Impact of lockdowns on household consumption - insights from alternative data sources

Cite
Print
Released
1/12/2021
Source
Australian National Accounts: National Income, Expenditure and Product, September 2021

Since the onset of the COVID-19 pandemic, lockdowns have been part of the Australian response to managing the health crisis. These have been introduced at various times to reduce transmission of the virus. Lockdowns have varied in nature, but generally included:

  • “Stay at home” orders where households were only able to leave their homes for essential reasons e.g. undertaking essential work and grocery shopping.
  • Closure or restricted activity of ‘non-essential’ businesses e.g. retail and hospitality.
  • Work from home arrangements for offices and a move to online education.
  • Other health measures such as social distancing, density limits and mask mandates.

Using supermarket scanner and bank transaction data, this spotlight provides additional insights into the impact of lockdowns on Household Consumption. 

Figure 1 shows a timeline of Australian lockdowns. Restrictions commenced with the “National lockdown” in March 2020, which lasted 6 weeks, followed by Victoria’s 16 week lockdown commencing in July 2020. From late 2020, the health response shifted towards shorter and more targeted lockdowns by State and Territory Governments to manage outbreaks within their respective jurisdictions. More recently, New South Wales (NSW), Victoria (VIC) and the Australian Capital Territory (ACT) have had extended lockdowns following the outbreak of the Delta variant in June 2021.  

Figure 1: Timeline of COVID-19 lockdowns in Australia

An image showing COVID-19 lockdown dates between March 2020 and December 2021.
This image presents a timeline of various lockdowns in Australia since the start of the COVID-19 pandemic. On 11 March, 2020, the World Health Organisation declared COVID-19 a world pandemic and Australia implemented a nationwide lockdown on 23 March, with different restrictions in place. These restrictions began to ease across the country between late April and May 2020. On 8 July, Victoria started a lockdown with Stage 3 "Stay at Home" restrictions reinstated and it did not end until 27 October. South Australia introduced a circuit breaker lockdown between 19 and 22 November, 2020. From January 2021 onwards, a series of state and territory- based lockdowns took place, including: 8-11 January Queensland lockdown 31 January - 5 February Western Australia lockdown 12-17 February Victoria lockdown 29 March - 1 April Queensland lockdown 24-27 April Western Australia lockdown 27 May - 10 June Victoria lockdown 26 June - 2 July Northern Territory lockdown The first positive case of the Delta variant was detected in Australia on 16 June, and further lockdowns were in place to reduce the possible transmission. These include: 26 June - 11 October New South Wales lockdown 12 August - 14 October Australian Capital Territory lockdown 5 August - 21 October the 6th lockdown in Victoria 29 June - 3 July Western Australia lockdown 29 June - 2 July Queensland lockdown 20-28 July South Australia lockdown 31 July - 11 August Queensland lockdown 15-27 July Victoria lockdown 8-11 August Queensland lockdown 16-20 August Northern Territory lockdown 15-18 October Tasmania lockdown 5-9 November Northern Territory lockdown 15 November - Northern Territory lockdown (currently ongoing)

The impact of lockdowns has been evident across a range of key economic statistics, particularly in Household Consumption. The National lockdown in the June quarter 2020, saw Household Final Consumption Expenditure (HFCE) decline by 12.5%; and in the September quarter 2021, HFCE declined by 4.8% with NSW, VIC and ACT in lockdown for substantial parts of the quarter (figure 2).

Insights from bank transactions data

In 2020, the ABS secured access to de-identified aggregated bank transactions data. Experimental bank data analysis indicates broad alignment with published National Accounts estimates. The bank transactions dataset has provided enormous value as a confrontation source and for quality assurance purposes. The timeliness, frequency and granularity of the dataset enabled additional insights into household consumption changes.

The analysis of bank data shows that during extended periods of lockdown, expenditure in discretionary categories declined significantly and there are noticeable impacts to total household consumption. A good illustration of this is with the Victoria July 2020 lockdown where a range of discretionary categories declined and remained low through the lockdown period (see figure 3). 

Index reference period: 31 May 2020 = 100.0

The shorter and sharper lockdowns implemented from late 2020 had markedly different impacts on household consumption. The weekly aggregated bank data showed expenditure in discretionary consumption categories declined when lockdowns were introduced but rebounded almost immediately when the lockdown ended. South Australia’s 3 day lockdown in November 2020 (figure 4) and Western Australia’s 5 day lockdown in January 2021 (figure 5) both demonstrate this rebounding behaviour. The same pattern is observable across short lockdowns introduced in other jurisdictions during the pandemic. From an aggregate perspective, analysis of household consumption indicates these short lockdowns had minor impacts. HFCE in South Australia rose 0.9% in December quarter 2020, despite a 3 day lockdown, while HFCE in Western Australia increased 0.3% in March quarter 2021 following its 5 day lockdown.

Index reference period: 1 November 2020 = 100.0

Index reference period: 17 January 2021 = 100.0

Insights from supermarket scanner data

The stockpiling of essential goods more commonly referred to as “panic buying”, was a regular occurrence throughout the pandemic in Australia. Weekly supermarket scanner data showed sales increasing dramatically when lockdowns were announced (figure 6). This trend has been evident in almost all lockdowns since the start of the pandemic. The supermarket scanner data showed that the magnitude of panic buying declined over successive lockdowns. This also meant that over time, the surge in consumption due to panic buying was no longer providing the same level of offset to falls observed in other categories of consumption during lockdowns.

Index reference period: 6 January 2020 = 100.0

Future uses of alternative data

The above analysis illustrates the enormous value that can be derived from both the supermarket scanner and bank transactions data. Supermarket scanner data has already been used extensively in the Consumer Price Index. The September quarter 2021 National Accounts publication includes a table with experimental estimates of HFCE food using this data. This can be found in the Data cubes section of Data downloads.

Development of a monthly household spending indicator using banks transactions data is underway with an update on this work due to be published in February 2022¹. There are also plans to explore opportunities to utilise bank data to directly compile parts of HFCE².

Footnotes

1. For more information see: Using bank transactions data to measure household spending.

2. For more information see: Enhancing household consumption measures in the National Accounts.