Show download pdf controls
  • Methodology

    We selected a model-based bottom-up methodology, employing channel analysis and micro-analytical techniques in order to estimate the petroleum resource rent tax (PRRT) gap.

    We selected this method given the lack of suitable external data to produce a top-down estimate and the size of the taxpayer base. These estimates are produced using assumptions based on operational intelligence and subject matter expertise.

    Find out about:

    See also:

    Selecting the methodology

    The selection of an estimate methodology depends heavily on the:

    • design of the tax system
    • characteristics of the population
    • data available.

    A number of options were researched and considered in order to determine the most appropriate method for estimation of the PRRT gap.

    Given that no external data sources are available, a top-down approach was not available. Further, a random enquiry program was not considered feasible given the size of the taxpayer base, and the large variance between individual projects.

    Following consultation with our independent expert panel, we produced our estimate using a bottom-up illustrative approach. This approach draws together detailed examination of data sources along with expert judgment and knowledge, and is used to estimate the tax gap across the PRRT population.

    We do not observe black economy or related fraud and evasion activity in this market. Therefore, we have not made allowance for the impact of the black economy. We find that petroleum resource rent taxpayers lodge PRRT returns as required and pay the liabilities that are due.

    Applying the methodology

    The high level of coverage and intensive engagements with taxpayers were used to estimate the tax gap over the key PRRT risks, including the calculation of assessable receipts, and the deductibility and classification of expenditure.

    We followed three steps in applying the bottom-up methodology to estimate the gap. These are summarised in Figure 3 and then described in detail.

    Figure 3: Steps to estimate the PRRT gap

    Figure 3. This image is a pictorial representation of the three calculation steps outlined in the text following this image. This image provides the names of the steps only.

    Find out about:

    Step 1: Determine the scope of the PRRT population

    We use operational data to determine the size and scope of the PRRT population, and the amount of PRRT paid.

    Step 2: Estimate the PRRT gap

    We identify the key areas of risk which are relevant to the PRRT. We then use our operational data and subject matter expertise to estimate the impacts of the risks, resulting in corresponding risk rates. These risk rates are applied to the population as defined in Step 1, to produce an estimate of the PRRT gap.

    Step 3: Estimate the theoretical liability

    We add the tax gap calculated in Step 2 to the amount of PRRT paid (determined in Step 1) to estimate the theoretical liability. We then divide the tax gap by the theoretical liability to determine the tax gap as a percentage.

    Summary of estimation process

    The steps for the estimation process and the results for each year as a dollar amount and percentage are shown in Table 3.

    Table 3: Applying the methodology, PRRT gap, 2013–14 to 2016–17








    PRRT paid ($m)






    Tax gap ($m)






    Theoretical liability ($m)






    Tax gap (%)





    Figure 4 displays these figures as a percentage.

    Figure 4: PRRT gap as a percentage of total theoretical liability

    Figure 4. This graph is a pictorial representation of the net gap percentages stated in Table 3 for the years 2013–14 to 2016–17.


    The following limitations apply to the three core components of the model used to estimate the PRRT gap:

    • We assume all project participants are registered and lodge returns as required, and do not participate in the non-observed economy.
    • We assume expert judgment (informed by ATO outcomes and engagement activities) is a reliable indicator of levels of non-compliance and the impact of law interpretation risks to assessable receipts, applied deductible expenditure and exploration expenditure, and transferred exploration expenditure in respect of lodged returns.
    • The extent of non-detection is unknown and is extremely challenging to measure. We use a figure based on operational data to estimate the impact of non-detection.
    • We assume final ATO adjustments represent the correct outcome at law.

    Accounting for non-detection in the gap

    Not all errors are detected through audit and assurance activity, so we account for this by applying a non-detection uplift to the unreported tax estimate.

    We apply different uplift rates depending on the level of assurance we have over the tax reported in each return. Where we have a high level of assurance, we apply a lower uplift rate to account for the confidence we have in that particular return. We apply a higher uplift to returns that we have not reviewed in detail.

    Updates and revisions to previous estimates

    Figure 5 displays the tax gap from our current model (2019) compared with the previous estimate (2018).

    Figure 5: Current and previous PRRT gap estimates, 2013–14 to 2016–17

    Figure 5. This graph provides a visual representation of the previous and current net gap estimates provided at Table 4.

    The PRRT gap estimates have been revised, which is standard practice in gap estimates. The revisions occur as a result of using updated information, including updated risk rates. While the estimates for 2013–14 to 2015–16 have been revised upwards, the estimates are still similarly low across all years.

    Table 4: Summary of published PRRT net gap percentages

    Gap release year















    Data sources

    The methodology draws on the operational data of the ATO.


    An independent expert panel has assessed the ATO estimate of the PRRT gap. This is described in Principles and approaches to measuring gaps.

    Based on advice from the independent expert panel, the reliability rating for the PRRT gap estimate is high.

    As with all gap estimates, the PRRT gap estimate remains sensitive to assumptions made. We will continue to monitor and test the overall approach as we improve it in future releases.

    Figure 6: Reliability rating scale from very low to very high – PRRT gap

    This image is a graphical representation of the reliability rating for the current large super funds income tax gap estimate. It graphically represents a rating of high, which is a score between twenty one and twenty five. The maximum possible score is thirty.

    See also:

    Return to:

      Last modified: 17 Oct 2019QC 57597