1270.0.55.001 - Australian Statistical Geography Standard (ASGS): Volume 1 - Main Structure and Greater Capital City Statistical Areas, July 2016  
Latest ISSUE Released at 11:30 AM (CANBERRA TIME) 24/07/2017   
   Page tools: Print Print Page Print all pages in this productPrint All RSS Feed RSS Bookmark and Share Search this Product

STABILITY AND CHANGE IN THE ASGS

This 2016 edition of the Australian Statistical Geography Standard (ASGS) is the first update to the original edition released in 2011. This chapter describes the degree and manner in which the Main Structure regions have changed in the five years since they were first released, as well as outlining other significant updates incorporated into the 2016 edition of the ASGS.

2016 ASGS STRUCTURES - SIGNIFICANT CHANGES

Significant changes introduced with the 2016 edition include the addition of a boundary and code for Australia, as well as the incorporation of Norfolk Island into the ASGS. The Non ABS structures are now primarily approximated using Mesh Blocks and this means that the boundaries for State Suburbs, Postal Areas, Natural Resource Management Regions and Australian Drainage Divisions are all approximated far more accurately than they were in 2011.

IMPROVED STABILITY FOR ASGS STATISTICAL AREAS

Separating the ABS defined statistical areas from other administrative areas, particularly Local Government Areas, was a key difference in the design of the ASGS compared to the previous Australian Standard Geographical Classification (ASGC). This was done to reduce the impact of Local Government Area changes on ABS designed statistical areas, allowing them to better meet the requirements of statistical datasets and have greater stability for improved comparison of statistical data over time.

Changes to the boundaries of ABS defined areas in the ASGS are still required to reflect changes occurring on the ground, such as new housing developments or transport infrastructure. However, these changes are carefully balanced against the strong desire to maintain stability and continuity with the previous versions of the ASGS. As a result of these changes there have been increases in the number of areas included in all the structures from Mesh Blocks up to Statistical Areas Level 4 (SA4s) as shown in the table below.

Table 1: Change in numbers of Statistical Areas of ASGS Main Structure from 2011 to 2016


ASGS Edition
Mesh Block
SA1
SA2
SA3
SA4
GCCSA

2011
347,627
54,805
2,214
351
106
34
2016
358,122
57,523
2,310
358
107
34
Increase between 2011 - 2016
10,495
2,718
96
7
1
0

The amount of change to the ASGS Main Structure statistical areas is small in comparison to the change that occurred in previous editions of the ASGC and this was a key reason for the implementation of the ASGS. This increased stability of statistical areas in the ASGS allows better comparison of data over time.

Increased stability in the ASGS is illustrated by Figure 1 below showing the percentage of Main Structure statistical areas and Mesh Blocks that have remained effectively unchanged between 2011 and 2016. This level of change has been measured using the 2011 to 2016 population based geographic correspondences, it does not take into account changes that impact area with no population. The larger the region the less likely it is to change as it is better able to absorb changes in population while staying within the defined criteria. The stability of the ASGS is highlighted by comparison with the Local Government Areas. Between 2011 and 2016, 9.5% of Local Government Areas changed, this is more than twice the level of change when compared to similarly sized ASGS Main Structure areas such as SA2s (4.5% change) and SA3s (3.1% change).

Figure 1: Percentage of 2011 ASGS Statistical Areas remaining effectively unchanged in the 2016 ASGS

Graph showing that 88% of Mesh Blocks, 94% of SA1s, 95.5% of SA2s, 96.9% of SA3s, 99.1% of SA4s and 100% of GCCSAs are effectively unchanged. Compared to 90.5% of LGAs.

REASONS FOR CHANGE IN ASGS STATISTICAL AREAS

Although stability in the ABS defined statistical areas is important, it is also necessary to change some areas to continue to meet the requirements of the statistical data that is published on them or to better reflect the changing settlement patterns they represent. These changes can occur in three main ways; splits, amalgamations and redesign. The nature and extent of these changes for the main structure is shown in the chart below.

Figure 2: Types of change to ASGS Main Structure Areas (Percentage of 2011 Areas changed)

Graph showing differences in the level of different types of changes for the different 2011 statistical areas.  The types of change identified are Redesign, Aggregations and Splits of 2011 Areas,
Where possible, changes to 2011 areas are made by splitting the areas into two or more 2016 areas. This minimises the extent of change to boundaries and allows the 2016 areas to be easily grouped together for comparison with the 2011 areas. Figure 2 highlights that for SA2s, SA3s and SA4s the majority of changes have been made through splits. For the smaller more detailed areas such as SA1s and Mesh Blocks the narrower population criteria means that changes by splits are not always possible and redesign changes are more likely to be required to manage changes to settlement patterns and infrastructure in these areas. For Mesh Blocks in particular, a large number of amalgamations have occurred. Many of these amalgamations were to reduce the number of Mesh Blocks covering zero population areas where these areas were excessively split in the 2011 edition of Mesh Blocks. To better illustrate why these changes occur a number of examples are shown below.

Figure 3: Image showing an example of SA1 splits to reflect growth

Image on left shows a large 2011 SA1 with housing developments within it. Image on the right shows 2016 SA1s that have been split out of the 2011 SA1 to capture the developmental growth appropriately.
Image 2017 DigitalGlobe

Where possible the boundaries are split to accommodate growth. In the example in Figure 3, the 2011 area had only started to show signs of development and was covered by one SA1 which enabled data to be released on the small population in the area at that time. However, in 2016 the whole area has been developed and the 2011 SA1 has been split into 13 new SA1s. This allows users to analyse data in greater spatial detail. The original 2011 SA1 boundary can still be seen, meaning surrounding SA1s are not impacted and the 2016 data can be compared with that from 2011.

Figure 4: Image showing an example of SA2 amalgamations to meet statistical criteria

Image on left shows two 2011 SA2s that are low in population for an effective SA2. Image on the right shows the 2 SA2s have been combined together to form a large 2016 SA2 with an appropriate population size.
Image 2017 DigitalGlobe

The original design of the ASGS statistical areas relied on data such as the Geocoded National Address File (G-NAF) and the cadastral (property) boundaries, which generally reflected settlement patterns. In a very small number of cases, the 2011 areas did not have the populations that were predicted by these data during the original design phase, which meant they did not meet the population criteria set for them. In the example in Figure 4, the 2011 SA2s on the left (Anketell – Wandi and Casuarina – Wellard (East)) have been combined to form the 2016 SA2 of Anketell – Wandi, shown on the right. This change ensures that the 2016 SA2 has a population over 3,000 people, which enables more accurate Estimated Resident Population (ERP) data to be produced for this area.

Figure 5: Image showing an example of Mesh Block amalgamations

Image on left showing 2011 Mesh Blocks that were inappropriately automatically created. The image on the right shows the Mesh Blocks have been combined to capture the whole area in only 1 Mesh Block for 2016.
Image 2017 DigitalGlobe

Many of the Mesh Block amalgamations have been done to correct issues with the automated process originally used to create Mesh Blocks. When Mesh Blocks were first created, G-NAF and cadastre were used to automatically create the areas. The combination of duplicate records in the early versions of the G-NAF and the lack of imagery in the early development of Mesh Blocks led to some Mesh Blocks representing non-existent settlements. As can be seen in the example in Figure 5, some of the 2011 Mesh Blocks include only one or zero dwellings. The relatively large number of amalgamations for Mesh Blocks reflects efforts to address issues like these. The level of change resulting from these issues will not occur across future editions of the ASGS Mesh Blocks.

Figure 6: Images showing an example of boundary redesign to account for new growth

Image on left shows a large 2011 SA1 that cuts through newly developed dwellings in a high growth area. Image on right shows 2016 SA1s that have been redesigned to follow new roads and cadastre and appropriately split to capture the new growth.
Image 2017 DigitalGlobe

When an area undergoes significant growth, as far as possible, the boundaries are split to accommodate the new growth. Occasionally a straight split is not appropriate when the original boundary is no longer sensible, particularly where the underlying growth has not occurred in the way that was anticipated. This is the case in the example in Figure 6, shown above, where the 2011 SA1 boundary is shown against the 2016 imagery (on the left) and cuts straight through dwellings. For 2016 the SA1 boundary was realigned to follow roads and other boundaries added to divide the growth into areas meeting the SA1 population criteria. Redesigning boundaries in this way in these high growth areas has minimal effect on the comparability of data over time as these areas generally contained little or no population in 2011.

CHANGES TO ASGS CODES

ASGS users should be aware that the codes and names that are associated with statistical areas can also change when an area changes. In addition, where changes occur in the larger Statistical Area regions this can result in changes to the code for the smaller areas contained within them, even if one of these smaller areas has not changed itself. This occurs because the ABS Structures within the ASGS have a hierarchical coding system, meaning the smaller areas carry the codes associated with the larger areas. For example an SA2 may not have changed but if it is within an SA3 that has changed, then both the SA3 and SA2 code will have changed along with all other SA2 and SA1 codes within that SA3.

For more information on ASGS code changes please refer to Appendix 2: ASGS Coding and Labelling Changes.

UNDERSTANDING CHANGE IN THE ASGS

The ABS provides several resources to help users understand these changes to boundaries and codes and to manage data across different ASGS editions and other geographies.

Geographic correspondences document the overlaps between two sets of regions. They provide a percentage for data allocation between the overlaps that are based on estimates of the population distribution with the two regions. Geographic correspondences between 2011 and 2016 editions of statistical areas are available in the downloads tab of the relevant ASGS publication. These can be used to understand the relationships between 2011 and 2016 areas and to convert data from 2011 ASGS regions to 2016 regions.

ABS Maps is an online mapping tool that allows users to visually compare two different sets of boundaries across different editions of the ASGS. This allows users to examine the nature and extent of individual boundary changes in their area of interest.