Skip to content
  • There are no suggestions because the search field is empty.

Why GRP estimates differ between data providers

Providers distribute the same state total in different ways, and the method explains the gap

There is no standard method for estimating Gross Regional Product below state level in Australia, so different providers can publish quite different GRP figures for the same LGA. In most cases both providers reconcile their estimates to the ABS Gross State Product total, so neither figure is missing activity - the difference is in how the state total is distributed between local areas. Understanding the main method differences lets a council read the figures together rather than choosing a 'right' one.

Where the differences come from

Top-down and bottom-up methods

A top-down model takes a state or national total and allocates it to regions using regional indicators. It holds the relativities between regions constant, and in particular it assumes every worker in an industry produces the state average - it cannot tell the difference between higher-paid workers generating high value added and lower-paid workers generating less. A bottom-up model builds each region up from local data instead, giving each region and industry its own production relationships. At .id, we use a bottom-up approach: the NIEIR model derives productivity for each LGA from taxation income data for that area's workers, so local wages, part-time shares and firm size show through in the result.

Productivity per job

This is the main driver of differences between providers. A provider applying state-average value added per job to every LGA will tend to sit below a locally derived estimate in metropolitan areas, where productivity is typically above the state average, and above it in regional areas. The state total is the same - it is distributed differently.

How employment is counted

.id's employment estimates are adjusted for the Census undercount of workplaces, so the same value added is spread across more workers. This makes .id's productivity per job figures lower overall than those of a provider using unadjusted employment, without either being wrong.

Where ownership of dwellings sits

The national accounts include a component for ownership of dwellings - the value accruing to dwelling owners, whether as rent received or as imputed rent for owner occupiers. Providers place it differently: some carry it inside the rental, hiring and real estate services industry, while Economic Profile reports it as a separate component outside the industry table, because no jobs or businesses attach to it. The value is the same - it just sits in a different place, so that one industry row is simply not comparable between providers.

What 'output' means

In addition to the difference in GRP, Economic Profile's output figures measure local production, consistent with Australian production in the ABS Input-Output tables. Some providers report a total-supply figure instead, which includes competing imports - for import-intensive industries such as manufacturing that can be close to double the local-production figure. This is why .id's output figures are almost always smaller than a total-supply figure for the same industry.

Where activity is attributed

Methods also differ on where some industries' activity is placed. In construction, for example, attributing activity to where the business is based and attributing it to where the building happens are both defensible choices - but they put the value in different LGAs. Only one of them answers the question 'what is happening in our LGA', and that is the basis .id uses.

How to read two figures together

When a council holds two GRP figures for the same area, work through them rather than defending one:

  • Check both figures are the same measure. Headline GRP, total value added and output are different measures, and comparing across them guarantees a gap.
  • Check where ownership of dwellings sits in each, before comparing any industry table.
  • Check whether an output comparison mixes local production with total supply.
  • Then read any remaining difference as a method difference, mostly driven by productivity per job.

Neither method is wrong. The question to ask of any GRP figure is whether its method answers the question the council is asking about its own area.

Frequently asked questions

Q: Another provider has our GRP at a different figure. Which one is right?

A: Both can be right, because they measure and distribute activity differently. Both typically reconcile to the ABS state total, so the difference is in how that total is shared between LGAs - mainly through productivity per job, the treatment of ownership of dwellings and how employment is counted. Compare methods, not just the headline numbers.

Q: Why is another provider's output figure for an industry so much higher than yours?

A: Check what each figure measures. .id's output is local production; a total-supply figure includes competing imports and can be close to double for import-intensive industries such as manufacturing.

Q: Why does the rental and real estate industry row differ so much between providers?

A: Usually because of ownership of dwellings. Some providers include it inside that industry, while Economic Profile reports it as a separate component outside the industry table. The row is not comparable between the two presentations, even though the underlying value is the same.

Q: We have moved from another data provider and our GRP has changed. Has our economy shrunk or grown?

A: Neither. A change of provider is a change of method, not a change in the economy. The two series distribute the same state total differently, so their levels are not directly comparable and the shift says nothing about whether the local economy has grown or shrunk. Published figures should always be quoted with their vintage and source, so any later comparison is made within one series.

Related articles