Related Experiment Videos
Small area population forecasting: some experience with British models.
Tijdschrift Voor Economische En Sociale Geografie = Journal of Economic and Social Geography = Revue De Geographie Economique Et Humaine = Zeitschrift Fur Okonomische Und Soziale Geographie = Revista De Geografia Economica Y Social
|January 1, 1983
Summary
Evaluating population forecasting models for planning, this study found that Holt-Winters, ratio-correction, and exponential-smoothing models offer the most accurate predictions for municipalities.
Area of Science:
- Demography
- Spatial Analysis
- Statistical Modeling
Background:
- Accurate population forecasts are crucial for effective urban and regional planning.
- Existing forecasting models vary in complexity and performance.
- A comprehensive evaluation of these models is needed to identify optimal methods.
Purpose of the Study:
- To evaluate the performance of various population forecasting models.
- To identify the most accurate models for municipal-level population projections.
- To provide evidence-based recommendations for planning purposes.
Main Methods:
- Utilized space-time population data for the Netherlands (1950-1979).
- Developed and evaluated 154 different forecasting models across 832 municipalities.
- Employed a data-driven approach, using the first 20 years for model building and the subsequent 10 years for performance evaluation.
Main Results:
- Identified Holt-Winters, ratio-correction, and low-order exponential-smoothing models as superior performers.
- Demonstrated significant variation in forecast accuracy among the evaluated models.
- Highlighted the importance of model selection for reliable population projections.
Conclusions:
- Holt-Winters, ratio-correction, and exponential-smoothing models are recommended for municipal population forecasting.
- The study provides valuable insights for demographers and urban planners.
- Accurate forecasting enhances the effectiveness of long-term planning and resource allocation.