Measuring and Tackling Indirect Tax Losses - 2004 An Update on the Government s Strategic Approach December 2004 1
Contents 1. Introduction 3 2. Measuring and Tackling VAT Losses 4 VAT Compliance Strategy 4 VAT Gap Methodology 4 VAT Gap Results 5 Missing Trader Intra-Community Fraud 7 Tackling Other VAT Losses 9 3. Measuring and Tackling Excise Losses 10 Alcohol 10 Tobacco 15 Hydrocarbon Oils 19 Annex A. Alternative estimates using the General Household Survey. 25 2
1 Introduction 1.1. This paper provides a further update of the Government s progress in tackling the problem of indirect tax losses, and updates estimates of the scale of revenue losses in each of Customs major tax regimes. Further details will be contained in the HM Customs & Excise Annual Report for 2003-04. 1.2. This paper should be read in conjunction with the technical papers published alongside the 2003 Pre-Budget Report entitled Measuring and Tackling Indirect Tax Losses 1, Measuring Indirect Tax Losses 2 published alongside the 2002 Pre-Budget Report and Measuring Indirect Tax Fraud 3 issued in 2001, which spell out in more detail the methodologies behind the estimates of indirect tax loss. 1.3. Measuring the scale of, and trends in, avoidance and evasion is not only inherently difficult, but also a relatively untested area of work for governments in the EU and around the world. Nevertheless, Customs have developed estimates for the UK that they believe are the best possible, based on all the information available. 1.4. As with previous technical papers, Customs would welcome views on the contents of this paper from businesses, academic experts and other interested parties, with a view to discussing and, where possible, improving these methodologies further. To comment, or to request further copies of these papers, please use the contact details below. Measuring and Tackling Indirect Tax Losses - 2004 Analysis Division Floor 2 HM Customs and Excise 1 Parliament Street London SW1A 2BQ 020 7147 0737 1 Measuring and Tackling Indirect Tax Losses HM Customs and Excise, December 2003. 2 Measuring Indirect Tax Losses HM Customs and Excise, November 2002. 3 Measuring Indirect Tax Fraud HM Customs and Excise, November 2001. 3
2 Measuring and Tackling VAT Losses Introduction 2.1. The VAT Compliance Strategy (VCS) was launched on 1 April 2003, with the aim of reversing the trend of an increasing difference between total theoretical VAT liability and actual VAT receipts the VAT gap. This document updates these figures for the year 2003-04. 2.2. The latest estimate of Missing Trader Intra-Community (MTIC) fraud - a significant component of the VAT gap - is also presented here. VAT Compliance Strategy 2.3. The VCS is designed to improve the service that Customs offers to the 1.8m VAT registered businesses. It aims to make it simpler and less costly for businesses to comply with the requirements of the VAT system while cracking down hard on those that continue to abuse the system. 2.4. The first year of the Strategy has been highly successful and is on track to meet its target of reducing the gap to 12 per cent of the theoretical liability by 2006. In 2003-04 net VAT receipts grew by 5.5 billion, and the VAT Gap fell from an estimated 15.8 per cent in 2002-03 to 12.9 per cent in 2003-04. VAT Gap Methodology 2.5. The methodology for constructing the VAT gap was first set out in Measuring Indirect Tax Losses in November 2002. It involves: assessing the total amount of expenditure in the economy that is theoretically liable for VAT; estimating the tax liability on that expenditure based on commodity breakdowns; deducting any legitimate reductions in the VAT liability occurring through schemes and reliefs to arrive at the net VAT theoretical tax liability (VTTL); deducting actual VAT receipts; and 4
assuming that the residual element - the gap - is the total VAT loss due to any cause, including general noncompliance, avoidance and fraud. 2.6. The tax loss figures for excise goods cited later in this paper also include the associated VAT lost on illicit sales. To avoid double counting an attempt has been made to remove VAT losses on alcohol and tobacco from the VAT gap. For the oils regime this is not currently possible, though further work is ongoing. VAT Gap Results 2.7. Table 2.1 presents the total theoretical liability (VTTL), VAT receipts and current estimates of the gap for all years from 1990-91 to 2003-04. Table 2.1: VTTL, actual receipts and VAT gap ( billion) Year (liability) VTTL Deductions Net VTTL Financial Year Net VAT Rcpts VAT Gap Percent Gap 1990 35.82 1.56 34.25 90-91 30.92 3.33 9.7% 1991 41.47 1.80 39.67 91-92 35.24 4.44 11.2% 1992 44.51 2.05 42.46 92-93 37.19 5.27 12.4% 1993 46.59 2.19 44.40 93-94 39.21 5.19 11.7% 1994 49.75 2.48 47.26 94-95 41.72 5.54 11.7% 1995 52.70 2.74 49.96 95-96 43.05 6.90 13.8% 1996 56.79 3.02 53.76 96-97 46.64 7.12 13.2% 1997 60.21 3.28 56.94 97-98 50.59 6.35 11.2% 1998 64.08 3.54 60.54 98-99 52.55 7.99 13.2% 1999 67.88 3.78 64.11 99-00 56.15 7.95 12.4% 2000 72.15 4.16 67.99 00-01 58.50 9.48 13.9% 2001 76.22 4.50 71.72 01-02 61.03 10.70 14.9% 2002 80.67 5.13 75.54 02-03 63.62 11.92 15.8% 2003 85.19 5.90 79.29 03-04 69.08 10.21 12.9% 2.8. Both in absolute and percentage terms, VAT losses declined sharply in 2003-04, the first full year of the VCS. Figure 2.1 shows that having climbed steadily for the previous three years, the gap in 2003-04 fell to 12.9 per cent from 15.8 per cent in 2002-03. 5
Figure 2.1: VAT gap as a percentage of net VTTL 2.9. Growth in VTTL is determined by the total value of transactions liable to VAT and any changes to the tax regime. As figure 2.2 shows, net VTTL has grown steadily in recent years, albeit at a gradually declining rate. However, the growth in net cash VAT receipts increased sharply in 2003-04, by 8.6 per cent. Figure 2.2: Percentage growth rates of net VTTL and net VAT receipts 2.10. Since publishing VAT gap outturns in December 2003, a number of small revisions have been made to estimates for earlier years. These arise from a combination of revisions to the underlying statistical 6
data and methodological changes. Figure 2.3 compares the latest VAT gap percentages with the previously published series. Figure 2.3: Revisions to VAT gap percentage. 18% 16% 14% 12% VAT Gap 10% 8% 6% PBR 04 PBR 03 4% 2% 0% 1990-91 1991-92 1992-93 1993-94 1994-95 1995-96 1996-97 1997-98 1998-99 1999-00 2000-01 2001-02 2002-03 2003-04 Financial Year MTIC Fraud 2.11. Missing Trader Intra-Community fraud (MTIC) represents a concerted and determined attack on the VAT system by organised criminals seeking to steal large sums. Estimates of the size of the fraud - presented as lower and upper bounds - have been published in previous Customs publications and are updated here. 2.12. For 2003-04, the methodology for calculating the upper and lower bound estimates has been revised. The improved methodology uses better information to identify legitimate transactions in the data, removing an element of over-estimation. Customs have removed some business transactions that are either considered highly unlikely to be fraudulent, or that are unlikely to result in a loss to the UK Exchequer. 2.13. However, the new estimate may still include some legitimate transactions. Additional work is underway to further reduce any overestimation bias in the measure and is expected to be completed in 2005. This further work will also explore the scope for making an allowance for revenue loss prevented through operational activity. 2.14. Table 2.2 shows the estimated level of MTIC fraud for each of the last five years. MTIC fraud fell for the second year running, standing at its lowest point since estimates were first published. This 7
builds upon previous years successes, stabilising the rapid growth of the fraud during 2001-02 and reducing fraud levels for the first time in 2002-03. 2.15. In 2003-04 the fraud is estimated to have fallen by over 25 per cent to between 1.06 billion and 1.73 billion. This is some 670 million to 800 million lower than the peak fraud levels seen in 2001-02, and has contributed between 480 and 610 million towards narrowing the overall VAT gap for this year. Table 2.2: Estimated MTIC Fraud ( billion) 1 Financial Year Upper Estimate Lower Estimate 1999-00 2.29 (2.40) 1.17 (1.21) 2000-01 2.47 (2.62) 1.31 (1.39) 2001-02 2.53 (2.75) 1.73 (1.77) 2002-03 2.34 (2.64) 1.54 (1.65) 2003-04 1.73 1]1 1.06 111 1 Figures published using previous methodology are shown in brackets. 2.16. The reduction in 2003-04 was achieved through a range of activities, including: undertaking nearly 7,000 pre-registration visits to new traders; identifying and refusing over 900 suspect registrations; identifying and cancelling nearly 500 existing missing-trader registrations; identifying MTIC debts and raising assessments totalling 83 million as a result of frauds detected; refusing over 200 million incorrectly claimed as VAT in fraudulent supply chains; securing injunctions to a value of over 20 million to meet VAT debts of missing traders; and, bringing prosecutions against 22 individuals resulting in jail sentences totalling 64 years, an average of almost 3 years for each offender. 8
Tackling Other VAT Losses 2.17. The VAT Compliance Strategy seeks to address the full spectrum of taxpayer behaviours that give rise to VAT losses. These include tax avoidance, fraud, other non-compliance and error. 2.18. Over 2003-04, through the VAT Compliance Strategy, Customs: successfully challenged avoidance schemes, identifying additional tax due of 473 million and brought forward legislation to protect future tax revenues; pursued taxpayers deliberately not complying with their tax obligations through improved risk assessment and targeted interventions, including a national campaign to tackle losses arising from the activities of illegal labour providers; reduced the total amount of debt outstanding for recovery by 113 million compared to the previous year and increased the total amount of payments collected by 197 million compared to the previous year; helped taxpayers to understand their VAT obligations through Business Advice Open Days, other awarenessraising events and direct contact with a wide range of businesses including 120,000 newly-registered businesses. 9
3 Measuring and Tackling Excise Losses Introduction Alcohol 3.1. Customs continue to tackle excise fraud on several fronts. This chapter reports on progress and presents the annually updated estimates of the scale of fraud against the main excise regimes. 3.2. Estimates of levels of fraud in excise goods are provided for the four most recent years for which they are available. The figures include and identify any revisions to previously published estimates where the underlying data have been revised or where the methodology has been refined and improved. 3.3. The figures presented are, in Customs view, the most robust and up-to-date estimates presently available of revenue losses from excise fraud and avoidance. However, the scale of illicit activity in excises is inherently uncertain, chiefly since the estimates rely on data from sample surveys. Therefore an indication of the uncertainty in the spirits fraud estimates has been included, and this will be extended to other excises, where possible, in the future. 3.4. This statistical uncertainty is presented here as confidence intervals. These intervals provide an indication of the range within which the true figure will probably lie. Methodological Changes for Estimates Spirits 3.5. The methodology for spirits estimates has been updated in two ways. The Department for the Environment, Food and Rural Affairs (DEFRA) have improved the adjustment used to ensure that National Food Survey data are comparable with Expenditure and Food Survey data 4. These data are used to estimate the consumption of spirits purchased through off-licence outlets. Previously there was a single adjustment factor calculated at a global level for all alcoholic products. Now it is available for 4 In 2001, the National Food Survey (NFS) was brought together with the Family Expenditure Survey (FES) as the Expenditure and Food Survey (EFS) under the auspices of the ONS although DEFRA remain responsible for publishing the NFS equivalent figures. 10
each type of alcoholic product separately 5. This allows revision of the fraud estimates for the latest and all previous years. At the end of April 2002 the duty classification for spiritsbased ready-to-drinks (RTDs) was changed. Until a new account code could be set up in the Customs systems in September 2002, RTDs were recorded under the imported and other spirits accounting code. Customs have now been able to estimate the amount of RTDs recorded under the spirits code and remove them from the spirits clearances for 2002-03. Presenting Spirits fraud estimates 3.6. All estimates of illicit economic activity are, by definition, subject to a degree of uncertainty. Some of this uncertainty is quantifiable namely that part caused by the random variation in sample data from the true values for the population. This uncertainty has always been present in Customs' estimates, but we are presenting an indication of the level of this for the first time in this paper. 3.7. To illustrate the uncertainty arising from sample data, Customs have produced confidence intervals for the spirits fraud estimates based on the estimates of spirits consumption that allow for random variation in the sample used. These confidence intervals capture the main source of random sampling uncertainty in the fraud estimates - the uncertainty for each annual consumption data point. The intervals presented are 95 per cent confidence intervals, used as a standard across statistical analyses. This means that there is a 95 per cent probability that the true value of the estimate falls within the given interval. However the distribution of values within these intervals means the true estimate is more likely to be near the centre of the interval. 3.8. The fraud estimates also make an adjustment for underreporting of spirits consumption, an inevitable feature of surveys of consumption. To correct for this it is assumed that there was no smuggling in 1992-93, and survey results for this year have been scaled to produce a level of consumption consistent with this assumption. Subsequent years estimates are produced by applying the trend in reported survey consumption between 1992-93 and the year in question to this 5 Further details can be found in http://statistics.defra.gov.uk/esg/statnot/familyfoodsn.pdf in the DEFRA S Family Food 2002-03 6 In 2001, the National Food Survey (NFS) was brought together with the Family Expenditure Survey (FES) as the Expenditure and Food Survey (EFS) under the auspices of the ONS although DEFRA remain responsible for publishing the NFS equivalent figures. 7 Further details can be found in http://statistics.defra.gov.uk/esg/statnot/familyfoodsn.pdf in the DEFRA S Family Food 2002-03 11
estimate. Therefore variation in the survey consumption estimate for 1992-93 also contributes to the confidence interval. 3.9. It should be noted that there are other uncertainties in the fraud estimate caused by biases in the source data and methodology. It is not possible to represent this uncertainty as a range. 3.10. The Office for National Statistics (ONS) are currently conducting work to explain and reconcile the differences between different estimates of alcohol consumption. Customs analysts are working with the ONS to ensure they have all necessary information at their disposal and the Government looks forward to the ONS s conclusions. Spirits Estimates 3.11. Estimates of the illicit market share for spirits rely on data from the Expenditure and Food Survey, which only become available around 18 months after the period surveyed. For this reason, estimates for 2003-04 are not yet available. 3.12. The results for 2002-03 are also provisional. The consumption data used to produce the fraud estimate are smoothed, which requires the estimation of some data for 2003-4. The use of these estimated data raise the uncertainty in the full year s estimate by around 1-2 percentage points. Table 3.1: Spirits: Revenue not collected, including Evasion ( million) 1 1999-00 2000-01 2001-02 2002-03 Cross-Border Shopping 2 100 150 150 150 Illicit (Fraud and Smuggling) s 350 450 450 250 Total non-uk duty paid 3 450 550 600 400 1 Includes duty and VAT although this will overstate losses to the extent that VAT is collected on sales of illicit alcohol through normal retail outlets. 2 Includes intra-eu duty free (for the first quarter of 1999-00) as well as EU duty paid. 3 Figures are independently rounded to the nearest 50m. As a result components may not appear to sum. 3.13. The confidence intervals associated with the estimate, described in paragraphs 3.6 to 3.9, are shown in Table 3.2. 12
Table 3.2: Spirits: Confidence Intervals based on variation in consumption estimates. ( million) 1 1999-00 s 2000-01 s 2001-02 s 2002-03 2 Illicit Spirits 0 to 750 50 to 850 50 to 850 0 3 to 650 1 Figures are independently rounded to the nearest 50m. 2 Does not include forecasting uncertainty. 3 The statistical lower bound is 150m. This illustrates the level of uncertainty, as a negative fraud estimate is not meaningful. 3.14. Table 3.3 shows the share of the total market that is estimated as being illicit. The estimates suggest that there was a fall in the level of spirits fraud between 2001-02 and 2002-03. 3.15. Customs use an array of operational evidence, data from trade sources and monthly tax receipts data to make judgements on latest trends in illicit markets. Customs are also looking increasingly to supplement that information with periodic testing exercises that over time will also help more precisely to gauge current trends if not absolute levels of fraud. Table 3.3: Spirits: Market Shares (Per Cent) 1999-00 2000-01 2001-02 2002-03 Illicit Market Share 11 13 13 7 Cross-Border Shopping 1 3 4 4 4 1 Includes intra-eu duty free (for first quarter of 1999-00) as well as EU duty paid. 3.16. Estimates of illicit market shares based on an alternative source of consumption estimates, the General Household Survey (GHS), are given in Annex A. Beer and Wine Estimates 3.17. Measuring Indirect Tax Fraud highlighted the difficulties associated with using a gap analysis to produce estimates of the scale of the illicit market in beer and wine. Over and above cross-channel passenger smuggling (which represents only a proportion of revenue losses and has previously been estimated using a bespoke survey) it has not been possible to estimate the overall size of this fraud using Customs established gap methodology. 3.18. Customs continue to explore alternative methods to estimate the overall scale of beer and wine fraud, although the available operational evidence indicates that neither the illicit beer nor wine market share is as large as that for spirits. As for spirits, Customs are looking at supplementing that information with periodic testing exercises that over 13
time will also help to gauge current trends, if not absolute levels of fraud. 3.19. Estimates of revenue lost on non-uk paid consumption through cross-border shopping of beer and wine are given in Table 3.4. Table 3.4: Cross-Border Shopping: Revenue not collected ( million) 1,2 1999-00 2000-01 2001-02 2002-03 Beer 100 100 50 50 Wine 200 200 200 200 1 EU duty paid. 2 Includes duty and VAT Operational Outcomes 3.20. Customs activity targeting alcohol fraud has yielded the following results: Table 3.5: Alcohol Seizures: 1999-2000 to 2002-03: litres 1999-00 2000-01 2001-02 2002-03 Spirits N/A 3.0m 2.1m 922,000 Beer N/A 5.6m 2.6m 2.9m Wine N/A 534,000 612,000 653,000 Table 3.6: Prosecutions Financial Year Spirits Mixed Excise 1 Total Successful Unsuccessful Successful Unsuccessful 2001-02 1 7 2 91 8 108 2002-03 21 3 34 28 86 1 Mixed cases include mixed alcohol or alcohol with other excise goods Table 3.7: Gangs Disrupted 2000-01 2001-02 2002-03 1 Gangs Disrupted 5 2 24 1 Gangs disrupted figure for 2002-03 reflects the outcome of one major criminal investigation "Fulcrum", which disrupted 15 gangs alone. 3.21. Customs efforts to tackle cross-channel passenger smuggling (one element of overall alcohol fraud) have proven effective. In 2001, revenue losses from both alcohol and tobacco cross-channel smuggling were down 74 per cent, with cross-channel passenger smuggling of beer almost eliminated and smuggling of wine and spirits halved. In 2002 there was a further reduction in alcohol smuggling by 14
Tobacco passengers of 45 per cent, meaning an overall reduction of 90 per cent in 2 years. Tackling Tobacco Smuggling 3.22. Customs Public Service Agreement (PSA) target is to reduce the illicit cigarette market to 17 per cent by 2005-06. Methodological Changes for Estimates 3.23. The methodology for smoothing cigarette and hand-rolling tobacco (HRT) clearances has been revised to take account of abnormal stock movements in the data, particularly for 2003-04. Tobacco companies build up stocks of cleared products in anticipation of duty changes at each Budget. With two Budgets falling in the financial year 2003-4, this created a significant discontinuity. These peaks have been smoothed over the two following months, and this has been applied to data for 2001 and onwards. 3.24. This change in methodology has no significant effect on previously published estimates. Cigarette Estimates 3.25. Estimates of the amount of non-uk duty paid cigarettes consumed in the UK, and associated revenue losses for the years 2000-01 to 2003-04 are set out in Tables 3.8 to 3.10. 3.26. The tables present revised estimates for all years to 2002-03, based on data from the General Household Survey (GHS) and the Omnibus survey. The estimate for 2003-04 remains provisional, as GHS data for that year are not yet available. This provisional figure is based on data from the Omnibus survey. 15
Table 3.8: Cigarettes: Revenue not collected, including Evasion ( billion) 1 2000-01 1 2001-02 2 2002-03 3 2003-04 5 Cross-Border Shopping 2,3 0.8 1.0 1.2 1.2 Illicit (Smuggling and Fraud) 2.7 2.5 2.0 1.9 Total non-uk duty paid 4 3.5 3.5 3.2 3.1 1 Figures include duty and VAT. 2 Include duty free as well as EU duty paid. 3 A definitional change was made to the distinction between cross border shopping and smuggling in October 2002, the change reduces the illicit market share by 1 percentage point for a full year. Further details can be found in Measuring and Tackling Indirect Tax Losses. 4 Figures are independently rounded to the nearest 0.1 billion. As a result components may not appear to sum. 5 Estimated using Omnibus data and may be subject to revision when GHS data become available. Table 3.9: Cigarettes: Market Shares (Per Cent) 2000-01 1 2001-02 2 2002-03 3 Illicit Market Share 21 20 15 15 Cross-Border Shopping 1,2 6 8 9 9 1 Include duty free as well as EU duty paid. 2 A definitional change was made to the distinction between cross border shopping and smuggling during 2002. 3 Estimated using Omnibus data and may be subject to revision when GHS data become available. Table 3.10: Cigarettes: Quantities (Billion Sticks) 1 2003-04 3 2000-01 2001-02 2002-03 2003-04 Total Consumption 76.5 75.0 73.0 *71.0 4 UK Tax Paid Consumption 56.0 54.5 54.5 54.0 Successfully Smuggled 16.0 14.5 11.5 *10.5 4 Cross-Border Shopping 2,3 4.5 6.0 7.0 6.5 Seized 2.8 2.6 1.9 1.8 1 Figures are rounded to the nearest ½ billion except seizures, which are rounded to the nearest 100 million. 2 Include duty free as well as EU duty paid. 3 A definitional change was made to the distinction between cross border shopping and smuggling during 2002. 4 Estimated using Omnibus data and may be subject to revision when GHS data become available. 16
Hand-Rolling Tobacco (HRT) Estimates 3.27. Estimates of the amount of non-uk duty paid HRT, and associated revenue losses, for the years 2001-02 to 2003-04 are set out in the Tables 3.11 and 3.12. As for cigarettes, final year estimates are based on the Omnibus survey, and are provisional. Therefore a revised estimate for 2002-03 is presented in this paper, with a provisional estimate for 2003-04. Table 3.11: HRT: Revenue not collected, including Evasion ( million) 2000-01 1 2001-02 2 2002-03 3 2003-04 3 Cross-Border Shopping 1 120 190 210 180 Illicit (Smuggling and Fraud) 680 580 630 750 Total non-uk duty paid 2 790 770 840 940 1 Include duty free as well as EU duty paid. 2 Figures are independently rounded to the nearest 10m. As a result components may not appear to sum. 3 Estimated using Omnibus data and may be subject to revision when GHS data become available. Table 3.12: HRT: Market Shares (Per Cent) 1 Include duty free as well as EU duty paid. 2 Estimated using Omnibus data and may be subject to revision when GHS data become available. Customs Strategic Approach to Tobacco Fraud 2000-01 1 2001-02 2 2002-03 3 2003-04 2 Illicit Market Share 63 52 53 57 Cross-Border Shopping 1 11 17 18 14 3.28. The Tackling Tobacco Smuggling Strategy 8 announced in March 2000, aims to reduce the illicit market share of smuggled cigarettes to no more than 17 per cent by 2005-06. Customs estimate that the market share of smuggled cigarettes was cut to 15 per cent in 2003-04, reducing the illicit market in cigarettes by nearly 30% on its peak. The Government is however concerned with the continuing significant smuggled share of the HRT market, and is examining the scope and case for further action to tackle smuggling of HRT. 3.29. In the first four years of the strategy, the total number of cigarettes smuggled into the UK each year has been reduced by over 5 billion sticks. This represents a reduction of over a third. Customs have also severely disrupted the gangs engaged in this type of fraud. 8 This document can be accessed on the Treasury s Internet Site at www.hm-treasury.gov.uk 17
3.30. Customs work with the UK tobacco manufacturers has successfully restricted the availability to smugglers of genuine UK tobacco products. Figure 3.1: Cigarette seizures by brand including counterfeit section Other (inc. Non UK Brand) 18% S Kings 9% Counterfeit 54% Genuine UK Brands 28% Sovereign 58% Dorchester 16% Other UK 16% 3.31. Figure 3.1 shows that smugglers are increasingly looking to counterfeit sources for their contraband. In 2003-04 54 per cent of seized cigarettes were counterfeit compared to 41 per cent in 2002-03 and 15 per cent in 2001-02. 3.32. Large-scale commercial smuggling continues to account for the vast majority of seizures. For the first time since the start of the Tobacco Strategy, more smuggled cigarettes have been intercepted before reaching the UK, with the assistance of a network of overseas officers, than at the UK seaports (see Figure 3.2). Figure 3.2: Cigarette seizures by mode Individual Seizures of over 500,000 sticks Seaports (passengers) 1% Seaports (freight) 35% Overseas 41% Inland 11% Airports 12% Overseas figure represents cigarettes seized en route to the UK in cooperation with overseas agencies. 18
3.33. During 2003-04, as part of the Tackling Tobacco Smuggling strategy Customs: seized 1.8 billion cigarettes - 1 billion in the UK at seaports, airports and inland and the remainder en route to the UK making a total of more than 9 billion cigarettes seized since the beginning of the strategy; broke up a further 69 gangs involved in the large-scale smuggling and supply of illicit cigarettes, bringing the total to 259 in four years; and successfully prosecuted 429 people, of which 229 received custodial sentences, averaging 14 months. Hydrocarbon Oils Tackling Oils Fraud 3.34. Customs PSA target is to reduce the illicit diesel market share in England, Scotland and Wales to no more than 2 per cent by 31 March 2006. Customs also have a supporting target to maintain the upward trend in the delivery of legitimate road fuel in Northern Ireland established in 2001. 3.35. The PSA target is underpinned by the UK Oils Fraud Strategy, launched at Budget 2002. Methodological Changes for Estimates 3.36. The methodology used to produce estimates of the non-uk duty paid share can be found in Measuring and Tackling Indirect Tax Losses. The estimates set out in this paper follow the same methodology, though Customs continue to consider ways to improve the estimates. 3.37. Since Customs last published estimates in PBR 2003, the Department of Trade and Industry (DTI) has revised its data for legitimate deliveries. As a result, Customs estimates for 2001 and 2002 have been revised, as shown in Tables 3.13 to 3.19. Great Britain: Diesel Estimates 3.38. The estimates of the diesel market in Great Britain are shown in Tables 3.13 and 3.14. As a result of Customs strategic approach the level of diesel fraud on the British mainland has reduced by one quarter between 2000 and 2003, and deliveries of legitimate road fuel into Northern Ireland were up by a further 6% in 2003 the third successive annual increase. 19
Table 3.13: Great Britain Diesel: Revenue not collected, including Evasion ( million) 1 *2000 0 *2001 1 *2002 2 *2003 3 Cross-Border Shopping 300 300 300 300 Illicit (Fraud and Smuggling) 1,000 950 750 850 Total non-uk duty paid 2 1,300 1,200 1,050 1,150 1 Estimates include duty and VAT. Estimating total revenue evaded for diesel is complicated by the fact that a significant amount is purchased by commercial users who could reclaim VAT incurred on legitimate purchases. However, it is assumed that lower fuel cost are reflected in the lower price further up the supply chain for the final products being transported and that VAT is therefore lost on these products. Hence, a simplifying assumption for the calculating total revenue evaded has been made that VAT is evaded for all non UK duty paid diesel. This assumption will mean that estimates of revenue loss are slightly overstated as some of the products being transported will be zero rated and therefore have no VAT implications. 2 Figures are independently rounded to the nearest 50m. As a result components may not appear to sum. 3 Estimates for this year are provisional. Table 3.14:Great Britain Diesel: Market Shares (Per Cent) 1 *2000 0 *2001 1 *2002 2 *2003 2 Illicit Market Share 8 7 6 6 Cross-Border Shopping 2 2 2 2 1 Figures are independently rounded to the nearest percentage point. 2 Estimates for this year are provisional. 3.39. As a result of the revisions to the DTI data discussed above, the current estimate for the illicit market share for 2002 is one percentage point higher than the previously published figure. 3.40. While there has been an increase in the estimate of losses due to illicit activity, a corresponding increase in consumption means that the illicit diesel market share in Great Britain remained constant at 6 per cent between 2002 and 2003. 3.41. The calculation of the illicit market relies upon data showing legitimate regional deliveries. There are some concerns about the quality of these regional data and consequently the latest illicit market assessment remains provisional while further work is undertaken to improve the quality of the data used. Great Britain: Petrol Estimates 3.42. Estimates of the revenue losses resulting from the use of non- UK duty paid petrol are shown in Table 3.15. 20
Table 3.15: Great Britain Petrol: Revenue and market share not collected, including Evasion 1 Figures are independently rounded to the nearest 50m. Estimates include Duty and VAT. 2 Figures are independently rounded to the nearest percentage point. 3 Estimates for this year are provisional. 3.43. Due to changes to the underlying data for legitimate deliveries, used to calculate the level of UK duty paid consumption, the 2002 estimate has been revised. As a result of these revisions, Customs consider that the share of the market for petrol non-uk duty paid is less than one per cent of the market. Northern Ireland Estimates *2000 0 *2001 1 *2002 2 Total non-uk duty paid 1,2-50m 50m 100m Total non-uk duty paid Market Share < 1% < 1% < 1% < 1% 3.44. Although the illicit market in Northern Ireland faces similar problems of rebated fuels fraud, including laundering, it also presents a number of distinct challenges. In particular the land boundary with the Republic of Ireland increases the risk of smuggling and provides a greater opportunity for legitimate cross-border shopping. Customs believe a large part of the Northern Ireland fuel market is therefore currently supplied by non-uk duty paid oils, through rebated fuels fraud and smuggling, as well as legitimate cross-border shopping. 3.45. Estimates of the revenue losses resulting from use of non-uk duty paid diesel and petrol and market share are shown in Tables 3.16 to 3.19. All estimates for Northern Ireland relate to total non-uk duty paid consumption, rather than just the illicit market. Data on consumers cross-border shopping behaviour, which would allow fraud and legitimate cross-border shopping to be separately identified, remain unavailable. Table 3.16: Northern Ireland Diesel: Revenue not collected, including Evasion ( million) 1,2 *2000 0 *2001 1 *2002 2 *2003 3 Total non-uk duty paid 235 230 220 230 1 Figures are independently rounded to the nearest 5m. 2 Estimates include duty and VAT, as explained in footnote 1 to table 3.13 3 Estimates for this year are provisional. *2003 3 21
Table 3.17: Northern Ireland Diesel: Market Shares (Per Cent) *2000 0 *2001 1 *2002 2 *2003 2 Total non-uk duty paid 1 67 66 61 58 1 Figures are independently rounded to the nearest percentage point. 2 Figures for deliveries are provisional. Table 3.18: Northern Ireland Petrol: Revenue not collected, including Evasion ( million) 1,2 *2000 0 *2001 1 *2002 2 *2003 3 Total non-uk duty paid 150 130 115 120 1 Figures are independently rounded to the nearest 5m. 2 Estimates include duty and VAT 3 Figures for deliveries are provisional. Table 3.19: Northern Ireland Petrol: Market Share (Per Cent) *2000 0 *2001 1 *2002 2 *2003 2 Total non- UK duty paid 1 38 35 31 32 1 Figures are independently rounded to the nearest percentage point. 2 Figures for deliveries are provisional. 3.46. Figure 3.3 shows that the non-uk duty paid share of Northern Ireland consumption continues to decrease for diesel, and has stabilised for petrol. This is in line with expectations, as increased Customs activity has contributed to a third successive annual increase in legitimate tax-paid road fuel deliveries into Northern Ireland up 6 per cent in 2003 - after five previous years of decline. This is shown in Figure 3.4 22
Figure 3.3: Northern Ireland Fuel Non-UK duty paid market share (Per Cent) 80% 70% Percentage of Market 60% 50% 40% 30% 20% Diesel Petrol 10% 0% 1997 1998 1999 2000 2001 2002 2003 Calendar Year Figure 3.4: Northern Ireland Fuel Deliveries market share (Per Cent) 100% 90% Percentage of Market 80% 70% 60% 50% 40% 30% 20% 10% 0% Petrol Diesel 1997 1998 1999 2000 2001 2002 2003 Calendar Year UK Oils Strategy 3.47. Budget 2002 launched the UK Oils Fraud Strategy with the aim of meeting the PSA target to reduce the illicit market share to no more than 2 per cent by 31 March 2006. This comprehensive strategy is focused on stopping large-scale criminal and commercial fraud by combining law enforcement action centred upon the large-scale supply and use of illicit fuel with a better control of the sale and distribution of red diesel, kerosene and "tied" oils (oils used in certain industrial processes, for example paint manufacture). 23
Operational Outcomes 3.48. As shown in Table 3.20 Customs have achieved the reduction in the illicit diesel market on the GB mainland between 2000 and 2003 by increasing pressure on the oil fraudsters, in particular through increasing: the number of vehicles challenged from nearly 30,000 in 2002-2003 to almost 85,000 in 2003-04 - up 190 per cent; the number of commercial detections from around 1,800 in 2002-2003 to over 2,000 in 2003-2004 - up 13 per cent; the number of convictions from 11 in 2002-2003 to 21 in 2003-2004 - up 91 per cent. Table 3.20: Oils Strategy 2000-01 2001-02 2002-03 2003-04 Vehicles Vehicles challenged 20,861 16,313 29,302 84,700 Vehicles detected 2,295 3,342 3,956 3,753 Laundering plants Total disrupted 20 38 46 31 Great Britain 3 30 25 18 Northern Ireland 17 8 21 13 Gangs disrupted Total disrupted 7 19 14 13 Great Britain 5 12 13 11 Northern Ireland 2 7 1 2 Assessments 6.1m 8.3m 13.2m 11.5m Prosecution figures (Great Britain) Number convictions 4 9 8 17 Average sentence 21 months 14 months 14 months 14 months Confiscation orders 143,000 660,000 374,000 167,387 Prosecution figures (Northern Ireland) Number convictions 5 15 3 4 Average sentence 3 months 16 months 8 months 7 months Confiscation orders Nil Nil 250,000 Nil 3.49. In 2002-03 Customs also worked hard with the industry to implement the Registered Dealers in Controlled Oils scheme. Over 5000 businesses are now authorised dealers with Customs making over 5500 visits to such dealers, and refusing applications to more than 50 businesses that were deemed to pose a significant fraud risk. 24
Annex A. Alternative estimates using the General Household Survey. A1. As highlighted in the NAO s report Estimating the level of Spirits Fraud 9, using different data sources for estimating spirits consumption produces different estimates of the level of spirits fraud. In particular the Scottish Whisky Association recommended using the General Household Survey data. This Annex uses a similar gap methodology to Customs standard practice, but instead uses the General Household Survey (GHS) to estimate total spirits consumption. The estimates produced are shown in Table A.1. Table A.1: Spirits: Revenue not collected, including Evasion: GHS Estimate ( million) 1 1999-00 4 2000-01 s 2001-02 s 2002-03 s Cross-Border Shopping 2 100 150 150 150 Illicit (Fraud and Smuggling) 150 200 100 *- 5 Total non-uk duty paid 3 300 300 250 *150 6 1 Includes duty and VAT although this will overstate losses to the extent that VAT is collected on sales of illicit alcohol through normal retail outlets. 2 Includes intra-eu duty free (for first quarter of 1999-00) as well as EU duty paid. 3 Figures are independently rounded to the nearest 50m. As a result components may not appear to sum. 4 No GHS alcohol question in this year. Estimate is based upon mean of 1998-00 and 2000-01 GHS consumption estimates. 5 It has not been possible to estimate the size of this fraud using a gap analysis because a comparison of UK duty paid consumption with estimates of total consumption does not generate meaningful results. 6 Bottom-up estimate of revenue loss based upon estimated level of cross border shopping only. Usual top down (gap analysis) approach does not produce a meaningful fraud estimate, hence loss from fraud is excluded from the estimate A2. The uncertainty intervals associated with the estimates are as shown in Table A.2. 9 Estimating the level of spirits fraud, memorandum by Comptroller and Auditor General, March 2004 25
Table A.2: Spirits: Confidence Intervals ( million): GHS Estimate 1 1999-00 2000-01 2001-02 2002-03 Illicit Spirits N/A 2-3 to 450-4 to 350-5 to 200 1 Figures are independently rounded to the nearest 50m. As a result components may not appear to sum. 2 Alcohol questions not included in GHS during this year. 3 The statistical lower bound is 50m. This illustrates the level of uncertainty, as a negative fraud estimate is not meaningful. 4 The statistical lower bound is 150m. This illustrates the level of uncertainty, as a negative fraud estimate is not meaningful. 5 The statistical lower bound is 350m. This illustrates the level of uncertainty, as a negative fraud estimate is not meaningful. A3. Table A.3 shows the associated market share of illicit purchases. Table A3: Spirits: Market Shares: GHS Estimate (Per Cent) 1 1999-00 2 2000-01 s 2001-02 s 2002-03 s Illicit Market Share 5 6 3 *- 3 Cross-Border Shopping 1 4 4 5 *- 3 1 Includes intra-eu duty free as well as EU duty paid. 2 No GHS alcohol question in this year. Estimate is based upon mean of 1998-00 and 2000-01 GHS consumption estimates. 3 It has not been possible to estimate the size of this fraud using a gap analysis because a comparison of UK duty paid consumption with estimates of total consumption does not generate meaningful results. 26