How accurate is forecast.id?
The accuracy of forecast.id usually associated with the quality of assumptions that underpin the forecasts. .id attempts to undertake a strong process of consultation to ensure that the assumptions that are used in the forecasts are validated by local government professionals, such as statutory and strategic planners, community service experts and other relevant parties, such as developers.
One of the benefits of the modelling techniques used by .id is the greater ability to scrutinise the assumptions and output, particularly after censuses. The key assumptions that are utilised in the forecasts include:
- residential development
- migration by age
- household formation by age
While the consultation processes ensure a greater degree of accuracy, the nature of urban development and demographic and household change is fluid, meaning that change and alteration to assumptions is always necessary over time. As such, a review of the forecasts is essential every one to two or so years (depending on the area) to check the assumptions and monitor the performance of the forecasts.
What economic assumptions are taken into account?
Economic assumptions are not explicitly part of the modelling process. They are implied in different ways, depending on the area being forecast.
In regional and rural areas, a close assessment of the local economic conditions must be undertaken, as they have a direct impact on migration patterns and levels of household and population growth.
In urban areas, the current state of the metropolitan and regional economy is assessed as an input to short-term assumptions about levels of residential development. As no economic cycle is assumed as a part of the forecasts, the levels of development may not typify year to year variations in residential and demographic change.
Why is there only one population forecast and not a number of scenarios?
When producing small area population forecasts, .id produces one standard set of outputs.
This is due to a number of reasons.
- Difficulty of differentiating small area assumptions
There are a multitude of different assumptions that are used in small area forecasts. Providing a sufficient scenario for each of the assumptions would lead to tens or even hundreds of sets of forecasts.
- Multiple forecasts lead to confusion
Combined experience of more than 20 years in conducting population forecasts and projections within the .id team has showed that multiple scenarios leads to confusion amongst users, who often ask which is the ‘best’ or ‘most appropriate’ scenario. Therefore, .id uses a single forecast approach, which allows a greater degree of consistency and more integrated approach by users of the data.
- Multiple forecasts leads to people using the best-case scenario depending on their viewpoint
The traditional ‘projection’ approach is to do a number of scenarios, based on varying assumptions. However, the different outcomes are often used by varied interest groups for the purposes of lobbying specific agenda. Therefore, .id uses a single forecast approach, which allows a greater degree of consistency and more integrated approach by users of the data.
- .id provides scenario analysis services
.id has developed a scenario service, which allows council to test the possible impact of (amongst other things) policy changes on population forecasts. This might include changes to growth corridors policy, medium density housing or diversifying housing markets.
How often should the forecasts be reviewed?
Population forecasts should be analysed and reviewed regularly. The need to update forecasts will vary, depending on the changes occurring at the local area.
Small areas with substantial residential growth, significant demographic change or changes to important institutions (non-private dwellings) need more regular updates (annual to bi-annual). This is also important for areas that are affected by global and national influences (e.g. tertiary student market or business migrants), notably inner city areas or coastal tourism areas. The main reason for updating relates to the fact that new information and new developments are coming to light at regular intervals.
This can be compared to middle suburban areas or rural areas, with stable or well-established housing markets that are less volatile and require less updating (every two to three years). The benefit of regular updating of forecasts is not only to maintain greater accuracy, but also the ability to build knowledge within organisations about key local residential developments and demographic changes.