Small Area Dwelling Stock Removals

Realised demolition approvals represent demolitions as they occur following an approval from a local council or other approving authority. They differ from demolition approvals as there is a lag between approval and actual demolition taking place, and because not all approvals go ahead. A dwelling demolition is defined as the complete and intentional dismantling of a dwelling, such that none of the structure remains on site. 

Estimates of realised demolitions have been calculated using a survival model based on the demolition outcomes of a sample of Victorian demolition approvals, which were investigated using aerial imagery. 
 

Demolition approvals are sourced from Building Approvals, Australia. Demolition outcomes (i.e. whether the demolition occurred and the demolition date) for a sample of Victorian demolition approvals were ascertained using aerial imagery from Nearmap Australia Pty Ltd. The sample consisted of approximately 12% of demolition approvals (893 dwellings) within the 2017-18 financial year from across Victoria. This period was chosen to allow enough time following approval for the demolition to have occurred.

The date of the first image where visual evidence of a demolition could be identified was recorded as the demolition date. It should be noted that frequency of Nearmap image captures will affect the accuracy of the demolition date. Nearmap image updates are prioritised for major Australian metropolitan areas and occur more frequently than in regional areas. 
 

For each approval in the sample, the time from approval to demolition was calculated in days. The average time between approval and demolition was 84 days in Greater Melbourne and 137 days in the Rest of Victoria.

The survival function for a dwelling surviving (i.e. not yet being demolished) was first estimated using a non-parametric (Kaplan-Meier) estimator. Based on this, the probability of a dwelling being demolished before 200 days after approval was estimated at 85%.

The survival model was next estimated using a parametric (log-normal) distribution as this was the best fit to the distribution of demolition outcomes based on the survival analysis. The survival model uses the parameters from this distribution to calculate the probability of a demolition occurring in a particular quarter. 

Adjustments were made to the survival model to account for the proportion of approvals where:

• demolition did not occur;
• demolition occurred prior to the approval date; or
• demolition occurred in excess of two years after the approval date.

As the vast majority of demolitions occurred within six quarters (546 days) after approval, this was used as the cut-off point after which demolition approvals cease to contribute to realised demolition estimates. 

To derive the number of realised demolitions, the probability of a demolition occurring was calculated for each demolition approval six quarters prior to the reference period and summed by quarter.

The graph below depicts demolition approvals and realised demolitions for the period between the September 2016 quarter and the June 2021 quarter. Realised demolitions over time are lower than demolition approvals due to approvals which do not go ahead. 
 

Unplanned stock losses refer to dwellings that are unintentionally and permanently destroyed (i.e. they cannot be repaired/restored). This includes dwellings destroyed in natural disasters such as bushfires, floods and cyclones. Temporary removals from stock where dwellings may be uninhabitable for a period of time following a natural disaster are not currently in scope.

Counts of destroyed dwellings are not currently available in a nationally consistent and timely manner. Dwelling losses from major disaster events are incorporated into estimates of dwelling removals where data is available from relevant Government authorities. The availability and quality of this data may improve in future, as data consolidation around disaster management and resilience progresses.

Estimates of unplanned stock losses for the period from the September 2016 quarter to the June 2021 quarter have been compiled based on currently available data from relevant Government authorities, which cover the 2019-2020 bushfire season only. Unplanned stock losses that occurred in Victoria as a result of this event were allocated to the March 2020 quarter.

The ABS is continuing to investigate alternatives to estimate unplanned stock losses where direct counts are not available (i.e. based on the location of dwellings and hazard exposure information).
 

Counts of dwellings destroyed by Local Government Area (LGA) and bushfire boundaries for the 2019-2020 bushfire season have been sourced from the National Recovery and Resilience Agency. Total dwelling counts by mesh block have been sourced from the Census of Population and Housing, 2016.

The LGA counts of dwellings destroyed were allocated to SA2s based on the proportion of exposed dwellings in each mesh block within an LGA.

The number of exposed dwellings in each mesh block and LGA was calculated by multiplying the proportion of the area that was affected by bushfires (based on the bushfire boundaries) by the 2016 Census dwelling count. This was used to calculate the proportion of exposed dwellings in each mesh block within an LGA, which was multiplied by the number of destroyed dwellings in the LGA. The estimate of dwellings destroyed in each mesh block was then aggregated to the corresponding SA2.
 

ABS building statistics (building approvals, building activity and the dwelling stock series currently under development) classify buildings according to the Functional Classification of Buildings. This defines a dwelling as a self-contained suite of rooms intended for long-term residential use, including cooking and bathing facilities. Regardless of whether they are self-contained or not, rooms within buildings offering institutional care (e.g. hospitals) or temporary accommodation (e.g. motels, hostels and holiday apartments) are not defined as dwellings. This differs from the definition used by the Census (a structure which is intended to have people live in it, and which is habitable on Census night), which can include non-permanent structures such as caravans, houseboats and tents, as well as communal or transitory accommodation such as cabins, hotels, prisons and hospitals.