







































AUSTRALIAN POPULATION STUDIES  2023 | Volume 7 | Issue 1 | pages 30-33 

 

© Grossman 2023. Published under the Creative Commons Attribution-NonCommercial licence 3.0 Australia (CC BY-NC 3.0 
AU). Journal website: www.australianpopulationstudies.org 

Visualising the impact of COVID-19 
on Australian SA3 populations 

Irina Grossman  The University of Melbourne 

 

Email: irina.grossman@unimelb.edu.au. Address: Melbourne School of Population and 

Global Health, Level 5, 207 Bouverie St, The University of Melbourne, Victoria 3010 Australia. 

 

Received 12 October 2022; accepted 6 March 2023; published 6 December 2023 

 

Introduction 

The COVID-19 pandemic, and the related health measures implemented to control its spread 

resulted in significant changes in Australian international and internal migration. For example, net 

overseas migration was 192,700 in 2019-20, but decreased to -84,940 in 2020-21 (ABS 2022). It is 

difficult to understand how this population decrease was distributed at the smaller area level, 

particularly as there were also changes in historic internal migration trends, including an increase in 

migration to regional areas from capital cities (Stephens et al., 2022). The key aim of this data 

visualisation is to investigate how COVID-19 impacted populations at the SA3 level by comparing 

2021 population estimates with possible populations if COVID-19 had not occurred. 

Data and methods 

Possible populations for a No-Covid Scenario were created for SA3 areas using Estimated Resident 

Populations (ERPs) from 2016 and 2019, which are found in ‘Table 2. Estimated resident population, 

Statistical Areas Level 3 (ASGS 2021), Australia’, of the ‘Population estimates by SA2 and above (ASGS 

2021), 2001 to 2021’ Data cube (ABS 2021a). The No-Covid Scenario populations were created by 

applying the average growth rate from 2016 to 2019 to the 2019 data in order to project the SA3 

populations to 2021. 

To investigate the potential impact of COVID-19 on SA3 populations the absolute percentage 

differences between the 2021 SA3 populations of the No-Covid Scenario and the 2021 SA3 ERPs were 

calculated. These differences were then visualised on a mosaic cartogram of Australian SA3 areas. 

Mosaic cartograms are maps where each area is represented by a simple shape (Cano et al., 2015). 

The maps presented here represent each SA3 area as an equally sized element, regardless of its 

physical size or the number of persons resident. The mosaic cartograms are presented layered on top 

of a standard map of Australian states and territories; the State and Territories – 2021 GDA94 ASGS 

Edition 3 digital boundary files were used to produce this map (Australian Bureau of Statistics, 

2021b).  

D
e

m
o

G
ra

p
h

ic
 

 

http://www.australianpopulationstudies.org/
mailto:irina.grossman@unimelb.edu.au


Australian Population Studies 7 (1) 2023  Grossman 31 

 

 

 

Figure 1: A mosaic cartogram visualising the impact of COVID-19 on 2021 Australian SA3 populations 

Source: SA3 ERPs were sourced from ABS (2021a). Digital boundaries of SA3 areas and of states and territories were 
obtained from ABS (2021b). The maps were created using the ONS Hexmap Automation tool (Mitchell & Tzelepis, 2017). 

Note: The percentage difference between the 2021 SA3 ERPs and the populations in the No-Covid Scenario is presented in 

the top map. For reference, the state or territory that each element belongs to is presented in the bottom map and 

selected SA3 areas in capital cities are labelled. 



32 Grossman  Australian Population Studies 7 (1) 2023 

 

To create the mosaic cartograms the 2021 SA3 GDA94 ASGS Edition 3 digital boundary files were 

downloaded (ABS 2021b). The boundary files were transformed into a mosaic cartogram shapefile 

using the ‘basic’ variant of the ONS Hexmap Automation tool (Mitchell & Tzelepis, 2017). Minor edits 

were made to the shapefile such that SA3 areas from different states were clustered together. The 

‘ggplot’ (Wickham, 2016) and ‘sf’ (Pebesma et al., 2018) R packages were used to create two maps. 

The minor edits were completed using Adobe Illustrator 2023 and Adobe Photoshop 2023. The first 

map (Figure 1, top map) visualises the absolute percentage change between the projected and 

estimated 2021 SA3 populations on a mosaic cartogram of Australian SA3 areas. The second map 

(Figure 1, bottom map) visualises which state or territory each element of the mosaic cartogram 

belongs to, and selected SA3 areas in capital cities are labelled. 

Key features 

The top map in Figure 1 visualises the 2021 changes in SA3 populations had COVID-19 not occurred 

and previous population trends continued. For reference, the bottom map of Figure 1 shows which 

states and territories each of the elements in the mosaic cartogram belongs to. The top map reveals 

that COVID-19 may have resulted in decreased populations in most SA3 areas. Nationally there was a 

1.9% reduction when compared to 2021 ERPs and the populations in the No-Covid Scenario; this 

corresponded to a potential national decrease of 478,314 persons. However, the impact varied 

between states and territories. According to the No-Covid scenario, Western Australia’s population 

was 10,074 persons, or 0.4%, greater than what could have been expected had pre-COVID-19 

population trends continued. This was the only state or territory with an increased population in the 

No-Covid Scenario relative to the 2021 ERPs. All other states saw reduced populations, particularly 

Victoria and the Australian Capital Territory which both had decreases of 3.7%, corresponding to 

243,144 and 16,622 persons less, respectively. 81.7% of SA3s in greater capital city regions had 

decreased populations, amounting to a reduction of 453,904 persons in the capital cities of Australia. 

Whilst regional populations were not as affected by COVID-19, they generally also saw population 

declines with 60.4% of SA3s outside of the capital city regions showing population decreases, 

resulting in a total of 24,410 fewer persons than what may have been expected in non-capital city 

regions in the absence of COVID-19.  

It is important to note that whilst short term projections, such as the one used here to create the No-

Covid Scenario, tend to be reasonably accurate, there are many factors that may impact SA3 

populations other than COVID-19, such as housing developments and local job opportunities. It is 

important to emphasise that the No-Covid Scenario presented here is just one of many potential 

scenarios. Furthermore, it is important to consider that the process of creating our mosaic 

cartograms moved the relative positions of the SA3 areas. Therefore, the SA3 areas which are in the 

same SA4, or the same Greater Capital City Statistical Area, did not always cluster together. This 

should be considered when interpreting the maps. 

Acknowledgements 

I am grateful to Tom Wilson who kindly provided helpful comments on earlier versions of this 

demographic. However, all errors and omissions are the responsibility of the author. 



Australian Population Studies 7 (1) 2023  Grossman 33 

 

References 

ABS (2021a). Regional population. https://www.abs.gov.au/statistics/people/population/regional-

population/latest-release  

ABS (2021b). Digital boundary files Australian Statistical Geography Standard (ASGS) Edition 3. 

https://www.abs.gov.au/statistics/standards/australian-statistical-geography-standard-asgs-

edition-3/jul2021-jun2026/access-and-downloads/digital-boundary-files  

ABS (2022). Overseas Migration, 2021-22. 

https://www.abs.gov.au/statistics/people/population/overseas-migration/2021-22-financial-

year 

Cano, R. G., Buchin, K., Castermans, T., Pieterse, A., Sonke, W., & Speckmann, B. (2015). Mosaic 

drawings and cartograms. Computer Graphics Forum, 34(3), 361-370. 

https://doi.org/10.1111/cgf.12648  

Mitchell, B., & Tzelepis, G. (2017). Automated generation of equal area cartograms (‘HEXMAPS’) at 

any scale. https://github.com/ONSgeo/HexMapping  

Pebesma, E. J. (2018). Simple features for R: standardized support for spatial vector data. The R 

Journal, 10(1), 439-446. https://doi.org/10.32614/RJ-2018-009  

Stephens, R., Cansdale, D., and Forbes L (2022). Increase in internal migration from capital cities to 

regional areas. https://www.abc.net.au/news/2022-02-18/migration-from-cities-to-regional-

areas-doubles-during-covid/100839386  

Wickham H (2016). ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. 

https://ggplot2.tidyverse.org  

 

https://www.abs.gov.au/statistics/people/population/regional-population/latest-release
https://www.abs.gov.au/statistics/people/population/regional-population/latest-release
https://www.abs.gov.au/statistics/standards/australian-statistical-geography-standard-asgs-edition-3/jul2021-jun2026/access-and-downloads/digital-boundary-files
https://www.abs.gov.au/statistics/standards/australian-statistical-geography-standard-asgs-edition-3/jul2021-jun2026/access-and-downloads/digital-boundary-files
https://www.abs.gov.au/statistics/people/population/overseas-migration/2021-22-financial-year
https://www.abs.gov.au/statistics/people/population/overseas-migration/2021-22-financial-year
https://doi.org/10.1111/cgf.12648
https://github.com/ONSgeo/HexMapping
https://doi.org/10.32614/RJ-2018-009
https://www.abc.net.au/news/2022-02-18/migration-from-cities-to-regional-areas-doubles-during-covid/100839386
https://www.abc.net.au/news/2022-02-18/migration-from-cities-to-regional-areas-doubles-during-covid/100839386
https://ggplot2.tidyverse.org/

