GOVERNMENT TOOLS
CONSULTING
COMPARE
Informed Decisions or Place Info: how the two forecasts compare
Informed Decisions and Place Info both publish population and dwelling forecasts for Australian communities. They are built on different premises.
Place Info's Place Forecast reflects back the assumptions of whoever is using it. It supplies a starting position, then hands the user the controls to change it, so local knowledge, and local expectations, go straight into the numbers.
Informed Decisions' National Forecasting Program is an independent assessment of the most likely future for a place. It rests on three things: the share of national growth each region is forecast to receive, a detailed assessment of existing dwellings and the development actually in the pipeline, and the reconciliation of those two against each other.
That difference matters most when a forecast goes in front of someone else. A number set in-house is the organisation's own view of its future, and a reviewer will read it that way. The value of an independent assessment is that the assumptions were not set by the party that has to defend them.
What follows compares the two on method, using each provider's published documentation.
BEFORE YOU COMPARE
What actually decides a population forecast
Forecasts of the same place can differ by tens of thousands of people. Almost all of that difference comes down to three assumptions, and they are worth understanding before comparing any two providers.
Where and when dwellings get built. People live in dwellings, so a small area cannot grow faster than homes appear in it. That makes the dwelling assumption the largest single driver of a local forecast, and it cannot be derived from history alone. It takes knowing which sites are coming, how many dwellings each will yield, when construction starts, how quickly they fill, and what happens when an area runs out of land to build on.
How many people live in each dwelling. Household size converts dwellings into people, and it is not a constant. It differs between a new apartment precinct and an established suburb of family homes, and it shifts as an area ages. Two forecasts with identical dwelling numbers will still disagree if they treat household size differently.
What the total is allowed to be. A forecast built for one area in isolation can say almost anything, because nothing outside it pushes back. A forecast built inside a national and state model cannot: the growth it gives an area has to be growth the country actually has.
Dwelling research is the largest single investment behind the National Forecasting Program, and that page covers how it is done.
The third assumption is the least visible and the most consequential, so it is worth its own section.
A place is forecast as part of the country, not on its own
Growth is not created locally. How many people Australia has in a given year is set by births, deaths and overseas migration. How many any one region has is set by where those people go. The real question a local forecast answers is what share of the country's growth lands in that place.
The National Forecasting Program is built that way round. It forecasts the nation, then the states, then the regions, then each local area's share within them, using one method everywhere in Australia. Every local forecast sits inside the same national picture, and none of them can assume a share of growth the country does not have to give.
That is also what makes a number hold up when someone else checks it. A forecast for a single area, built without that context, is difficult to defend against a state projection that was built with it.
The Informed Decisions knowledge base explains what top-down forecasting is and how the constraint is applied.
At a glance
| Informed Decisions: National Forecasting Program | Place Info: Place Forecast | |
|---|---|---|
| How future dwellings are projected | Researched site by site: the location, capacity, timing and sequence of individual development sites, maintained as an ongoing research program | The last 12 months of building approvals held steady, converted to completions using observed construction timelines |
| National and state context | Every local area is forecast as a share of a national and state picture, using one method across Australia | No national or state control total described in the published methodology |
| Average household size | Forecast to change over time, by dwelling type and by area | Carried forward from the jump-off year unless a council supplies its own assumption |
| Who supplies local development knowledge | Informed Decisions' development researchers, as part of the product | The council, through add, reduce, override and ceiling adjustments |
| Who produces the forecast | Specialist demographers and development researchers | The user, on top of a mechanical baseline |
| Update cycle | Monthly, a region at a time. Top-down model reviewed annually | Not stated publicly |
| Written analysis | Published with every region release | Not stated publicly |
| Geography | Custom micro-geography derived from SA1s, rolling up to small area, LGA, region, state and national | Local government areas and the small areas within them |
WHAT THE METHOD BUYS YOU
Why organisations choose the National Forecasting Program
The table above compares method. What follows is what the method is for.
An independent assessment. Most of the moments that matter involve someone questioning the number: a state department, a planning panel, a funding body, a developer, a court. A forecast set by the organisation presenting it is an argument. One produced independently, inside a national model, is evidence.
Research rather than settings. The development pipeline behind each area is maintained as ongoing research, using council data, building approvals, aerial photography, property databases and land-use change. It is the assumption that moves the answer most, so it takes the most work.
A forecaster who owns the area. Forecasts are released monthly, a region at a time, and each release carries written analysis from the forecaster who produced it, explaining what changed and why.
One method everywhere. The same methodology covers every location in Australia, so forecasts are comparable across council boundaries. That matters for anyone planning services, networks or investment across more than one area.
A published methodology. How the forecast is built is documented and public, so the assumption driving growth in any small area can be produced on request.
Where a self-serve tool fits. An organisation with demographic capability in-house, wanting to test its own scenarios quickly for an internal audience, is well served by a tool it drives itself. The National Forecasting Program is built for the other case, where the number has to be defensible to someone who did not produce it.
Questions people ask
Informed Decisions' National Forecasting Program is used across industries with quite different interests in the rate of development. Some would prefer to see more of it, some would prefer less. In most cases what they want is an honest assessment of the most likely outcome, produced by someone with no position to argue.
Place Info's Place Forecast is the better fit for a different job. It models a set of assumptions and shows what that growth would look like, with the user in control of the assumptions. Where the task is to put a case, or to demonstrate what follows if a particular level of development is achieved, a tool you drive yourself is the right one.
The question worth settling first is which of those two jobs is actually in front of you.
Three questions worth asking.
How to test any forecast
What are the dwellings in this forecast based on, and has anyone looked at the development pipeline in this area? Is the area forecast as a share of a national and state picture, or on its own? And what is assumed about household size over the forecast period?
Those three answers explain most of the difference between any two forecasts of the same place. Informed Decisions publishes its answers to all three, and will walk through them for a specific area on request.