Note: Where available, the PDF/Word icon below is provided to view the complete and fully formatted document
Tax and Revenue—House of Representatives Standing Committee—The Tax Expenditures Statement—Report, November 2015


Download PDF Download PDF

The Parliament of the Commonwealth of Australia

The Tax Expenditures Statement

House of Representatives Standing Committee on Tax and Revenue

November 2015 Canberra

© Commonwealth of Australia 2015

ISBN 978-1-74366-415-5 (Printed version)

ISBN 978-1-74366-416-2 (HTML version)

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Australia License.

The details of this licence are available on the Creative Commons website: http://creativecommons.org/licenses/by-nc-nd/3.0/au/.

Contents

Chair’s foreword ................................................................................................................................. vii

Membership of the Committee ............................................................................................................ ix

Terms of reference .............................................................................................................................. xi

List of abbreviations ........................................................................................................................... xii

List of recommendations ................................................................................................................... xiii

THE R EPO RT

1 The inquiry ........................................................................................................... 1

Origins of the inquiry ................................................................................................................ 1

Inquiry overview ........................................................................................................................ 2

Committee comment ................................................................................................................. 3

2 Background on tax expenditures ....................................................................... 5

Introduction ............................................................................................................................... 5

History ......................................................................................................................................... 5

Advantages and disadvantages of tax expenditures ................................................................... 5

Estimating methods ..................................................................................................................... 7

Tax expenditures in Australia .................................................................................................. 9

Legal requirements ..................................................................................................................... 9

The Statement........................................................................................................................... 11

Past reviews ............................................................................................................................ 14

The 2008 ANAO audit ............................................................................................................... 15

The Parliament’s follow-up inquiry ............................................................................................ 16

iv

The Australia’s Future Tax System (AFTS) Review .................................................................. 16

The 2013 ANAO audit ............................................................................................................... 17

Committee comment ................................................................................................................. 18

3 Issues raised during the inquiry....................................................................... 19

Is the Statement effective? ..................................................................................................... 19

The positives ............................................................................................................................. 19

The limitations ........................................................................................................................... 20

Policy reviews of tax expenditures ........................................................................................ 23

Background ............................................................................................................................... 23

Analysis ..................................................................................................................................... 25

Committee comment ................................................................................................................. 26

Aggregating tax expenditures ................................................................................................ 27

Background ............................................................................................................................... 27

Analysis ..................................................................................................................................... 28

Committee comment ................................................................................................................. 29

The benefits of superannuation tax concessions ................................................................ 31

Committee comment ................................................................................................................. 32

Benchmark for savings .......................................................................................................... 34

Background ............................................................................................................................... 34

Analysis ..................................................................................................................................... 36

Committee comment ................................................................................................................. 37

Benchmarks and structural features of the tax system ....................................................... 38

Background ............................................................................................................................... 38

Analysis ..................................................................................................................................... 41

Committee comment ................................................................................................................. 42

Caveats in the Statement........................................................................................................ 43

Background ............................................................................................................................... 43

Analysis ..................................................................................................................................... 44

Committee comment ................................................................................................................. 45

Quality of the estimates .......................................................................................................... 46

Background ............................................................................................................................... 46

Analysis ..................................................................................................................................... 48

v

Committee comment ................................................................................................................. 49

Longer term estimates ............................................................................................................ 50

Committee comment ................................................................................................................. 51

Revenue gain estimates ......................................................................................................... 52

Background ............................................................................................................................... 52

Analysis ..................................................................................................................................... 54

Committee comment ................................................................................................................. 55

Memorandum items ................................................................................................................ 56

Committee comment ................................................................................................................. 57

AP PE NDI CE S

Appendix A - Submissions ..................................................................................... 59

Appendix B - Exhibits .............................................................................................. 61

Appendix C - Public hearings ................................................................................. 63

Appendix D - Implementation of past recommendations ..................................... 65

vi

Chair’s foreword

Tax expenditures calculate the effects of tax concessions provided to particular groups of taxpayers. Often, a tax concession has a particular policy goal, such as encouraging a valuable activity or improving equity.

In the late 1960s, the observation was made that tax expenditures are similar to government spending and deserve comparable scrutiny. Following a report by the House of Representatives Standing Committee on Expenditure in 1982, Australia’s first comprehensive Tax Expenditures Statement (the Statement) was published in 1986.

Publishing a regular listing of tax expenditures is now a legal requirement under the Charter of Budget Honesty Act 1998.

The Tax Expenditures Statement has been reviewed several times in the past decade by the Australian National Audit Office (2008 and 2013), the Joint Committee of Public Accounts and Audit (2009), and as part of the Australia’s Future Tax System Review (2009).

The Committee conducted a thorough review of the Tax Expenditures Statement, building on the work of these previous inquiries. We found that the Statement achieved its stated goal of facilitating scrutiny of tax expenditures and informing debate on the tax system. However, there were areas in which efficiencies could be found, and the Statement improved.

For example, we supported Treasury’s suggestion that it devote fewer resources to estimating smaller, more technical expenditures, and shift these resources to other parts of the Statement. There is also scope to conduct related work, such as prioritised reviews of tax expenditures, calculating a meaningful aggregate of tax expenditures, and calculating the longer term costs and benefits of superannuation.

viii

Further, we found that producing the Statement can be labour-intensive, and that there are significant costs to enhancing it. While Treasury draws on highly skilled staff to compile the Statement, greater consultation with stakeholders may have benefits. These may include: improving the data; establishing benchmarks; and determining which expenditures should receive longer-term estimates, or should be calculated using the more sophisticated revenue gain method.

The Committee has also made several recommendations that would improve the explanation of data and methods used in the preparation of the Statement, as well as increasing the visibility of its limitations.

The Statement is held in high regard around the world. It is an integral part of budget transparency, and it informs the national debate on taxation. While it can be improved in places, I commend Treasury on producing a document of high quality and technical complexity.

I would like to thank my fellow Committee Members for their assistance in the inquiry and the support of the Secretariat. I would also like to thank the parliamentary agencies of the Australian National Audit Office and the Parliamentary Budget Office. Their expertise and access to data mean that parliamentary committees are now well-equipped to undertake technical inquiries of this nature.

Finally, I extend my thanks to the agencies, organisations and individuals that contributed to the inquiry through submissions or evidence at public hearings.

Bert van Manen MP Chair

Membership of the Committee

Chair Mr Bert van Manen MP

Deputy Chair Dr Jim Chalmers MP (until 22/10/15)

Hon Bernie Ripoll MP (from 22/10/15)

Members Mr John Alexander MP (until 12/10/15) Ms Clare O’Neil MP

Hon Bronwyn Bishop MP (from 9/11/15) Mr Michael Sukkar MP

Ms Terri Butler MP Mr Angus Taylor MP (until 20/10/15)

Mr Ian Goodenough MP (until 19/10/15) Mr Tim Watts MP

Mr Andrew Hastie MP (from 19/10/15) Mr Matt Williams MP

Mr Steve Irons MP (from 19/10/15)

x

Committee Secretariat

Secretary Mr David Brunoro (from 29/10/15)

Ms Susan Cardell (to 28/10/15)

Inquiry Secretary Mr David Monk

Research Officer Mr Shane Armstrong

Administrative Officer Ms Tamara Palmer

Terms of reference

The Committee will inquire into the Tax Expenditures Statement, including the recommendations made by:

 the Australian National Audit Office in its reports of May 2008 and May

2013

 the Joint Committee of Public Accounts and Audit in its report of June

2009

 the Australia’s Future Tax System Review of December 2009 (recs 137-139).

The Committee will not examine the policy arguments around individual tax expenditures.

List of abbreviations

ABS Australian Bureau of Statistics

AFTS Australia’s Future Tax System

ANAO Australian National Audit Office

ATO Australian Taxation Office

CAANZ Chartered Accountants Australia and New Zealand

CIS Centre for Independent Studies

DPL Department of the Parliamentary Library

DSICA Distilled Spirits Industry Council of Australia

FSC Financial Services Council

GST Goods and Services Tax

JCPAA Joint Committee of Public Accounts and Audit

MYEFO Mid-Year Economic and Fiscal Outlook

OECD Organisation for Economic Cooperation and Development

PBO Parliamentary Budget Office

SMSF Self-managed superannuation fund

TES Tax Expenditures Statement

List of recommendations

3 Issues raised during the inquiry

Recommendation 1

That Treasury devote fewer resources to estimating smaller, technical tax expenditures. This could involve reviewing them less frequently and reporting them as a range.

Recommendation 2

That the Government give Treasury an ongoing mandate to conduct prioritised reviews of tax expenditures and publish the results — building upon Recommendation 1 from the ANAO’s 2008 audit report.

Recommendation 3

That Treasury develop a method of aggregating tax expenditures to support its obligation under section 12 of the Charter of Budget Honesty Act 1998 to give an overview of tax expenditures in the Budget Papers. The overview can be tailored to the needs of the Budget Papers, rather than exactly reflecting the Statement. Consultation may assist on this.

Recommendation 4

That Treasury model the long run interactions between superannuation and the age pension, develop present value estimates of the future costs and benefits of superannuation and its tax concessions, and publish the results. The Tax Expenditures Statement should reference or briefly summarise this research in the context of the superannuation estimates.

This should occur at least once every five years and also following significant changes in policy.

xiv

Recommendation 5

That Treasury include in the Statement information on revenue collected through superannuation taxation, which could constitute a reference in the table of large tax expenditures and further information later in the document.

Recommendation 6

That Treasury retain the comprehensive income tax benchmark for savings in the Tax Expenditures Statement.

Recommendation 7

That Treasury:

 incorporate the capital gains tax exemption for the main residence into the benchmark

 develop a transparent process and criteria to assist consideration of benchmarks which reflect the practical possibility of a tax concession being abolished

 consult with stakeholders on the benchmarks used in the Tax Expenditures Statement.

Recommendation 8

That Treasury consider ways of increasing the visibility or warnings in the Statement to better draw the attention of readers to the Statement’s limitations.

Recommendation 9

That Treasury include in the Statement an informative explanation of data and methods for large estimates and when estimates are substantially revised, either in terms of size or reliability.

Recommendation 10

That Treasury and the Australian Taxation Office consult with stakeholders on possible data sources and apply a prioritised approach to data improvements.

Recommendation 11

That Treasury produce longer term estimates for tax expenditures where they are expected to differ significantly from the forward estimates presented in the Statement. Deciding which tax expenditures should receive this treatment could be subject to materiality considerations and consultation.

xv

Recommendation 12

That Treasury gradually expand the number of revenue gain estimates that it calculates for the Tax Expenditures Statement, with a focus on larger tax expenditures and those relevant to public debate. Stakeholder consultation may assist in this process.

Recommendation 13

That Treasury include selected memorandum items in the Tax Expenditures Statement. The decision on which items to include could be based on their size, relevance to public debate, and stakeholder consultation.

xvi

1

The inquiry

Origins of the inquiry

1.1 The inquiry was in response to two audit reports by the Australian National Audit Office (ANAO) into the Tax Expenditures Statement (the Statement) in 2008 and 2013. In the first report, the ANAO made six overall recommendations, comprising 11 sub-recommendations. Treasury accepted these, albeit with qualification in four cases. In 2009, the Joint Committee of Public Accounts and Audit (JCPAA) conducted a follow-up inquiry and made three recommendations that built on those by the ANAO.1

1.2 In the second report, the ANAO found that only a few of the earlier recommendations had been fully implemented, and that one had been implemented but then discontinued. The 2013 ANAO report also covered recent developments in the Statement, including the JCPAA report and the Australia’s Future Tax System Review. It made one additional recommendation.

1.3 The Statement has an important role in improving the transparency of the tax system. Given that some ANAO recommendations were not implemented, the Committee considered that an inquiry into the Statement was warranted.

1.4 Under Standing Order 215(c), House committees can inquire into Auditor-General’s reports and agency annual reports. The main proviso is

1 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08; JCPAA, Report 414: Review of Auditor-General’s Reports tabled between August 2007 and August 2008, 2009; ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13.

2 THE TAX EXPENDITURES STATEMENT

that the Speaker must refer the relevant agency to the Committee under a schedule tabled in the House. In this Parliament, all schedules to date have referred Treasury and the Australian Taxation Office (ATO) to the Committee.

1.5 A number of other conditions apply. For example, a House committee can only inquire into the current annual report for a particular agency. Treasury’s 2014 annual report referred to the Tax Expenditures Statement on page 40.

1.6 Further, a House committee can only inquire into an audit report if it has notified the JCPAA of its intention to do so. The Committee Chair wrote to the Chair of the JCPAA along these lines on 17 June 2015.

1.7 House committees do not need both avenues to self-refer an inquiry. The Committee raises them to indicate that it had two separate ways of initiating an inquiry.

1.8 On 17 June 2015, the Committee resolved to inquire into the Tax Expenditures Statement and adopted terms of reference for the inquiry. The full terms of reference were:

The Committee will inquire into the Tax Expenditures Statement, including the recommendations made by:

 the Australian National Audit Office in its reports of May 2008

and May 2013

 the Joint Committee of Public Accounts and Audit in its report

of June 2009

 the Australia’s Future Tax System Review of December 2009

(recs 137-139).

The Committee will not examine the policy arguments around individual tax expenditures.

Inquiry overview

1.9 The inquiry was advertised by media release, social media, and direct mail. The Committee specifically sought submissions from the Assistant Treasurer, parliamentary agencies, academics, tax representative bodies, and non-government organisations, including think-tanks.

1.10 The Committee received 13 submissions and two supplementary submissions. The submissions are listed at Appendix A.

1.11 The Committee received 15 exhibits. These are listed at Appendix B.

THE INQUIRY 3

1.12 The Committee held six public hearings in Canberra, which included three teleconferences. Public hearing details are listed at Appendix C.

1.13 The remainder of the report is organised as follows:

 Chapter 2 gives an overview of tax expenditures, the information

provided in the Statement, and the outcomes of past reviews. Further information about past reviews is in Appendix D.

 Chapter 3 examines the issues raised in the inquiry and includes the

Committee’s recommendations.

Committee comment

1.14 In undertaking the inquiry, the Committee appreciated the assistance it received from the ANAO and the Parliamentary Budget Office. These parliamentary agencies have significant technical expertise and their support means that parliamentary committees are well-placed to conduct this type of inquiry. The Committee looks forward to working with these agencies in the future.

4 THE TAX EXPENDITURES STATEMENT

2

Background on tax expenditures

Introduction

History

2.1 The concept of tax expenditures was initially proposed by Mr Stanley Surrey of the United States Treasury Department in 1967. He wished to make the point that tax concessions provided to particular groups, industries, or classes of taxpayers, was similar to giving them a government grant or other payment.1

2.2 Surrey’s initial goal was to limit the number of tax expenditures. The idea was that, once the tax expenditures had been brought to the attention of Congress, it would either remove them or replace them with a spending program. The ultimate aim was to improve the equity, efficiency and simplicity of the tax system. However, this did not eventuate. Tax expenditures have become more numerous in the United States, although tax expenditure reports are seen as an important tool in tax analysis and tax policy.2

Advantages and disadvantages of tax expenditures

2.3 Tax expenditures can achieve policy goals. These include:

1 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 9. 2 Edward D. Kleinbard, Rethinking Tax Expenditures, 2008, Address to the Chicago-Kent College of Law Federal Tax Institute, pp. 1-5, accessed from https://www.jct.gov/publications.html on 26 June 2015.

6 THE TAX EXPENDITURES STATEMENT

 encouraging the private sector to move into a field where the

government is a major service provider

 reducing government supervision of such spending3

 promoting desired economic behaviour, such as encouraging saving

through superannuation

 improving equity, such as the GST exemption on unprepared food.4

2.4 The criticisms made against tax expenditures are that they can:

 be restricted in their operation by the design of the tax system

 be ineffective if there are conflicting tax provisions or government

programs

 reduce economic efficiency through economic distortions

 reduce equity where they depend on the taxpayer’s marginal tax rate

and this rate varies across taxpayers

 erode the revenue base if they are large

 increase tax complexity and increase compliance and enforcement costs

 more easily evade monitoring and analysis, compared with spending

that must be approved by the legislature or listed in an agency annual report.5

2.5 The effectiveness of tax expenditures was raised in the inquiry. Associate Professor Julie Smith stated that most research indicates that ‘tax expenditures are generally not justifiably delivered through the tax system.’ The Australian National Audit Office (ANAO) noted that past

3 Zhicheng Li Swift, Hana Polackova Brixi, and Christian Valenduc, ‘Tax Expenditures: General Concept, Measurement, and Overview of Country Practices,’ p. 3, in World Bank, Tax Expenditures - Shedding Light on Government Spending through the Tax System: Lessons from Developed and Transition Economies, 2004; AFTS Review, Report to the Treasurer, Part Two - Detailed Analysis, 2009, pp. 725-28. 4 Robert Carling, Right or Rort: Dissecting Australia’s Tax Concessions, 2015, Research Report 2,

Centre for Independent Studies, p. 2. 5 Zhicheng Li Swift, Hana Polackova Brixi, and Christian Valenduc, ‘Tax Expenditures: General Concept, Measurement, and Overview of Country Practices,’ p. 5, in World Bank, Tax Expenditures - Shedding Light on Government Spending through the Tax System: Lessons from

Developed and Transition Economies, 2004; AFTS Review, Report to the Treasurer, Part Two - Detailed Analysis, 2009, pp. 725-28.

BACKGROUND ON TAX EXPENDITURES 7

reviews in Australia have recommended that up to half of tax expenditures be either abolished or modified.6

2.6 It would appear that tax expenditures have their place and can achieve useful policy goals, but that they also have risks which need to be managed. The publication of the Tax Expenditures Statement is an important first step in understanding and managing these risks.

Estimating methods

2.7 Estimating the size of a tax expenditure is difficult because it is an attempt to determine what might have happened if a tax law was not in place. There are few past examples to draw on.

2.8 The usual approach for estimating a tax expenditure, both in Australia and overseas, is the revenue forgone method.7 It reflects existing use of the tax expenditure and assumes that taxpayer behaviour will not change if the measure is revoked, and the rest of the tax system will also not change. Therefore, if $10 billion were spent on a GST-exempt good or service in a year, the tax expenditure would be $1 billion. It is the simplest method.

2.9 A more sophisticated approach is to estimate taxpayers’ reaction to the removal of a tax expenditure and this is what the revenue gain method seeks to do. The method also seeks to include second order effects such as interactions with other parts of the tax and transfer system. However, the method does not consider the interactions with other tax exceptions —it assumes all other tax exceptions are unchanged. It also requires better data than the revenue forgone method.8

2.10 Generally, the revenue gain method produces a smaller tax expenditure than revenue forgone. This is due to the good or service becoming more expensive and taxpayers purchasing less of it. For example, in the 2014 Statement, the two estimates for the exemption from interest withholding tax on certain securities are $1.3 billion and $1.9 billion respectively. Where behaviour is not expected to change much, the estimates can be

6 Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 4; Mr Andrew Morris, ANAO, Committee Hansard, 9 September 2015, Canberra, p. 6. 7 Zhicheng Li Swift, Hana Polackova Brixi, and Christian Valenduc, ‘Tax Expenditures: General Concept, Measurement, and Overview of Country Practices,’ p. 12, in World Bank, Tax

Expenditures - Shedding Light on Government Spending through the Tax System: Lessons from Developed and Transition Economies, 2004. 8 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 54-55.

8 THE TAX EXPENDITURES STATEMENT

close. The two estimates for the GST exemption for food are $6.3 billion and $6.4 billion.9

2.11 Revenue gain estimates can be complex.10 The Grattan Institute gave the Committee an overview of the process for superannuation. If the tax on contributions were abolished, the factors to take into account would include:

 super guarantee contributions would continue, since they are a legal

obligation

 Treasury assumes that most of the voluntary concessional contributions

and non-voluntary concessional contributions would no longer be made

 Treasury assumes that these funds would be diverted to other

investments without the generous concessions of superannuation

 This would increase taxpayers’ income and be taxed at the highest

marginal rate, resulting in an estimate close to the revenue forgone estimate.11

2.12 Despite their complexity, the advantage of revenue gain estimates is that they are easier to interpret. They represent the best estimate of the cost of a particular concession on the Budget because they are an estimate of what government would gain if the concession were removed.

2.13 The third method is outlay equivalent, where revenue forgone estimates are used to generate an estimate of what it would cost to generate the same effect if the tax expenditure were a spending program. Often, direct expenditures are taxable, but tax expenditures are not. Therefore, a tax expenditure will have a larger effect than direct spending of the same size.12 It would be expected that outlay equivalent method would often give larger estimates than the other two methods.

2.14 In 1986, the Economic Planning and Advisory Council stated that the outlay equivalent method has the potential to improve the comparability

9 Treasury, Tax Expenditures Statement 2014, 2015, p. 7. 10 OECD, Tax Expenditures in OECD Countries, 2010, pp. 151-52. 11 Mr Brendan Coates, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 6. 12 Colin Brown, ‘Tax Expenditures in Australia,’ pp. 55, 59, in World Bank, Tax Expenditures -

Shedding Light on Government Spending through the Tax System: Lessons from Developed and Transition Economies, 2004.

BACKGROUND ON TAX EXPENDITURES 9

of tax expenditures and spending programs. It could also allow for the better integration of tax expenditures in the Budget process.13

2.15 In 2010, the OECD reported that Sweden provides outlay equivalent estimates. The United States used to report outlay equivalent estimates.14

2.16 Although some individual tax expenditures may be simple to estimate, the large number of them can mean that preparing a comprehensive report each year is labour intensive.15 The 2014 Statement lists 297 tax expenditures.

Tax expenditures in Australia

Legal requirements

2.17 The Charter of Budget Honesty Act 1998 requires the Government to publish detailed information on tax expenditures. Division 2 of the Act requires the Treasurer to publish and table a mid-year economic and fiscal outlook report, which must ‘contain a detailed statement of tax expenditures, presenting disaggregated information on tax expenditures.’16 The report is titled the Mid-year Economic and Fiscal Outlook (MYEFO).

2.18 Treasury does not meet this obligation. The Statement is presented as a separate document, rather than being included in MYEFO. However, it is published within the time-frame required for MYEFO. Treasury explained that MYEFO and the Statement are both resource intensive, and it would be difficult to produce the two together:

… I think the publication of a very detailed tax expenditure statement at the time of the MYEFO is unlikely to be practical, in our view, particularly as some policy decisions may not be taken until late in the process of preparing that statement. We have little time to actually prepare any estimates on the basis of those decisions. That has not been done by any government to date, and

13 Referenced from ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 63. 14 OECD, Tax Expenditures in OECD Countries, 2010, p. 151. 15 Zhicheng Li Swift, Hana Polackova Brixi, and Christian Valenduc, ‘Tax Expenditures: General

Concept, Measurement, and Overview of Country Practices,’ p. 12, in World Bank, Tax Expenditures - Shedding Light on Government Spending through the Tax System: Lessons from Developed and Transition Economies, 2004; Treasury, Tax Expenditures Statement 2014, 2015. 16 Clause 16(1)(b) of the Charter of Budget Honesty Act 1998.

10 THE TAX EXPENDITURES STATEMENT

I think fundamentally it reflects the fact that MYEFO and tax expenditure statements are both highly resource intensive to produce. Producing them simultaneously would be quite difficult for us. Nevertheless, we do produce the TES by the end of January each year and shortly after MYEFO. That is based on decisions taken up to the last MYEFO statement.17

2.19 Statements have been published every year since the Act commenced. The exception was 1998-99, due to the implementation of major tax reforms and the introduction of accrual budgeting.18

2.20 Another obligation under the Act is that MYEFO, and hence the statement on tax expenditures, must be based on external reporting standards. This implies that they would be independently set. Further, different groups could take the same raw data and develop similar tax expenditure estimates. However, no such standard currently exists.19

2.21 Finally, the Act requires the Treasurer to table a budget economic and fiscal outlook report. The report must contain ‘an overview of the estimated tax expenditures for the budget year and the following 3 financial years.’20 The 2015 Budget reproduced four years of information from the 2014 Statement in relation to the 25 largest measured tax expenditures and the two largest negative tax expenditures. The information was not updated for the effects of Budget decisions.21

2.22 Previous Budgets have included more information. Up to 2012, they included aggregate information of tax expenditures.22 On the one hand, this is technically incorrect because of the interactions between tax expenditures. On the other hand, aggregate data gives an indication of their size. Previous Budgets have occasionally discussed tax expenditures in relation to new measures.23 The practice of excluding Budget decisions

17 Mr Steve French, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 2. 18 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 41. 19 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 13; ABS, Submission 1, p. 1; Clause 16(3)(a) of the Charter of Budget Honesty Act 1998. 20 Clause 12(1)(d) of the Charter of Budget Honesty Act 1998. 21 The Hon. J.B. Hockey MP, Treasurer, and Senator the Hon. Mathias Cormann, Minister for

Finance, Budget Strategy and Outlook: Budget Paper No. 1 2015-16, 2015, pp. 4-20, 4-21. 22 For example, The Hon. Wayne Swan MP, Deputy Prime Minister and Treasurer, and Senator the Hon. Penny Wong, Minister for Finance and Deregulation, Budget Strategy and Outlook: Budget Paper No. 1 2012-13, 2012, p. 5-49. 23 For example, the Hon. Wayne Swan MP, Deputy Prime Minister and Treasurer, and Senator

the Hon. Penny Wong, Minister for Finance and Deregulation, Budget Strategy and Outlook: Budget Paper No. 1 2012-13, 2012, p. 5-15.

BACKGROUND ON TAX EXPENDITURES 11

from the tax expenditure information in the Budget Papers appears to be long standing.24

The Statement

2.23 The Statement reflects Government policy up to and including the matching MYEFO. The Statement’s intention is to inform debate and facilitate parliamentary scrutiny:

The TES is intended to facilitate scrutiny of tax expenditures by Parliament and parliamentary committees, the media and general public. Transparent reporting of tax expenditures also helps inform debate on the efficiency and equity of the tax system.25

2.24 In evidence, the ANAO argued that Treasury needed to make further improvements to the Statement to meet these objectives.26

2.25 The Statement explains the revenue forgone and revenue gain methods and lists a number of cautions in relation to the tax expenditure estimates:

 judgment is required in determining the benchmark; it may be

contentious and may change over time

 revenue gain estimates are problematic because:

⇒ abolishing the tax expenditure in question may be implausible

⇒ there may not be useful information on how taxpayers would react if

the tax expenditure were abolished

⇒ judgments are required about what the likely policy settings would

be if a tax expenditure were abolished

 tax expenditures are not comparable over time because benchmarks

change, new ones are identified, and data or modelling methods change

 direct expenditure estimates and tax expenditure estimates are often

not directly comparable due to tax effects or because they measure different things

 tax expenditures cannot be added together because removing one item

can change how others are used.27

24 For example, The Hon. Peter Costello MP, Treasurer, and Senator the Hon. Nick Minchin, Minister for Finance and Administration, Budget Strategy and Outlook: Budget Paper No. 1 2005-06, 2005, p. 5-33. 25 Treasury, Tax Expenditures Statement 2014, 2015, p. 3. 26 Mr Mark Simpson, ANAO, Committee Hansard, 9 September 2015, Canberra, p. 3. 27 Treasury, Tax Expenditures Statement 2014, 2015, pp. 3-5.

12 THE TAX EXPENDITURES STATEMENT

2.26 The Statement then turns to large tax expenditures. It lists five unquantified expenditures, all of which are positive (in favour of taxpayers) except the last, which is negative (in favour of the Government). They are classed as large because they are all predicted to be over $1 billion in size. The Statement lists them as:

 income tax exemption for Commonwealth, State and Territory

public authorities, and State and Territory entities (B3);

 exemption for foreign branch profits from income tax (B11);

 off-market share buy-backs (B28);

 philanthropy — income tax exemption for registered charities,

public educational, scientific and community service entities (B56); and

 quarantining of capital losses (E27).

2.27 The letters and numbers represent the reference code used in the Statement for each tax expenditure. The letter refers to a different tax or benchmark: ‘B’ stands for business income, and ‘E’ refers to capital gains tax. The number refers to its position in the list for each benchmark.28

2.28 The Statement then lists large quantified tax expenditures, 25 of which are positive and two negative. The top 10 positive expenditures and two largest negative expenditures are reproduced on the next page.

28 The other letter codes are: ‘A’ for personal income; ‘C’ for retirement income; ‘D’ for fringe benefits tax; ‘F’ for commodity and other indirect taxes; ‘G’ for natural resources taxes; and ‘H’ for the GST.

BACKGROUND ON TAX EXPENDITURES 13

Table 1.1 Largest measured tax expenditures for 2014-15

Tax expenditure Estimate $m

Revenue

forgone

Revenue gain

Large positive tax expenditures

E6 Capital gains tax main residence exemption - discount component 25,500 n/a

E5 Capital gains tax main residence exemption 20,500 n/a

C3 Concessional taxation of employer superannuation contributions 16,300 15,550

C6 Concessional taxation of superannuation entity earnings 13,400 11,750

H28 GST - Food 6,400 6,300

E11 Capital gains tax discount for individuals and trusts 5,800 n/a

H16 GST - Education 3,950 3,550

H2 GST - Financial supplies - input taxed treatment 3,550 3,550

H19 GST - Health - medical and health services 3,550 3,500

C5 Concessional taxation of non-superannuation benefits 2,700 2,700

Large negative tax expenditures

F24 Customs duty -2,550 -2,550

F12 Higher rate of excise levied on cigarettes, ≤ 0.8 grams of tobacco -1,840 n/a

Source Treasury, Tax Expenditures Statement 2014, 2015, p. 7.

2.29 Chapter 2 of the Statement lists each individual tax expenditure with explanatory information. Of the 297 tax expenditures in the 2014 Statement, 145 are unquantified.29 The remainder have eight years of estimates using the revenue forgone method. An example is provided below.

Table 1.2 Example of a revenue forgone entry in the Tax Expenditures Statement

H16 Education Education ($m)

2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 2016-17 2017-18

2,800 3,050 3,350 3,600 3,950 4,300 4,700 5,150

Tax expenditure type: Exemption 2013 TES Code: H17

Estimate reliability: Medium

Commencement date: 1 July 2000 Expiry date:

Legislative reference: Section 38-50 of the A New Tax System (Goods and Services Tax) Act 1999

Certain education supplies are GST-free. These include education courses, directly related administrative services, curriculum-related goods, student accommodation for students attending a primary, secondary or special education course, excursions and field trips and supplies related to the recognition of prior learning.

Source Treasury, Tax Expenditures Statement 2014, 2015, p. 115. ‘H’ in the title refers to the GST.

29 CAANZ, Submission 3, p. 5.

14 THE TAX EXPENDITURES STATEMENT

2.30 Chapter 3 gives estimates using the revenue gain method for 10 tax expenditures. The corresponding revenue gain entry for the education estimate above is in table 1.3.

Table 1.3 Example of a revenue gain entry in the Tax Expenditures Statement

H16: GST - Education

Estimates Revenue forgone estimate ($m) Revenue gain estimate ($m)

2014-15 2015-16 2016-17 2017-18 2014-15 2015-16 2016-17 2017-18

3,950 4,300 4,700 5,150 3,550 3,850 4,250 4,600

Reason for difference Removing the GST exemption for education would be expected to decrease demand for education services. This is primarily due to an expected fall in demand for private education and ‘discretionary

courses’.

Source Treasury, Tax Expenditures Statement 2014, 2015, p. 124.

2.31 Appendix A in the Statement comprises technical notes explaining the reliability of the estimates, the calculation of the various benchmarks, and modelling techniques. Appendix B lists tax expenditures that have been abolished or changed since the previous Statement.

2.32 Up until 2012, the Statement included aggregate information of tax expenditures.30 As noted above, this has the disadvantage of being technically incorrect because tax expenditures interact with each other. However, aggregate data gives a rough indication of the overall scale of tax expenditures.

Past reviews

2.33 In its 2008 report, the ANAO noted that there had been six public sector reviews of tax expenditures between 1970 and 2000. The reviews’ recommendations were rarely implemented and, therefore, they tended to make similar findings and recommendations. The ANAO summarised the reviews’ findings as:

 poorly defined aims [for tax expenditures];

 inadequate methods, information and data with which to

estimate the cost and evaluate the effectiveness and efficiency of tax expenditures;

 insufficient budgetary scrutiny and consideration, both within

government and by Parliament; and

30 Treasury, Tax Expenditures Statement 2012, 2013, pp. 4-5.

BACKGROUND ON TAX EXPENDITURES 15

 lack of regular and systematic review.31

2.34 The ANAO stated that the only report to have some recommendations implemented was that by the House of Representatives Standing Committee on Expenditure in 1982.32 The inquiry found that the information Parliament was receiving on tax expenditures was improving, but it was still inadequate. The Committee recommended improvements to reporting, including:

 a comprehensive list of tax expenditures, including costs and objectives

 tax expenditures be cross-classified with the departments responsible

for direct spending equivalents

 ministers’ second reading speeches that relate to a tax expenditure

should include an explanation why it is preferred over direct spending.33

2.35 The report was partially implemented. The listing of tax expenditures that it recommended is now the current Tax Expenditures Statement.34

The 2008 ANAO audit

2.36 The ANAO noted that the Charter of Budget Honesty Act 1998 was designed to establish an integrated fiscal framework. One element of this would be for MYEFO to include a detailed statement on tax expenditures based on external reporting standards. However, as a matter of practicality, Treasury had not been able to include the Statement in MYEFO.

2.37 Treasury advised the ANAO during the audit that the best way of controlling tax expenditures was at policy development, where they should be considered in the same way as spending proposals. However, the ANAO noted that past practice had been inconsistent. It also stated that:

 the benchmarks used to calculate tax expenditures are based on

judgments and can be arbitrary or vary over time

 there were no reporting standards for tax expenditures

31 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 33-34. 32 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 34. 33 House of Representatives Standing Committee on Expenditure, Taxation Expenditures, 1982,

pp. vii, 12. 34 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 34; Julie Smith, Tax Expenditures: the $30 Billion Twilight Zone of Government Spending, 2003, Department of the Parliamentary Library Research Paper No. 8 2002-03, p. 7.

16 THE TAX EXPENDITURES STATEMENT

 there was a substantial number of unidentified tax expenditures

 over 40 per cent of identified tax expenditures were not quantified, and

of those that were, two thirds were not based on reliable estimates.35

2.38 The ANAO stated that a program of ongoing reviews would be beneficial. It made six recommendations (comprising 11 sub-recommendations) to address these issues.36 The recommendations and the extent of their implementation are provided in Appendix D.

The Parliament’s follow-up inquiry

2.39 One of the statutory functions of the Joint Committee of Public Accounts and Audit (JCPAA) is to examine ANAO reports.37 The JCPAA conducts this work through regular hearings with the Auditor-General and agencies subject to audit. That Committee held a hearing with the ANAO and Treasury in 2008 and tabled its report the next year.

2.40 The Committee made recommendations around reporting that supplemented those of the ANAO. The recommendations and their implementation are also given in Appendix D.38

The Australia’s Future Tax System (AFTS) Review

2.41 This inquiry was a broad review of the tax system. It mainly covered tax policy, but it also examined tax administration. Treasury completed the report in December 2009 and it was released the next year.

2.42 Two of the report’s recommendations mirrored those of the ANAO. They covered better integrating tax expenditures in the Budget process and developing reporting standards for tax expenditures. It also recommended legislative change in relation to MYEFO and that the Council of Australian Governments (COAG) should encourage the States and Territories to standardise their reporting on tax expenditures. These recommendations are also listed in Appendix D.

35 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 12-14. 36 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 21-23. 37 Section 8(1)(c) of the Public Accounts and Audit Committee Act 1951. 38 JCPAA, Report 414: Review of Auditor-General’s Reports tabled between August 2007 and August

2008, 2009, pp. 43-57.

BACKGROUND ON TAX EXPENDITURES 17

2.43 The inquiry discussed benchmarks for measuring tax expenditures, but did not make recommendations on this.39

The 2013 ANAO audit

2.44 This follow-up audit noted that progress in implementing the ANAO and JCPAA recommendations had been ‘slow’ with only four of the 11 sub-recommendations being fully implemented. One recommendation, to systematically review tax expenditures, had been implemented then discontinued. The ANAO’s view was that the recommendations were still relevant.40

2.45 Treasury advised the ANAO that, in the face of reduced resources, it had focussed on Government priorities such as the AFTS Review.41

2.46 The ANAO made a recommendation to improve the robustness of the reliability ratings of estimates in the Statement. It also made a number of suggestions that Treasury could implement that would either improve the Statement, or improve how Treasury managed tax expenditures. They were not formal recommendations, but are listed here for information. The ANAO stated that Treasury could:

 quantify tax expenditures for new policy proposals in the Budget

Papers

 with other jurisdictions, develop external standards for reporting tax

expenditures

 include additional factors to consider when reviewing tax expenditures,

in particular:

⇒ the interaction of the tax expenditure with other State and

Commonwealth government programs

⇒ its distributive impact

⇒ compliance costs for the taxpayer and ATO

⇒ alternative mechanisms for delivering the benefit

 identify tax expenditures in other portfolios itself

39 AFTS Review, Report to the Treasurer, Part Two, Detailed analysis, 2009, pp. 724-33. 40 ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13, pp. 16-18. 41 ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13, p. 16.

18 THE TAX EXPENDITURES STATEMENT

 with the Australian Taxation Office (ATO), first identify the data

problems with the estimates and then use this to guide information requests

 give reasons in the Statement where it does not publish a revenue gain

estimate for each of the largest tax expenditures

 implement a documented system to improve the reliability of the

estimates, including timeframes and targets.42

Committee comment

2.47 In following up previous recommendations during the inquiry, the Committee noted the statement by Treasury that it no longer supported either the ANAO’s recommendations from its 2008 audit, or the JCPAA’s follow up inquiry in 2009. In particular, Treasury stated:

With the benefit of further consideration, Treasury now considers that the recommendations contained in these reports are no longer relevant or appropriate.43

2.48 The Committee accepts that an agency may decide not to implement an ANAO or parliamentary committee recommendation that it has previously accepted if circumstances have changed so as to allow a reasonable explanation. The Committee also appreciates the circumstances under which Treasury operates and its resource constraints.

2.49 However, the Committee is concerned that Treasury has now rejected these recommendations without explanation. The Committee’s opinion is that an obligation lies with Treasury to communicate and explain this change. It should not have to be discovered by a follow-up inquiry.

2.50 The ANAO and JCPAA are important institutions that safeguard public finances and protect the interests of the Parliament. Further, as this inquiry showed, there are alternative ways in which the recommendations could have be implemented or progressed. Finally, it would have been a courtesy to the ANAO and JCPAA to advise them of this change, given that Treasury had formally accepted the recommendations.

2.51 The Committee will raise this issue with the JCPAA.

42 ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13, pp. 45, 51, 55, 64, 66, 72, 76. 43 Treasury, Submission 7, p. 1.

3

Issues raised during the inquiry

Is the Statement effective?

The positives

3.1 The goals of the Statement are to facilitate scrutiny of tax expenditures and inform debate on the tax system.1 The Committee certainly received evidence during the inquiry that the Statement achieved these outcomes. The Parliamentary Budget Officer stated ‘it makes a significant contribution to budget transparency.’2 Associate Professor Julie Smith put the same view.3 The Grattan Institute commented that ‘annual reporting of tax expenditures is a standard part of good budget process.’4

3.2 The Parliamentary Budget Office stated that it used the Statement in its work. The Australian Statement has been recognised overseas as going furthest in meeting best practice standards.5

3.3 The Committee finds that the Statement is generally effective in meeting its stated goals.

1 Treasury, Tax Expenditures Statement 2014, 2015, p. 3. 2 Mr Phil Bowen, PBO, Committee Hansard, 14 October 2015, Canberra, p. 1. 3 Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 2. 4 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 1. 5 Mr Phil Bowen, PBO, Committee Hansard, 14 October 2015, Canberra, p. 7; PBO, Submission 9,

p. 4.

20 THE TAX EXPENDITURES STATEMENT

The limitations

3.4 Despite this general finding, the Committee heard criticisms of the Statement. Robert Carling of the Centre for Independent Studies made the general point that it is open to misinterpretation, which can produce policy risks:

I agree that there needs to be a tax expenditures statement, but I think that what began as a good idea has developed into something that is doing as much harm as good to the policy debate, because there is frequent misinterpretation and misuse of the data in public policy analysis and commentary, which carries the risk of suboptimal policy prescriptions and outcomes.6

3.5 The Committee also heard criticisms about the benchmarks used, data quality, and the period that it covered. Most of these criticisms can be traced back to some of the Statement’s key features. These are discussed below.

Only a starting point

3.6 One limitation with the Statement is that, although it is designed to be a starting point for analysis, it at times becomes the end point because more sophisticated information is difficult to obtain. The Parliamentary Budget Officer commented that this misuse of the Statement indicates a ‘significant unmet demand’ for more sophisticated tax analysis.7

3.7 The Grattan Institute agreed that ‘we cannot ask too much of the Statement.’8 It also noted that the ‘Statement does not exist in a vacuum’ and that there may well be more advanced analysis in the public domain that also informs debate.9

3.8 This problem was specifically raised in relation to superannuation tax concessions. The question of what caveats should be placed in the Statement is discussed later in this chapter.

6 Mr Robert Carling, CIS, Proof Committee Hansard, 11 November 2015, Canberra, p. 1. 7 Mr Phil Bowen, PBO, Committee Hansard, 14 October 2015, Canberra, p. 1. 8 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 2. 9 Mr Brendan Coates, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, pp. 2-3.

ISSUES RAISED DURING THE INQUIRY 21

It involves judgment

3.9 Another issue with the Statement is that it involves judgment to prepare, for example in developing the benchmark. This can mean that the estimates are contentious. The Statement itself acknowledges this.10

3.10 The Committee discussed this at the hearings. The general view was that there was always going to be debate about estimating tax expenditures, at least at the margins.11

3.11 In this environment, the Committee believes Treasury can use consultation to test the decisions it makes in preparing the Statement. Chartered Accountants Australia and New Zealand (Chartered Accountants) advised the Committee that they had not been consulted on the Statement.12 The Parliamentary Budget Office commented that consultations tend to occur during policy development, but there was nonetheless scope for further discussions:

Most of the consultation around the TES, in terms of identifying what tax expenditures are and the like, is something that is very much internal to Treasury. I do not know that there is a huge amount of consultation with taxpayer groups or the like at present, and that is certainly something that could happen more. I suppose the issue with that is that there is a comprehensive benchmark there. The identification of tax expenditures occurs against that. Often it happens as part of the policy development process. When you are looking at changes in tax policy that are going through in budgets, at the time something goes through, it would be identified as whether or not it creates a new tax expenditure or abolishes a tax expenditure.13

3.12 Therefore, many of recommendations in this report include a suggestion that Treasury consult with stakeholders.

Resource intensive to prepare

3.13 The final general issue with the Statement is that it is resource intensive. Treasury made this point in evidence.14 The Parliamentary Budget Officer

10 Treasury, Tax Expenditures Statement 2004, 2015, p. 3. 11 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 1; Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 5; Mrs Susan Franks, CAANZ; Committee Hansard, 19 August 2015, p. 6. 12 Mr Michael Croker, CAANZ, Committee Hansard, 19 August 2015, Canberra, p. 3. 13 Mr Colin Brown, PBO, Committee Hansard, 14 October 2015, Canberra, p. 8. 14 Mr Steve French, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 1.

22 THE TAX EXPENDITURES STATEMENT

confirmed this from his own organisation’s experience and commented that devoting a large quantity of additional resources to the Statement may not be worthwhile, because the estimates would still not be additive:

It is well recognised in the international community that the costing function of organisations like the PBO is the most resource-intensive function that these organisations perform … It is very resource intensive to prepare sophisticated costings, which is what we do all the time, but it would require much more resources if that were going to be the basis for preparing this document, and I think the document would still have limitations, because if you cost individual elements without costing an integrated package then you cannot add them up.15

3.14 The Committee received evidence that Treasury was not resourcing the Statement as much as desired by some stakeholders. The Financial Services Council stated that Treasury had invested heavily in making innovations to Budget estimates, but there was ‘an underinvestment in improving the Tax Expenditures Statement’ and ‘scant innovations.’16

3.15 The resource impact of recommendations in this inquiry therefore became very important. The Parliamentary Budget Officer stated that recommendations to expand the scope of the Statement ‘should be mindful that these improvements are likely to come at a cost either to Treasury or to taxpayers—or to both.’17

3.16 One way of addressing resource issues is to seek greater efficiencies. Treasury commented during the hearing that some of the smaller estimates took effort to prepare. If they were not material or significant, then Treasury suggested that it could provide a less accurate estimate for them:

There are a lot of very small tax expenditures and they are often quite technical in nature. I think for the annual updating of some of those estimates the value of that we would question at times. If they are not changing very frequently … reporting as a range may well be the most sensible thing to do and to review them less frequently than annually.18

3.17 The Committee regards this as a common-sense suggestion and would produce a risk-based approach. On average, tax expenditures with larger

15 Mr Phil Bowen, PBO, Committee Hansard, 14 October 2015, Canberra, p. 7. 16 Mr Andrew Bragg, FSC, Proof Committee Hansard, 11 November 2015, Canberra, p. 8. 17 Mr Phil Bowen, PBO, Committee Hansard, 14 October 2015, Canberra, p. 1. 18 Mr Steve French, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 8.

ISSUES RAISED DURING THE INQUIRY 23

estimates will attract more attention and debate. It would be preferable for Treasury to allocate more resources to them. Then the Statement’s inputs and outcomes would be more closely aligned to its purpose. The Committee appreciates Treasury’s input on this point.

Recommendation 1

3.18 That Treasury devote fewer resources to estimating smaller, technical tax expenditures. This could involve reviewing them less frequently and reporting them as a range.

3.19 The other way of addressing resource issues is to consider whether Treasury should receive additional funding to enhance the Statement. Stakeholders during the inquiry acknowledged that Treasury is currently working with fewer resources. Chartered Accountants stated:

We do have a Treasury that is under a lot of pressure, a Treasury that has been somewhat reduced in size …

We are supportive of Treasury being given funding to do its job. We are a longstanding supporter of Treasury trying to get the resources it needs to do its job.19

3.20 The Committee is mindful of the need for fiscal responsibility. Therefore, recommendations in the report have been designed to have a lower budget impact and to focus on the more important tax expenditures. However, if Treasury believes that it cannot implement some of the report’s recommendations without additional resources, and the resources are warranted, then the Committee would support a funding bid for this purpose.

Policy reviews of tax expenditures

Background

3.21 In its 2008 audit report, the ANAO recommended that Treasury should establish a program of regular and systematic reviews of tax expenditures. The recommendation stated:

ANAO recommends that the Department of the Treasury:

19 Mr Michael Croker, CAANZ, Committee Hansard, 19 August 2015, Canberra, p. 2.

24 THE TAX EXPENDITURES STATEMENT

 develop an approach for the conduct of an ongoing prioritised

review of the existing program of tax expenditures; and

 publish for each tax expenditure information on the timing and

outcome of the review.20

3.22 The ANAO made this recommendation because it was concerned that, over time, individual tax expenditures may no longer be effective, or be relevant to government policy. They could also compromise the Budget. The ANAO expressed concern that governments were inheriting tax expenditures from their predecessors without subjecting them to critical assessment.21

3.23 Treasury accepted the recommendation, but added a qualification in relation to the second part. Treasury stated that the publication of the reviews would be a matter for government.22

3.24 In its 2013 report, the ANAO stated that Treasury had conducted a series of reviews of tax expenditures between November 2007 and early 2011. Treasury’s aim was to conduct a series of rolling reviews so that each tax expenditure would be reviewed every five years.23

3.25 Treasury developed a template for the reviews, which the ANAO described as a ‘reasonable approach.’ However, some reviews were conducted in a simplified manner without the template. Another procedural problem was that Treasury did not establish a formal register of the reviews. Such a register would have assisted with planning and safeguarded against items being reviewed multiple times.24

3.26 Out of 123 tax expenditures reviewed, 44 were recommended for modification or abolition. Treasury proposed that 58 be retained, 11 were deferred, and 10 contained no recommendation, or the tax expenditure was subject to repeal.25 In other words, over a third of tax expenditures reviewed had issues that were worth addressing.

3.27 In its submission to the inquiry, Treasury noted the recommendations from the 2008 audit report and the 2009 follow-up by the Joint Committee of Public Accounts and Audit (JCPAA). Treasury stated that, ‘With the

20 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 40. 21 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 38-39. 22 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 40. 23 ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13, p. 54. 24 ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13, pp. 55-

59.

25 ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13, pp. 56-57.

ISSUES RAISED DURING THE INQUIRY 25

benefit of further consideration, Treasury now considers that the recommendations contained in these reports are no longer relevant or appropriate.’26

3.28 In relation to this particular recommendation, Treasury commented that tax expenditures were already considered in the Budget process and that the reviews it conducted were not worthwhile because they were not consistent with the agenda of government. It stated:

Treasury considers that there would be little value in undertaking stand-alone reviews of tax expenditures in addition to the consideration of tax expenditures undertaken in the annual Budget process.

Past reviews provided little value, as they tended to not be aligned with the policy agenda of the government of the day.27

Analysis

3.29 The ANAO’s submission recommended reviews of tax expenditures as an area of focus for Treasury.28 At the hearing, the ANAO made some comments on Treasury’s position, in particular that it did not prioritise its reviews and that the Budget process does not focus on existing tax expenditures. Instead, it focusses on new or modified measures:

I think that separate review process can be useful. A lot of the issue is the priority process, however we did [not] really see the prioritisation of the previous Treasury review. It was not clear what was given priority and what was not …

We are not exactly sure how that would happen through the budget process. There have been improvements, but there is not the focus on the existing stock of tax expenditures. Really the focus usually is on the modified and new measures.29

3.30 The ANAO also noted that the Government’s policy agenda of the day could be incorporated into the prioritisation process.30

26 Treasury, Submission 7, p. 1. 27 Treasury, Submission 7, p. 5. 28 ANAO, Submission 4, pp. 5-6. 29 Mr Andrew Morris, ANAO, Committee Hansard, 9 September 2015, Canberra, pp. 6-7.

30 Mr Mark Simpson, ANAO, Committee Hansard, 9 September 2015, Canberra, p. 7.

26 THE TAX EXPENDITURES STATEMENT

Committee comment

3.31 The Committee notes that the reviews that Treasury undertook between 2007 and 2011 recommended that over a third of reviewed tax expenditures should be abolished or modified. This demonstrated concerns about the effectiveness of tax expenditures. A review in 1997 recommended that approximately half be abolished or modified.31

3.32 The Committee accepts that Treasury may have had high expectations for the reviews that did not eventuate. And with hindsight, it appears that reviewing all of the hundreds of tax expenditures every five years was ambitious. The Committee would be satisfied if Treasury completed a modest number of reviews each year. The important feature is that they should be prioritised, as suggested by the ANAO.

3.33 The ANAO’s recommendation in 2008 also stated that the reviews should be published. The Committee notes that the Canadian equivalent of the Tax Expenditures Statement for 2014 includes an evaluation of the federal charitable donation tax credit and an analysis of the extent to which companies take advantage of differing tax rates across Provinces.32 Given this precedent, the Committee believes that there is scope for Treasury to at least publish a summary of its work. Publication of reviews need not be as part of the Statement itself, but could form a closely related work stream.

3.34 During the inquiry the Committee heard that the Statement was an essential first step towards increasing the transparency of tax expenditures. However, it also heard that the Statement was not enough — that gaps remain in terms of the depth and reliability of policy analysis available to inform robust public debate. Treasury-led policy reviews would form an important second step towards filling this gap. Treasury’s analysis is highly respected for its credibility and impartiality, and there are many historical examples of Treasury’s analysis improving the quality of public debate. Reviews would only need to focus on the underlying analysis of a particular tax concession — attempting to clearly lay out the costs, benefits and considerations. In this respect the reviews would be ‘government neutral’.

3.35 The Committee understands that without a clear mandate there may be uncertainty about Treasury’s role in terms of which tax expenditures it

31 Mr Andrew Morris, ANAO, Committee Hansard, 9 September 2015, Canberra, p. 6. 32 Department of Finance, Canada, Tax Expenditures and Evaluations 2014, 2015, pp. 30, 71, accessed from http://www.fin.gc.ca/taxexp-depfisc/2014/taxexp-depfisc14-eng.pdf on 20 November 2015.

ISSUES RAISED DURING THE INQUIRY 27

selects to review and the material it publishes. Therefore, the Committee supports the Government giving Treasury an ongoing mandate to conduct and publish this work. Such a mandate could also give Treasury some guidance on how the process would operate.

3.36 Finally, the Committee notes that the ANAO’s 2013 report listed a number of suggested improvements to the process that Treasury conducted in 2007-2011. These are listed in Chapter 2 and the Committee would be pleased if Treasury were to adopt them in future.

Recommendation 2

3.37 That the Government give Treasury an ongoing mandate to conduct prioritised reviews of tax expenditures and publish the results — building upon Recommendation 1 from the ANAO’s 2008 audit report.

Aggregating tax expenditures

Background

3.38 Previous editions of the Statement have included aggregated information on tax expenditures. For example, the 2004 Statement included the following tables:

 trends in total tax expenditures, including as a percentage of GDP

 aggregated tax expenditures by function (defence, health etc)

 comparing aggregated tax expenditures against direct spending by

function

 trends in total tax expenditures and direct spending.33

3.39 This information flowed through to the Budget Papers. Budget Paper No. 1 for 2004-05 included the total trends table with brief comments in an appendix.34 Section 12 of the Charter of Budget Honesty Act 1998 requires the Budget economic and fiscal outlook report to include ‘an overview of

33 Treasury, Tax Expenditures Statement 2004, 2015, pp. 8-13. 34 The Hon. Peter Costello MP, Treasurer, and Senator the Hon. Nick Minchin, Minister for Finance and Administration, Budget Strategy and Outlook: 2004-05 Budget Paper No. 1, 2004, pp. 5-32, 5-33.

28 THE TAX EXPENDITURES STATEMENT

the estimated tax expenditures for the budget year and the following 3 financial years.’

3.40 Despite this framework, there have been concerns that tax expenditure data has not been applied in policy decisions. Associate Professor Julie Smith stated in 2003 that there has not ‘been any public evidence that the Commonwealth Government has met undertakings to integrate tax expenditure analysis into its fiscal management practices.’35 The ANAO’s 2008 audit argued that tax expenditures could be better integrated into the Budget process. Its 2013 audit found that progress had been made on individual measures, but there was still room for improvement.36

3.41 The 2012 Statement was the last to have aggregated data. Instead of an aggregate table, the Budget Papers now have an appendix with the list of the 25 largest tax expenditures from the current Statement.37

Analysis

3.42 Treasury explained in evidence that it removed the aggregates from the Statement because: such information is not provided by other countries; aggregation uses the less sophisticated revenue forgone estimates; the aggregates were misinterpreted; and the number of unmeasured large tax expenditures meant that aggregation lost much of its meaning:

… that is consistent with the approach taken by other major countries. It essentially reflects the fact that the standard approach for preparing the tax expenditure estimates is the revenue forgone approach, and that approach estimates the existing utilisation of a tax expenditure. It does not represent the impact on the budget if the tax expenditure were removed because it does not take into account the changes in behaviour of taxpayers in response.

I think we have had some concerns that this is often poorly understood, and by aggregating the tax expenditure estimates we amplify the potential for the misunderstanding. Those figures tended to be represented in the press as the aggregate impact on the budget if tax expenditures were removed. One of the

35 Julie Smith, Tax Expenditures: the $30 Billion Twilight Zone of Government Spending, 2003, DPL Research Paper No. 8 2002-03, p. 1. 36 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 40-44; ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13,

pp. 38-46. 37 The Hon. J.B. Hockey MP, Treasurer, and Senator the Hon. Mathias Cormann, Minister for Finance, Budget Strategy and Outlook: Budget Paper No. 1 2015-16, 2015, pp. 4-20, 4-21.

ISSUES RAISED DURING THE INQUIRY 29

difficulties here is that for quite a number of tax expenditures it is quite unlikely that some of them would ever be removed.

A further issue is the one I mentioned earlier, that many tax expenditures are unquantified. Even though many of those tax expenditures are likely to be relatively small, there are some large ones, and so it is very unclear what the aggregation of the quantified measures actually represents.38

3.43 The ANAO did not support the removal of aggregated information. It argued this has made tax expenditures less visible. Its submission stated this has:

… resulted in reduced visibility of the overall magnitude and relative importance of tax expenditures. To better inform policy development and Budget processes, there would be benefit in the Treasury considering ways of providing aggregate and comparative information about tax expenditures in future TES.39

3.44 There was no clear view on aggregation during the inquiry. Robert Carling from the Centre for Independent Studies supported Treasury’s approach due to the ‘fallacy of aggregation.’40 Associate Professor Julie Smith argued that aggregate information should be restored to the Statement.41

Committee comment

3.45 A number of factors are involved in this issue. On the one hand, a simple aggregation of revenue forgone tax expenditures almost certainly overstates their effect on the Budget due to interactions between them. On the other hand, strategic analysis of the Budget is hampered without a sense of their total size. Further, when aggregate information was included in the Budget Papers, it generally was not integrated into the document. This suggests that this information did not result in tax expenditures being integrated into the Budget process.

3.46 Nonetheless, the Committee believes the Charter of Budget Honesty Act 1998 requires the Budget to include information on tax expenditures for a constructive, strategic purpose. The Committee is of the view that this

38 Mr Steve French, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 2. 39 ANAO, Submission 4, p. 6. 40 Robert Carling, CIS, Submission 8, p. 5. 41 Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 6.

30 THE TAX EXPENDITURES STATEMENT

includes monitoring the scale of tax expenditures compared with revenue and the overall economy.

3.47 The Committee accepts that such a result is a long way off. However, if this project is approached with sufficient focus and/or on an iterative basis, it should eventually yield meaningful data and improve debate.

3.48 Treasury has commented that it believed some tax expenditures would never be removed and that it made little sense to include them in an aggregated figure. The Committee acknowledges this point and recommends later in the chapter that some tax expenditures be removed from the Statement on this basis.

3.49 The Committee can also see that the Statement and an overview for the Budget Papers could serve different purposes and may conceptually be different to each other. For example, if the Statement retained some tax expenditures that are likely to be removed then the Budget overview could exclude them. Alternatively, an aggregation methodology may involve some level of consideration or modelling of the interactions between expenditures. Several variations are possible. Developing an accepted and sustainable approach for a Budget overview is not necessarily straightforward, and should involve wide consultation by Treasury.

3.50 Although development of an aggregation methodology will not be simple it is clearly worth the effort, and the Committee would be pleased to give ongoing support to Treasury in this endeavour.

Recommendation 3

3.51 That Treasury develop a method of aggregating tax expenditures to support its obligation under section 12 of the Charter of Budget Honesty Act 1998 to give an overview of tax expenditures in the Budget Papers. The overview can be tailored to the needs of the Budget Papers, rather than exactly reflecting the Statement. Consultation may assist on this.

ISSUES RAISED DURING THE INQUIRY 31

The benefits of superannuation tax concessions

3.52 The superannuation concessions result in large tax expenditures. The revenue forgone estimate for 2014-15 for contributions is $16.3 billion and for earnings is $13.4 billion.42

3.53 During the inquiry, superannuation groups expressed concern that these figures only tell half the story. In particular, they only describe the current costs of superannuation tax concessions and omit the long term benefits. Mercer (Australia) stated:

My primary suggestion is that we also have a number in there outlining the benefits to future government cost savings of these concessions. If we see these concessions as an investment for the future, we have no measure of what the return on that investment is. The return on that investment is going to be billions of dollars. I am not suggesting that is easy. You need a model to suggest what that saving is, and there will need to be assumptions as to how superannuation is rundown over years. I think, even if it is an estimate, it will give us a better feel for the cost to government over the longer term.43

3.54 Other organisations made similar recommendations to the Committee, including the Grattan Institute, the SMSF Association and the SMSF Owners’ Alliance.44 The Committee heard that the benefit to the Commonwealth Budget is mostly through shifting retirees from the full pension to the part pension.45

3.55 Treasury has already conducted some modelling on this topic. The Charter Group report in 2013 included a graph demonstrating the effects of an increase in the superannuation guarantee to 12 per cent. This is reproduced in figure 3.1. It shows the loss in tax revenues initially outweighing the pension savings, but the pension savings then grow to approach the size of the tax losses towards the end of the modelling period. This analysis does not incorporate the benefits of superannuation and the related tax concessions to individuals.

42 Treasury, Tax Expenditures Statement 2014, 2015, pp. 64-65. 43 Dr David Knox, Mercer (Australia), Proof Committee Hansard, 21 October 2015, Canberra, p. 3. 44 Grattan Institute, Submission 6, p. 7; SMSF Association, Submission 12, pp. 6-7; SMSF Owners’ Alliance, Submission 10, p. 6. 45 Mr Robert Carling, CIS, Proof Committee Hansard, 11 November 2015, Canberra, p. 6.

32 THE TAX EXPENDITURES STATEMENT

Figure 3.1 Reduction in taxes and Age Pension outlays from increasing the superannuation guarantee to 12 per cent

Source Charter Group, A Super Charter: Fewer Changes, Better Outcomes: A report to the Treasurer and Minister Assisting for Financial Services and Superannuation, 2013, p. 11.

Committee comment

3.56 The Committee accepts that superannuation will have significant, long run benefits. However, the Committee also accepts that Treasury is yet to quantify these results at the same time as it is publishing the present costs of superannuation’s tax concessions. The Committee is of the view that the Statement’s aim of informing debate implies that it will do so in a balanced way.

3.57 Therefore, the Committee recommends that Treasury develop present value estimates of the costs and benefits of superannuation tax concessions and publish the results.

3.58 The next question is where it should be published and how often. The Committee received various suggestions, such as the Government’s current tax review, the Intergenerational Report, and the Statement.46 The

46 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 3; Mr Brendan Coates, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 3; Dr David Knox, Mercer (Australia), Committee Hansard, 12 August 2015, Canberra, p. 3.

ISSUES RAISED DURING THE INQUIRY 33

Committee acknowledges that the Statement may not be the best place in which to publish this work. However, the Statement should at the least reference this research, or briefly summarise it, so that readers gain balanced information and can easily locate more detailed analysis if required.

Recommendation 4

3.59 That Treasury model the long run interactions between superannuation and the age pension, develop present value estimates of the future costs and benefits of superannuation and its tax concessions, and publish the results. The Tax Expenditures Statement should reference or briefly summarise this research in the context of the superannuation estimates.

This should occur at least once every five years and also following significant changes in policy.

3.60 One aspect of superannuation and its tax concessions that is often overlooked is that individuals do pay tax on their contributions and earnings. Provided various criteria are met (eg concessional caps), the usual tax rates are15 per cent for employer contributions and fund earnings. Capital gains can be taxed at 10 per cent if assets are held for over 12 months. If the criteria are not met, the tax rates can increase to the taxpayer’s marginal rate.

3.61 These taxes produce significant revenue, estimated at $9.2 billion for 2015-16, and projected to increase to $11.9 billion in 2018-19.47 This feature of superannuation tax concessions could usefully inform debate and the Committee would like to see some information about this included in the Statement.

Recommendation 5

3.62 That Treasury include in the Statement information on revenue collected through superannuation taxation, which could constitute a reference in the table of large tax expenditures and further information later in the document.

47 The Hon. J.B. Hockey MP, Treasurer, and Senator the Hon. Mathias Cormann, Minister for Finance, Budget Strategy and Outlook: Budget Paper No. 1 2015-16, 2015, p. 9-24.

34 THE TAX EXPENDITURES STATEMENT

Benchmark for savings

Background

3.63 The appropriate benchmark for savings was an important issue in the inquiry. It was mainly raised in the context of superannuation tax concessions, but is applicable across all types of savings.

3.64 The Statement currently uses a comprehensive income tax benchmark. ‘Comprehensive’ means that income from all major sources - labour, capital and natural resources - is taxable. ‘Income’ means an entity’s increase in assets plus consumption in a period. Under this benchmark, superannuation contributions and earnings should both be taxed at an individual’s marginal rate of tax. However, both of these aspects of superannuation are concessionally taxed, triggering tax expenditure estimates in the Statement.

3.65 The alternative benchmark proposed during the inquiry was a consumption, or expenditure benchmark. This comprises an entity’s consumption in a period and excludes changes in their assets. The precise outcome depends on details. Under a pre-paid expenditure benchmark, superannuation contributions come from income and should be fully taxed. Any concessions will trigger a tax expenditure similar to a comprehensive income tax benchmark. But an increase in wealth from savings would be tax exempt. So any tax on superannuation earnings will result in a negative tax expenditure under such a benchmark. In other words, the size of a tax concession for earnings would not be measured. It would be disregarded. This result comes from the very definition of the benchmark.

3.66 The 2013 Statement gave an example of this. That Statement published revenue forgone estimates under both income tax and pre-paid expenditure benchmarks. The results are summarised in table 3.1.

ISSUES RAISED DURING THE INQUIRY 35

Table 3.1 Comparison of superannuation tax expenditures, 2013-14, $m

Expenditure tax

benchmark

Income tax benchmark

Taxation of employer contributions 16,000 16,000

Taxation of personal/self-employed contributions 670 670

Taxation of unfunded superannuation 490 490

15% tax on earnings in accumulation phase -4,700 16,100

0% tax on earnings in pension phase 0

10% tax on capital gains in accumulation phase -1,100

Measures for low income earners 130 130

Tax on funded lump sums -250 -250

Source SMSF Association, Submission 12, p. 4. The three entries for earnings for the income tax benchmark were not disaggregated in the original document and are presented as a single $16.1 billion figure here.

3.67 Contributions are treated the same under the two different benchmarks. However, the returns from superannuation are treated much differently. Under the comprehensive income tax benchmark, they result in a tax expenditure of $16.1 billion (a comparison against each taxpayer’s marginal rate of tax). Under the expenditure tax benchmark, the superannuation returns result in a negative tax expenditure of $5.8 billion. This is because some tax is paid, whereas the benchmark assumes it to be tax free.

3.68 Superannuation groups in the inquiry supported the expenditure tax benchmark, either on its own, or shown together with the income tax benchmark. Robert Carling from the Centre for Independent Studies also took this view. The Grattan Institute was the only organisation to support the income tax benchmark.48

3.69 Other parts of the tax system are affected by this debate, including the two large tax expenditures for the capital gains tax exemption for the main residence. Under a pre-paid expenditure benchmark, they would report a nil tax expenditure because they are purchased with after tax income (buying one’s home cannot be treated as a deduction) and the investment return is expected to be tax free.49 This tax expenditure is discussed further in the next section relating to benchmarks.

48 For example, Mr David Knox, Mercer (Australia), Proof Committee Hansard, 21 October 2015, Canberra, p. 3; SMSF Owners’ Alliance, Submission 10, pp. 4-5; SMSF Association, Submission 12, p. 5; Robert Carling, CIS, Submission 8, p. 4; Grattan Institute, Submission 6, pp. 3-4. 49 Treasury, Tax Expenditures Statement 2013, 2014, p. 194.

36 THE TAX EXPENDITURES STATEMENT

3.70 Another example is bank interest. Under the comprehensive income tax benchmark, it is not a tax expenditure because it is taxed at full marginal rates. However, if an expenditure benchmark were applied, it would be a large negative tax expenditure. A similar treatment would apply to savings income generally.

Analysis

3.71 The discussion around the suitable benchmark for savings comes back to the purpose of the Statement. Some superannuation groups and the Centre for Independent Studies argued that, since they believe that a concessional tax treatment for savings is becoming recognised as optimal, then this should also be reflected in the benchmark and the Statement.50 For example, the AFTS Review adopted the principle that savings in the form of superannuation and owner occupied housing should be concessionally taxed or tax free in line with an expenditure benchmark.51

3.72 Some superannuation groups took this initial position in relation to tax rates and then argued that this should be reflected in the Statement. The SMSF Owners’ Alliance commented:

We feel that, if the TES is to provide a helpful estimate, it should ideally provide an estimate of the deviation from an optimal taxation of savings. It is now widely argued that the optimal taxation savings is not the income tax scale; then it should be compared with a zero or low rate of tax. In other words, the current figure which compares it with taxation at progressive scales, provides a distorted view … it does not provide a figure that helps a debate towards an efficient tax system …52

3.73 The alternative argument, made by the Grattan Institute, was that the Statement should demonstrate the cost of tax concessions to inform debate.53 It may be more economically efficient to have lower tax rates for various forms of savings, but the goal should be to determine the cost of encouraging those savings and then to use this information to determine whether the outcomes are supported on a cost/benefit analysis.

50 For example, Mr Robert Carling, CIS, Submission 8, pp. 4, 6; SMSF Association, Submission 12, p. 5. 51 AFTS Review, Report to the Treasurer, Part Two, Detailed analysis, 2009, p. 13. 52 Mr Malcolm Clyde, SMSF Owners’ Alliance, Proof Committee Hansard, 21 October 2015,

Canberra, p. 9. 53 Grattan Institute, Submission 6, p. 4.

ISSUES RAISED DURING THE INQUIRY 37

3.74 The AFTS Review also took this approach. It noted that incorporating policy into the benchmark reduces the Statement’s effectiveness in informing the community:

The purpose of reporting tax expenditures is so the community can understand how the tax system affects the economy and society more broadly. Benchmarks should allow an objective evaluation of the effects of government policy, rather than represent that policy. For example, if a tax concession is set up to assist a particular industry the benchmark should not incorporate this objective, but should provide a basis for identifying and valuing the concession. This allows the community to judge whether this form of assistance is appropriate.

Currently, many of the most important economic and distributional effects of taxes are incorporated in the benchmarks and so are not reported in the Tax Expenditures Statement.54

3.75 The AFTS Review took a nuanced approach in relation to superannuation tax concessions. It argued that tax rates should reflect a tax expenditure benchmark, but then said that benchmarks for the purpose of the Statement should not reflect policy to allow the cost of concessions to be identified and inform debate. Consistent with this, the Review’s recommendations in relation to tax expenditures did not propose that the expenditure benchmark be used for superannuation tax concessions.55

Committee comment

3.76 One of the features of superannuation tax concessions is that they cost a great deal in terms of current tax revenue, but deliver substantial benefits in the future by encouraging people to save for their retirement. This could reduce reliance on the age pension and allow people to enjoy their retirement more. This model of superannuation tax concessions has broad support in the community. The Financial Services Council stated that superannuation ‘is our signature policy for addressing the cost of our ageing population.’ 56

3.77 The Committee believes that the Statement, in its current form, produces figures on the current costs of superannuation tax concessions without any reflection or analysis about future benefits, either to the Government or to retirees. If one of the purposes of the Statement is to inform debate, then

54 AFTS Review, Report to the Treasurer, Part Two, Detailed analysis, 2009, p. 731. 55 AFTS Review, Report to the Treasurer, Part Two, Detailed analysis, 2009, p. 729. 56 Mr Andrew Bragg, FSC, Proof Committee Hansard, 11 November 2015, Canberra, p. 8.

38 THE TAX EXPENDITURES STATEMENT

the Committee believes that it is paramount that the Statement and related policy analysis should do so in a balanced way.

3.78 The Committee took evidence that the Statement could present both benchmarks.57 Elsewhere in the report, the Committee has recommended that Treasury should more comprehensively model and publish the benefits of superannuation tax concessions.

3.79 The expenditure benchmark incorporates the policy judgment that savings have wider benefits to society than simply those that accrue to the saver. It is of itself a simplified cost/benefit analysis with the assumption that the wider benefits of savings are so great that they should be tax free.

3.80 Given the size of superannuation tax concessions and their public interest, a more sophisticated approach, which publicly reports both the current estimates based on the income tax benchmark and the results of the new recommended modelling, is preferred. This work should be conducted outside the Statement.

Recommendation 6

3.81 That Treasury retain the comprehensive income tax benchmark for savings in the Tax Expenditures Statement.

Benchmarks and structural features of the tax system

Background

3.82 The Statement uses eight benchmarks. Five of them are based on a comprehensive income tax using the Schanz-Haig-Simons definition of income. The income tax benchmark is used for business tax, personal income tax, fringe benefits tax, retirement savings taxation, and capital gains tax. The other three benchmarks are based on the design features of those taxes. They are for the GST, natural resource taxes, and commodity and other indirect taxes.58

57 Dr David Knox, Mercer (Australia), Proof Committee Hansard, 21 October 2014, Canberra, p. 3; Mr Robert Carling, CIS, Submission 8, p. 6. 58 ‘Comprehensive’ refers to taxing returns on labour, capital, and natural resources; AFTS Review, Architecture of Australia’s tax and transfer system, 2008, p. 331. The Schanz-Haig-Simons

definition of income for an entity in a period is ‘the increase in the entity’s economic wealth

ISSUES RAISED DURING THE INQUIRY 39

3.83 The choice of benchmark and the decision of whether to exempt some features of the tax system from the benchmark are key decisions in estimating a tax expenditure. In the 2014 Statement, Treasury outlined its twin principles that the benchmark should provide a very general tax treatment, but that it can be adjusted for structural parts of the tax system:

The framework for defining the benchmarks used in this statement is based on two principles.

 The benchmark should represent the standard taxation

treatment that applies to similar taxpayers or types of activity. Consequently, a benchmark taxation treatment should neither favour nor disadvantage similar taxpayers or activities.

 The benchmark may incorporate certain elements of the tax

system which depart from a uniform treatment of taxpayers where these are fundamental structural elements of the tax system. Such elements could include integral design features; for example, the progressive income tax rate scale for individual taxpayers.59

3.84 The issue is that, if a particular feature of the tax system is included in the benchmark, then it will not be reported as a tax expenditure. It will then be less visible and transparent. In many cases, this is appropriate or required as a matter of practicality. For example, the benchmark for capital gains is based on realisation, rather than accruals. The latter would have significant administrative problems.60

3.85 The Statement lists the departures from the various benchmarks. Examples of the departures from the comprehensive income tax benchmark are:

 nominal income is assessed, rather than real income

 taxpayers deduct expenses at historical cost

 imputed rent from owner-occupied housing is not included in income

 gains from compensation are not included

 gains from gambling are not included, unless the taxpayer conducts this

as a business

(stock of assets) over the period, plus the entity’s consumption in the period, where consumption includes all expenditures except those incurred in earning or producing income.’ PBO, Submission 9, p. 8. 59 Treasury, Tax Expenditures Statement 2014, 2015, p. 129. 60 Treasury, Tax Expenditures Statement 2014, 2015, p. 135.

40 THE TAX EXPENDITURES STATEMENT

 losses are deductible against income in later years.61

3.86 The question over what to include or exclude from the various benchmarks has generated much debate. In 2003, Julie Smith raised questions about some of the exclusions from the benchmark, including the non-taxation of imputed rent from owner-occupied housing. In the 2008 audit, the ANAO discussed the treatment of alcoholic beverages and customs duty.62

3.87 The treatment of alcoholic beverages was also brought to the attention of the Committee by the Distilled Spirits Industry Council of Australia (DSICA). Noting the 2008 ANAO audit, their concern was that the different ways in which alcohol is taxed is incorporated in the excise benchmark for alcohol, resulting in no tax expenditures.63

3.88 Alcoholic beverages are taxed in many different ways, depending on the beverage, packaging, and value of the product. The PBO recently produced analysis that converted these different taxation methods to a single effective measure of the tax on a litre of alcohol. In other words, if a consumer purchased enough of the beverage so that the alcohol content comprised a litre, the following numbers represent the tax that the consumer would pay.

3.89 Depending on the packaging, the rate for commercial full-strength beer is either $25.41 per litre, or $36.08. The effective tax rate on a litre of alcohol for a $15 bottle of wine is $17.07. The rate for spirits is $79.38.64 None of this variation is portrayed in the Statement’s estimates.65

3.90 DSICA recommended that a single benchmark for alcoholic beverages be introduced.66

3.91 Treasury has acknowledged that the decisions about what to include and exclude from the benchmark can involve judgement. The Statement does not provide an explanation for the various benchmarks and exclusions, but acknowledges that there may be disagreement about them:

Reconciling these two criteria often involves an element of judgment. In particular, there may be different views on which

61 Treasury, Tax Expenditures Statement 2014, 2015, pp. 131-32. 62 Julie Smith, Tax Expenditures: the $30 Billion Twilight Zone of Government Spending, 2003, DPL Research Paper No. 8 2002-03, pp. 8-9; ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, pp. 54-57. 63 DSICA, Submission 11, p. 2. 64 PBO, Alcohol Taxation in Australia, 2015, Report no. 03/2015, p. 6. 65 Treasury, Tax Expenditures Statement 2014, 2015, pp. 137-38. 66 DSICA, Submission 11, p. 3.

ISSUES RAISED DURING THE INQUIRY 41

structural elements to include in the benchmark. Consequently, benchmarks vary over time and across countries and can be arbitrary.67

Analysis

3.92 Stakeholders in the inquiry noted the importance of correctly identifying the appropriate structural elements of the tax system in developing the benchmarks. Mr Robert Carling, from the Centre for Independent Studies, noted that the tax expenditures for capital gains tax on owner-occupied housing were not realistic because it was very unlikely that such a tax would ever be applied. He stated that New South Wales took the opposite approach and made the exemption the benchmark. He recommended that all tax expenditures should be examined to determine whether they should remain, or be incorporated in the benchmark.68

3.93 The PBO noted that many aspects to the benchmarks’ structural elements had a historical element to them. The PBO made a general comment that these structural benchmarks should be carefully scrutinised:

In practice, the benchmarks actually applied in the TES include exceptions for certain longstanding and entrenched elements of the tax system referred to as ‘structural’ elements. These structural elements of the benchmarks warrant careful scrutiny as they can mean that provisions that give a substantial benefit to some taxpayers are not identified as tax expenditures.69

3.94 In addition, Associate Professor Julie Smith noted that some of the structural elements of the tax system were not particularly long standing and that a debate about the benchmarks was warranted:

For example, there is a statement in the latest one that justifies the wine equalisation tax expenditure simply by saying that the differentiation of the tax treatment of wine has long been an element of the tax system. Well, actually, that came in only in 2000, with the GST, so it is not a justifiable statement. I think those sorts of methodological things should be open for debate … 70

67 Treasury, Tax Expenditures Statement 2014, 2015, p. 129. 68 Mr Robert Carling, CIS, Submission 8, pp. 3-4, 6. 69 PBO, Submission 9, p. 9. 70 Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 7.

42 THE TAX EXPENDITURES STATEMENT

Committee comment

3.95 Generating the Statement is a complex analytical task and the resulting report affects many competing interests. The Committee heard evidence that there will never be full agreement on all benchmarks used, and the Committee accepts this.71

3.96 The Committee also understands that Treasury must use its best judgment in compiling the Statement. Treasury has a great deal of expertise and for many of the questions involving tax policy, Treasury is the pre-eminent authority in Australia. The Committee is of the view that Treasury has made a large number of accurate, well-considered decisions in developing the Statement’s benchmarks.

3.97 However, the Committee is of the view that Treasury may in some cases be directing resources to the tax expenditures that are not relevant to the Statement’s purposes of facilitating scrutiny and informing debate. The best example is the capital gains tax exemption on the principal residence. If removing this exemption have not on the public agenda for a significant period, then estimates of it are unlikely to be of much use to stakeholders and the public. Given Treasury’s resource constraints, the Committee considers that it would be appropriate to divert effort away from estimating tax expenditures that have little policy relevance and towards issues more relevant to debate.

3.98 The Committee believes that there is clear evidence that the capital gains tax exemption on the principal residence should now be removed from the Statement through its inclusion in the benchmark. For other exemptions of a similar nature Treasury should consider the same approach.

3.99 The Committee is also of the view that Treasury can supplement its expertise more through consultations with stakeholders. The Committee notes the evidence, referred to earlier in this chapter that Treasury does not generally consult on the Statement. If any consultations are held, they tend to occur when a new measure is being introduced.72 To the Committee, it appears that a good way of testing one’s judgment would be to ask the opinion of others. The Committee also recommends that Treasury consult on the benchmarks used in the Statement.

71 Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 5. 72 Mr Michael Croker, CAANZ, Committee Hansard, 19 August 2015, Canberra, p. 3; Mr Colin Brown, PBO, Committee Hansard, 19 August 2015, Canberra, p. 8.

ISSUES RAISED DURING THE INQUIRY 43

Recommendation 7

3.100 That Treasury:

 incorporate the capital gains tax exemption for the main residence into the benchmark

 develop a transparent process and criteria to assist consideration of benchmarks which reflect the practical possibility of a tax concession being abolished

 consult with stakeholders on the benchmarks used in the Tax Expenditures Statement.

Caveats in the Statement

Background

3.101 The Statement includes a number of caveats. At the start of the 2014 Statement, there is a discussion of the differences between the revenue forgone and revenue gain methods and that the latter is more sophisticated by taking into account taxpayer reactions. There is also a list of caveats, including:

 revenue gain estimates can be difficult to calculate, given the lack of

data on taxpayer reactions

 to calculate an estimate, the likely policy settings around the abolition

of the estimate need to be considered, and this requires judgment

 estimates over time are not comparable due to differences in data,

benchmarks, and methods

 estimates are not additive because the removal of one tax expenditure

will affect the use of others.73

3.102 During the inquiry, stakeholders commented that the Statement was often used in public debate without regard to these caveats. Most of these comments revolved around the aggregation of estimates and superannuation tax concessions.74

73 Treasury, Tax Expenditures Statement 2014, 2015, pp. 4-5. 74 For example, SMSF Owners’ Alliance, Submission 10, p. 7; SMSF Association, Submission 12, p. 7; Financial Services Council, Submission 13, p. 3.

44 THE TAX EXPENDITURES STATEMENT

3.103 The complaint was that stakeholders would add together the revenue forgone estimates for contributions and earnings to develop a total cost of superannuation tax concessions to the budget. This would be inaccurate for two reasons: it would not take into account the behavioural effects; and it would not account for interactions between the two tax concessions and other possible investments. Mercer (Australia) stated in evidence:

… many commentators and, indeed, journalists will add the revenue forgone numbers and say, ‘Superannuation costs X dollars’. In a sense that is right, but people imply that that is the potential revenue gain, which it is not.75

3.104 The SMSF Owners’ Alliance stated that, although Treasury had commented generally that such statements were wrong, they could take a more active role in correcting the record. Their argument was based on an analogy with the role of the directors of public companies, who have a responsibility to correct misleading information in a timely manner. They also recommended that the Statement could do more to draw the readers’ attention to its caveats.76

Analysis

3.105 The Committee notes that the Statement is a very technical document. It has a specific interpretation, similar to a financial audit report. The Committee can see a risk of misunderstanding because of this technical and specific nature and also the fact it is not widely used throughout the community.

3.106 Around this context, the Committee received two points of view at the hearings on the use of the Statement. The first view, discussed above, was that Treasury and/or the Government had ownership of the document. Since the Statement is designed to inform debate, Treasury and/or the Government have a role in ensuring that its correct interpretation is promoted. The SMSF Owners’ Alliance argued that more could be done:

If I, as a party, publish a document that is intended to be helpful— it is there for a reason—and it is being misused or misunderstood, we just feel there should be a culture of a greater sense of

75 Dr David Knox, Mercer (Australia), Proof Committee Hansard, 21 October 2015, Canberra, p. 3. 76 SMSF Owners’ Alliance, Submission 10, p. 7; Mr Duncan Fairweather, SMSF Owners’ Alliance, Proof Committee Hansard, 21 October 2015, Canberra, p. 8.

ISSUES RAISED DURING THE INQUIRY 45

responsibility than there appears to be at the moment, to correct that misconception—not two years down the track.77

3.107 The alternative view, put by the Grattan Institute, was that those who quote from the Statement are responsible for ensuring that they use the right context.78 The Committee notes that this view is more consistent with the idea of open debate. If a person uses the Statement inconsistently with its warnings, they are open to criticism.

3.108 However, the fact that people seem to do so with limited sanction suggests the warnings in the document are not sufficiently clear to inform enough of its users.

Committee comment

3.109 It is worth keeping this issue in perspective. No matter the details discussed in this section, superannuation tax concessions are generous79 and a large number of people take advantage of them. If an income tax benchmark is used, the numbers will be large regardless of the type of estimate. Large tax expenditures are also consistent with a large number of Australians saving to enjoy their retirement, reducing pressure on the pension, and creating a pool of funds that can be productively invested in the economy.

3.110 Broadly, the Committee is of the view that policy outcomes are best served in the long run by a robust debate. The Committee acknowledges that a technical document such as the Statement may not always be used in accordance with the caveats included in it. But if a comment is made and reported in the media, it is always open to others to dispute it.

3.111 From the perspective of government, the Committee does not see a general role for Treasury to ‘intervene’ when the Statement has been incorrectly applied. The Committee does not accept that Treasury or the Government have a legal obligation to the public similar to what company directors have to shareholders or potential shareholders.

3.112 However, the Committee does support the idea of Treasury increasing the visibility of the warnings in the Statement about its limitations. It is a very technical document that has a specific interpretation. Its unique nature means that its specific interpretation is often not the common usage.

77 Mr Malcolm Clyde, SMSF Owners’ Alliance, Proof Committee Hansard, 21 October 2015, Canberra, p. 13. 78 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 2. 79 Grattan Institute, Submission 6, p. 5.

46 THE TAX EXPENDITURES STATEMENT

Further, the table of tax expenditures at the front of the document makes it easy for users to collect information, without requiring them to read the introductory text and caveats.

3.113 The Committee accepts that users of the Statement do not necessarily work through the document in a systematic manner. The Committee also accepts that its highly technical interpretation may not be readily understandable to first-time readers. Therefore, the Committee is of the view that increasing the visibility of the warnings in the document would assist users of the Statement and place more responsibility on them in how they use it. Possible methods include text boxes or regular footer references.

Recommendation 8

3.114 That Treasury consider ways of increasing the visibility or warnings in the Statement to better draw the attention of readers to the Statement’s limitations.

Quality of the estimates

Background

3.115 The availability of data is the key issue in generating estimates in the Statement. In evidence, Treasury stated that, if the ATO is not collecting data for a particular tax expenditure, then it probably will not be able to measure it:

Fundamentally one of the challenges for us is that virtually all of the tax expenditures are in the Treasury portfolio and are reliant on data sources which the ATO has. To the extent that the ATO is not collecting data … then we are fundamentally constrained.80

3.116 Chartered Accountants raised two data issues during the inquiry. The first was that the Statement does not discuss the data and methods used for each tax expenditure. Related to this is that Treasury has in the past substantially revise its estimates without explanation. Chartered

80 Mr Steve French, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 4.

ISSUES RAISED DURING THE INQUIRY 47

Accountants provided the following example in relation to the tax expenditure for GST on imported services.

Table 3.2 Estimates for tax expenditure of GST on imported services, various Statements, $m

TES year Estimate year 2007/8 2008/9 2009/10 2010/11 2011/12 2012/13 2013/14 2015/16

2014 150 150 170 180 190

2013 120 150 150 160 170 190

2012 110 120 150 160 180 210 230

2011 900 950 900 950 1,050 1,100 1,200 1,250

2010 900 950 900 1,000 1,100 1,200 1,300

2009 900 950 1,050 1,150 1,250 1,350

2008 800 850 900 1,000 1,100

Source CAANZ, Submission 3, p. 8.

3.117 The 2012 and later Statements gave estimates up to eight times higher for the tax expenditure, compared with previous Statements. Chartered Accountants noted that the only explanation provided in the 2012 Statement was that there were ‘improvements to modelling methodology and data.’81

3.118 Chartered Accountants recommended to the Committee that the Statement should include more information about how estimates are calculated, ‘including the data used, the reliability of that data, the assumptions made and sensitivity analysis.’ 82

3.119 The second issue that Chartered Accountants raised was that Treasury and the ATO were not necessarily making use of all possible data sources, or were considering discontinuing possible data sources. It also offered to work with Treasury and the ATO on measuring tax expenditures, noting that its membership may have valuable insights:

We see, for example, the ATO considering whether the business activity statement should have as many boxes on it. That is a work in process. We have not heard a final decision by the ATO. If those boxes were to be removed from the business activity statement, that may deprive some other agency—for example, the Australian Bureau of Statistics—of the relevant data they need to provide the community with information of a tax-related nature …

If it be the committee’s recommendation, we would be welcoming of working with Treasury on trying to engage with our members

81 CAANZ, Submission 3, p. 8. 82 CAANZ, Submission 3, p. 8.

48 THE TAX EXPENDITURES STATEMENT

or to provide further insights from our members about the particular tax expenditures that are being considered as part of policy reform. We are also conscious of the fact that some of our members are in a position to give the ATO and Treasury better data than perhaps is currently available—for example, the recently announced small business tax concession for $20,000 asset write-offs.83

Analysis

3.120 Good data is a key factor in developing good estimates. The Committee heard that Treasury would very much like better data, but is constrained in the costs it can impose on taxpayers and the resources that Government may need to expend as well:

… with tax data, fundamentally it is the answers that are provided by people or companies on their tax returns. So, it becomes a real question of how many questions can you put on a survey or on a tax return, what detail they can go into and at what point do the costs of additional questions or more complex questions start to outweigh the benefits.

As people who need data … we would love, from that perspective alone, more and more questions, but the reality is that data collection is not the only consideration when you are designing a survey or designing a tax return form. The costs of people having to fill out the survey and having to fill out the tax return are also very important and there are also … ATO or ABS costs that come into it. So, the perfect world where we would get a lot more data is not the only consideration. We have to live within an imperfect world where there are inherent limits on the data that we can get. It is very good data, but there are limits.84

3.121 The ATO made a similar comment, noting that it collects data primarily to calculate tax liability, as well as for the ATO to understand the risks to the tax system:

The other point is about the data that we collect. We principally collect it for the purpose of assisting taxpayers to calculate their liability for tax, and for the tax office to understand where the risks are in their system. There is always a tension between the collection of data and imposing red tape costs to our taxpayers

83 Mr Michael Croker, CAANZ, Committee Hansard, 19 August 2015, Canberra, pp. 2-3. 84 Mr Paul Palisi, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 4.

ISSUES RAISED DURING THE INQUIRY 49

and the third parties that we get data from. That is always going to be a kind of limitation on the quality that you can get.85

3.122 At the same hearing, the ANAO commented on its recommendations from the 2008 audit report. It stated that one of its goals has been to encourage Treasury and the ATO to take a more structured approach to collecting data for the Statement, rather than imposing compliance costs without regard to taxpayers:

… an element of recommendation 6 was for the development of an approach to prioritise reliability—to take a structured approach. What we are not saying is to go out tomorrow and increase the burden on taxpayers simply to provide a data stream for this purpose but start to work out where the priorities are, where you want to devote your efforts, where you get your best return for imposing a collection obligation and have that as a structured approach.86

Committee comment

3.123 During the inquiry, Treasury did not directly comment on the suggestion from Chartered Accountants that it should publish an explanation in the Statement for the data and methods used in compiling each estimate. Although Treasury would have the information to hand, inserting an explanation for each estimate would add to the cost of the document. Treasury stated during the inquiry that the Statement is already resource intensive.87

3.124 In principle, the Committee sees merit in the suggestion by Chartered Accountants because it will add to the credibility and usefulness of the Statement. It may also help users understand the document’s limitations. However, the cost implications mean that improvements should be made on a prioritised basis. Examples of such priorities are large tax expenditures and where an estimate is substantially revised, either in size or reliability.

85 Mr Andrew England, ATO, Committee Hansard, 9 September 2015, Canberra, p. 7. 86 Mr Mark Simpson, ANAO, Committee Hansard, 9 September 2015, Canberra, p. 7. 87 Mr Steve French, Treasury, Committee Hansard, 9 September 2015, Canberra, p. 1.

50 THE TAX EXPENDITURES STATEMENT

Recommendation 9

3.125 That Treasury include in the Statement an informative explanation of data and methods for large estimates and when estimates are substantially revised, either in terms of size or reliability.

3.126 In relation to data sources, the Committee accepts that Treasury and the ATO must be careful when collecting data from taxpayers not to indiscriminately impose compliance costs on them. This applies to all data, not just in relation to the Statement. However, the Committee did not receive evidence during the inquiry that these agencies are taking the sort of structured approach advocated by the ANAO. The Committee’s perception is that Treasury does the best it can with whatever data is produced through other processes.

3.127 The Committee also sees an opportunity for consultations to assist. If government is wary of increasing compliance costs for taxpayers, then part of a structured approach would be to talk these issues over with stakeholders. If a particular sort of data would be useful for the Statement, and stakeholders advise that taxpayers and their tax agents have it readily to hand, then the collection costs could be negligible. Treasury and the ATO gave no evidence of such consultations.

3.128 The Committee encourages Treasury and the ATO to take a prioritised, structured approach to data sources, as recommended by the ANAO. Consultations with stakeholders should provide information on what sort of data requirements will impose significant costs on taxpayers.

Recommendation 10

3.129 That Treasury and the Australian Taxation Office consult with stakeholders on possible data sources and apply a prioritised approach to data improvements.

Longer term estimates

3.130 Where practicable, the Statement provides eight years of estimates for each tax expenditure. These are the four prior years, the current year, and three forward years. The practice of providing three years of future

ISSUES RAISED DURING THE INQUIRY 51

estimates is consistent with the forward estimates system used in the Budget Papers.

3.131 In evidence, the PBO recommended to the Committee that longer term estimates could be useful where the effect of a tax expenditure is expected to change over the medium term, for example in 10-15 years. This behaviour would not be captured over the forward estimates:

… many tax expenditures have an impact over the estimates period that is not representative of the impact beyond the forward estimates period. As a result, from looking at the tax expenditures estimates that are presented you would get a misleading impression of what the significance is going to be over the medium term. Examples of that are things like depreciation provisions. If you introduce a provision for accelerated depreciation, the impact of that mounts over the years, will peak at some point and then decline. Most of that pattern occurs outside of the forward estimates. There are other provisions which, because of phasing in provisions or the like, may not mature in that forward estimates time frame.88

3.132 The PBO gave the example of an instant asset write-off and a benchmark that depreciates assets against effective life. In this case, taxpayers will have higher levels of deductions in the early years, producing a positive tax expenditure. But once assets are depreciated, the pattern will reverse as deductions fall to zero and are less than if the assets were depreciated over their effective life and they still had a notional value. This would result in a negative tax expenditure in later years.89

3.133 The Committee heard in evidence that the superannuation system is yet to mature.90 Therefore, medium term estimates for these tax expenditures may also be different from current estimates.

Committee comment

3.134 The Committee considers that extra analysis of this type would be a useful addition to the Statement. The document cannot cover every variation on a tax expenditure, but the Committee considers that, where the future behaviour of a tax expenditure is sufficiently different, and the tax expenditure has particular relevance, then more analysis is warranted. A

88 Mr Colin Brown, PBO, Committee Hansard, 14 October 2015, Canberra, p. 2. 89 Mr Colin Brown, PBO, Committee Hansard, 14 October 2015, Canberra, p. 2. 90 Mrs Andrea Slattery, SMSF Association, Proof Committee Hansard, 11 November 2015, Canberra, p. 9.

52 THE TAX EXPENDITURES STATEMENT

tax expenditure could be relevant from its size, precedent value, or other features.

3.135 The Committee also notes that extra resources may be involved. This is why the Committee suggests that Treasury be selective. The decision of what tax expenditures should receive this additional analysis could also be subject to consultation.

Recommendation 11

3.136 That Treasury produce longer term estimates for tax expenditures where they are expected to differ significantly from the forward estimates presented in the Statement. Deciding which tax expenditures should receive this treatment could be subject to materiality considerations and consultation.

Revenue gain estimates

Background

3.137 Expenditures in the Statement are generally estimated using the revenue forgone approach. This measures the use of the concession and the method is simple. For example, if $1 billion was spent on a GST-exempt service, then the tax expenditure would be an additional 10 per cent, or $100 million.

3.138 The revenue gain approach is more sophisticated. It seeks to take into account taxpayer behaviour. Therefore, if a tax concession is hypothetically removed, taxpayers might direct their investments or spending to other tax-preferred locations. The greater the scope for this to occur, the lower the revenue gain estimate. The advantage of this approach is that the estimate is a better representation of what the tax concession costs the Budget.

3.139 In its 2008 report, the ANAO recommended that Treasury ‘develop estimates of large or otherwise significant tax expenditures using the revenue gain method.’91 Treasury has implemented this recommendation by publishing revenue gain estimates for 10 large tax expenditures, which

91 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 66.

ISSUES RAISED DURING THE INQUIRY 53

includes seven of the 10 largest revenue forgone estimates. They are shown in table 3.3.

Table 3.3 Revenue gain estimates included in the Tax Expenditures Statement

Tax expenditure Estimate $m

Revenue

forgone

Revenue gain

Positive tax expenditures

C3 Concessional taxation of employer superannuation contributions 16,300 15,550

C6 Concessional taxation of superannuation entity earnings 13,400 11,750

H28 GST - Food 6,400 6,300

H16 GST - Education 3,950 3,550

H2 GST - Financial supplies - input taxed treatment 3,550 3,550

H19 GST - Health - medical and health services 3,550 3,500

C5 Concessional taxation of non-superannuation benefits 2,700 2,700

A38 Exemption of Family Tax Benefit payments 2,220 2,220

B14 Exemption from interest withholding tax on certain securities 1,860 1,310

Negative tax expenditures

F24 Customs duty -2,550 -2,550

Source Treasury, Tax Expenditures Statement 2014, 2015, p. 7.

3.140 Six of the revenue gain estimates are lower than the revenue forgone estimates. The extent to which they are lower is consistent with basic economics. The estimate for GST on food is almost the same as the revenue forgone estimate because consumers are not expected to change their consumption by much if GST were applied. The estimate for interest withholding tax is substantially lower, reflecting other tax-effective opportunities for investors. Four revenue gain estimates are the same as the revenue forgone estimates.

3.141 The Statement provides warnings about revenue gain estimates, in particular, that there is little information about taxpayer reactions. Therefore, assumptions are required, which ‘can be difficult to meaningfully substantiate.’92

3.142 In its follow-up inquiry, the JCPAA supported the publication of revenue gain estimates, recommending that estimates should be published for the 20 largest tax expenditures.93

92 Treasury, Tax Expenditures Statement 2014, 2015, p. 5. 93 JCPAA, Report 414: Review of Auditor-General’s Reports tabled between August 2007 and August 2008, 2009, p. 52.

54 THE TAX EXPENDITURES STATEMENT

3.143 In its response to the JCPAA, Treasury committed to publishing a ‘limited’ number of revenue forgone estimates. It also raised the problem of the assumptions required about taxpayer behaviour:

Estimates of six tax expenditures using the revenue gain, approach were published in the 2008 Tax Expenditures Statement (TES). The tax expenditures were chosen to highlight the impact of behavioural assumptions on estimates … It is also intended that the 2009 TES include a limited number of estimates of tax expenditures using the revenue gain approach. This time it is intended to focus on the largest tax expenditures …

It should be noted that there remain practical difficulties in making revenue gain estimates including the information or assumptions needed for the behavioural responses of taxpayers to policy changes and the assumptions must be made regarding the policy specifications for removing each tax expenditure.94

Analysis

3.144 In the inquiry, the Committee received a range of views about revenue gain estimates. The PBO commented that adding more revenue gain estimates to the Statement would significantly improve its value, although the method still has limitations:

The key thing about revenue gain items is that they are not a panacea. They give more information in terms of the impact on the budget of abolishing tax expenditures, but they are not necessarily additive. They have the same problems as the revenue forgone estimates that are included in the tax expenditure statement but, in respect of the individual measures, they do perhaps give a better idea of the budget impact.95

3.145 Associate Professor Julie Smith was more critical of revenue gain estimates. She argued that the assumptions required to support the estimates were inaccurate because people can be neither predictable nor rational. The witness stated:

As an economist I would have to say that the expertise and the information we contribute on this is pretty poor, because people are fairly unpredictable and the research that has been done around the world on behavioural responses to taxation is not very

94 Treasury, Executive Minute, received 3 February 2010, p. 1. 95 Mr Colin Brown, PBO, Committee Hansard, 14 October 2015, Canberra, p. 2.

ISSUES RAISED DURING THE INQUIRY 55

informative. There is a huge range in the estimates of behaviour change. They are highly contested estimates … people actually do not behave the way economists expect. For example, even though it is to people’s benefit to shift out of default superannuation schemes, they stay in them. It is not rational.96

3.146 The Grattan Institute supported Treasury’s current approach of publishing estimates for the larger and more debated tax concessions, noting that they are ‘extremely useful.’97 It also stated that revenue forgone estimates are often a good proxy for revenue gain:

We think there may be a case for Treasury to continue to refine its methodology and to continue to publish those revenue gain estimates but focusing on the largest and most contentious tax expenditures, where those are likely to be of value to the debate. It is also interesting to note that where Treasury has published those revenue gain estimates in the past, we have seen that their estimates of behavioural change are often smaller than people that criticise the revenue forgone estimates might suggest.98

Committee comment

3.147 The Committee supports Treasury’s work to date in preparing and publishing a small number of revenue gain estimates. Notwithstanding the difficulties in calculating them, they provide additional insight into the tax system.

3.148 The Committee notes the JCPAA’s 2009 recommendation for revenue gain estimates for the 20 largest tax expenditures. The Committee does not believe such a blanket expansion is warranted at this stage. This is because of the resources involved in calculating them, the difficulty in substantiating the assumptions, and the fact that they are often close to the revenue forgone estimates.

3.149 However, the Committee is of the view that the revenue gain estimates assist in explaining to users how the tax system works. They are also consistent with how most people interpret the Statement, namely they present the best estimate of what the Budget would gain if a concession were removed. Being mindful of resources, the Committee recommends that Treasury gradually increase the number of revenue gain estimates in

96 Associate Professor Julie Smith, Committee Hansard, 19 August 2015, Canberra, p. 6. 97 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, p. 1. 98 Ms Danielle Wood, Grattan Institute, Committee Hansard, 12 August 2015, Canberra, pp. 1-2.

56 THE TAX EXPENDITURES STATEMENT

the Statement on a priority basis. The Committee also sees value in Treasury consulting with stakeholders on this point.

Recommendation 12

3.150 That Treasury gradually expand the number of revenue gain estimates that it calculates for the Tax Expenditures Statement, with a focus on larger tax expenditures and those relevant to public debate. Stakeholder consultation may assist in this process.

Memorandum items

3.151 Memorandum items are features of the tax system that are not deviations from the benchmark and do not qualify as tax expenditures. However, they are of interest to the public and reporting on them would assist debate on the tax system. The commonly cited example of this is the Canadian tax expenditures document.

3.152 The ANAO’s 2008 report discussed the Canadian practice but did not make a specific recommendation on it. However, the ANAO did make a more general recommendation about reporting standards:

ANAO recommends that the Department of the Treasury develop standards to govern the integrated reporting of outlays and tax expenditures under the Charter of Budget Honesty, drawing on international developments in this area.99

3.153 In 2009, the JCPAA considered the Canadian approach. The JCPAA noted that it erred on the side of disclosure and that memorandum items covered aspects of the tax system that were preferential, even if they were within the benchmark. The report recommended that ‘Treasury further investigate the merits of the Canadian model of taxation expenditure reporting.’100

3.154 Treasury declined this recommendation, noting that the Australian document provided some information that was not available in overseas equivalents. It also commented that the ATO already published data on

99 ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08, p. 47. 100 JCPAA, Report 414: Review of Auditor-General’s Reports tabled between August 2007 and August 2008, 2009, p. 55.

ISSUES RAISED DURING THE INQUIRY 57

features of the tax system that could be regarded as memorandum items. Treasury stated:

Treasury also notes that Australia takes a broad view of what constitutes a tax expenditure and incorporates a substantial amount of information regarding benchmarks and individual tax expenditures. This information is more comprehensive than that provided by most other countries. Information on a number of the memorandum items identified under the Canadian system, such as tax deductions, are published in detail in the ATO’s annual Taxation Statistics report.101

3.155 The Committee notes that Treasury’s comment about the Statement taking ‘a broad view of what constitutes a tax expenditure’ is not necessarily consistent with the comments of the AFTS Review. Its final report stated that ‘many of the most important economic and distributional effects of taxes are incorporated in the benchmarks and so are not reported in the Tax Expenditures Statement.’102

3.156 During the inquiry, the PBO recommended to the Committee that memorandum items would be a useful addition to the Statement. It defined them as ‘things that are not so much tax expenditures but are of interest in the common debate around taxes.’103 The PBO gave the examples of negative gearing and work-related deductions:

Some things are relatively simple to get now—work-related expenses, for instance, is one that comes up frequently. Negative gearing is within the existing tax expenditures benchmark—it is not a tax expenditure per se, because the expenses associated with

an investment are deductible. That is really all that is happening— you are allowed a deduction against all income rather than just the income of the investment.104

Committee comment

3.157 The main reasons that Treasury gave for not including memorandum items in the Statement are that:

 the Australian benchmark is already broad and picks up a large number

of tax expenditures

101 Treasury, Executive Minute, received 3 February 2010, p. 2. 102 AFTS Review, Report to the Treasurer, Part Two, Detailed analysis, 2009, p. 731. 103 Mr Colin Brown, PBO, Committee Hansard, 14 October 2015, Canberra, p. 2. 104 Mr Colin Brown, PBO, Committee Hansard, 14 October 2015, Canberra, p. 2.

58 THE TAX EXPENDITURES STATEMENT

 some information on memorandum items is already publicly available

through the ATO.

3.158 The Committee does not accept these arguments. The AFTS Review did not find that Australia has a broad benchmark. Further, there will always be features of the tax system that attract debate. This is because complexity will lead to aspects of the tax system seen as preferential, and tax is inherently of public interest. So the benchmark, no matter how broad, will not pick up all areas of the tax system that are debated.

3.159 One of the purposes of the Statement is to inform debate. It has become the ‘go to’ document for information about concessions in the tax system. The community debates the effects of concessions relevant to the GST or superannuation tax concessions because they are large and they are in the Statement. But if other preferential aspects of the tax system are excluded from the Statement, the Committee is of the view that they are also excluded from the debate.

3.160 In one respect, much of the work for memorandum items has already been completed by the ATO through compiling tax statistics. Nonetheless, the Committee acknowledges that publishing data on memorandum items would involve additional work and resources. Therefore, the Committee supports a selective, risk based approach that provides the most useful information at the least cost. A consultation process would also assist.

Recommendation 13

3.161 That Treasury include selected memorandum items in the Tax Expenditures Statement. The decision on which items to include could be based on their size, relevance to public debate, and stakeholder consultation.

Bert van Manen Chair

A

Appendix A - Submissions

1. Australian Bureau of Statistics

2. Mercer

3. Chartered Accountants Australia and New Zealand

4. Australian National Audit Office

5. Associate Professor Julie Smith

6. Grattan Institute

7. Treasury

7.1 Supplementary Submission

8. Robert Carling, Centre for Independent Studies

8.1 Supplementary Submission

9. Parliamentary Budget Office

10. SMSF Owners’ Alliance Limited

11. Distilled Spirits Industry Council of Australia Inc

12. SMSF Association

13. Financial Services Council

60 THE TAX EXPENDITURES STATEMENT

B

Appendix B - Exhibits

1. Lidia Xynas and Dr Steve Jaynes: Revenue Law Journal: From tax expenditures to rebates: An ‘output based equity’ approach for Australia's retirement policies (provided by Associate Professor Julie Smith)

2. Dr Julie Smith: The Australian Economic Review: Gambling Taxation: Public Equity in the Gambling Business (provided by Associate Professor Julie Smith)

3. Eric J. Toder, Joseph Rosenberg, and Amanda Eng: National Tax Journal: Evaluating Broad-Based Approaches for Limiting Tax Expenditure (provided by Associate Professor Julie Smith)

4. Dr Julie Smith: Tax Expenditures: the $30 Billion Twilight Zone of Government Spending (provided by Associate Professor Julie Smith)

5. Dr Julie Smith: The Medicare levy surcharge arrangements Tax penalty or hidden tax subsidy? (provided by Associate Professor Julie Smith)

6. Eric J Toder: Responses to Tax Incentives in a Complex and Uncertain Tax Law (provided by Associate Professor Julie Smith)

7. Hana Polackova Brixi, Christian MA Valenduc, Zhicheng Li Swift: Tax Expenditures - Shedding Light on Government Spending through the Tax System (provided by Associate Professor Julie Smith)

8. SMSF Owners’ Alliance: Correspondence 11 February 2013 to Treasury (provided by the SMSF Owners’ Alliance)

62 THE TAX EXPENDITURES STATEMENT

9. SMSF Owners’ Alliance: Correspondence 26 February 2013 from Treasury (provided by the SMSF Owners’ Alliance)

10. SMSF Owners’ Alliance: Correspondence 10 March 2013 to Treasury (provided by the SMSF Owners’ Alliance)

11. SMSF Owners’ Alliance: Correspondence 11 February 2014 to Assistant Treasurer (provided by the SMSF Owners’ Alliance)

12. SMSF Owners’ Alliance: Media Release 28 February 2015 (provided by the SMSF Owners’ Alliance)

13. SMSF Owners’ Alliance: Media Release 22 May 2015 (provided by the SMSF Owners’ Alliance)

14. Robert Gottliebsen: ‘Treasury’s hoax is tormenting the super debate’ Business Spectator (provided by the SMSF Owners’ Alliance)

15. NATSEM: Increasing Superannuation Contributions for future health and aged care needs (provided by the Financial Services Council)

C

Appendix C - Public hearings

Wednesday 12 August 2015 - Canberra

Grattan Institute

Ms Danielle Wood, Fellow

Mr Brendan Coates, Senior Associate

Wednesday 18 August 2015 - Canberra

Chartered Accountants Australia and New Zealand

Mr Michael Croker, Tax Leader Australia

Ms Susan Franks, Senior Tax Advocate

Private capacity

Associate Professor Julie Smith

Wednesday 9 September 2015 - Canberra

Australian National Audit Office

Mr Mark Simpson, Acting Group Executive Director, Performance Audit Services Group

Mr Andrew Morris, Executive Director, Performance Audit Services Group

Treasury

Mr Steve French, Division Head, Tax Analysis Division

Mr Anthony King, Manager, Retirement Income Modelling Unit

Mr Paul Palisi, Manager, Personal Tax and Transfers Unit, Tax Analysis Division

Mr Bruce Bastian, Senior Adviser, Retirement Income Modelling Unit

Mrs Emma Nation, Analyst

64 THE TAX EXPENDITURES STATEMENT

Australian Taxation Office

Mr Andrew England, Deputy Commissioner, Integrated Tax Design

Ms Julia Neville, Assistant Commissioner, Integrated Tax Design

Wednesday 14 October 2015 - Canberra

Parliamentary Budget Office

Mr Phil Bowen, Parliamentary Budget Officer

Mr Colin Brown, First Assistant Parliamentary Budget Officer, Budget Analysis Division

Mr Tony McDonald, Assistant Parliamentary Budget Officer, Revenue Analysis Branch

Wednesday 21 October 2015 - Canberra

SMSF Owners’ Alliance

Mr Duncan Fairweather, Executive Director

Mr Malcolm Clyde, Finance and Research Director

Mercer (Australia) Pty Ltd

Dr David Knox, Senior Partner

Wednesday 11 November 2015 - Canberra

Centre for Independent Studies

Mr Robert Carling, Senior Fellow

Financial Services Council

Mr Andrew Bragg, Director of Policy & Global Markets

SMSF Association

Mrs Andrea Slattery, Managing Director/CEO

Mr Jordan George, Head of Policy

D

Appendix D - Implementation of past recommendations

Table D.1 is provided on the next page.

66 THE TAX EXPENDITURES STATEMENT

Table D.1 Summary of recommendations and issues raised by the Auditor-General, JCPAA and AFTS inquiry and their implementation

Issue Recommendation or observation Agency response Agency action or comment Status

Reviews of tax expenditures Treasury conduct ongoing prioritised reviews: ANAO 2008, rec 1(a)

Agreed Reviews commenced in 2007 and discontinued in 2011. Treasury stated they provide little value when the Government has a different policy agenda.

Discontinued

Treasury publish the reviews: ANAO

2008, rec 1(b)

Qualified - A matter for Government

No reviews were published. Not

implemented

Budget processes

Treasury advise Ministers on better integrating tax expenditures and outlays in the Budget process: ANAO 2008, rec 2

Agreed Changes to the Budget Process Operational Rules mean that the two are largely treated the same. However, in 2011, about half of new or changed tax expenditures did not go through the Expenditure Review Committee of Cabinet (but complied with the Rules). Treasury stated that tax expenditures are considered during the Budget, depending on the Government’s policy agenda.

Partly implemented

Budget rules should treat spending and

tax expenditures symmetrically: AFTS, rec 135

None Not mentioned to the Government’s response to the AFTS Review in 2010. However, see above. Implemented

Treasury develop standards to govern

the integrated reporting of outlays and tax expenditures: ANAO 2008, rec 3

Qualified - The two terms have inconsistent definitions

Treasury stated that preparing the Statement is already resource-intensive. Developing a new reporting method may not be the best use of resources. Further, the key issue is the benchmark, and developing standards will not resolve this problem.

Not implemented

Reporting standards are independently

developed for identifying and measuring tax expenditures: AFTS, rec 137

None Not mentioned in the Government response to the AFTS Review in 2010. However, see above. Not implemented

Accuracy of the estimates Treasury liaise with agencies that collect revenue to identify all those that

administer relief from taxes: ANAO 2008, rec 4(a)

Agreed Treasury correspondence with agencies to identified four new tax expenditures in 2008-12. In the same period, Treasury found 38 new tax expenditures in the Treasury portfolio. Treasury stated that the benefits were marginal.

Implemented

Treasury develop annual arrangements

to obtain data from agencies outside its portfolio: ANAO 2008, rec 4(b)

Agreed Treasury and the ATO approached other agencies to improve their data. However, there was little improvement to the reliability of the estimates. Treasury stated that most tax expenditures are in the Treasury portfolio.

Implemented

APPENDIX D - IMPLEMENTATION OF PAST RECOMMENDATIONS 67

Issue Recommendation or observation Agency response Agency action or comment Status

Treasury include information in the

Budget Papers on improvements to estimates due to obtaining data from other agencies: JCPAA, rec 9

Qualified - Would instead monitor the quality of estimates over time

Up to the 2013 audit, Treasury had no systematic approach to improving reliability. A ‘post-mortem’ was conducted after each Statement, but it instead focussed on formatting and production. There was no improvement in reliability from 2008 to 2011. Treasury states that they key issue around estimates is the availability of data and collecting better data would increase compliance costs for taxpayers.

Not implemented

Treasury and the ATO develop estimates

of large tax expenditures using the revenue gain method: ANAO 2008, rec 5

Agreed The 2014 Statement includes estimates for 10 large tax expenditures. Implemented

Treasury develop revenue gain

estimates of the 20 largest tax expenditures: JCPAA, rec 7

Qualified - Practical difficulties in these estimates

The 2014 Statement includes estimates for 10 large tax expenditures. Treasury considers that reporting estimates for 10 tax expenditures is sufficient.

Partly implemented

Treasury investigate the Canadian

model, which uses the broader concept of ‘memorandum items’: JCPAA, rec 8

Agreed Treasury investigated the Canadian model but decided not to move further in that direction. Treasury is also concerned that extra estimates of this nature would increase misunderstanding of the Statement.

Implemented

Treasury prioritise improvements to the

reliability of the estimates: ANAO 2008, rec 6(a)

Agreed Up to the 2013 audit, Treasury had no systematic approach to improving reliability. A ‘post-mortem’ was conducted after each Statement, but it instead focussed on formatting and production. There was no improvement in reliability from 2008 to 2011. Treasury states that they key issue around estimates is the availability of data and collecting better data would increase compliance costs for taxpayers.

Not implemented

Treasury include reliability information for

each expenditures in the Statement: ANAO 2008, rec 6(b)

Agreed Treasury adopted the framework used by the ANAO in the 2008 audit in each Statement up to the 2013 audit. Implemented

Treasury and the ATO develop estimates

for large, unquantified tax expenditures: ANAO 2008, rec 6(c)

Agreed Of 12 items identified in 2008, three had been quantified by the 2013 audit. Treasury and the ATO meet annually to discuss unquantified expenditures, but little documentation kept on why expenditures were unquantified. Treasury states that they key issue around estimates is the availability of data and collecting better data would increase compliance costs for taxpayers.

Partly implemented

68 THE TAX EXPENDITURES STATEMENT

Issue Recommendation or observation Agency response Agency action or comment Status

Treasury assess options for reliably

measuring a new expenditure during policy development: ANAO 2008, rec 6(d)

Qualified - taxpayer compliance costs should be low

Treasury reiterated its aim of minimising the information sought from taxpayers to reduce compliance costs. No information

Treasury and the ATO standardise how

they allocate reliability ratings to estimates: ANAO 2013, rec 1

Agreed Treasury announced a new system in the 2013 Statement. The reliability of the data, assumptions, and other factors such as volatility, are scored on a 0-3 ranking, and these are added together to give an overall 0-9 score.

Implemented

MYEFO The Charter of Budget Honesty Act 1998 requires the analysis of tax expenditures to be included in MYEFO, rather than as a stand-alone document: ANAO 2008, p. 41

NA As of the 2013 audit, Treasury had not explicitly raised this

with Government. Act still requires Statement to be in MYEFO.

NA

Amend Charter of Budget Honesty Act

1998 so the Statement does not need to be in MYEFO; AFTS, rec 136

None See above. Further, the issue was not covered in the

Government response to the AFTS Review in 2010. Not implemented

States and Territories

COAG encourage the States and Territories to uniformly report tax expenditures using independent standards: AFTS, rec 138

None Issue not included in COAG communiques. Not mentioned in the Government response to the AFTS review in 2010. Not implemented

Source ANAO, Preparation of the Tax Expenditures Statement, 2008, Audit Report No. 32 2007-08; ANAO, Preparation of the Tax Expenditures Statement, 2013, Audit Report No. 34 2012-13; Treasury, Tax Expenditures Statement 2014, 2015; Treasury, Australia’s Future Tax System - Report to the Treasurer, Part Two - Detailed Analysis, 2009, p. 729; Council of Australian Governments, www.coag.gov.au, accessed 3 August 2015; Treasury, Submission 7, pp 5-10; The Hon. Kevin Rudd MP, Prime Minister, and the Hon. Wayne Swan MP, Deputy Prime Minister and Treasurer, ‘Stronger, Fairer, Simpler: A Tax Plan for Our Future, 2 May 2010, Media Release No. 28 (for the Hon. Wayne Swan).