We use a 5-step top-down method to estimate the PAYG withholding gap estimate.
Step 1: Estimate salary and wages subject to withholding
We start with external data on salary and wages published by the ABS as part of its annual Australian National Accounts. We adjust that data in the following way:
- We apply industry-specific uplift factors averaging 2.3% to account for our best internal estimate of hidden wages.
- We subtract benefits and payments not subject to withholding, including:
- payments in kind (fringe benefits)
- salary sacrifice into super
- employee share schemes
- defence exempt income.
- The resulting amounts are our best estimates of salary and wages subject to withholding for each industry.
Step 2: Calculate the theoretical PAYG withholding amount
In this step, we multiply the amounts calculated in step 1 by an average rate of withholding for each industry to calculate the withholding amount that should be reported and remitted to us.
We calculate the average rate of tax withheld from tax return data for each year. This includes:
- tax withheld from salary and wage income
- lump sum payments made by employers to their employees
- allowances over the total salary-and-wages-type income.
We then re-calculate the average withholding rate for each industry allowing for the uplift to average wages and consequent higher average withholding rate after recognising hidden wages. The estimated average withholding rate for each industry also recognises that hidden wages are likely to be more prevalent amongst individual taxpayers with low wages than those with higher wages.
Step 3: Calculate PAYG withholding reported to the ATO
In step 3, we calculate total PAYG withholding reported from our internal data on an accrual basis. Non-resident dividends, interest and royalty withholding tax amounts are excluded.
We subtract other non-salary amounts subject to withholding, including:
- personal services income for sole traders
- tax withheld where an Australian business number is not quoted
- voluntary agreements
- benefits and pensions.
We are then left with PAYG withholding reported minus excluded amounts.
Step 4: Calculate the PAYG withholding net gap and gross gap
In step 4, we subtract PAYG withholding reported (calculated in step 3) from the theoretical PAYG withholding amount (calculated in step 2). We then add non-pursuable debt amounts to determine the net gap. Non-pursuable debt is debt deemed irrecoverable at law or not economical to pursue.
Finally, we add compliance outcomes and employer self-adjustments to the PAYG withholding net gap to determine the gross gap.
Then we report net and gross gaps as a share of theoretical withholding.
Summary of the estimation process
Table 2 provides a summary of each step of the estimation process and the results for each year.
Step | Description | 2015–16 | 2016–17 | 2017–18 | 2018–19 | 2019–20 | 2020–21 |
---|---|---|---|---|---|---|---|
1.1 | Salary and wages ($m) | 725,508 | 743,608 | 780,758 | 822,053 | 853,918 | 887,096 |
1.2 | Shadow economy uplift (2.3%) ($m) | 16,600 | 17,053 | 17,972 | 18,906 | 19,534 | 20,236 |
1.3 | Earnings not subject to withholding ($m) | 29,024 | 25,893 | 26,539 | 30,190 | 31,007 | 30,179 |
1.4 | Salary and wages subject to withholding (A) ($m) | 713,084 | 734,768 | 772,191 | 810,769 | 842,445 | 877,153 |
2.1 | Average rate of withholding (B) | 25.1% | 25.2% | 25.5% | 25.8% | 26.0% | 25.4% |
2.2 | Multiply (A) and (B) to calculate theoretical liability (C) ($m) | 179,228 | 185,330 | 196,911 | 209,049 | 219,162 | 222,861 |
3.1 | Total PAYG withholding reported ($m) | 174,238 | 180,429 | 193,389 | 206,398 | 216,771 | 222,166 |
3.2 | Subtract excluded amounts ($m) | 1,714 | 1,813 | 1,899 | 2,070 | 2,200 | 2,329 |
3.3 | Expected PAYG withholding collections (excluding excluded amounts) (D) ($m) | 172,524 | 178,616 | 191,490 | 204,329 | 214,571 | 219,837 |
4.1 | Non-pursuable debt (E) ($m) | 575 | 594 | 632 | 671 | 703 | 715 |
4.2 | (C) minus (D) plus (E) to equal net gap (F) ($m) | 7,279 | 7,308 | 6,052 | 5,391 | 5,294 | 3,738 |
4.3 | Add amendments to net gap (F) ($m) | 1,592 | 2,406 | 2,444 | 2,793 | 1,924 | 1,362 |
4.4 | Equals gross gap ($m) | 8,871 | 9,714 | 8,497 | 8,184 | 7,218 | 5,100 |
4.5 | Gross gap (%) | 4.9 | 5.2 | 4.3 | 3.9 | 3.3 | 2.3 |
4.6 | Net gap (%) | 4.1% | 3.9% | 3.1% | 2.6% | 2.4% | 1.7% |
Find out more about our overall methodology, data sources and analysis used for creating our tax gap estimates.
Limitations
We use ABS National Accounts data to estimate theoretical salary and wages amounts. Our data from tax returns does not apply the same definition as the series compiled by the ABS. While we make adjustments to improve consistency, there are some items in the ABS data series we cannot quantify.
There are limitations with this approach, which means there will be margins of error with the data, including:
- sampling and non-sampling errors
- National Accounts data are subject to revisions, which can be material and result in changes to the estimated gap.
The model is not heavily assumption-driven but is sensitive to some key assumptions, including the:
- estimate of the extent of hidden wages (which is now based on the findings of a comprehensive internal study)
- distribution of hidden wages, including across individual taxpayers at different income levels
- distribution of employees across various PAYG withholding scales to calculate average withholding rates, including the increase due to the inclusion of hidden wages.
Updates and revisions to previous estimates
Each year we refresh our estimates in line with our annual report. Changes from previously published estimates occur for a variety of reasons, including:
- improvements in methodology and assumptions
- revisions to data
- additional information becoming available.
Figure 2 displays the net gap estimates from our current model compared to the previous estimates.
Figure 2: Current and previous PAYG withholding gap estimates, 2008–09 to 2020–21
The data is presented in Table 3.
Table 3: Current and previous PAYG withholding net gap estimates (percentage), 2008–09 to 2020–21
Year published | 2008–09 | 2009–10 | 2010–11 | 2011–12 | 2012–13 | 2013–14 | 2014–15 | 2015–16 | 2016–17 | 2017–18 | 2018–19 | 2019–20 | 2020–21 |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
2023 | n/a | n/a | n/a | n/a | n/a | n/a | n/a | 4.1%. | 3.9% | 3.1% | 2.6% | 2.4% | 1.7% |
2022 | n/a | n/a | n/a | n/a | n/a | n/a | 3.9% | 3.6% | 3.5% | 2.6% | 1.9% | 1.5% | n/a |
2021 | n/a | n/a | n/a | n/a | n/a | 2.9% | 2.7% | 2.4% | 1.7% | 1.8% | 1.6% | n/a | n/a |
2020 | n/a | n/a | n/a | n/a | 3.0% | 2.9% | 2.7% | 2.3% | 1.7% | 1.8% | n/a | n/a | n/a |
2019 | n/a | n/a | n/a | 4.2% | 3.0% | 2.8% | 2.7% | 2.2% | 1.7% | n/a | n/a | n/a | n/a |
2018 | n/a | n/a | 4.3% | 4.1% | 2.9% | 2.7% | 2.6% | 1.9% | n/a | n/a | n/a | n/a | n/a |
2017 | n/a | 3.7% | 4.0% | 4.1% | 2.9% | 2.7% | 1.9% | n/a | n/a | n/a | n/a | n/a | n/a |
2016 | 2.5% | 2.5% | 2.9% | 3.0% | 2.1% | 1.4% | n/a | n/a | n/a | n/a | n/a | n/a | n/a |
2015 | 2.0% | 2.1% | 2.2% | 2.1% | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a | n/a |