lag period, were tabulated and compared to actual enrollment. Calculations for percentage error of the economic forecasting model and other two methods in use were also made and tabulated. The sign test was used to determine which prediction method had the least error. Findings and conclusions. This study found that the economic model was a more accurate predictor of kindergarten enrollments than were either of the other two methods tested for all three lag periods. It also found that regional economic factors were usually better predictors of kindergarten enrollment than were local economic factors or the combined regional and local economic variable set. For a majority of the 36 school districts, factor analysis revealed that the regional economic factors, total residential units, and total single family units accounted for 94.1 percent of the variation in kindergarten enrollments. For those few districts whose kindergarten enrollment was best predicted by local economic factors, the local economic factors, units sold within the district and percentage of regional market activity, accounted for approximately 80 percent of the variation in kindergarten enrollment.