4.1 The first term of reference for this review has required the Inspector-General to identify and consider the adequacy and effectiveness of the Tax Office's policies and procedures relating to the manner in which it processes GST refunds arising from the lodgment of credit BASs. This chapter sets out the Inspector-General's findings and recommendations in relation to this term of reference.

Nature of Tax Office's policy for processing GST refunds

4.2 The Tax Office states that it applies risk management methodologies to its administration of the tax system. This involves identifying risks and analysing their likelihood and potential consequences. The Tax Office states that it works in an environment where it is not practical to materially treat all risks, so its approach is to identify and take steps to properly manage the highest risks and to monitor lesser risks to prevent them from increasing.5

4.3 The approach which underlies the Tax Office processing of GST refunds reflects the Tax Office view that GST returns which show a net credit due to the taxpayer are the biggest risk in the GST system. This is based on overseas experience.6 The Tax Office has also advised that the Financial Management and Accountability Act 1997 requires that it ensure the integrity of such refunds.

4.4 As a result of this approach, the Tax Office has specialised examination procedures which apply to GST refund returns only. These procedures do not apply to GST payable returns. Of the refund returns that are subject to these examination procedures, the majority will be examined prior to, rather than after, the refund is paid. This policy has not changed since the time when the GST was first introduced.7

4.5 The main risks associated with the payment of GST refunds are as follows:

  • The first risk is that GST refunds may be paid to taxpayers in circumstances involving criminal fraud. For this purpose, the Tax Office uses the definition of fraud which has been provided in guidelines that have been issued to all Commonwealth Government agencies. These guidelines define fraud as 'dishonestly obtaining a benefit, both tangible and intangible, by deception or other means.'8
  • The second risk is that the quantum of a GST refund that is paid to a taxpayer may be inflated because mistakes have arisen in the calculation of that refund.

4.6 According to the Tax Office, GST refund cases involving mistakes could involve recklessness, deliberate evasion and dishonest claims as well as errors. The Tax Office considers that the range of behaviours leading to mistakes means that, unless there is clear evidence of an intent to defraud, it is more appropriate and faster to resolve these cases through the administrative powers of the Commissioner.

4.7 The Tax Office principally mitigates the above risks by intercepting GST refunds prior to their payment through the use of a control known as the risk rating engine (RRE).

4.8 The Tax Office has provided the Inspector-General with an overview of the function/purpose of the RRE and has further advised that the RRE is in the process of being significantly changed.

4.9 It is inappropriate for this report to discuss in detail the manner in which the RRE operates. However, in essence the RRE is applied at a particular point in the automated processes which the Tax Office uses to process GST refunds. It operates to flag those GST refund cases which the Tax Office considers should not be automatically processed and paid but which should be subject to manual examination by a tax officer.

4.10 The RRE was initially designed to operate on the basis of a number of risk characteristic tests and exception tests. However, a number of these tests were suspended with effect from October 2000.9

4.11 The Tax Office has advised that currently there are two levels of risk profiling within RRE processing. The first level is focused on identifying potential 'serious non-compliance' (fraudulent behaviour) and the second, on identifying potential general non-compliance (although serious non-compliance can originate from general non-compliance cases). It advises that all taxpayers are subjected to both levels of processing. However, the serious non-compliance processing focuses on characteristics of the taxpayer and whether the taxpayer is of concern to the Tax Office, while general compliance profiling focuses principally on the transaction, that is, details on the lodged activity statement.

4.12 Up until July 2004, there were three main criteria which would operate to halt a particular taxpayer's GST refund under the RRE process. They were as follows:

  • the refund involved a new business;
  • the refund exceeded certain dollar value thresholds; or
  • the refund originated from a taxpayer with a high fraud risk profile or who was known to the Tax Office.

4.13 In the case of the second criterion, the level of the threshold depended on the size of the relevant enterprise.

4.14 From the beginning of July 2004, as discussed later in this chapter, the Tax Office has implemented more focused checks relating to unusual behaviours over time and has released cases with a good compliance history.

4.15 The features of the RRE that operated prior to July 2004 led to approximately 90 per cent of the total value of GST refunds claims being stopped for checking during the 2003-04 year.10 These stopped refunds represented 4.3 per cent of the total number of GST refunds.11

4.16 This result suggests that during 2003-04 the RRE may principally have targeted refunds based on their size. This is a rather narrow and unsophisticated risk management approach. Due to the policy design of the GST system, a relatively high level of refunds is to be expected.

Effect of manual intervention on prompt payment of refunds

4.17 The Tax Office's approach of subjecting the vast majority of the dollar value of GST refunds to manual checking has the obvious consequence of holding up these refunds for the time it takes these manual processes to be performed.

4.18 As such, this approach creates tension with one of the principal design features of Australian GST law — that refunds should be paid promptly. This feature is reflected in the current legislative provisions governing GST refunds which require the Tax Office to pay interest to taxpayers if GST refunds are not made within 14 days of lodgment of the relevant BAS or the provision of all relevant information.12

4.19 The principle that GST refunds should be paid promptly has been embedded in Australia's GST law for three main reasons. The first reason is that it is a core principle of GST laws worldwide that business-to-business transactions should generally bear no GST burden.13 The second reason is that any delay in paying GST refunds will cause cash flow issues for entities entitled to those refunds. The third reason is that a failure to pay GST refunds promptly to businesses and other entities entitled to those refunds will create economic distortions. For example, new firms may face a competitive disadvantage if they are subject to delayed refunds. Likewise, the competitiveness of GST free sectors (such as the export sector) may be harmed by significantly delayed GST refunds.14

Need to balance risk to revenue against prompt payment of refund claims

4.20 The Tax Office's approach to administering GST refunds must be one which achieves an appropriate balance between the competing priorities of protecting the revenue from mistaken or fraudulent claims and the necessity of ensuring that refunds are paid promptly.

4.21 In the view of the Inspector-General, the Tax Office's current approach to administering these refunds, through the operation of the risk rating engine and subsequent processes, has involved an inadequate application of risk management processes. The Tax Office has identified fraud and mistakes as risks that have to be managed but has adopted a simplistic method of assessing these risks. This approach has had the effect of delaying the vast majority of GST refunds in monetary terms and has involved the investment of close to one-quarter of its overall GST audit program effort.15 However, this approach may have yielded only minimal protection to the revenue, in terms of either the number of GST fraud cases detected or the value of mistaken claims detected.

4.22 The Tax Office advises that it has had a program operating since July 2003 to analyse BAS refund data and the state of the refund system. As discussed later in this chapter, from July 2004 it has implemented more sophisticated tests, together with an override facility in relation to certain GST refunds.

Refund cases involving fraud

4.23 Material provided by the Tax Office to the Inspector-General for the purposes of this review indicates that the risk rating engine processes do not result in the identification of a significant number of GST refund fraud cases.

4.24 The Tax Office was unable to provide the Inspector-General with details of the number of fraud cases involving GST refunds which have been completed over the four years since the commencement of the GST on 1 July 2000. It was able to provide the number of prosecutions for all GST-related offences for these years. The total number of such prosecutions was small, ranging from four in 2000-01 to 80 in 2003-04.

4.25 In addition, the Australian National Audit Office (ANAO) found in its 2003 review into the Tax Office's management of GST fraud that, at the time of its report, only 1.9 per cent of the GST fraud cases registered on the Tax Office's reporting systems for fraud identified the RRE as the referral source. The ANAO further stated that it recognised that a further 53 per cent of identified GST fraud cases 'would in all probability' have originated from work generated by the RRE. However, it also noted that in the sample of cases it examined the RRE was not identified as the referral source for any cases.16

Refund cases not involving fraud

4.26 During 2003-04, a number of the GST refunds stopped by the RRE were found to be partly payable because of mistakes. These mistakes were detected through manual desktop checking procedures.17

4.27 The total dollar value of adjustments made to these GST refunds was $275 million. This is small compared to the $20.574 billion total dollar value of all GST refund cases that were initially stopped by the RRE.18 This suggests that the vast majority of GST refund cases that were initially stopped by the RRE during 2003-04 were later released by the Tax Office.

4.28 The Tax Office has not been able to advise the Inspector-General of the split of all of the above $275 million of GST adjustments into adjustments which are of a permanent nature or of a (technical) timing nature.

4.29 Adjustments of a permanent nature are adjustments where the taxpayer has underpaid their GST or overclaimed their input tax credit entitlement and the underpayment or overclaim will not be reversed in a subsequent time period. Examples include input tax credits claimed by a purchaser where the supplier of the relevant item is not registered or required to be registered for GST.

4.30 GST adjustments which are of a (technical) timing nature are those adjustments which may be 'reversed' in a subsequent tax period. Examples include cases where a taxpayer has claimed an input tax credit for a particular transaction in one BAS, but is not in possession of a valid tax invoice for that transaction by the time of lodgment of that BAS. In this case, the credit is not claimable in the earlier BAS, but is claimable in the tax period when the taxpayer has obtained a valid tax invoice.

4.31 The Tax Office was able to advise the Inspector-General how much of the above adjustments relates to permanent and timing adjustments for large business taxpayers. It has advised that $14.5 million of the above adjustments related to large business taxpayers and that, of this amount, $12 million related to permanent adjustments and $2.5 million to timing adjustments.

4.32 In many cases, the Tax Office's desktop checks on the refund related to only part of the claimed refund. However the entire amount of the refund was delayed until these checking processes were completed. This had the effect of unnecessarily delaying the portion of the refund which the Tax Office ultimately decided to release.

GST refunds involving net payable returns

4.33 The Tax Office through its overall compliance program undertakes a mix of pre- and post-issue checks on the correct reporting of GST liability. Refund cases are reviewed prior to the issue of the refund. In debit cases (some of which may also involve mistakes in relation to GST refunds) the Tax Office undertakes a post-issue review. It does not hold up the processing of cases where payments to the Tax Office are involved.

4.34 The weighting of the Tax Office's overall GST audit activities between examining GST refund returns and other audit activities (including the examination of GST payable returns) is illustrated in Appendix 4. This appendix contains a table provided by the Tax Office which shows its planned GST audit activities for 2004-05. In examining this table, it should be noted that the number of GST payable returns is about four times the number of GST refund returns.19

Obligations to ensure the integrity of refunds

4.35 The Tax Office has justified its approach to checking GST refunds on its obligations to ensure the integrity of refunds under the Financial Management and Accountability Act 1997.

4.36 The Inspector-General notes that the efficiency and effectiveness of the Tax Office's processes for disbursing Government monies are not matters which are within the terms of reference of this review.

Nature of operation of the Risk Rating Engine on particular taxpayers

4.37 Once flagged by the RRE for verification, GST refund cases have been streamed to one of the following verification areas of the Tax Office prior to the refund being paid:

  • the Serious Non Compliance branch (SNC) (for cases which appear to involve fraud);
  • the Interpretation and Large Enterprise Compliance area (ILEC) (for cases involving large business taxpayers, being taxpayers who are part of an economic group with an annual turnover of over $100 million);
  • the Government and Community sector branch (GCS) (for refund cases involving large taxpayers in the government or community sector);
  • 25a Compliance Verification Centre (CVC) (for all other small business cases, including small government and community sector taxpayers). A CVC will then refer a case to the General Field area where the refund is to be verified by detailed field activity.

Refund verification processes for large companies and government and non-profit bodies

4.38 During 2003-04 the risk rating engine operated to flag GST refunds exceeding certain dollar thresholds for manual checking.

4.39 This approach did not take into account the tax risk profile of the taxpayers being subjected to these processes. As a result, GST refunds which exceed certain dollar values for low-risk taxpayers, such as large companies and government entities, were delayed for the time that it took the Tax Office to perform these verification processes.

4.40 These processes may have taken only a few days to complete for a majority of these taxpayers. However, for a number of such taxpayers these processes took longer than a few days, through no fault of the taxpayer concerned. During 2003-04, 39 per cent of all refunds for large enterprise taxpayers that were stopped for verification were delayed for longer than 14 days. For government and community sector taxpayers, 3 per cent of such cases were delayed for longer than 14 days.

4.41 This review has also established that these verification processes generated $14.5 million of GST adjustments for large enterprise taxpayers for the 2003-04 year. For the government and non-profit sector, $0.1 million of tax adjustments could be attributed specifically to this activity.

4.42 The Tax Office has advised the Inspector-General that in 2003-04, $7.35 billion of GST refunds claimed by large enterprise taxpayers were flagged for manual checking by the risk rating engine and $7.16 billion of GST refunds were flagged for checking by this process for large government and community sector taxpayers. The adjustments made in respect of these manual interventions therefore represent 0.2 per cent of the total refunds held for large taxpayers and 0.0014 per cent of the total refunds held for large government and community sector taxpayers.

4.43 The Tax Office has advised that during the 2003-04 year, it raised $1.1 billion of GST tax adjustments (excluding penalties and interest) from all GST audit activities.20 It has also advised that the total amount of GST collected in 2003-04 was $33.2 billion.21 These figures give a ratio of GST adjustments from all audit activity to net GST collected of 3.3 per cent. This percentage is 16.5 times that which was achieved from RRE-generated verifications for large enterprise taxpayers and 2,357 times that which was achieved from RRE-generated verifications for government and community sector taxpayers.

4.44 The above figures suggest that the Tax Office should consider minimising or reducing the level of pre-issue checks applied to these forms of entity and redirecting the resources that are applied to this activity to other forms of audit activity.

Tax Office response made during the review

4.45 The Tax Office has stated during the course of this review that both its Interpretation and Large Enterprise Compliance (ILEC) and Government and Community Sector (GCS) branches propose to introduce an override process into the operation of the RRE in relation to large publicly listed companies, and any large taxpayer that has a minimum of a five-year history with the Tax Office (that is, has been registered for any tax for a minimum of five years) that indicates a satisfactory compliance record. It is estimated that 1,453 refunds, from some 4,114 ILEC taxpayers (not all will lodge refund claims), amounting to $5 billion would be released before being verified. ILEC and GCS will select a relevant sample of these refunds to verify after the refund has been paid to the taxpayer (post-issue checks). This policy has been progressively implemented from July through to December 2004.

4.46 The Tax Office has also advised that it is developing appropriate criteria for Government and Community Sector (GCS) in relation to 197 Government market segment entities, the majority of which are entitled to claim refunds. It is expected that close to all of these entities will become subject to the override policy. This will result in a reduction of $4.6 billion in stopped refunds for this sector.

4.47 GCS also manages 372 large market entities in the health sector. Some of these entities are entitled to claim refunds. This sector is currently being evaluated to determine the number of entities that would meet the criteria for inclusion in the override process.

Inspector-General's comments

4.48 The Tax Office's response indicates that $5 billion of GST refunds for ILEC taxpayers will no longer be potentially subject to manual pre-refund checks.

4.49 The Tax Office's current approach to this issue means that the balance of GST refunds for ILEC taxpayers may continue to be subject to the existing pre-refund checking processes.

4.50 The Inspector-General notes that the Tax Office's proposed processes for government organisations means that in the future the GST refunds for 'close to all' such taxpayers may now be issued without verification prior to payment. This reflects the low likelihood of any overpaid amounts being difficult to recover.

Verification activities for small business taxpayers

4.51 This review has also established that the risk rating engine and subsequent verification processes (other than those involving field audit activity) have resulted in $178.5 million of GST adjustments in the 2003-04 year for small business taxpayers. A further $82 million of GST adjustments was recovered from small business taxpayers as a result of more detailed audit activity on refunds stopped for checking.

Failure of RRE criteria to recognise the nature of the taxpayer's industry or prior compliance history

4.52 Both taxpayers and tax practitioners raised concerns that the Tax Office's 2003-04 RRE processes for checking GST refunds for both large and small business taxpayers did not allow either the nature of the taxpayer's industry, the life cycle stage of the taxpayer's business or their past history in terms of previous compliance checks to halt the risk rating engine flagging the case for the application of manual checking procedures.

4.53 As a result of this, taxpayers asserted that they were having successive refunds delayed despite being in an industry where GST refunds were the norm or where the taxpayer's previous compliance history indicated that such a procedure was inappropriate. Businesses and other enterprises that stated that they were subject to these repeated verification process included property developers, exporters and business making other forms of GST free supplies, such as food.

4.54 The Inspector-General notes that the risk rating engine's pre-July 2004 process of halting all refunds over certain dollar values supports these submissions made by taxpayers. This process meant that the payment of a GST refund could be delayed even in cases where the refund was not unusual for the taxpayers concerned.

4.55 This type of approach for processing GST refunds is not consistent with that which has been recommended by the International Monetary Fund (IMF) and other international parties. The IMF recommends that revenue authorities should have processes in place to check GST refunds prior to payment when the refunds involve new businesses and unusually large refunds.22 The term 'unusually large' is interpreted as being abnormal in view of the expected circumstances of the taxpayer.

4.56 The Tax Office has also provided material to the Inspector-General which confirms taxpayers' concerns that, in certain cases, their BAS refunds were halted by the RRE, despite their earlier refunds having been recently checked.

4.57 The Tax Office has advised that during the period from 1 July 2003 to 29 February 2004, a total of 20,020 taxpayers had multiple refund activity statements stopped for verification activity. Overall, this represented 77,128 activity statements, or an average of four activity statements per taxpayer which had to be verified before the funds were released.23

4.58 The above approach to processing GST refunds is not in accordance with the Tax Office's Compliance Model. Under this model, the nature and scale of any Tax Office audit activity needs to take into account the taxpayer's attitude to tax compliance. A GST refund policy which halts refunds for all taxpayers on the basis of the size of the relevant refund does not take into account a taxpayer's compliance profile.

Tax Office response made during the review

4.59 The Tax Office advises that it has had significant new approaches in train to review the impact of the RRE on particular taxpayers since July 2003. The Tax Office advises that it was well advanced in the process of improving the RRE tests to take account of earlier BAS reviews of businesses as well as industry characteristics prior to the commencement of this review. It states that these new measures have been implemented from July 2004.

4.60 Under these new measures, pre-refund checks will no longer be conducted where the relevant taxpayer has previously been the subject of a verification check within a certain time period and the amount of the refund claimed is not unusual for that particular taxpayer, for example, because of the nature of its industry24 or because of the season during which the refund is claimed.25

4.61 The Tax Office has advised that this strategy will account for around 40 per cent of the work items generated by the risk rating engine.26

Inspector-General comment

4.62 The Inspector-General considers that these initiatives should result in Tax Office practices for pre-refund checks being more closely aligned with the objective of ensuring that pre-refund checks are conducted only on unusually large GST refunds.

4.63 However, these new measures have not been in place for a sufficient period of time for the Inspector-General to assess at this stage the extent to which this measure might meet this objective.

Key recommendation

4.64 The findings referred to above in relation to the first term of reference for this review lead to the following key recommendation.

Key recommendation 1

The Inspector-General recommends that the Tax Office improve its systems to better match the risk issues associated with paying GST refunds. These systems need to achieve a better balance between paying GST refunds in a timely manner and preventing fraudulent or incorrect refunds from issuing.

Tax Office Response

4.65 The Tax Office agrees with this recommendation. Since the introduction of the GST the Tax Office has had effective systems in place to prevent the payment of fraudulent or incorrect GST refunds and its performance in paying GST refunds in a timely way compares favourably with other tax administrations.

4.66 The Australian National Audit Office, in its performance review of Goods & Services Tax Fraud Prevention & Control in 2002-03 acknowledged this effectiveness in its statements:

'The ATO has well structured compliance programs that, although not specifically directed at GST fraud, incorporate fraud prevention strategies, detection capabilities and enforcement policies'27.

'The ATO has implemented a highly visible GST compliance program with broad coverage across business that aims to:

  • identify incorrectly calculated/or avoided tax;
  • provide assurance that the GST system is working as intended; and
  • assess the effectiveness and targeting of education programs.'28

4.67 The ANAO found that the GST General Compliance program is based on the ongoing assessment of the GST compliance risks facing the Tax Office and forms part of an overarching compliance program.

4.68 With the benefit of our experience in administering the GST, the Tax Office has been able to refine its refund integrity program. A program of change was set in train in July 2003 with the outcomes implemented from July 2004. The program in particular focuses on assessing the likelihood and consequences of GST refunds being fraudulent or incorrect. This program has resulted in additional more sophisticated case selection rules, as well as an override facility which provides more automated streamlining of cases with an historical record of good compliance and regular refund claims. As a result of the changes made to date, it is anticipated that the number and value of refunds delayed for verification will be substantially reduced in 2004-05 and future financial years. This program will continue with further monitoring of refunds.

Other concerns with Tax Office processes for GST refunds

Third party verifications

4.69 Submissions to this review raised concerns that GST refunds were delayed, prior to payment, while Tax Office verification staff made checks with third parties to confirm the veracity of tax invoices of the taxpayer claiming a refund of input tax credits. It was submitted that, under these procedures, the Tax Office also seeks to verify that the supplier has accounted for their part of the transaction and has remitted the GST collected to the Tax Office.

4.70 One submission noted that:

'It is only when both sides of the transaction are to the satisfaction of the Tax Office that the GST refund is remitted to the taxpayer claiming the input tax credit.'29

4.71 Concerns were also raised that there were no administrative or legal reasons for the delay arising from these checks. It was asserted that, in cases where a taxpayer claiming an input tax credit has a legitimate tax invoice and there is no obvious perception of collusion between the supplier and purchaser to subvert the GST regime, then there was no reason to withhold the GST refund simply because the supplier had not fulfilled its obligation.30

4.72 Submissions also noted that third parties who were contacted under these procedures may not process these requests for confirmation on a timely basis. This is because third parties are generally aware that these requests do not relate to their own tax affairs but those of other taxpayers. This awareness exists even though Tax Office staff are not permitted, under the secrecy provisions of the taxation law, to disclose to the third party the nature of the review they are conducting or the tax affairs of the recipient of an invoice.

Tax Office response made during the review

4.73 The Tax Office has made the following response to these concerns.

4.74 Firstly, it has stated that its procedures do not support claims that it is only when both sides of a transaction are to the satisfaction of the Tax Office that the GST refund is remitted to the taxpayer claiming the input tax credit.

4.75 Secondly, the Tax Office has stated that it will continue to undertake third party checks in some situations. It advises that it has concerns about the integrity of some refund claims. For example, it notes that there have been a number of fraud cases where large refunds have been claimed for heavy equipment for which tax invoices have been produced. The Tax Office further advised that third party checks have subsequently shown these tax invoices have been artificially generated and the reported transactions did not occur.

4.76 The Tax Office believes that removal of these checks would have a severe impact on the integrity of the refund system and encourage even greater levels of fraudulent behaviour.31

Inspector-General comment

4.77 The Inspector-General notes that the above Tax Office response does not address the concerns raised about delays arising from third party verifications in cases which do not involve fraud.

Delayed refund interest

4.78 Tax Office policy is that where a GST refund has been delayed and the delay is due to Tax Office processing procedures, or to Tax Office verification procedures which involve merely the verification of entries made on a BAS, delayed refund interest will be paid to the taxpayer. This payment should be made automatically, that is, without any request for payment by the taxpayer. The interest is calculated from the day that is 14 days after the date of lodgment of the BAS up until the day the relevant refund is paid.32

4.79 The rate of delayed refund interest is the same as that which applies to overpayments of tax, which is currently 7 per cent less than the General Interest Charge. It is calculated on a simple interest basis. For the quarter ended December 2004 the rate of delayed refund interest was 5.44 per cent.33

4.80 In a number of submissions made to this review, taxpayers who experienced delays of more than 14 days from the date their BAS was lodged (or relevant information was provided) until the date when their GST refund was paid claimed that delayed refund interest was not automatically paid to them. These taxpayers claimed that they were required to specifically contact the Tax Office and request that this interest be paid.

4.81 The non-automatic payment of delayed refund interest was also observed in a number of GST refund cases examined by staff of the Inspector-General during fieldwork conducted for this review.

Tax Office response made during the review

4.82 The Tax Office has advised that delayed refund interest will not be automatically calculated when a refund is processed manually rather than by a computer. According to the Tax Office, there are only a limited number of circumstances where a refund will be processed manually. These include cases where there is an indicator on the taxpayer's account which prevents the refund from being automatically issued, cases where the refund is to be paid into a taxpayer's bank account and it exceeds a very large threshold, and cases where the refund may need to be issued urgently.34

4.83 These comments lead to the following recommendation.

Subsidiary recommendation 1

The Inspector-General recommends that the Tax Office establish systems which identify all cases where delayed refund interest should be paid.

Tax Office response

4.84 The Tax Office agrees with this recommendation. We are confident that we now have business processes in place to identify and pay delayed refund interest in all applicable cases. We acknowledge that prior to system enhancements in July 2003 there were limited instances in which delayed refund interest was not routinely paid.

5 Commissioner of Taxation, Annual Report 2003-04, at pp. 195-6.

6 Russell, Barrie, Deputy Commissioner, Goods and Services Tax, quoted in Hansard, Joint Committee of Public Accounts and Audit, Review of Auditor-General's reports 4th quarter 2002-3, 13 October 2003 at PA 12.

7 The GST started on 1 July 2000.

8 Attorney-General's Department, Commonwealth Fraud Control Guidelines, Guideline 2, available from www.ag.gov.au.

9 Australian National Audit Office, Goods and Services Tax Fraud Prevention and Control, Audit Report No. 55, 2002-03 Performance Audit, at paragraph 23 on p. 16.

10 This figure has been publicly stated by the Tax Office. See, for example, Granger, J., quoted in Hansard, Joint Committee of Public Accounts and Audit, Review of Auditor-General's reports 4th quarter 2002-03, 13 October 2003, at PA 15, and Monaghan, M., 'Making identity management work' Speech presented at the Information Security World Conference — 23-25 August 2004, downloaded from www.ato.gov.au on 7 September 2004.

11 ATO Minute No IGT-GST09-2004 at paragraph 2.1 on p. 3.

12 Sections 12 AA to 12 AD of Taxation (Interest on Overpayments and Early Payments) Act 1983. Note that the GST refund provisions contained in A New Tax System (Goods and Services Tax) Act 1999 originally required the Tax Office to pay GST refunds within 14 days unless one of a number of exclusions were satisfied. These provisions were later amended, with effect from the start date of the GST, so as to exclude any legislative time frame within which a GST refund must be paid.

13 Ebrill, L., Keen, M., Bodin, J. and Summers, V., The Modern VAT, International Monetary Fund Washington D.C. 2001 at p. 157.

14 ibid.

15 ATO Minute No IGT-GST11-2004 at paragraph 1.2 on p. 13.

16 Australian National Audit Office, Goods and Services Tax Fraud Prevention and Control, Audit Report No. 55, 2002-03 Performance Audit at paragraph 4.28 on p. 69.

17 In this report the term 'desktop' applied before a term refers to activities which are carried out inside the Tax Office and which do not involve a field visit to a taxpayer.

18 ATO minute No IGT-GST15-2004 at paragraph 9.1 on p. 6.

19 For the 2003-04 year, the Tax Office has advised, in ATO Minute No IGT–GST 09-2004 at paragraph 5.1 on p.7, that there were 7,904,276 original BAS statements with net amounts payable. By contrast, there were 1,890,128 original BAS refund returns.

20 Commissioner of Taxation, Annual Report 2003-04 on p. 65.

21 ibid at p. 40.

22 Ebrill, L., Keen, M., Bodin, J. and Summers, V., The Modern VAT, International Monetary Fund Washington D.C., 2001 on p. 162.

23 ATO Minute No IGT-GST06-2004 on p. 12.

24 ATO Minute No IGT-GST02-2004 at paragraph 8.4.3.

25 ATO Minute No IGT-GST03-2004 at paragraph 5.3.1.

26 ATO Minute No IGT-GST03-2004 at paragraph 5.6.3.

27 Australian National Audit Office, Goods and Services Tax Fraud Prevention and Control, Audit Report No. 55, 2002-03 Performance Audit at paragraph 19 on p. 15, and paragraph 3.34 on p. 61.

28 ibid at paragraph 3.21 on p. 57.

29 Taxpayers Australia Inc, Review of ATO's Administration of GST refunds arising from the lodgment of Business Activity Statements, Submission to Inspector-General of Taxation, dated 30 April 2004.

30 ibid.

31 ATO Minute No IGT-GST11-2004 at paragraph 8.1 on p. 11.

32 ATO Receivables Policy at paragraph 84.2.9.

33 Australian Taxation Office, General interest charge (GIC) rates, Fact Sheet, downloaded from www.ato.gov.au on 16 November 2004.

34 ATO Minute No IGT-GST18-2004 at paragraph 2.2 on p. 4.