Introduction

In 2002–03, the Australian Taxation Office (ATO) collected revenue of $185 billion.¹ Self-assessment and a broad range of compliance strategies support collection of this revenue. The ATO relies upon risk management approaches and the integrity features built into the tax system to identify instances where taxpayers fail to comply. The ATO's compliance strategies aim to optimise collections and instil confidence in the community that the taxation system is operating effectively.

The ATO's use of third party investment income information is an important part of the ATO's compliance strategies. There has been a legislative requirement for banks and other financial institutions to provide detailed information about certain common financial transactions since 1936. Parliament's requirements for this information have been widened over the years in terms of the transactions reported upon and the institutions required to provide the information.

The Annual Investment Income Report (AIIR) was a significant feature of the TFN legislation introduced by Parliament in 1988² and was, amongst other things, ‘designed to improve the efficiency of the ATO's income matching system'.³ The AIIR is a legislatively required data reference set used to validate certain classes of investment income that have to be reported on income tax returns. Since 1992, investment bodies⁴ have been providing the ATO with an AIIR containing the details of investment income paid to all investors⁵ each financial year and any tax that may have been withheld on these accounts.

The Audit

The objective of the performance audit was to report to Parliament our assessment as to how well the ATO manages and uses the AIIR data in taxation administration. The ANAO considered the following four key areas in addressing the audit objective.

• Governance arrangements within the ATO, focussing on whole of ATO and whole-of-government aspects of the AIIR data, as distinct from solely business line applications.
• Receipt of AIIR data and how well the ATO facilitates the collection of complete and valid AIIR data from investment bodies.
• Management of AIIR data through the construction by the ATO of valid entity records by using the AIIR data in conjunction with existing ATO client identification master files.
• Use of the AIIR data on a systematic basis to inform active compliance activities.

Overall conclusion

The efficient and effective use of AIIR data is an important part of taxation administration. The AIIR legislation creates substantial reporting obligations for investment bodies. The latter are required to provide the ATO with timely and valid data that covers a range of investments and investor categories.

Overall, we conclude that the ATO needs to improve the management and use of the AIIR data. While the ATO's AIIR data matching program has proven to be an effective compliance tool with regard to the individual taxpayer market segment, a more strategic approach that is focussed on greater optimisation of the compliance value of the AIIR data is now warranted. The ANAO considers that relatively low cost initiatives taken at the data receipt stage would have significant beneficial consequences of improved compliance, and assist the ATO to better optimise the collection of tax revenue.

The ATO has recently introduced several initiatives to improve the management and use of the AIIR data. These include the creation of a Data Matching Steering Committee to develop an overall approach to, and the better targeting of ATO data matching activities; the implementation of a number of steps aimed at improving AIIR lodgement compliance and awareness amongst investment bodies; the creation of data integrity teams responsible for improving the quality of data sent in by investment bodies; and the implementation of initiatives to improve use of AIIR data in the non-individual market segments.

The ANAO has also identified a number of improvements to strengthen the ATO's management and use of the AIIR data. These include:

• the development of performance information to better inform the ATO about the management and use of the AIIR data;
• a review of AIIR reporting requirements and opportunities for changes to the approved form of AIIR reporting;
• the implementation of data quality standards and guidelines;
• opportunities to improve the compliance value of AIIR data through ongoing analysis of unmatched and low¬quality matched data; and
• measures to make better use of the data, particularly in the non-individual market segments.

Key Findings

Background and Context (Chapter 1)

The ATO seeks to optimise revenue collection by achieving, in any particular instance, the most appropriate mix of five key components of tax administration (self-assessment, the Taxpayers' Charter, the compliance model, market segmentation and risk management) relative to taxpayer behaviour. The strategies that the ATO uses to optimise revenue in this way, in turn, determine ATO's resourcing strategies. Getting the mix of the components ‘right' across the entire spectrum of taxpayers' behaviour is not a simple matter. The mix is complex and ever changing. It is determined by where risks (not simply revenue) lie, their comparison and what the Parliament and community would reasonably expect. ATO's decisions about the mix, and the consequent resource deployment, depend crucially on the quality and completeness of the data used to inform decision-making.

Governance Arrangements (Chapter 2)

Historically, the business group within the ATO that has responsibility for the administration of the taxation of individuals has had responsibility for the administration of the AIIR data and its predecessors. The management and use of AIIR data is provided by Compliance Systems, a business directorate within the Personal Tax (PTax) business line. During the last two years, the
Compliance Systems Directorate has achieved improved efficiencies, including: the rationalisation of a number of duplicated systems and work groups; documenting the Directorate's many systems, databases, practices and procedures; improving internal reporting and supervision arrangements; and raising the Directorate's profile throughout the ATO. The Directorate has also implemented enhanced IT systems that have improved the ATO's use of the AIIR data.

The ANAO found that PTax is the only business line in the ATO that uses the AIIR data on a systematic basis. PTax is responsible for the administration of the tax affairs of individual taxpayers. PTax carries out sophisticated data matching, using the AIIR data to verify that individual taxpayers have correctly reported some classes of investment income in their tax returns. ANAO considers that the AIIR data contains valuable information that can assist in the review of the more complex tax affairs of individuals with tax relevant links to non-individual entities such as partnerships, trusts and private companies, including superannuation funds.

We also found the AIIR data contains investment income information that is relevant to the tax payable by non-individuals. The AIIR data has other potential uses apart from the forgoing. For example, it has a role in the identification of individuals, as well as non-individuals, that may have lodgement responsibilities with which they have not complied. The AIIR data also has a whole-of-government role. The Child Support Agency (CSA) and the Department of Immigration and Multicultural and Indigenous Affairs (DIMIA) use the AIIR data. The ANAO considers that there is potential for increased whole-of-government use of the AIIR data in the future. This could arise from heightened concerns about the protection of Australia's security and a stronger interest in the verification of recipients' continuing eligibility for financial payments from the Commonwealth.

The ATO has recently established arrangements, including a Data Matching Steering Committee and an AIIR Steering Committee, to improve the co-ordination of activity that has a bearing on the work of Compliance Systems, and to review the management and wider use of the AIIR data across the ATO.

Receipt of AIIR data (Chapter 3)

The receipt of AIIR data refers to the receipt by the ATO of all AIIRs from all investment bodies that have an obligation to provide them. It includes the extent to which the AIIRs received by the ATO comply with the ATO's reporting specifications as well as the extent to which the specifications themselves comply with legislation. Steps taken by the ATO to assist investment bodies understand their reporting obligations are also examined. The ANAO found that the ATO does not have effective strategies to ensure that investment bodies with a legislative obligation to provide an AIIR to the ATO are doing so. Further, there is no systematic lodgement enforcement approach designed to ensure that known investment bodies provide their AIIR to the ATO by the due date. ANAO considers that an effective and adaptable approach to engaging all relevant investment bodies so that they understand and fulfil their reporting responsibilities is required.

The ATO has advised that it now has initiatives underway to improve reporting compliance amongst investment bodies and to improve the integrity and timeliness of the AIIR data sent to the ATO. These include an investment body visit program and a data quality audit that will provide baseline metrics for data quality issues associated with reported AIIR data.

We found that there is scope for a review of the current AIIR reporting requirements that would identify the costs and benefits associated with the ATO continuing to receive information from investment bodies relating to certain classes of investment income. This review could also determine the existence of other investment income types that would usefully be covered by the AIIR reporting requirements.

We also found that there may be scope to move away from the current technology of receipt of data based on computer tapes, disc, and paper, to a technology based on the ATO's secure access to secure websites and the direct down-loading of AIIR information from them. Adoption of this technology could achieve important efficiencies for the ATO and investment bodies.

Management of AIIR data (Chapter 4)

The management of AIIR data refers to the ATO's conversion of data received from investment bodies into a format suitable for income matching and other administrative activities. ANAO found the ATO's identity matching engines, particularly their phonetic scoring capacity,⁶ to be highly effective. Further, they are essential in an environment where the quality of source data continues to be a problem. This is likely to be the case with the AIIR data for some time and is an intrinsic problem with source data provided by the general public, especially to a revenue collection agency.

As part of our review of the ATO's management of AIIR data we looked at the capacity of the ATO to differentiate variations in the quality of the identity matching that is applied to AIIR records. We found that up to 82 per cent of all AIIR records will receive a high quality match. With regard to the lower grades of identity match, we consider that, through implementing ongoing analysis of unmatched and low quality matched AIIR data, the ATO will gain valuable insights into specific data quality areas requiring its attention.

The ATO has recently established initiatives to use the AIIR data to assist in the identification, and bringing to account, of those individuals and other entities that are operating outside the tax system.

ANAO considers that the improvements now required in the ATO's management of the AIIR data are more likely to come from improvements in the quality of the data, rather than refinements in the power of the identity matching engines. In this regard, we consider there is scope for the ATO to implement mandatory data quality standards.

Use of AIIR data (Chapter 5)

Given the large volume of the AIIR data, the more that the ATO can use it systematically in computer-based income matching systems, the greater the value that the ATO might achieve from the data. The AIIR data can also be used in other ways as well, such as for lodgement enforcement and compliance investigations.

We found that only the PTax business line had established a systematic income matching capability. Whilst the Small Business business line and Ptax had conducted pilot studies of a non-individual matching capability in the 1990s, the ATO did not proceed to establish this capability. ANAO accepts that the ATO, in carrying out the non-individual matching capability pilot projects in the SB and Ptax business lines, made a reasonable effort to test the usefulness of income matching of non-individual taxpayers.

The ATO considers that the absence of a non-individual matching capability does not mean that the ATO is leaving significant risks untreated. The ATO has explained that the tax affairs of non-individual entities are more extensively risk assessed, and that the ATO's active compliance coverage in these markets is also more extensive and intensive than it is for individuals. The ATO considers that the pilot projects justified the ATO not proceeding with a non-individual matching capability at that time. The ATO has pointed out that the decisions it makes about the redevelopment of its Information Technology capability are also made on a risk management basis. In this regard, the ATO notes that there is always a much bigger demand for IT development than can be met with available resources. The ATO advises that of highest priority in determining investments, is the implementation of new Government policy and sustaining current systems. The ATO considers that the issues associated with the establishment of a non-individual matching capability cannot be easily addressed with a simple resource shift and some systems redevelopment.
If the ATO is to make the best use of the AIIR data, it will be necessary to achieve at least a satisfactory proportion of the compliance value of the data.⁷ To date, the ATO has not been able to estimate this value.

In the absence of valid estimates of the AIIR's compliance value, it is difficult for the ATO to manage and use the AIIR data strategically; that is, on a whole of ATO basis and including all categories of investment income and taxpayer. The ANAO considers that a strategic approach is necessary to obtain the best value from the data. The extent to which the ATO is able to achieve, cost-effectively, the compliance value inherent in the AIIR data, depends on the ATO's appreciation of, and capacity to measure, the compliance value of the data in terms of total Potential Revenue Gain (PRG)⁸ for all AIIR entity records. As part of the audit, the reporting gap associated with the interest income of individuals was examined more closely for the AIIR data for the year 2002, the latest year for which there is full set of AIIR data. The aim was to gain a reliable estimate of the compliance value attached to this subset of the AIIR data.

We found that the existing case selection methodology of the IMS case selection team which aims to separate available high quality cases into primary case categories and exclusion categories, prohibits any such estimation. We also found that a proportion of primary cases selected for actioning was shown to be unproductive. In addition, we consider that there may be significant amounts of PRG associated with several of the exclusion categories.

The ATO accepts that, within its exclusion categories, there will be cases that are productive and worth pursuing by case actioning teams. The ATO advised that it plans to undertake fresh scoping audits on each exclusion category aimed at updating its risk assessment perspective. This will be used to update existing case selection practices.

The ANAO notes that additional revenue may be recoverable from individual taxpayers in relation to discrepancies for other classes of investment income, such as dividends and distributions from partnerships, trusts (including unit trusts), and private companies.

The ANAO also notes that, during the current financial year, the ATO will consider a proposal from the relevant business lines for increased funds for additional case actioning of matched discrepant individuals and non¬individuals, along with not lodged and not registered cases.

Recommendations

The ANAO made eight recommendations aimed at improving the ATO management and use of the AIIR data in taxation administration. The ATO has agreed with all of these recommendations.

Summary of ATO's response

We are pleased that the ANAO concludes that our AIIR data-matching program is an effective compliance tool with regard to the individual taxpayer market segment. It is important to have this confirmation given that the prime purpose for which we developed our data matching capability was to treat risk in the individual market.

We also note ANAO acknowledgement of the effort we are making to refine that capability and possibly extend it into other compliance markets to complement our other important compliance efforts. Any decision to invest in an extension of the current capability will need to take account of the total compliance risks that warrant treatment, as well as the choices we face with other systems investment decisions, especially given our current Change Program.

While we view our data capture/matching processes and systems as already highly effective compliance tools, we believe the ANAO's recommendations set an appropriate direction to assist us to build on that capability and proficiency.

Footnotes

1 Australian Taxation Office, Annual Report 2002–03, p.43. This amount includes collections of $30.7 billion for the Goods and Services Tax and $21.6 billion for Excise and other indirect taxes.

2 Refer to Taxation Laws Amendment (Tax File Numbers) Bill 1988, and House of Representatives Hansard pps 858 to 861, 1 September 1988.

3 Refer to House of Representatives Hansard p. 860, 1 September 1988

4 The most significant of these are large financial institutions and public companies. However, the legislation also applies to unit trusts, loans made to government bodies or a body corporate, solicitors who invest client monies, some real estate agents, and betting investment bodies.

5 The six categories of investor identified in tax legislation are: individuals, partnerships, trusts, companies, superannuation funds and government agencies. The legislation requires that the AIIR data include the details of all of the investment income (as legislatively defined) of all entities in each of these six categories that receive the investment income from any of the investment bodies covered by the legislation.

6 This is achieved through the use of a fully customised edit list (name dictionary) capable of identifying typing and spelling errors, ethnic name anglicisation issues, orthographic and phonetic misinterpretation, truncation and abbreviation, missing words and extra words and word sequence variation.

7 “Compliance value” has both quantitative and qualitative dimensions. Quantitatively, it is the estimated amount of additional revenue that may be collected from the recovery of tax due in relation to investment income discrepancies disclosed by the AIIR data. An aggregate measure is the compliance rate, which is the percentage of the total investment income that should have been disclosed on tax returns that was actually disclosed on those returns. Qualitatively, it includes improvements to compliance that may be attributed to the AIIR data. The ATO uses Potential Revenue Gain to estimate additional revenue collectable from discrepant income, which is the result produced by comparing the Employment, Welfare, Investment and other amounts reported about the taxpayer by a relevant reporting body against the tax return amounts disclosed by a taxpayer.

8 The Potential Revenue Gain (PRG) is the net amount of revenue that could be recovered if the ATO were to take effective administrative action in relation to the transaction. The PRG is the calculated net tax payable after the omitted income is added, less the amount already paid. The PRG is calculated as the difference between tax that would be paid on the income resulting from adding the discrepant income to the individual's taxable income as determined from the individual's tax return and the tax that was actually paid prior to the identification of the discrepancy.