Background

1. The Australian Taxation Office (ATO) administers the tax and superannuation systems that support and fund services to Australians. As part of its ongoing operations to ensure that all taxpayers pay their fair share of tax, the ATO undertakes risk and compliance activities. Most of these activities are undertaken as part of the ATO’s ongoing operations. The ATO is also often provided with substantial additional Budget funding to supplement ongoing compliance activities.¹ The supplementary funding enables the government to explicitly address specific compliance risks, notwithstanding that the ATO could also address the risks through its business-as-usual compliance activities.

2. Between 2010–11 and 2015–16, the Government announced 18 specific Budget-funded compliance measures, of which 16 measures had a start date of earlier than 1 July 2015.² These 16 compliance measures had a total commitment to raise liabilities of approximately $7.4 billion and collect approximately $8.0 billion³ at a cost of some $1.4 billion from 2010–11 to 2014–15. For some measures, the additional revenue included so-called ‘indirect revenue’, that is the sustained or ‘flow-on’ amounts of tax that are paid in later years as a result of the initial compliance activity.

Audit objective and criteria

3. The objective of the audit was to assess the effectiveness of the Australian Taxation Office in achieving revenue commitments established under specific Budget-funded compliance measures. To form a conclusion against this objective, the Australian National Audit Office adopted the following high-level audit criteria:

• the ATO has established sound processes and practices to support the development of specific Budget-funded compliance measures;
• the ATO effectively monitors the revenue raised and expenses incurred in implementing these compliance measures; and
• revenue commitments and other benefits from these compliance measures have been met.

Conclusion

4. It is unclear if the ATO achieved the revenue commitments made to government for all Budget-funded compliance measures over the period 2010–11 to 2014–15. The main shortcoming is the absence of an accurate measure of business-as-usual baseline revenue raised from comparable pre-existing compliance activities. The ATO also did not accurately monitor the additional expenses incurred in implementing compliance measures.

5. The lack of clear business-as-usual baseline levels to enable accurate attribution of revenue and expenses to compliance measures, along with inaccurate attribution undermines the assurance to government that the agreed amount of additional revenue is actually being raised and that the additional resources provided are used as agreed. The lack of an accurate baseline resulted in the significant overstatement of revenue attributed to one relatively large measure. It is therefore important that the ATO develops better ways of measuring additional revenue and expenses from its compliance measures, and improves checking of the accuracy of estimated revenue outcomes.

Supporting findings

Developing compliance measures

6. Each year, the ATO identifies and develops potential compliance measures, based on existing compliance risk management processes and consideration of a range of factors, including estimated revenue outcomes and additional resource requirements. Justification for the additional resource requirements is provided to the Department of Finance (Finance), and the Department of the Treasury (Treasury) examines the revenue estimates, prior to new policy proposals being considered by the Government. However, there is a lack of clarity in the new policy proposals about the compliance activity relating to the base funding, and therefore about the overall and additional levels of compliance activity following the implementation of the measure. Documenting in funding proposals how the additional revenue will be calculated and reported and the base or pre-existing level of activity would help to ensure that the committed additional level of compliance activity was occurring.

7. The robustness of the ATO’s revenue models could be improved by clearly specifying the basis of the assumptions used. Maintaining a register of assumptions (to provide a record of their derivation) would help to ensure that the assumptions are suitable, and would promote consistency in their application across models. Greater assurance of the accuracy of the models would be provided by requiring the relevant business lines to formally agree to the revenue estimates and assumptions used in the models. The ATO should also include penalties and interest in the revenue models, where possible, particularly when they are likely to be material. These charges have usually been excluded from the models, but included in revenue outcomes, even where they are relatively large.

8. In addition to direct revenue commitments, seven of the 16 measures announced by the Government and implemented between 2010−11 and 2014−15 included indirect revenue commitments. It can take the ATO a number of years to develop, test and approve methodologies to estimate the indirect revenue effects of new measures and, for some measures, indirect revenue estimates have not yet been determined. To provide a sound basis for Treasury to include indirect revenue in a proposed Budget measure, the ATO should clearly indicate how it will be determined and then reported in the year following receipt of the revenue.

Monitoring compliance measures

9. The ATO’s monitoring of revenue and expenses associated with compliance measures could be improved. It has not had methodologies in place to accurately calculate additional revenue and expenses for all measures when they are implemented and does not have a comprehensive set of performance indicators for both business-as-usual and measure-related compliance activity that covers revenue, operational information, estimated expenses and rates of return. The ATO could also improve checking of the accuracy of the calculation of revenue outcomes from compliance measures against the approved methodologies for their calculation. However, the ATO has put in place satisfactory arrangements for reporting to the states and territories on its compliance activities for the GST.

10. Closure reports have not been routinely prepared for closed measures. Closure reports were prepared for only three of six closed measures, none of which satisfactorily addressed whether revenue commitments had been met. Closure reports should be prepared for all measures, and include an evaluation of the additional revenue resulting from the measures and lessons learnt.

Meeting compliance commitments

11. It is unclear if the ATO has achieved the additional revenue commitments for all compliance measures implemented over the period 2010–11 to, as there were shortcomings in the methodologies underpinning its estimates for some specific measures. This is mainly because the ATO does not set a comparable baseline of business-as-usual compliance activity and related revenue that are expected to be maintained over the life of a measure. For many measures revenue is calculated using a percentage of total compliance revenue and this percentage is not always applied consistently. Baseline and measure data can also be affected by large and abnormal cases, and there is a need for a consistent treatment that recognises where revenue would be expected to result from the ATO’s ongoing business activities.

12. The ATO has indicated that for some measures it would be difficult to identify the base level of revenue, and requiring it to maintain a base level of compliance activity under a measure would limit its ability to reallocate resources to realise higher revenue in other areas. This suggests the need for the ATO to seek funding for additional compliance activity on a different basis and to measure the additional revenue against a reliable overall base level of compliance activity and revenue. However, the base levels of revenue that the ATO has been using to estimate aggregate level compliance revenue (that is, for all compliance activity) have been based on an unreliable estimate of business-as-usual revenue—a flat ‘base’ level of revenue for income tax and GST, which has not been adjusted since 2010–11. This baseline did not account for important factors, such as increases in the Budget forward estimates, the impact of one-off changes in Budget funding to the ATO and changed outcomes from the ATO’s business-as-usual compliance activities.

13. There is a need for greater timeliness in the reporting of indirect revenue, which should be reported in the closure reports of measures.

14. The projected rates of return on the investments made in the compliance measures since 2010–11 have varied considerably—ranging from $3.29 to $11.76 for every dollar invested, based on liabilities expected to be raised. For most measures examined, the ATO’s revenue outcome figures indicated that higher rates of return had been achieved than projected. However, rates of return of measures would be more meaningful if based on additional cash collected, a consistent approach to treating abnormal compliance activities and the amount of Budget funding of departmental and capital expenses.

Recommendations

Summary of Australian Taxation Office response

15. The ATO’s summary response to the report is provided below. The ATO also provided a covering letter (Appendix 1) and response to recommendations, and more detailed comments against the audit conclusion and findings (Appendix 2).

ATO response

The ATO thanks the ANAO for the many opportunities to comment on the development of this report, however has been unable to agree fully with the recommendations. This is because the ATO does not agree that it is unclear whether the ATO met its revenue ‘commitments’, arising from compliance measures over the period 2010–11 to 2014–15.

Nevertheless, the ATO does agree that it could have better documented assumptions for estimating revenue. The ATO has agreed to provide more information in funding proposals about levels of pre-existing activity and ‘indirect’ revenue, where relevant.

Similarly, the ATO acknowledges that there are some situations where the reporting of the outcomes of compliance measures could be made more consistent or improved. The ATO will not depart from the case-by-case approach it takes to the reporting of abnormal outcomes, but will prepare guidance on this topic.

Introduction

1.1 The Australian Taxation Office (ATO) administers the tax and superannuation systems that support and fund services to Australians. Its responsibilities include: collecting revenue; administering the Goods and Services Tax (GST) on behalf of the Australian states and territories; managing a range of programs that result in transfers and benefits back to the community; regulating major aspects of Australia’s superannuation system; and custodianship of the Australian Business Register. In 2014–15 gross tax collections by the ATO totalled $432.3 billion ($336.8 billion in net tax collections).⁴

1.2 The ATO’s compliance model (Figure 1.1) shows a continuum of taxpayer attitudes towards compliance. At the base of the continuum, taxpayers have the desired attitude of being ‘willing to do the right thing’. At the other extreme are taxpayers who have decided not to comply. The ATO promotes compliance with the taxation system primarily by encouraging and helping people to do the right thing. Nonetheless, there is still a need to target those who have chosen not to comply.

Figure 1.1: ATO compliance model

Source: ATO.

1.3 Most risk and compliance activities are undertaken as part of the ATO’s ongoing operations. The ATO is also often provided with additional funding in the Budget, Mid-Year Economic and Fiscal Outlook (MYEFO) Statement or other economic statement⁵ to address risks to the operation of Australia’s tax and superannuation systems. Many of these Budget measures focus on compliance activities (such as taxation audits) directed at those who engage in higher risk tax behaviour, and are intended to supplement other compliance activities undertaken by the ATO.

1.4 Between and 2015–16, the Government has announced 18 specific Budget-funded compliance measures, of which 16 measures had a start date of earlier than 1 July 2015.⁶ These compliance measures⁷ are listed at Appendix 3. These 16 compliance measures had a total commitment to raise liabilities of approximately $7.4 billion and collect approximately $8.0 billion⁸ at a cost of some $1.4 billion from 2010–11 to 2014–15. For some measures, the additional revenue included ‘indirect revenue’, that is the sustained or ‘flow-on’ amounts of tax that are paid in later years as a result of the initial compliance activity.

1.5 While the ATO has made a commitment to collect the additional revenue from compliance measures, with the exception of the GST measures for which annual GST Administration Performance Reports are prepared in accordance with the GST Administration Performance Agreement between the ATO and the Council on Federal Financial Relations⁹, there is limited external reporting or scrutiny of the extent to which these targets are met each year. For instance, the ATO does not provide in its annual report a statement on the base and additional revenue flowing from the measures.

1.6 At an aggregate level, the ATO has reported that taxation liabilities from compliance activities (both business-as-usual and Budget measure activities) have grown from approximately $7.0 billion in 2006–07 to approximately $13.5 billion in 2014–15. ATO internal records indicate that in 2014–15 around 55 per cent of income tax and GST liabilities were measure-related.¹⁰ In the same year, funding for Budget measures represented around 31 per cent of funding of the Client Engagement Group (formerly the Compliance Group), which undertakes most of the ATO’s compliance activities.

1.7 Between 2010–11 and 2013–14, the annual funding for the Group was approximately $1.1 billion. In 2014–15 and funding fell to $1.0 billion and $0.94 billion respectively (see Figure 1.2). This was mainly due to the ATO reducing internal ‘base’ funding allocations for compliance activities by $164 million in.¹¹ This reduced base compliance funding was made in the context of lower overall funding for the ATO and placing greater emphasis on arrangements to support the willing participation of taxpayers. There was no forecast of reduced revenue being collected by the ATO associated with the reduced base funding for compliance activities.

Figure 1.2: Client Engagement Group (formerly Compliance Group) sources of funding, 2009–10 to 2015–16

Note: GST measure funding is provided by the states and territories. Amounts for 2009–10 to 2014–15 are final allocations, while the allocation for 2015–16 was as at January 2016.

Source: ATO.

Previous audit coverage

1.8 Other ANAO audits have examined matters related to some of the specific Budget-funded compliance measures listed at Appendix 3, although they did not all specifically examine the ATO’s performance in meeting the revenue commitments for the measures.¹² The 2015–16 audit of Strategies and activities to address the cash and hidden economy examined two measures that were in the scope of this audit.¹³ That audit made two recommendations, one of which was that the ATO report on the achievement of indirect revenue targets associated with Federal Budget funding for ATO cash economy activities.

Audit approach

1.9 The objective of the audit was to assess the effectiveness of the Australian Taxation Office in achieving revenue commitments established under specific Budget-funded compliance measures. To form a conclusion against this objective, the Australian National Audit Office adopted the following high-level audit criteria:

• the ATO has established sound processes and practices to support the development of specific Budget-funded compliance measures;
• the ATO effectively monitors the revenue raised and expenses incurred in implementing these compliance measures; and
• revenue commitments and other benefits from these compliance measures have been met.

1.10 The audit methodology included discussions with ATO staff and examination of ATO records relating to the 16 specific compliance measures announced between 2010–11 and 2014–15 with a start date prior to 1 July 2015, and the extent to which the announced revenue targets had been met. It also included testing and analysis of the ATO’s systems and processes as they relate to the development, funding and revenue outcomes of those compliance measures.

1.11 While all compliance measures announced and implemented between 2010−11 and 2014−15 were examined, more detailed analysis was undertaken of the following measures, which were selected to cover a range of tax types, Budget years, measure values and responsible ATO business lines, while not seeking to duplicate work undertaken through the ANAO’s cash economy audit and other previous ANAO audits:

• GST compliance program—working together to improve voluntary compliance () and GST compliance program—two year extension (2012–13 Budget);
• Tax compliance—increased data matching resources (2011–12 Budget)¹⁴;
• Tax compliance—maintaining the integrity of the tax and superannuation system (2012–13 MYEFO);
• Tax compliance—managing tax debt in challenging times: a balanced and differentiated approach (2012–13 Budget); and
• Tax compliance—Australian Taxation Office trusts taskforce (2013–14 Budget).

1.12 The audit has been conducted in accordance with the ANAO’s auditing standards, at a cost to the ANAO of approximately $330 000.

Does the Australian Taxation Office have sound processes to identify and develop compliance measures?

Identifying and developing potential measures

2.1 Business lines develop possible new policy proposals to address identified compliance risks. The merits of these proposals are then assessed internally, at the senior executive level, taking into account factors such as the seriousness of the non-compliance, the likely revenue outcome, the benefits of any law change and the need for additional resources.

2.2 Once the need for additional funding to address a compliance risk has been identified, the ATO prepares a statement justifying the request to Finance for additional resources. For the measures listed at paragraph 1.11, except for the GST compliance program, this justification did not include the revenue estimates for the measures. These are separately reviewed and approved by Treasury. Following approval of the estimates by Finance and Treasury, new policy proposals including the approved estimates are submitted to the Government for consideration in the Budget process.

2.3 Finance indicated that its primary focus is on the justification of the request for additional funding for compliance activities. Finance also advised that, where changes are made to resource levels for measures under development, it reaches agreement with the ATO on the adequacy of the revised resources and, where necessary, seeks advice from Treasury on the continued achievability of the revenue commitments. Treasury advised the ANAO that it has a high level of confidence in the ATO’s capacity to prioritise the highest risks and formulate the most appropriate compliance strategies.

2.4 As outlined in Table 2.1, some of the compliance measures implemented between and 2014–15 supplemented funding for business-as-usual activities or extended previous measures where there was considered to be continued benefit from addressing the compliance risks under the measure. Other measures provided additional funding to address new or emerging risks. Where the ATO identifies a new or emerging risk, it may investigate the risk (including undertaking preliminary compliance activity) or extend the approaches of existing compliance activities.

Table 2.1: Nature of compliance measures commenced from 2010–11 to 2014–15

Note: Some measures, such as the 2012–13 MYEFO measure, Tax compliance—maintaining the integrity of the tax and superannuation system, had elements that addressed new or emerging risks, but the primary nature of the measures was an extension of previous measures and are therefore categorised as such in the table. While increased data matching resources were provided in the 2011–12 MYEFO measure, the later measure, Tax compliance—improving compliance through third party reporting and data matching, has been categorised as addressing a new or emerging risk because it expanded the range of data for data matching purposes.

Source: ANAO analysis.

2.5 Finance advised the ANAO that compliance activity is often funded through new measures, rather than by supplementation of the ATO’s base funding, to enable the effectiveness of this additional activity to be evaluated before further funding is provided. In practice, however, there is a lack of clarity in new policy proposals about what compliance activity is expected to be maintained from base funding and therefore on the overall and additional levels of compliance activity following the implementation of the measure.

Revenue estimates modelling

2.6 The ATO’s Revenue Analysis Branch prepares the revenue estimates for all new policy proposals. This enables the ATO to use the expertise of experienced modellers, aims to promote consistency in preparing revenue estimates and provides a single point of contact for all revenue commitments. The revenue estimates are reviewed by Treasury in finalising new policy proposals. Treasury advised that it seeks to satisfy itself as to the reasonableness and coherence of the approach taken in estimating revenue. It also advised that it undertakes a detailed analysis of the revenue models, but places reliance on the ATO’s knowledge and data informing the assumptions and rates of return in the revenue models.

2.7 There are a number of complexities in estimating revenue to be generated from compliance activities, and the estimates are dependent upon a number of assumptions. Key issues include:

• multiple compliance activities may affect the same taxpayer groups and heads of revenue, such as income tax or the GST;
• there may be diminishing rates of return over time from some compliance activities;
• there are often lags in collecting taxation revenue (it may take two years or more from the time a tax assessment is raised to the time the tax is collected); and
• there can be significant differences between the initially assessed tax liabilities, revised assessments (or net tax liabilities that are included as revenue in the Budget as they affect the fiscal balance¹⁵) and the amounts that are actually collected.¹⁶ This is because initially assessed liabilities are subject to objection and review and can be amended. Also, taxpayers may not be able to pay the assessed amounts, at least within the prescribed time periods.

2.8 To assess the robustness of the modelling of the ATO’s revenue estimates for proposed compliance measures, the ANAO examined the modelling for the five measures listed at paragraph 1.11.

2.9 Most revenue estimates for the compliance measures are based on the estimated number of additional activities to be performed, the percentage of those activities that are expected to result in additional revenue, the projected revenue that will be raised for each successful activity and the number of activities to be performed by each additional staff member. However, the ATO does not generally set a clear baseline for compliance activity, based on pre-measure performance and target activity for any pre-existing measures, when estimating additional revenue from measures. Where possible, setting a clear base level of compliance activity for the measure would provide assurance that the estimated revenue is additional and will not be offset, at least in part, by subsequent reductions in baseline activity. This would also make it easier to report on the ATO’s performance in meeting its compliance revenue commitments. To provide assurance that the estimated revenue is additional, the ATO should document in its funding proposals how it proposes to measure the additional revenue and the level of pre-existing activity related to the risks being addressed by the measure. This issue is discussed further at paragraphs 4.6 to 4.11.

2.10 The Revenue Analysis Branch uses spreadsheets in the preparation of models to estimate the revenue from proposed compliance measures. The Institute of Chartered Accountants for England and Wales has published twenty principles for good spreadsheet design. Consistent with these better practice principles, good models should, among other things:

• be able to be clearly understood by those who use them. This includes incorporating adequate instructions, validation, help and other documentation to promote ease of use and avoid input errors; and
• clearly identify inputs, workings and outputs.

2.11 The revenue models do not provide underpinning evidence of the key assumptions used, such as evidence for the number of cases to be reviewed by each full-time equivalent (FTE) compliance officer, the strike rate or percentage of cases that result in additional revenue and the average revenue generated per successful compliance activity. The ATO advised that the assumptions are decided in consultation with business lines but, with few exceptions, it was unable to provide details of those consultations or the basis of the assumptions. To help provide assurance on the accuracy of the estimates, there would be merit in business lines formally agreeing to the assumptions used in the models and the revenue estimates, given their ownership of the measures. The starting point for assumptions to determine the revenue estimates for new compliance measures, particularly for those measures that represent an extension of existing activity, should be the data obtained from similar existing compliance activities (such as the number of cases per FTE compliance officer, the number of FTE compliance officers, the number of cases, the strike rate and the rate of return per case).

2.12 The justification statements that were prepared to support the provision of additional resources for the measures (see paragraph 2.2) did not always provide detailed assumptions for application in the revenue models. There were also unexplained differences in revenue assumptions used in justification statements and revenue models¹⁷, within models and between models. Some differences in the treatment of revenue outcomes in the revenue models are summarised in Table 2.2.

Table 2.2: Application of revenue model outcome assumptions

Note 1: This was the ‘complex audit’ component (three elements in this component). Other components comprised: refund integrity; property; excise; integrity of business systems; lodgment; debt—compliance expansion; and extended debt collection.

Source: ANAO analysis.

2.13 Specifying the basis of the assumptions used in revenue models and maintaining a register of such assumptions (to provide a record of their derivation) would help to ensure that assumptions reflect the results of recent similar audit activity and would promote consistency in their application across models. Maintaining a central list of assumptions should also reduce the time taken to prepare revenue models. Where assumptions are determined in consultation with the business lines, there would also be merit in maintaining a record of those consultations. The record of the assumptions could include the last date of review, the owner of the assumption and, where the measure is an extension of a previous measure, the reasons for any changes to previous assumptions.¹⁸ Ideally, it should be possible to clearly match the number of compliance staff (as distinct from support staff) in the revenue models to those in the expense estimates. In four of the five measures examined, it was not readily possible to match the resources funded (average staffing level) between the cost model and the revenue model.¹⁹

2.14 The ATO does not include penalties and general interest on a consistent basis in the revenue estimates, but does include them when reporting revenue outcomes (the liabilities and collections that were achieved).²⁰ Penalties and interest can be significant. For example, GST penalties and interest liabilities totalled $1.14 billion out of total GST liabilities of $4.54 billion in 2013–14. This introduces a significant conservative bias in the estimates. While it may be prudent to include a conservative bias in the estimates, the ATO should include penalties and interest in the revenue estimates where they are likely to be material. The ATO also does not consistently separately identify penalties and interest when reporting on outcomes and doing so would provide greater transparency on the revenue outcomes.

Forecasting indirect revenue

2.15 Seven of the 16 measures announced by the Government and implemented between 2010−11 and 2014−15 included both direct and indirect revenue commitments (see Table 4.1). Direct revenue is that which results directly from the compliance activity, that is, the additional amounts of tax paid as a result of ATO audits and reviews of taxpayers. Indirect revenue is based on changes in taxpayer behaviour as a result of the compliance activity, that is, the sustained tax compliance in later years of those who had been audited.

2.16 Indirect revenue may also include the effects on taxpayers not directly affected by the compliance activities (called ‘ripple effects’). For example, a compliance activity with some members of a taxpayer group might have ‘ripple’ effects on the behaviour of other members of the taxpayer group. Members of the broader group may become aware of the compliance activity of the ATO and modify their behaviour voluntarily. This might be because many of the group use the same tax agents as those who were audited or because they otherwise become aware of the ATO compliance activity and modify their behaviour accordingly. These effects can be more difficult to measure than the sustained changes in behaviour of those who may have been audited and, for this reason, are not generally included in the Budget estimates.

2.17 Where included in a measure, the assumptions used by the ATO to estimate indirect revenue vary and the basis for the assumptions is not explained. The 2010–11 GST measure included ‘ongoing’ revenue from interactions with late or non-lodgers at a rate of 50 per cent one year after contact by the ATO, 25 per cent two years after contact by the ATO and 13 per cent three years after contact by the ATO. By contrast, the Trusts measure assumed that indirect revenue would be 60 per cent of net direct revenue (after deduction of bad debts) one year after contact by the ATO, 40 per cent two years after contact by the ATO and 20 per cent three years after contact by the ATO.

2.18 The ATO has experienced difficulties in reporting on indirect revenue outcomes in a timely way. These difficulties and the ATO’s reporting of indirect revenue outcomes are discussed at paragraphs 4.20 to 4.22. As shown at Table 4.1, in some cases, indirect revenue has not been calculated or a methodology developed to determine the indirect revenue, even where a measure has been operating for a number of years. Where the ATO estimates that indirect revenue will be received in a particular year, it should report on those revenue outcomes by the end of the following year.

2.19 Where there is no sound basis for determining and measuring indirect revenue outcomes (reflected in the development of an approved methodology for doing this at the start of the measure), and where it is unlikely that actual indirect revenue can be accurately reported in a timely way (the year after the projected year of receipt of the indirect revenue), it is questionable whether indirect revenue should be proposed for inclusion in compliance measures. Where the ATO proposes that indirect revenue be included in a measure, it should clearly document how the revenue will be calculated and reported in a timely way (that is, if the ATO estimates that indirect revenue will be received in a particular year, it should be able to report on the revenue outcome in the following year). This would assist Treasury in deciding whether or not to include the indirect revenue in the Budget.

Does the Australian Taxation Office effectively monitor revenue and expenses associated with compliance measures?

Arrangements for monitoring compliance measures

3.1 As part of annual planning processes, business lines usually request the Revenue Analysis Branch to endorse the compliance targets for the measures to provide assurance that they are consistent with the ATO’s revenue commitments. The results of the ATO’s compliance activities are available to managers through the ATO’s online management information system in ‘Heartbeat’ reports. Monthly reports are also prepared by business lines and used to help identify where corrective action may need to be taken. Business line reports generally focus on the current financial year and are used to help monitor performance against the revenue commitments. The inclusion of information in reports at the end of a financial year on the performance of the measure across all financial years to that date would provide a more complete picture on the performance of the measures and help to identify where any corrective action may be required.

3.2 Business lines also provide information on their performance against compliance revenue commitments to the Tax Practitioner Lodgment Strategy and Compliance Support business line, which then compiles consolidated performance reports for all compliance measures.²¹

3.3 For transparency and accountability, the methodology used to calculate revenue from a measure should accurately reflect the outcomes of the activities funded under the measure. Instances were identified where revenue outcomes were not calculated accurately or were not determined at all (see, for example, paragraphs 4.2 and 4.23 to 4.26). The ANAO also found that there were arbitrary changes from one year to another in the activities that were deemed to have been funded by the 2011–12 measure, Tax compliance—increased data matching resources.²² This could have been prevented by putting in place rigorous and consistent methodologies for the calculation of liabilities and collections at the start of a measure²³ and quality assuring the data against these methodologies. While the Tax Practitioner Lodgment Strategy and Compliance Support business line compiles the consolidated reports, it relies on senior business line management clearance of the information provided to assure the quality of the data. However, to provide additional assurance on the accuracy of the data it receives, there would also be value in the Tax Practitioner Lodgment Strategy and Compliance Support business line quality assuring the data against approved methodologies for their calculation, applying a risk-based approach. For example, this could be done where there are significant variations in the estimated revenue from one year to another or in comparison to the Budget estimates, as was the case with the data matching revenue outcome figures, the accuracy of these figures could be checked.

3.4 The way the ATO calculates revenue from its compliance measures is largely dependent on the type of measure involved. Where a measure supplements resources for an existing activity, and the compliance activity affects a large number of taxpayers, the ATO estimates the revenue from the measure and business-as-usual activity, with compliance revenue being identified through the use of accounting project codes. For example, it may take a percentage of the compliance revenue for an activity based on the shares of measure and business-as-usual funding for an activity. The data matching activity measures and the Goods and Services Tax (GST) compliance measures were examples of where this approach was used to estimate revenue from the measures.²⁴

3.5 Where a measure targets specific risks and only affects a relatively small number of taxpayers, the ATO generally uses specific accounting project codes for the activity. Outcomes can then be tracked more accurately and reports prepared based on the recorded results of the measure’s compliance activity.²⁵ This approach is being used, for example, to determine the revenue raised by the Trusts Taskforce and for elements of the measure Extension of Strategic Compliance Initiative—engaging newly identified high wealth individuals (part of the 2012−13 MYEFO measure, Tax compliance—maintaining the integrity of the tax and superannuation system). While direct tracking of outcomes is more accurate than estimating outcomes, care needs to be taken that only measure-related cases are recorded against the measure, with recorded cases being quality assured as measure-related.

3.6 Both approaches for deriving revenue outcomes will not accurately indicate whether the revenue commitments were met if they do not assess the additional revenue generated by a measure. This requires having an accurate base level of compliance activity and revenue to assess the additional activities and revenue generated by the measure. While the Revenue Analysis Branch assures the revenue targets for compliance measures, it currently does not assure targets for business-as-usual compliance activities. The measurement of revenue from compliance measures is examined further in Chapter 4.

Monitoring GST revenue

3.7 The states and territories provide the funding for the GST compliance measures as GST revenue is paid to the states and territories. It is therefore important that the ATO has effective arrangements in place to report on the outcomes of the GST measures.

3.8 There is a GST Administration Performance Agreement between the ATO and the Council on Federal Financial Relations (see footnote 9), which provides for the establishment of the GST Administration Sub-Committee of state and territory Treasury representatives with responsibility for monitoring all aspects of the operation and administration of the GST. The Sub-committee has delegated oversight of aspects of its role to the GST Policy and Administration Sub-group. The ATO each year publishes end-year performance reports on GST administration. The ATO also holds an annual ‘Melbourne Meeting’, at which it briefs state and territory representatives on GST issues and responds to issues raised.

3.9 Between 2006 and 2015, a representative of the Queensland Treasury also worked in the ATO to provide the states and territories with more direct involvement in the administration of the GST. This arrangement has not been continued, following the retirement of the Queensland Treasury representative in November 2015, on the basis that the states and territories are satisfied with the level of oversight provided through existing arrangements. The former Queensland Treasury representative advised the ANAO that he regarded the efforts being made by the ATO to address GST compliance risks as ‘excellent’.

3.10 While the arrangements for monitoring the ATO’s administration of the GST are appropriate, as indicated at paragraph 4.11, there is scope for greater clarity over the base level of GST revenue, used to determine revenue outcomes from the GST measures.

Expenses incurred on compliance measures

3.11 Additional funding for implementing measures is allocated to the relevant business lines. However, for most measures, the ATO does not monitor the allocation of those resources by the business line, so providing scope for business lines to reduce the number of staff previously working on business-as-usual activities, instead of increasing resources in line with the agreed compliance measure. For example, if the ATO had maintained the same level of FTE in 2012–13 that it had before the implementation of the 2011–12 MYEFO measure, Tax compliance—increased data matching resources (233 FTE in 2011–12), the ATO would have had a total FTE of 386 (instead of 346) working on data matching in that year (details are at ‘Other information’ in Table 4.2). The ATO advised that, following the measure, it would have reallocated some staff to other business-as-usual areas where more effective outcomes could have been achieved. This indicates that the ATO did not require all of the additional resources provided for this measure and that the measure was used in part to enable the ATO to fund non-measure activity.

3.12 As discussed previously, some measures are extensions of business-as-usual compliance activities, and others are managed as a separate project. Where a project is established to implement a measure, it should be relatively straightforward to determine the overall cost of the project. However, resources are not always specifically allocated to a project and for most measures the ATO has advised that the cost is not able to be separately tracked and reported. While it may not be practical to separately track the actual cost of some measures, where feasible, doing so (by setting up project codes in the ATO’s financial management system to track actual expenditure) would provide valuable feedback on the cost of the measures. If expenses cannot be separately identified, the business line could estimate them (for example, using the average cost of the audit activities attributed to a measure). Knowing the cost of compliance activities would also enable the efficiency of those compliance activities to be evaluated. This could include comparisons between the average cost of measure-related and business-as-usual funded activities and between different compliance activities.

Evaluating the performance of compliance measures

3.13 The performance of compliance measures cannot be evaluated without defining a baseline level of activity by the head of revenue and business line and measuring performance against a set overall target. It is expected that compliance measures will increase the level of resources deployed on the relevant compliance activities and increase the total revenue generated. To provide effective monitoring of a measure, the ATO therefore should monitor total revenue for the compliance activity as well as the shares of business-as-usual and measure-related revenue. Exception analysis should explain the reasons for any material variation. Relevant performance indicators are summarised at Table 3.1.

Table 3.1: Compliance performance indicators

Note 1: Liabilities are a point in time figure and may be reduced after further consideration by the ATO. Cash collected should be the primary indicator of the ATO’s effectiveness in raising and collecting additional revenue. Like liabilities, cash collected is often estimated and is not an actual figure.

Note 2: Operational information currently monitored by the ATO includes timeliness of processing, completeness of registrations, cases completed within defined timeframes and prosecutions. Commentary is also provided in some performance reports on changes in behaviour that improve the level of voluntary compliance.

Source: ANAO.

3.14 While data on these indicators can be obtained from some business line reports, they are not presented in a coherent way, precluding ready analysis of the effectiveness of the measures. There would be merit in the ATO reviewing the indicators it uses to monitor the effectiveness of the measures, having regard to the indicators shown at Table 3.1.

3.15 As noted at paragraph 2.7, the net liabilities raised by the ATO affect the overall fiscal balance. However, a critical indicator of the success of a measure is the extent to which the ATO achieves the projected cash collections for the measure. This not only measures the effectiveness of the ATO in identifying and raising the amounts owing to it, but also considers its effectiveness in recovering these amounts.

3.16 While it is also important to know the cost of a measure and how this compares with the Budget allocations, it would still be appropriate to use the Budget allocation to determine the return on investment of a measure since this is the cost to the government of the measure.

Closure reports

3.17 Seven of the 18 measures listed at Appendix 3 were completed by the end of the 2014–15 financial year. One of these measures, the 2010–11 Budget measure, GST compliance program—working together to improve voluntary compliance was extended by the 2012–13 Budget initiative, GST—compliance program—two year extension, and the ATO has argued that the measure is therefore ongoing. Closure reports had been prepared for three of the remaining six completed measures. ²⁶ None of these reports satisfactorily addressed whether or not the revenue commitments had been met (see Table 3.2).

Table 3.2: Completion/non-completion of closure reports for completed measures

Note: Excludes the 2010–11 Budget measure, GST compliance program—working together to improve voluntary compliance, which was extended by the 2012–13 Budget initiative, GST—compliance program—two year extension.

Source: ATO.

3.18 Two of the three completed closure reports focused on the satisfactory transfer of responsibility for any outstanding matters, but did not indicate the revenue achieved. The third closure report on the cash economy reported on the achievement of direct revenue, but did not report on the achievement of indirect revenue in the Budget estimates (the ATO has since calculated the indirect revenue for this measure). It also did not assess whether additional revenue from the measures had been achieved.

3.19 Whether or not a later measure continues (such as with the various GST measures), there would be merit in the ATO completing a report that evaluates performance against the revenue commitments and other efficiency and effectiveness indicators at the conclusion of the measure period. These reports could usefully indicate the:

• activities undertaken and their effectiveness (for example, the comparative effectiveness of the types of compliance activities conducted);
• additional revenue from the measure and how it was calculated; and
• lessons learned and opportunities for improvement and any further action required, such as the estimation of indirect revenue, where applicable, and follow-up action as part of ongoing (business-as-usual) operations.

3.20 Two of the three closure reports included information on the cost of the measure, comparing the actual cost with the additional funding provided. The third report merely indicated that costs for the measure were detailed in the Budget papers and that these costs had been allocated to different projects within the measure based on the staffing for those projects. All closure reports should include an estimate of the total cost of the measure, compared with the budgeted cost of the measure²⁷, to assist in the development of new measures and provide a more complete summary of the performance of the measure.

Has the Australian Taxation Office achieved the additional revenue commitments for the compliance measures?

ATO reporting of achievement of revenue commitments

4.1 In its annual reports, the ATO provides information on performance in meeting individual revenue commitments for compliance measures. The level of information varies from year to year and different measures may be considered under the relevant heads of revenue.

4.2 There are difficulties and delays involved in accurately estimating indirect revenue at the end of a financial year (described in paragraphs 4.19 to 4.22, where it is noted that the ATO has yet to determine indirect revenue for some measures). For this reason, direct revenue (liabilities and cash collections), rather than total direct and indirect revenue, is a more reliable indication of the revenue outcomes of a measure. The ATO’s annual report 2014–15 includes information on the direct revenue commitments and outcomes for income tax and GST compliance activities, as shown in Figure 4.1.²⁸ These reports indicate that, except in 2010–11 and 2011–12 for collections, the ATO met its compliance commitments for each year in the past five years and for the period as a whole. However, Figure 4.1 includes outcome amounts for the Not-for-profit reforms, for which the ATO has apportioned the outcome figures to equal the total Budget estimate of $41 million over different years. Figure 4.1 also does not include collections for the 2012–13 Budget measures, Tax compliance—managing tax debt in challenging times: a balanced and differentiated approach (see case study at paragraph 4.35) and Addressing the level of unpaid tax and superannuation in the community, for which the ATO also does not have accurate collection figures, and the Reporting taxable payments measure, for which no direct revenue was forecast.²⁹

Figure 4.1: Direct measure-related income tax and GST heads of revenue liability and collection compliance commitments and outcomes, 2010–11 to 2014–15

Note: Many compliance measures contributed to both the income tax and GST heads of revenue.

Source: ATO.

4.3 Figure 4.2 shows the ATO’s reported direct revenue commitment outcomes for 12 of the 16 compliance measures announced and implemented between 2010−11 and 2014−15 (excluding the four measures listed in the previous paragraph).

Figure 4.2: Total direct income tax and GST collection outcomes on compliance measures, 2010–11 to 2014–15

Note: Both the 2010–11 and 2012–13 GST measures are shown on one line. The 2011–12 Budget measure, Reporting taxable payments, the 2012–13 Budget measures, Tax compliance—managing tax debt in challenging times: a balanced and differentiated approach and Not-for-profit reforms and the 2013–14 Economic Statement measure, Addressing the level of unpaid tax and superannuation in the community, are excluded from the chart for the reasons mentioned at paragraph 4.2.

Source: ATO.

4.4 Figure 4.2 indicates that the ATO achieved the direct collection commitments for all measures, except the GST compliance program, which raises both GST and other revenue (see case study at paragraphs 4.27 to 4.30). However, ANAO analysis of the increased data matching resources measure indicates that the ATO’s calculation of the revenue outcomes for this measure was incorrect and that the commitment for it was also not met (see case study at paragraphs 4.23 to 4.26).

Factors affecting the accuracy of reported revenue outcomes

4.5 There are a number of factors that can affect the accuracy of reported revenue outcomes. These include:

• the ‘base’ level of ongoing revenue for measurement of additional revenue generated by the measure;
• the methodology that is used to determine the revenue outcome from the measure (see paragraphs 3.3 to 3.6);
• whether or not penalties and interest are included in the reported revenue (see paragraph 2.14);
• the treatment of large and abnormal cases (for example, major fraud investigations) that result in the raising of very large liabilities and potential recoveries; and
• the calculation of indirect revenue.

Base or ongoing revenue

4.6 Compliance measures are expected to realise additional revenue for the government. This means that they are expected to increase revenue above amounts that would have been achieved had the measure not been implemented. This revenue is the ‘base’ revenue from business-as-usual activity and other previously approved activity (and may be expected to increase over time, consistent with changes in the forward estimates).

4.7 As proposals for new measures include estimates of additional revenue, it is useful to know what additional revenue is generated, and where commitments have not been met, whether this was because of a reduction in revenue from business-as-usual activities (and the reasons for this) or because the measure itself was not as successful as predicted. This information would also enable the Revenue Analysis Branch to provide business lines with assurance on the total compliance revenue to be achieved each year for the compliance activity to meet ATO revenue commitments and inform future new policy proposals.

4.8 Revenue from measures that supplement resources for business-as-usual compliance activity should amount to total revenue for the compliance activity less revenue from business-as-usual activity (and other previously approved additional activity). However, the ATO does not consistently calculate revenue for compliance measures in this way. Rather, after the introduction of a compliance measure, the ATO generally attributes revenue from relevant compliance activities to a measure based on its share of total resources for those activities. The remaining share is deemed to be business-as-usual revenue. However, this approach means that the business-as-usual revenue may not be maintained and that the committed additional revenue may not be realised.

4.9 Without estimates of base revenue, the ATO is unable to provide assurance that the additional revenue commitments for specific measures have been met. It has also indicated that, for some measures, it would be difficult to identify the base level of revenue and would limit its ability to reallocate resources to realise higher revenue in other areas and there is no necessary requirement for it to do so. However, in seeking additional resources and committing to raise additional revenue for a specific compliance activity, there is the implication that the revenue for that activity comprises previous (base) revenue and the additional revenue from the measure. If there is a need to retain flexibility to allocate resources to areas of highest risk, it would be prudent for the ATO to review the basis on which it seeks additional resources for compliance activities. For example, instead of seeking additional funding for a specific compliance activity, the ATO could seek additional funding for a group of compliance activities or for its compliance program as a whole. This would still require the ATO to maintain a reliable overall measure of base revenue for compliance and to report reliably on the additional revenue that was realised from the additional funding provided, but would provide greater flexibility on the composition of its base compliance activity.

4.10 The ATO does currently maintain estimates of base compliance revenue (for the purpose of estimating total—measure and non-measure—revenue from compliance activity), but these are not reliable. For income tax, it assumed that baseline income tax compliance commitments and outcomes were the same for each year from 2010–11 to 2014–15 ($4.65 billion in liabilities and $3.0 billion in collections). The ATO last estimated a base amount for its income tax related compliance activity for the 2009–10 financial year. The approach took a five-year average for the period 2003–04 to 2007–08 (after adjusting for outcomes relating to measures³⁰) to allow for changes in compliance activities and outcomes from one year to another and then adjusted this to obtain a base figure for ³¹ The base revenue for income tax has not been adjusted since that time, meaning that it does not reflect increases in the nominal value of the revenue base and may therefore not be accurate.

4.11 A similar approach was taken with GST base compliance revenue (including large, abnormal risks and penalties). For GST revenue from all compliance activity, the ATO set a base of $1.3 billion for liabilities and $1.1 billion for collections for 2010–11.³² This base level of revenue has been maintained to the present time in dollar terms, notwithstanding increases in the forward estimates and the growth in the economy since then. Realistic base levels should be set for compliance activity.

Treatment of large abnormal cases

4.12 In determining base revenue, decisions may be required about the treatment of large abnormal cases (which the ATO calls ‘outliers’) and other major one-off exercises, which can distort revenue from one year to another. The ATO has calculated actual direct liabilities of $300.5 million in 2013–14 and $241.6 million in 2014–15 for the Trusts Taskforce compared with the Budget estimates of $10.6 million in 2013–14 and $62.8 million in 2014–15. The substantial difference in the estimated and actual outcomes was due to a major crime investigation that started in 2012. Excluding this major investigation, the actual result would have been $37.9 million in 2013–14 and $28.2 million in. Given the importance of this investigation, the ATO would clearly have continued to work on it in the absence of the measure, highlighting the need to more clearly identify what additional revenue will be generated from measures, having regard to ongoing work. It also highlights the sensitivity of compliance outcomes to ‘lumpy’ outcomes. A major investigation might result in very large revenue in one year, distorting the outcomes that might reasonably be expected over a longer period of time.

4.13 It is also noted that serious tax fraud cases can result in the ATO raising tax liabilities, but other agencies, such as the Australian Federal Police, may be responsible for collecting this revenue following prosecution of the taxpayer for the fraud.

4.14 Similar issues regarding the treatment of large and abnormal revenue were identified with GST revenue. For example, in 2010–11 the ATO raised $487 million from outlier cases and undertook a one-off major letters campaign that resulted in additional GST liabilities of $287 million. The ATO, in its revenue outcomes, apportioned some (but not most) of the GST liabilities from these outlier cases to the measure.

4.15 As the ATO has traditionally resourced major investigations and exercises as required, and given their variability from one year to another, it would be reasonable to remove them from both the base and measure revenue when calculating the additional revenue from a measure. Alternatively, they could be averaged over a number of years when determining the base level of revenue. The ATO has used both approaches in the past, and there would be merit in it developing guidelines for their consistent treatment that recognises the variability of this revenue and where the activity would normally be expected to be undertaken as part of the ATO’s ongoing business.

Indirect revenue

4.19 Where indirect revenue is included in revenue commitments, the Revenue Analysis Branch approves calculation methodologies developed by business lines. It can take the ATO a number of years to develop, test and approve methodologies to estimate the indirect revenue effects of new compliance measures. This seems unacceptably long and likely to make it difficult to extract the information needed in a timely way. As at April 2016, the ATO had not estimated indirect revenue for three of the seven measures that included both direct and indirect revenue commitments and announced by the Government and implemented between 2010−11 and 2014−15 (see Table 4.1). The estimation of indirect revenue is not always completed, even when a measure has operated for some years, as was the case with the cash economy audit, where the measure was finalised but indirect revenue estimates were not derived and were therefore not reflected in the closure report for that measure (although the ATO has since completed its analysis of the indirect revenue from the measure). For one of the seven measures (Tax compliance—maintaining the integrity of the tax and superannuation system), indirect revenue had been estimated for some but not all elements of the measure.

Table 4.1: Calculation of indirect revenue outcomes for the seven compliance measures that included indirect revenue commitments implemented since 2010–11

Source: ATO advice.

4.20 The estimation of indirect revenue can be complex. A simple comparison of the current year tax payable to the baseline tax paid (that is, year 0) for a taxpayer may not take account of changes in the level of taxable income of the taxpayer because of changes in economic or business conditions and in the taxpayer’s personal circumstances. Further, where a taxpayer is impacted by a number of measures, it may be difficult to determine which measure, if any, resulted in the changed revenue. For example, a non-lodging taxpayer might be contacted by the GST measure team and also be the subject of a Trusts Taskforce audit.

4.21 It was clear from the ANAO’s examination of the performance reports of business lines that they focus on the direct revenue commitments, and largely ignore the indirect revenue commitments in their performance reporting, because indirect revenue estimates are unavailable. While it is recognised that there are lags in the receipt of indirect revenue³³, to enable full transparency, indirect revenue should be calculated and reported in a timely way (the year after the date of their estimated receipt). Indirect revenue should also be reported in the closure reports of measures, where relevant.

4.22 As noted in Table 4.1, where the ATO has not yet determined indirect revenue from a measure, in its revenue reporting it often uses the commitment estimates as the outcome figures. It would be preferable not to include outcome figures and note that indirect revenue for some measures has still to be calculated.

Case studies

Data matching

4.23 The importance of knowing the base level of activity and having a consistent methodology to determine revenue outcomes is illustrated in the approach used by the ATO to estimate revenue from the 2011–12 MYEFO measure, Tax compliance—increased data matching resources. The Government provided the ATO with $58.3 million over the forward estimates period ( to 2015–16) to improve its data matching capability to improve compliance with the tax system. The measure was expected to increase liabilities (and the fiscal balance) by $436.3 million (see Table 4.2) and cash collections by $328.4 million over the forward estimates period.³⁴ The measure essentially represented an extension of business-as-usual activity and data matching included in other Budget measures, and was implemented on that basis.³⁵

Table 4.2: Additional liabilities generated under the measure Tax compliance—increased data matching resources compared to measure commitments and claimed ATO outcomes, 2012–13 to 2014–15

Note 1: The pre-measure figure of $790.0 million for 2011–12 compares with a figure of $350.7 million for 2010–11. The increase in liabilities between 2010–11 and 2011–12 was explained partly by an increase in business-as-usual activity and partly by an increase in activity from other (earlier) compliance.

Note 2: ‘Other measure approved FTE’ is additional FTE provided by the Improving compliance through third party reporting and data matching measure.

Note 3: ‘Base FTE’ includes measure-related FTE that commenced before the start of the Increased data matching resources measure and is not related to the two data matching measures (for example, FTE related to the Reporting government grants and payments measure).

Source: ANAO analysis of ATO data and ATO annual reports.

4.24 In 2011–12, the financial year before the measure was implemented, the Data Matching Compliance Strategies team raised liabilities of $790.0 million through the finalisation of 544 140 cases completed by 233 FTE. In the following year (2012–13) when the measure was implemented, this increased to $939.0 million, an overall increase in revenue of $149.0 million, compared to the Budget estimate of $174.2 million (net liabilities), $214.6 million calculated by the ATO as measure-related revenue and the ANAO’s estimate of $135.2 million.³⁶ In 2013–14, the second year of the measure, measure-related liabilities were calculated by the ATO at $400.6 million and in 2014–15 at $118.5 million, and by the ANAO at $100.6 million in 2013–14 and $129.2 million in 2014–15. The ANAO estimates are based on the liabilities from the two data matching compliance activities (new and emerging risks and foreign source income) that the ATO had identified as being measure-related.

4.25 The reason for the widely varying outcomes reported by the ATO was due to arbitrary changes in the methodology used to calculate the revenue outcomes:

• The original methodology for the calculation of the revenue outcome for this measure was based on applying percentages of estimated measure-related activity for two data matching compliance activities mentioned in paragraph 4.24, that is, liabilities of $149.4 million in 2012–13. However, the ATO increased this to $214.6 million, but was unable to provide an explanation for this change.
• In 2013–14, when overall liabilities increased by only $150.0 million from the 2011–12 baseline, the ATO calculated measure-related liabilities of $400.6 million. There was an error in the calculation of this amount, which resulted in an overstatement of the revenue outcome. Instead of using the average of the percentages of the two previously mentioned compliance activities and applying that percentage to those activities, it used the sum of those percentages and applied that to total data matching compliance revenue (adjusted to exclude liabilities for the Reporting Government Grants and Payments measure).
• In 2014–15, the percentages were again reduced to arrive at a claimed figure of $118.5 million.

4.26 This indicates that the ATO did not have a consistent and objective approach to the derivation of the measure-related revenue. It would also have been more appropriate for the ATO to have focused on the total additional revenue that was achieved and not on its estimate of the share of revenue that resulted from the resources provided by the compliance measure. As shown in Table 4.2, on the basis of pre-measure liabilities of $790.0 million for 2011–12, additional liabilities generated by data matching totalled $483.0 million for the three years 2012–13 to 2014–15, compared with the measure commitments for these years of $436.3 million and the $733.7 million claimed by the ATO and $365.0 million estimated by the ANAO.³⁷ However, as indicated in note 1 of Table 4.2, liabilities in 2011–12 were considerably higher than in, partly because of an increase in business-as-usual activity and partly because of the implementation of other Budget measures. If business-as-usual activity fluctuates greatly from one year to another, it may be reasonable to average the business-as-usual component over three years, for example, and adjust the 2011–12 pre-measure liabilities base amount accordingly.

GST compliance program

4.27 The 2010–11 Budget measure, GST compliance program—working together to improve voluntary compliance, funded additional activities to promote voluntary GST compliance and provide a level playing field for Australian businesses. In accordance with the GST Administration Performance Agreement, the states and territories provide the funding for this measure. The 2010–11 measure has since been extended twice, first in the 2012–13 Budget (GST compliance program—working together to improve voluntary compliance) and later in the 2015–16 Budget (GST compliance program—working together to improve voluntary compliance). This analysis focuses on the 2010–11 measure.

4.28 The 2010–11 measure provided for additional staff of up to 744 FTE to complete a range of compliance activities³⁸, addressing risks such as fraudulent GST refunds, under-reporting of GST liabilities, non-lodgment and non-payment, and reducing a growth in aged debt. It included commitments for increased GST, income tax and excise revenue of $2691 million ($1110 million in GST revenue, $1557 million in income tax revenue and $24 million in excise revenue). This comprised direct revenue of $2024 million and indirect revenue of $667 million. It was also expected to result in increased cash collections of $3298 million (direct collections of $2681 million and indirect collections of $617 million). The increased cash collections included an amount of $929 million for the collection of outstanding debt.³⁹

4.29 Excluding the commitments and outcomes for the collection of outstanding debt, the ATO’s calculations of the income tax and GST revenue outcomes for the GST compliance program—working together to improve voluntary compliance for 2010–11 to 2013–14 are shown at Figure 4.3. It shows that total liability and collection outcomes for these heads of revenue exceeded the revenue forecasts ($3503 million, compared to the commitment of $2567 million for liabilities⁴⁰ and $2626 million, compared to the commitment of $2346 million⁴¹ for collections). The allocation of the total GST and income tax revenue generated from compliance activities was based on the proportion of staffing expenses on those activities that were funded by the measure. This was not a robust basis for allocating revenue to the measure.

Figure 4.3: GST and income tax revenue outcomes (excluding debt collection) for the GST compliance program—working together to improve voluntary compliance measure, 2010–11 to 2013–14

Source: ANAO analysis of ATO data.

4.30 As noted in paragraph 4.28, the measure also envisaged that there would be increased collections from the collection of outstanding debt. There were two components to the collection of outstanding debt under the measure—‘Debt-post Global Economic Crisis’ (comprising external referrals and aged debt) and ‘Debt-Strategic’. The outcomes advised by the ATO, compared to the commitments in the Budget measure, are shown in Table 4.3. These indicate that the ATO exceeded the commitments by $2643 million. However, the ANAO considers that these figures are unreliable. This is because of the low likelihood that such outcomes could have been achieved and problems that were identified with other debt collection measures, which led the ATO to conclude that reliable outcome figures were not available for these measures. There are similar doubts about the reliability of the debt figures for this measure. For example, in its revenue model for the measure, the ATO identified that:

• there would be additional collections from external referrals of $109.5 million in, increasing to $126.7 million in the following three years. However, the reported outcomes were $213.6 million in 2010–11, increasing to over $800 million in each of the following three years; and
• there would be additional collections of $40 million over the four years (that is, $10 million a year), but the reported amounts totalled $168 million, including $126 million in the fourth year).

Table 4.3: Debt collection commitments and ATO advised outcomes for the GST compliance program—working together to improve voluntary compliance measure, 2010–11 to 2013–14

Note: ‘Difference’ amounts reflect rounding of the commitment and outcome figures in this table.

Source: ATO.

Trusts Taskforce

4.31 In 2013–14 the Government provided $67.9 million over four years to the ATO to undertake compliance activity in relation to taxpayers who have been involved in egregious tax avoidance and evasion using trust structures. This measure was estimated to increase revenue by $379.0 million over the forward estimates period. In underlying cash terms, the estimated increase in receipts was $217.1 million. For the first two years of the measure (2013–14 and 2014–15), the ATO has calculated total liabilities raised of $542 million, compared to the commitment for those years of $73 million (see Table 4.4). As noted at paragraph 4.12, most of the liabilities in these years related to a major crime investigation that started in 2012.

Table 4.4: Liability outcomes for the Trusts Taskforce measure, 2013–14 and 2014–15

Source: ANAO analysis of ATO data.

4.32 Where a measure is designed to target a new risk or a risk where the level of business-as-usual activity is minimal, it may be reasonable to assume that all revenue from the measure would be additional and can be counted towards the measure.⁴² However, in making this assessment, it is necessary to be clear about the level of pre-existing activity in estimating additional revenue.

4.33 The Trust Taskforce measure was intended to address a risk that had been newly identified, and compliance activity on trusts was minimal before the ATO started investigating the compliance risk. For this reason, the ATO assumed that the revenue raised by the Trusts Taskforce was measure related. The ATO also argued that, because trusts represent just one element of the business structures established by companies, its compliance work has not generally separately identified trust related outcomes and it would have been too difficult to identify a revenue ‘base’ for trusts. Nonetheless, the ANAO notes that, for the 2011–12 and 2012–13 financial years, the two years preceding the measure, the ATO internally reported direct trusts-related liabilities of $25 million and $73 million respectively.⁴³ Further, in 2012–13, the year before the measure, 261 audits and reviews were finalised, but in the first year of the measure, 193 cases were completed, compared to the estimated 32 cases for the measure. As these are sizeable amounts, it is clear that there was a significant level of pre-existing activity, which should have been taken into account in determining additional revenue outcomes.

4.34 In terms of receipts, collections of direct revenue totalled $42.6 million, compared to the Budget estimate of $31.1 million. As noted at Table 4.1, indirect revenue has not yet been estimated for the Trusts Taskforce measure. Further, as noted at paragraph 4.12, most of the liabilities raised for this measure related to a large abnormal case.

Managing debt in challenging times

4.35 In the 2012–13 Budget measure, Tax compliance—managing tax debt in challenging times: a balanced and differentiated approach, $106 million over four years (2012–13 to 2015–16) was provided to the ATO to improve the management of outstanding taxation debts and superannuation guarantee charges. It was mainly an extension of existing business-as-usual debt collection activities and was estimated to increase cash collections by $997.5 million over the first three years, 2012–13 to 2014–15. However, the ATO advised of actual cash collections of $8.9 billion and other collections of $0.6 billion, but later advised that these figures included business-as-usual collections, and it was unable to correctly attribute additional collections to the measure.

Maintaining the integrity of the tax and superannuation system

4.36 The 2012–13 MYEFO measure, Tax compliance—maintaining the integrity of the tax and superannuation system, was an amalgamation of 10 elements or sub-measures, most of which extended and refined earlier measures. It was expected to have a net gain to the Budget of $2.0 billion, including a GST component, over the forward estimates period.⁴⁴ In underlying cash terms, the estimated net increase was $1.0 billion, including a GST component. The ATO’s calculated total (direct and indirect) outcomes for the first two years of the measure are shown at Figure 4.4. It indicates that overall the ATO is meeting its commitments for the measure, with total outcomes exceeding commitments for both liabilities ($2.6 billion, compared to commitment of $1.5 billion) and collections ($1.6 billion, compared to commitments of $1.2 billion).

Figure 4.4: Maintaining the integrity of the tax and superannuation system—total income tax and GST liability commitments and outcomes, 2013–14 and 2014–15

Source: ATO.

4.37 As noted at Table 4.1, the ATO has not yet calculated the outcome figures for indirect revenue for all components of the measure. Where indirect revenue outcomes were not available, the ATO included the commitment figures. Had the ATO excluded indirect revenue from these components until they became available (which would be preferable), the ATO would have still met its revenue commitments for 2013–14 and 2014–15.

Rates of return on investment of compliance measures

4.38 The projected rates of return on investment of the various compliance measures funded since 2010–11 (calculated by the ANAO as the average amount of estimated liabilities raised per dollar of estimated departmental and capital expenses) has ranged from around $3.29 for the 2015–16 Budget measure, Serious Financial Crime taskforce—addressing financial and tax fraud, to $11.76 for the 2011–12 MYEFO measure Tax compliance—countering fraudulent Phoenix activities by company directors. As noted at paragraph 3.15, while liabilities raised are important, particularly in identifying the amount of tax payable to the Commonwealth, a critical indicator of the success of a measure to the Budget is the cash collected. If the rates of return were calculated just on the cash collected, the projected rates of return would have ranged from around $2.48 for the 2010–11 Budget measure ATO compliance program—dealing with the cash economy to $13.46 for the 2015–16 Budget measure GST—compliance programme—three year extension.⁴⁵

4.39 It should be noted that some measures involve a range of activities, all of which may not involve active compliance. For example, the GST compliance program includes targeted communications strategies publicising the outcomes of important prosecutions stemming from the program, so helping to promote greater self-compliance. A strict comparison of rates of return on investment would include only those resources that were involved in the compliance activity (for example, tax audits). Nonetheless, different rates of return on investment can be expected from different types of compliance activity.

4.40 For most measures examined by the ANAO, the ATO’s revenue outcome figures indicate that much higher rates of return had been achieved than projected (see Table 4.5).

Table 4.5: Rates of return on investment from selected compliance measures

Note: The selected measures are those listed at paragraph 1.11. Outcomes include penalties (not included in the estimates of any of the measures) and interest (only included in the data matching measure). For some measures, such as the GST measure, the outcome figures for returns of investment are based on estimated ‘actual’ liabilities (with collections being a percentage of the estimated liabilities).

Source: ANAO analysis of ATO data.

4.41 These rates of return use the ATO’s estimates of measure-related liabilities and collections. However, as noted earlier, the ATO’s estimates of measure-related liabilities may not accurately reflect the additional liabilities from the measure. It is important that additional liabilities and the departmental and capital expenses provided by the measure are used to estimate actual rates of return on investment, since these represent the actual returns on the investment and the level of investment by the Government respectively.

4.42 The actual rate of return includes penalties and interest not included in the revenue estimates. Revenue can also vary significantly from year to year depending on whether or not the measures result in very large revenue from isolated investigations. Calculating rates of return that include and exclude penalties and interest and abnormal revenue would provide more valid comparisons of projected and actual rates of return. It would also be more meaningful for final rates of return to be based on cash collections and not on liabilities.

Appendix 1 ATO summary response

Appendix 2 ATO detailed response

Appendix 3 List of compliance measures announced by the Government between 2010–11 and 2015–16

Note a: This measure was an amalgamation of the following 10 elements or sub-measures:

• A fairer system through removing the profit from serious and organised crime;
• Extension of Strategic Compliance Initiative—continuing to ensure individuals with high incomes and income from foreign sources pay their fair share of tax;
• Extension of Strategic Compliance Initiative— engaging newly identified high wealth individuals;
• Extension of Strategic Compliance Initiative— ensuring wealthy citizens pay their fair share of tax;
• Extension of Strategic Compliance— global profit shifting and business location decisions;
• Extension of Strategic Compliance Initiative— maintaining a level playing field–small business benchmarks;
• Extension of Strategic Compliance Initiative— maintaining a level playing field for tax and superannuation–contractor compliance arrangements;
• Protecting the community from domestic promoters of tax avoidance and evasion schemes;
• Strengthening Foundations— short term compliance debt and lodgment; and
• Strengthening Foundations (Tax Practitioner and Lodgment Strategy).