DETERMINING THE 41st LEVY SCHEME

OCP Final Report

17 December 2001

CONTENTS

  1. INTRODUCTION
  2. ISSUES
  3. THE LEVY QUANTUM
  4. OPTIONS FOR THE LEVY MECHANICS
  5. RISKS
  6. NEXT STEPS

Appendix – Alternative margin calculation

 

  1. INTRODUCTION

    2. Background

    1. The Horserace Betting Levy Board was unable to accept the Bookmakers’ Committee proposal for the Forty-First Levy Scheme covering the year commencing 1 April 2002. The determination of the levy has therefore fallen, under statute, to the Secretary of State for the Department for Culture, Media and Sport.

      2. Though no longer reflected in statute the criteria to be used by the Secretary of State in making the determination are well established and accepted and these are to balance the needs of racing with the capacity of the bookmakers to pay.

      Process and Approach

    2. OCP was successful in bidding to undertake a report to assist DCMS in the determination of the scheme and our engagement commenced on 26th November with a remit to report in three weeks. The consultant team consisted of Rod Morgan, Geoff Horton and Mike Thomas. The short time scale prohibited detailed independent economic studies of the industry. We were given access to a number of studies, most of which had been commissioned by the main parties involved.

    3. Our role was not to identify a single quantum or mechanism, but to explore arguments for and against the various options and to present our findings.

    4. Our starting point was to review the written submissions made by the Horserace Betting levy Board (HBLB), The British Horseracing Board (BHB) and the Bookmakers Committee (BC) and to hold meetings with each of these parties to explore those submissions and to access additional data underpinning them. Further meetings were held to address issues of detail emerging from our analysis, and to explore further arguments and options.

    5. This report has been written for the Secretary of State to assist her in determining the 41st levy scheme. We report, but do not express an opinion, on several technical matters such as State Aid and VAT where Counsel’s opinions have been sought by the some of the above parties and DCMS.

      6. Structure of Report

    6. In Section 2 we set out the issues that have arisen during the course of the negotiations over the 41st levy.

      7. In Section 3 we address the levy quantum by considering the needs of racing the capacity of the bookmakers to pay and identify the options.

      In Section 4 we describe the alternative types of levy schemes.

      In Section 5 we summarise the risks.

      In Section 6 we present an outline of the work that remains to complete the determination.

      Acknowledgements

    7. We wish to record our thanks for the considerable assistance we have been given by DCMS, HBLB, BHB and BC, and their advisors in our work.

 

2. ISSUES

Complexity of this year’s negotiations

    1. This year’s negotiations on the levy were more complex than usual because they took place against a background of:

Impact of Gross Profits Tax

    1. Because GPT has only recently been introduced its impact on turnover and the relationship between turnover and gross profit over the next year cannot be predicted with any degree of certainty. This makes estimating the amount that is affordable and setting the parameters of the levy to achieve that target more difficult than in previous years.

      2. Basis of the Levy

    2. For the last 28 years the levy has been set as a percentage of bookmakers’ turnover, and there has generally been assumed to be a reasonably stable relationship between turnover and gross profits over that period. The uncertainty over the impact on turnover and gross profits of GPT has led the BC to recommend a levy system for the 41st scheme based on gross profits, which would share the risk of a levy based on turnover rising (possibly much) faster than profits between the bookmaking and racing industries.

      3.  Impact of proposed commercial arrangements

    3. The BHB has proposed new licences for data and picture rights for bookmakers to succeed the existing SIS contracts that expire on 30 April 2002. The proposed charges for these licences are substantially higher than previously. As of now a small number of licences have been taken up by the bookmakers and it is uncertain how many will be by 30 April, by which time the levy determination will need to have been made.

      4. This creates additional uncertainty in quantifying the needs of racing and the ability of bookmakers to pay, and because of this the BHB has proposed to offset any levy payments made by bookmakers for periods that overlap with the charges arising under any licence contracts which are agreed. There is also uncertainty as to what the VAT position regarding such an arrangement would be. An alternative suggestion from the BC is that the levy should stop being paid when the licences come into effect, even if contracts have not been agreed.

      Legal Challenges

    4. Bookmakers have a number of legal challenges to the BHB’s proposed commercial arrangements underway, the final outcomes of which are unlikely to be clear before the determination has to be made. These challenges (to the OFT and European Court) could affect both the nature of the proposed arrangements and the level of charges.

      5. Summary of position

    5. The Secretary of State is required to determine the 41st levy scheme following the failure of parties to reach a negotiated agreement. Applying the needs and capacity criteria is made harder this year by the uncertainties identified above which necessitate a judgement over not only the target levy quantum but also over the mechanics by which the levy is to be administered.

3. LEVY QUANTUM

Introduction

    1. This section considers the needs of racing and bookmakers’ ability to pay with a view to determining the sum that it would be appropriate for the 41st levy to aim to yield.

      2. The Needs of Racing

    2. The concept of the needs of racing is not a firm one. At one extreme, whatever racing’s income, it is likely that there will still be expenditure that could be made and that would be of some value, and so needs could be regarded as almost infinite. At the other extreme, racing functions on a significant scale and would probably continue to do so without levy support, and so it could be argued that no funding is needed because none is necessary for racing to exist.

    3. It is a general economic principle that additional expenditure or activity is worthwhile if its benefits exceed its costs. In most cases benefits can be judged from what prospective purchasers are prepared to pay and costs from what sellers require and so the normal operation of the market can be relied on to achieve a sensible outcome. However, there may be circumstances in which benefits or costs accrue to persons other than the purchaser or seller. In this case, betting on horses relies on the existence of horseracing and many activities of the horseracing industry will yield benefits for betting and punters.

    4. Although the levy is not a payment for a product, in assessing the needs of racing, it is useful to consider whether activities have overall benefits that exceed their costs and the revenue they can obtain within racing and, in particular, whether they yield benefits for betting.

      5.   

      BHB – 1998 Plan and 2001 Memorandum 

    5. The BHB’s 1998 Financial Plan (BHB 1998) listed funding needs totalling £105 million additional to the levy yield of £50 million expected in that year. These are listed in the table below.

      6. BHB 1998 Financial Plan: Expenditure Category

      Additional sum required (£m)

      Race prize money

      66

      Marketing

      14.25

      Reducing race entry fees

      3.3

      Breeders’ prizes

      10

      "Integrity services"

      1.8

      Other – Veterinary research, training, R&D

      5

      Reversal of levy reduction

      5

      Total

      105.35

       

       

    6. The BHB’s November 2001 memorandum to DCMS referred to the 1998 Financial Plan when discussing racing’s needs and used similar headings, but with the addition of other items such as racecourse investment, and industry employment and wage levels.

      7.  Owners Losses and Prize Money 

    7. Owners receive in prize money only about a quarter of their expenses of owning, training, and racing horses and spend about £170 million more than they receive (£230m costs net of prize money £55m and other income). This is the main component of the claim (e.g. BHB 2000 3.3.1) that racing loses around £200 million a year.

    8. The BHB wished to increase prize money by £66 million to a total of £125 million which would pay £25 million to trainers, jockeys and staff and leave £100 million to fund roughly half of owners’ costs. Prize money has actually increased to £72 million in 2000 of which the levy funds £33 million.

    9. However, survey evidence suggests that if owners’ prize money were increased they would spend virtually all the extra money in buying more or better horses. Owners’ losses are not indications of disequilibrium or of imminent contraction, as they would be in most other industries, but the result of other benefits that owners derive from owning horses. Indeed the number of horses in training rose 9% between 1996 and 2000 and racehorse auction prices by 23% in the same period.

    10. If owners received another £50 million and spent it as they say they would the proportion of costs covered by winnings would rise from 24% (£55m of £230m) to 37% (£105m of £280m) rather than the 50% mentioned by BHB. There would be a further significant increase in racing activity.

    11. There seems no obvious reason to prefer one figure rather than another. Both parties have drawn attention to international comparisons to support their cases. However, important differences in the circumstances of racing and betting between the various countries cited mean that little can be concluded from them.

    12. The key question would seem to be whether the increase in racing activity that is implied by a 20+% increase in owners’ spending is required. The evidence presented (BHB 1998) that this is desirable in that some races in the summer months have too few runners and that the quality of fields may be too low as evidenced by an increase in British-trained horses competing abroad and a decrease in foreign-trained horses competing in Britain. However, the Bookmakers Committee points to 11 of the top 15 prize-winning owners and this season’s champion trainer being not resident in the UK. It claims that summer race entries are deterred by the hard ground and cites the need to ballot would-be entrants to limit race fields at other times as evidence of, if anything, an oversupply.

    13. The need for additional prize money is not established but neither is the converse, that additional spending would be inefficient.

      14. Marketing

    14. BHB wishes to spend an additional £14.25 million on marketing, which it estimates would yield a further £57.5 million revenue. Although there will be costs associated with the additional revenue it would at first sight seem that the marketing programme would be self-funding and would not need support. Indeed, as a general rule, one would expect a marketing programme to be self-funding and to need support only insofar as it produced benefits for parties other than those paying for it – in this case bookmakers and punters.

    15. It therefore does not seem reasonable for the levy to support the entire programme.

      16. Race Entry Fees and Appearance Money

    16. Reductions in race entry fees and expansion of appearance money are sought to improve the quality of races. The arguments for funding these are very similar to those for funding additional prize money. It is for those who are expert in the industry to assess which combination of larger prizes, lower entry fees, and appearance money would be the most effective.

    17. However, as for the increase in prize money, justification of this finance relies on a need to encourage more entry in races, which we consider not to have been established. 

      18. Breeders’ Prizes

    18. The argument for subsidy of horse breeding is that other countries do so and that the UK needs to move towards matching the subsidy to maintain its competitive position.

    19. We have not analysed statistics on breeding to see if it is in decline. The rise in horse auction prices would suggest otherwise, as does anecdotal evidence we have encountered. The BHB 1998 plan mentions that "most of the world’s major stallions, apart from those that Middle Eastern owners stand in Britain, are based in Japan, America or Ireland" but it is not surprising for the two largest economies in the world to have large shares in a market or unusual for much of a UK industry to be foreign-owned.

    20. It is not in a country’s interest to subsidise an industry purely in order to compete with an industry that is subsidised in another country. It is only sensible to do so where there is evidence of predatory intent by the other country (i.e. when it does so in order to raise prices after it has eliminated competition) or where there are demonstrable external benefits from the existence or scale of the industry.

    21. We are not convinced that a case has been made for a large increase in subsidy to breeding.

      22. Integrity Services

    22. Services to police the fairness of racing provide obvious benefits to those betting on the outcome. It is probable that an efficient level of provision will require substantial funding from outside racing.

      23. 1998 Other

    23. The 1998 plan sought additional funding for research in veterinary science and racecourse management and in training and education. Insofar as these increase the quality of racing they may also provide external benefits that would not be funded by racing and racegoers.

      24. 2001 Other

    24. The November 2001 BHB memorandum also mentioned racecourse investment and the level of employment of racing staff together with the wages they receive.

    25. Racecourse investment may affect the quality and quantity of external benefits provided by racing but not capable of being funded by it. It is therefore possible that external funding is appropriate.

    26. The memorandum expresses concern that there is a shortage of stable staff at the wages offered and that each member of staff now looks after more horses. Unemployment has fallen since the early 1990s and the labour market has tightened, perhaps particularly so in traditional racing areas. However, the increase in the ratio of horses to staff of 2.2% a year could be a sign of normal technical progress and of increasing capital equipment per worker.

    27. Overall running costs per horse have risen faster than general inflation and so there is no presumption that racing has been falling behind the economy as a whole, but rather the reverse.

      28. HBLB expenditure

    28. The HBLB itself makes a call on the funds raised by the levy and is planning to do so exceptionally in 2002-03 in order to finance its own winding up costs. It presently plans to stop payments to racing at the end of the calendar year 2002 and to retain revenue from the final quarter of the financial year on the assumption that racing can obtain other means of finance.

    29. The approved package of winding up costs is £1.4 million and we understand that £1½ million has been accrued against it in the present levy scheme. However, there is a further uncertain liability under the board’s pension scheme that is presently estimated in the range £10-14 million, although previous estimates have been considerably smaller. 

    30. The HBLB considers that its unfettered capital assets, a £50 million capital fund and the Horseracing Forensic Laboratory, should be transferred to the racing industry on its termination and not used for this purpose. The hypothecation of one quarter’s revenue (£22.7 million on the HBLB expectation of 41st levy yield) plus a contribution from the Tote would be likely to generate a surplus that would presumably also be transferred.

    31. The pension-funding requirement (as opposed to the board’s proposed revenue hypothecation) represents a need of racing. Although there may be a need for a 42nd levy (if one accepts the BHB and HBLB argument that levy income is needed while commercial income is subject to legal challenge and expects the ECJ not to conclude before April 2003), we do not consider that it would be prudent to assume so and assume the entire sum to be a need of racing in 2002-03.

      32. Conclusion on the Needs of Racing 

    32. The arguments that a large increase is needed in payments to racing are difficult to assess but are in our view far from self-evident. There was a £9m increase in budgeted revenue in the 39th scheme. Furthermore, racing’s other income has increased by more than the 1998 or 2000 plans anticipated. Sales to Attheraces and SIS alone are likely to yield in excess of £40 million compared to the existing revenue of £13 million.

    33. We do not believe that it has been demonstrated that the needs of racing cannot be accommodated within this rise in other income.

      34. Bookmakers’ Ability to Pay 

    34. Bookmakers’ ability to pay has dramatically improved since October 2001 because taxation of betting has changed. A betting tax of 6.75% on betting turnover has been replaced by a tax of 15% on gross profits – defined as bets placed less the amount returned in winnings.

    35. In discussing the ability to pay we consider first the scope of income to be considered and then the forecast of income and of any costs associated with it.

      36. Definition of Profits 

      Non-Horseracing Income

    36. The parties differ in their treatment of betting income from sources other than horseracing. Europe Economics ignore it in their work for BC but the Henley Centre include it in the work quoted by BHB in their initial memorandum (BHB November 2001).

    37. Betting tax has been removed generally, including on the (roughly) one third of betting that does not relate to horseracing. Insofar as this benefit is not passed to punters it improves bookmakers’ ability to pay a levy.

    38. In a competitive industry a reduction in tax would normally lead to a reduction in prices rather than an increase in profitability. However, not all the benefit would be passed on if there were a compensating increase in costs that affects all members of the industry and potential entrants to it or if there is market power.

    39. There is no obvious mechanism to increase competitors’ costs artificially in order to make the benefit of the non-horseracing tax reduction remain with bookmakers. We do not envisage that the horserace betting levy can be imposed to increase the costs of all non-horserace betting or that, if it could, that it would be reasonable to impose its costs on other punters.

    40. However, there may be market power in the provision of non-horserace betting services and the replacement of betting tax by gross profit tax may increase bookmaker’s retained profits, particularly in the short term. This might therefore affect the ability to pay but the extent to which it might do so is not clear. Moreover, it seems illogical to set a levy on one type of income – horseracing – but to justify it on the basis of a wider definition of income.

    41. We do not propose to take non-horseracing income into account and do not consider that the Henley Centre’s calculations for the BHB are appropriate in this respect.

      42. Forecast of Profits 

    42. In forecasting bookmakers’ net profits, and so their ability to pay, there are four main aspects to consider:

Margin

    1. Both Europe Economics’ study for the BC and Henley’s for the BHB assumed that bookmakers would remove the 9% charge that they then imposed on betting shop bets and otherwise leave their margin unchanged. Thus a margin of 22.5% on cash horseracing bets is regarded as a basic margin of 15.5% plus the 9% on the turnover, which is itself inflated by 9%. When the 9% deduction disappears the margin will return to 15.5%.

    2. However, this is not necessarily what economic theory would suggest would be rational behaviour by bookmakers and is not what was assumed by Paton, Siegel, and Vaughan Williams of Nottingham University in their September 2000 report – An Economic Analysis of the Options for Taxing Betting - to HM Customs and Excise.

    3. They assumed that bookmakers would take into account the fact that their 15.5% margin would now be subject to 15% tax (against which they could not offset operating costs) and set a margin that takes into account the needs to fund both the continuing levy part of the 9% deduction and the gross profit tax. This might result in a margin of about 20%.

    4. In practice, bookmakers’ marginal costs are very low and so the economic theory assumed in the Nottingham study, which is based on marginal cost, may not apply. Moreover, it is argued that bookmakers may have difficulty in changing their underlying margin since the majority of bets are settled at odds determined in the on-course betting market. On the other hand, off-course bookmakers could offer different odds, affect the overall margins through special offers etc., or even affect on-course odds by participating in that market.

      5. Response by Punters

    5. Both studies used an assumption of behaviour by punters that meant that gross profits would be the same whatever margin was set although the amount of bets placed (or turnover) would vary. They assumed "constant drop", i.e. that punters would continue to lose the same amount in total. In economic terms this is an elasticity of (minus) one; if price rises by a given percentage the quantity purchased falls by the same amount, with the result that total spending remains the same.

    6. Both Europe Economics and the Nottingham University team considered other elasticities, but estimated and sought to apply them (in our view wrongly) not to the total margin but to the rate of deduction only. This led them to consider a reduction in price of 100% - complete removal of the 9% deduction, rather than one of 31% - fall of margin from 22.5% to 15.5%. EE used an elasticity of –0.63 and NU one of –1.09, both with respect to deductions only, which led them to contemplate turnover increases of 63% and 109% respectively.

    7. There is as yet little evidence of movements in turnover or margins since October. It seems clear that turnover has substantially increased, perhaps by as much as 30%, but experience differs between firms.

    8. There is clearly uncertainty over punters’ response and in conversation both BC and BHB have indicated that they would expect that the full response would occur only after some time has elapsed. We have therefore considered elasticities of –0.5, -1, and –1.5 on the total margin.

    9. The elasticity of –1, where there is a "constant drop" and punters’ total losses remain unchanged on an increased betting turnover, is our baseline assumption. Elasticities of –0.5 (where punters increase betting by only half as much as the price has fallen and so in this case reduce their total spending on losses by 16% and increase turnover by 19%) and –1.5 (where punters increase betting by half as much again as the price has fallen and so in this case increase their total spending on losses by 19% and increase turnover by 69%) are used in scenarios to examine the sensitivity of the findings.

      10. Telephone and internet

    10. The response on turnover other than in shops is likely to be proportionately much larger (although small in total quantity) because the tax that has been removed represents a larger proportion of the total margin, the potential for switching to or from offshore betting is much greater and members of the BC have undertaken to "repatriate" business as a result of the tax change. To the extent that business is repatriated ability to pay may fall as it incurs additional tax and levy costs. However, capture of business from UK to other providers or of the global betting market would increase ability to pay. We have assumed the same elasticity which, given the larger price change, produces about a doubling of turnover in the base case.

      11. Natural Growth

    11. The data used in the studies, and the last data unaffected by the foot and mouth crisis, relate to 2000. One would expect some growth to have occurred between that time and 2002/03 even in the absence of tax changes. The Henley report hypothesises an increase of about 5% in the two years in total and we see no reason to dissent from that.

      12. Calls on Profits 

      Operating Costs

    12. Increasing turnover will raise operating costs. Europe Economics assume the 45% increase in shop turnover will raise operating costs by 9%. Henley assumes a 9% cost increase between 2000 and 2002 but only a 2.5% increase in response to a 42% increase in shop turnover. The Bookmakers’ Committee has supplied us with figures suggesting that, if the increase in turnover was entirely in the volume of betting slips a 45% increase in turnover would raise operating costs 6% in cashier staff costs alone.

    13. We have assumed the 9% increase between 2000 and 2002 and that, in the short run within 2002, costs increase 1½% for every 10% increase in turnover. As a base level we have assumed £550 million costs, a figure slightly above that for shops only, in order to make some allowance for credit betting costs.

      14. "Repatriation Costs"

    14. Bookmakers’ tax and levy payments will increase as betting that is presently offshore is repatriated and becomes subject to levy and tax. We have assumed a cost of £13 million as calculated by Europe Economics.

      15. Pictures and data

    15. New licences for the sale of data and pictures by racing to bookmakers are due to come into force in May 2002 and so may determine payments for data and pictures for all but one month of the 41st levy period. These new commercial arrangements will clearly affect the ability to pay but it is intended that the arrangements should eventually replace the levy and that, while the two operate in tandem, the total amount paid by bookmakers to racing should be independent of the levy. Mechanisms for achieving this are discussed in the next section and the amount to be paid under the new commercial arrangements is not taken into account in assessing the ability to pay.

    16. However two possibilities are considered further in this section – that the levy proceeds may exceed payments for data and pictures and that the size of levy rate determined may affect the commercial agreement.

      17. Baseline projection

    17. The combination of these assumptions produces the illustrative figures in the table below. These show numbers for 2000, for a hypothetical 2002-03 where betting tax is still levied, and our baseline assumption for 2002-03. The baseline case assumes a 15.5% margin and a unit elasticity (or constant drop).

    18. Our baseline assumption shows bookmakers’ net profits from horseracing at £312 million, £94 million higher than the £218 million in 2000.

      19.   

      Equilibrium Position

    19. This modelling results in an increase in bookmakers’ profitability but, in a competitive industry, the operation of competition would result in profits returning to a normal level through competition between existing firms or as new firms entered the market. Bookmaking is not a fully competitive market, although international competition may be making it increasingly so, and therefore some of this profit increase would be likely to be retained by bookmakers with the rest being passed to punters in lower prices.

    20. To the extent that levy rises pass the increase in profit to racing the gains to both punters and bookmakers would be reduced.

      21. Alternative scenarios

    21. We also show six scenarios that combine high, base or low elasticities of punter response (-0.5, -1 or –1.5) with high or slightly lower margins. A seventh scenario combines a 17% margin with a "constant drop" (or –1) elasticity.

    22. The various scenarios show possibilities, albeit perhaps small ones, of net profits being significantly different, with net profits ranging from £157 million to £493 million (albeit in admittedly extreme cases).

      23. Principles of Division

      Beneficiaries 

    23. We assume that, when balancing the needs of racing with the bookmakers’ ability to pay, we are considering a division of income between two parties – the betting and racing industries. However, the former contains two interests – punters and bookmakers – and the latter perhaps three – the industry, owners, and racegoers.

    24. Punters have already received a benefit from the removal of the 9% deduction, part of which stems directly from the removal of Gross Betting Duty. However, this was in response to a leakage of business overseas and it might be argued that the punter remains a valid beneficiary for the further development of the industry.

    25. It may be helpful, although it is certainly oversimplistic, to regard the industries primarily as intermediaries and to consider a transfer between punters and owners. One might then consider a division of the reduction in taxation between punters and owners and whether punters might wish to pay owners to provide more or better racing.

      26. Principles

    26. The principles to be applied are not clear. The position is further complicated by the BHB intention to introduce a commercial arrangement for data and pictures in May 2002 and the intention on both sides (by either terminating or netting off the levy) that the arrangement will determine payments from betting to racing and that they will be invariant to whatever levy rate is set.

      27. Equal Shares between Betting and Racing

    27. Perhaps the most intuitively obvious approach is to estimate the extra that is now available and to divide it equally. Our baseline scenario shows an increase in net profit of £94 million relative to the 2000 baseline. Adding half of this to the intended 2000/01 yield of £59 million would produce a figure of £106 million.

      28. A Three-way split

    28. If one assumes a three-way split, between punters, bookmakers and racing, £63 million would be left with bookmakers (on the assumption that competition would force at least half to be returned to punters), and a third of the total given to racing so that levy proceeds would rise to £90 million.

      29. Status quo

      Needs of Racing

    29. Another approach might say that the needs of racing used to exceed bookmakers’ ability to pay and so the latter effectively determined the levy but that now the reverse is true and an assessment of the needs of racing should limit the total. We considered that it has not been demonstrated that the needs of racing cannot be accommodated within the rise in other horseracing income and so proceeds should remain at around £60 million.

      30. Relationship with the commercial arrangements

    30. The same conclusion could result from taking into account the impact of the levy on commercial negotiations. By providing a payment from bookmakers to racing in the event of failure to conclude a commercial arrangement it changes the bargaining position of both parties (raising the price at which racing is likely to agree and lowering the additional cost to bookmakers of any negotiated price) and would be likely to affect the outcome, unless the outcome is determined by competition authorities. Indeed, since the levy is likely to operate alone for only one month, it might be argued that it has almost no effect other than on the parties’ negotiating position.

    31. It might then be argued that the levy should be held constant to reduce influence on the commercial negotiations, which might be expected to themselves provide the balance between the needs of racing and the ability of bookmakers to pay. It would also reduce the risk that the levy might exceed the eventually agreed or determined commercial arrangement. 

      32. HBLB Exceptional Payments

    32. Alternatively one might argue that the status quo should apply to funds paid to racing from the levy. In this case the additional HBLB expenditure would be added to the total making a levy target of £72m. 

      33. Minimum Funding

    33. An extension of the argument relating to the impact on the commercial negotiations would be to reduce the levy to a minimum funding sufficient to avoid disruption to the industry with the intention that the commercial arrangements should be influenced as little as possible. If a commercial agreement is best able to strike a balance between the needs of racing and the ability of bookmakers to pay, it follows that any distortion to the agreement beyond what is strictly necessary would achieve a worse balance.

    34. In the time available we have been unable to determine what level of funding would prevent severe disruption. The HBLB’s winding up costs would need to be addressed plus a further amount to counter the difficulties the BHB would experience from advancing funds from proceeds still open to challenge in the European Court and perhaps also from the OFT or a subsequent appeal to the Competition Commission.

      35. Levy rates

    35. The table below shows the average levy rates required on turnover or gross profits to achieve the target yields. It is beyond the scope of this project to calculate the various rates actually to be applied, which requires prior determination both of the levy quantum and the method of imposition.

      36.   

    36. We have also calculated the yields that would result were the other scenarios to turn out to be the case. These are shown in the table below together with some statistics showing the average error, the average absolute error, and the standard deviation (a statistical measure of the size of the absolute error).

      37.  

    37. The average error with a turnover tax is about £20 million while that with a gross profits tax is around £10 million.

 

4. OPTIONS FOR THE LEVY MECHANICS

    1. From the submissions and our discussions we have identified four possible choices for the levy mechanics, which are set out further ahead together with an appraisal of their key features. These choices arise from combinations of:

    1. Taking each of these in turn, we describe here our understanding of how these proposed arrangements could work, together with the arguments made by the various parties in support of them, and other points that have emerged during our work.

Turnover vs Gross Profits

Turnover

Operation:

The scheme has operated for the last 28 years on the basis of a levied percentage of bookmaker turnover. There is a company scheme, and a shop based scheme which provides some amelioration of the charge for smaller bookmakers. The fortieth scheme had a cap applied to it to provide a measure of safeguard against turnover rising sharply on the introduction of GPT.

This basis (but with no cap) has been put forward by both HBLB and BHB as their preferred basis for the 41st levy.

Arguments for

Commentary
  • Familiar to all 
  • Machinery in place 
  • This basis is aligned with the commercial arrangements proposed by BHB for the supply of data and pictures from 1 May 2002
  • the turnover based scheme is familiar to all
  • the HBLB and bookmakers have existing systems in place to implement a turnover scheme
  • we are not persuaded of the relevance of this argument, given that the levy is not a payment for a product

    •  

    Arguments against

    Commentary
    • The relationship between turnover and gross profit (where the latter provides a basis for assessing ability to pay) has become unstable and uncertain, particularly in the short term, since the introduction of GPT. This makes predicting the outcome of a turnover based scheme very difficult
    • As submitted the scheme is particularly unfair to smaller bookmakers
    • Without a cap there is a risk of turnover growth being levied when there is no growth in gross profits
    • The application of the current year’s cap has proved to be difficult, and confusing to bookmakers
  • all parties agree that the introduction of GPT will have a significant effect on turnover, however their views vary on the size of the impact, its speed, and how gross profits will be affected. Initial indications from bookmakers are that turnover has risen by perhaps 30% since 1 October, but that margins have been more volatile than previously. They cite the increase in large bets on short odds as a possible explanation for the latter. Against this background we find the setting of a turnover based levy that reflect bookmakers’ capacity to pay problematic.
  • We agree that likely outcomes could affect small bookmakers adversely without further safeguards
  • We agree that there is such a risk.

    •  

    • We note these comments made by HBLB and BC

    Other points:

    Without a cap and without protection for small bookmakers, it is hard to see how a turnover based scheme can now be certain of producing a levy which reflects bookmakers ability to pay. The reintroduction of a cap could in theory protect bookmakers but with the current uncertainty over turnover it would be very difficult to set and according to the HBLB would be "unsound". In subsequent conversations HBLB has suggested a different way of reducing the impact of the levy on smaller bookmakers, but this does not provide the same protection as a cap. The BC have indicated that the betting industry would take legal action over any determination based on turnover.

     

     

    Gross Profits

    Operation:

    The operation of this scheme is that proposed by the Bookmakers Committee.

    The levy would be set to tax the gross profits of bookmakers attributable to horseracing. There would be a lower rate for shops with a GP under £160,000pa. The top 9 companies (representing 85-90% of the levy by value) would agree a fixed percentage based upon EPOS data of the larger bookmakers to represent the proportion of their gross profits that are leviable, and this would also be the default arrangement for all other bookmakers. (The percentage would be reviewed after six months). The levy returns would therefore show the GPT figure as returned to HMC&E, the horseracing element (calculated by this fixed percentage), and the appropriate levy charge. Bookmakers other than the top 9 companies would have the right to commission (at their own expense) a chartered accountant to calculate and certify a different level of GP attributable to horseracing and to submit their return on that basis.

    Arrangements for on course bookmakers who are unaffected by the tax changes because of exemption would continue as now.

    Arguments for:

    Commentary
    • this basis for the levy will more closely reflect ability to pay
    • the uncertainties over turnover become unimportant
    • the gross profits sum used will be the same as that returned to HMC&E

     

    Arguments against:

    • untried and untested levy methodology
    • new methodology for (possibly) just one year, with associated costs





    • gross profit margins will vary enormously month by month and bookmaker by bookmaker . Policing of returns will be more difficult with the opportunity for inaccurate returns.
    • possible manipulation/distortion of horseracing percentage



    • appeals from smaller bookmakers could clog the administrative system






    • racing and bookmakers have a similar interest in growing gross profits

    • The racing industry calculate that an increase of 45% of turnover is necessary to protect the bookmakers gross margin. Should this increase not occur racing will allegedly be "worse off" under the GP system
    • GP based levy "rewards" inefficient bookmakers whose gross profits are lower
    • whilst taxes may often based on profits, other royalty rates are usually linked to turnover not profits
    • BHB argues that the normal criteria of efficiency, equity (fairness) and transparency are not satisfied by a GP levy
    • the commercial mechanisms proposed by BHB involve turnover based payments for the use of pictures and data. A simple and seamless transition would be accomplished if the levy were to continue on a turnover basis
  • we agree with the view that gross profits provides a more certain method of ensuring that the levy reflects bookmakers’ ability to pay
  • movements in turnover become irrelevant to this basis of levy
  • the basic gross profit quantum will be identical to that submitted to HMC&E, who have their own procedures for ensuring that this sum is reliable for GPT purposes

    • Commentary:

    • this form of levy has not been operated before
    • there is the possibility that a 42nd levy may be required. Indeed both BHB and BC have made arguments for this, in which case the levy methodology would have a 2-year life
      HBLB have estimated the costs of implementing the scheme to be £50,000 for once-off systems changes and £20,000 for temporary administrative support. We do not consider these costs to be material to the debate.
    • the GP quantum is transparent in that returns of this number are made to HMC&E. The introduction of a fixed percentage horseracing percentage with a transparent methodology for determining this (and reviewing it after 6 months) in our view significantly reduces the risks of inaccurate returns.
    • the methodology proposed by the BC for agreeing the percentage of horseracing turnover appears satisfactory. There is a possibility that bookmakers could depress horseracing profits by, for example, pricing other betting products more attractively. We have no basis to express an opinion on this risk, which the BC dismissed in our conversations with them.
    • the system proposed actually changes the appeals procedure to one of self-certification which requires no administrative "appeals" mechanism. In our conversations with HBLB they appeared to accept that the potential burden that they had expressed concern about in their submission was significantly reduced by what the BC was proposing.
    • The gross profits arrangements might be taken to imply that both racing (from a levy perspective) and bookmakers have a similar interest in increasing gross profits, at the expense of punters. Some described to us possible ways in which this might happen and others have dismissed these as implausible, and whilst recognising this risk would exist we have no basis for evaluating it.
    • whilst there is some force to this argument, it remains the case that a GP based levy will more closely reflect bookmakers’ ability to pay

     

    • we do not accept this argument because lower gross profits may actually reflect greater efficiency, we take notice of the fact that HMC&E felt able to accept gross profits as an appropriate basis for taxation
    • as the levy is not a royalty we do not see the relevance of this argument
    • we disagree. On the efficiency criteria we do not believe that the costs of changes to administrative systems (£70k) are material, a GP based levy more closely reflects ability to pay and is therefore arguably more fair, and the mechanism for determining the percentage of gross profit attributable to horseracing appears transparent.
    • whilst this may be true, we are of the view that as the levy is not a payment for a product the most appropriate mechanism for it may be determined on different grounds as indicated above

    Other points:

    The fixed horseracing percentage would be reviewed over the six months to 30 June, and could then be adjusted, both for this and for any subsequent levy scheme.

    The HBLB, in subsequent conversations, have suggested that the BC’s assurances on the mechanism go a long way to satisfying the Board’s concern about the practicalities of the GP basis.

    Offset vs Termination

    Offset

    Operation:

    The offset mechanism described is that currently promoted by BHB.

    The new contracts will run from 1 May 2002 – 30 April 2004. Any bookmaker signing by 1 May 2002 will be charged the amount due under the contract less levy amounts paid under the 41st scheme. Any bookmaker signing later than 1 May (and who has been using the data and/or pictures) will be unable to offset levy payments for the period from 1 May until the date he signs.

    Arguments for:

    • Removes the risk of bookmakers "paying twice" (ie levy plus amounts due under commercial arrangements)





    • The legal challenges to BHB’s contracts mean than BHB/racing cannot be certain of all of the monies paid under the contracts. BHB will "ring fence" such sums (as they determine) and they will therefore be unavailable to racing. In these circumstances the continuation of the levy will mean that racing has some certainty over funding.

    Commentary:

    • There remains confusion over whether the restricted offset (where a bookmaker signs after 1 May 2002) applies both to data and picture payments or just the former. In either case, the restricted offset means that some bookmakers could "pay racing" twice, both under the levy and under the commercial arrangements. We recommend that DCMS seeks clarification in writing from BHB as this interpretation is not clear from the contract documentation we have seen.
    • it is true that the offset mechanism will provide racing with some certainty of income whilst the commercial arrangements are put in place. The BC has argued that this is not the purpose of the levy, nor is it the responsibility of the betting industry to underwrite racing whilst the legal issues are resolved.

    Arguments against

    Commentary
    • the BC argues that this mechanism turns the levy into the price or part of the price for a product, which has never been the levy’s purpose
    • This mechanism adversely affects the negotiating "power" of the bookmakers over the commercial terms of the contract, as racing has certainty of funding via the levy
    • the proposed mechanism means that any bookmaker signing the contracts late (having continued to use the data/pictures) will not be able to offset his levy payments for that period
    • the BC has put DCMS on notice that this mechanism will be challengeable under European state aid legislation
    • the outcome of the OFT investigation may make the operation of the offset impossible, for example by ruling that BHB may not make the contracts, but that these should be made with the racecourses
    • as very few contracts have been signed the offset is not yet legally enforceable
    • VAT may become chargeable on the full commercial price rather than the balance payable after the offset
  • DCMS has taken legal advice on this

  • It is true that the certainty of a levy affects the negotiating positions of both BHB and bookmakers
  • see our comments above under "paying twice"

    •  

    • we note that the BC has received legal advice to this effect and that DCMS has obtained its own advice on this issue
    • this potential risk will remain until the OFT makes its ruling. If BHB’s commercial arrangements are prohibited or restricted by its ruling, then levy payments may not be readily offset.
    • as indicated above, we recommend that DCMS seeks written clarification of the application of the offset from BHB
    • this issue will remain unresolved until HMC&E gives a definitive ruling once contracts have been signed. BHB have however received legal advice that this will not be the case, and a letter from HMC&E on 10 December which supports this view in principle. In the light of the latter, we are inclined to view the risk of an adverse VAT ruling to be relatively low.

    Other points:

    Current BHB contracts make no mention of offset of any 42nd Levy payments (if there is to be such a scheme).

    The submission provided by the BHB speaks of a "full offset". The non recoverable element is made explicit only in the licences now in front of bookmakers.

    BHB is willing to give DCMS confirmation that it will continue to offer the offset terms after the determination.

    The BC has indicated that the industry will pursue legal action in the event of a determination relying on parallel payments.

     

    Termination

    Operation:

    The operation of this scheme is that proposed by the Bookmakers Committee.

    Basically the levy would be set at an annual rate that would be leviable monthly, but that would fall to a peppercorn when the commercial contracts come into effect. As described above, these contracts are currently timed to come into effect on 1 May 2002.

    Arguments for:

    Commentary
    • avoids any possibility of "double payment" of levy and under the commercial contracts
    • provides for a clean break between levy and commercial arrangements
    • removes State aid issue
    • does not affect negotiating powers of the parties over the terms of the contracts
  • we agree
  • we agree
  • we agree
  • the termination does still affect the negotiating position of the parties , although in a different way to the offset

    • Arguments against

    Commentary
    • as the current BHB contracts are set to come into effect on 1 May 2002, the levy would come to an end at 30 April 2002 ie it would be for one month
    • As there can be no certainty over the total (and level of "free") income that BHB will have secured under the contracts by 1 May, racing can only count on funding for this one month. HBLB have indicated that in these circumstances it would need to begin to wind up its affairs now, crystallising a pension liability of between £10m and £14m. In these circumstances, racing would effectively cease from 31 December.
  • we agree, but note that the BC’s view that it would be within BHB’s gift to delay the introduction of contracts until the legal uncertainties are resolved. BHB insists that they must start on 1 May 2002
  • HBLB has made clear its view in this regard and we have no basis on which to challenge this.

    • Other points:

    The clean break model has been assumed in all earlier discussions of the abolition of the levy. This termination option differs only in that:

      • there is (arguably) insufficient time for the HBLB to organise a smooth handover

      • the break would take place within a levy period.

    The problem would disappear if the racing and betting industries could agree a date later and more practical than 1 May 2002 for the full implementation of new commercial arrangements.

     

      1. The four options for the levy mechanics are therefore combinations of the above and are set out in the following sections, together with the main points as we see them derived from the above analysis.

        2. Option A – Turnover based with offset of commercial charges

      2. This is the recommended option in the proposals of both the HBLB and BHB.

      3. The turnover basis is familiar to all, however,

      1. The offset mechanism does:

    however,

    Option B – Turnover base with cap, with offset of commercial charges

      1. Apart from the introduction of the cap, the features of this option are the same as Option A. The purpose of the cap would be to remove the risk of the levy burden rising disproportionately relative to the bookmakers’ capacity to pay if turnover increases but profits remain flat.

      2. The cap mechanism was introduced for the first time in the 40th Levy Scheme, but is proving to be very complex to administer. The HBLB is unhappy at the prospect of carrying on such a mechanism in the 41st scheme, particularly in the light of the increased uncertainty about future turnover.

        3. Option C – Gross profits based levy, with offset of commercial charges

      3. Setting the levy as a percentage of gross profits:

    however,
      1. The points about offset are as under Option A.

        2. Option D – Gross profits based levy, with termination rather than offset

      2. This is the recommended option in the proposals of the BC.

      3. The issues relating to the Gross Profits basis are as under Option C.

      4. Termination of the levy means that the rate of monthly payment would fall to a peppercorn on the inception of the commercial arrangements by BHB.

      5. This mechanism would:

      1. However, with termination occurring on the 30 April 2002, when existing licences lapse,

     

    5. RISKS

      1. The tone of the submissions and the readiness of protagonists to pursue litigation suggest a material risk of undesirable outcomes.

      2. These include one or more of the following:

      1. The risk of these outcomes follows from the points made earlier in Section 4 regarding the various parties’ perceptions of weaknesses of the options for the levy mechanics:

    6. NEXT STEPS

      1. Our terms of reference were to review the various arguments put forward by the protagonists and to consider the options available to the Secretary of State. We were not asked to propose a solution. The process of completing the determination will involve the following steps.

      2. Decisions of principle are required in relation to the target quantum and the mechanisms by which the levy will be collected. Since the two are closely related it is only when both have been identified that it will be possible to develop the further detail of the determination. These decisions will doubtless be informed by expert advice on issues such as the risk of European intervention in relation to State Aid.

      3. The Secretary of State may seek written assurances from one or more parties, e.g.

      1. Further calculation will be required to translate an overall quantum figure into its components parts. (We would expect the HBLB to assist with this task, in conjunction with the betting industry.)

      2. The further detail required in the determination will need to encompass, in particular:

    APPENDIX – alternative margin calculation

     

    1. The Nottingham University study assumes that the betting industry will set a margin by ensuring that its post-tax revenue at the margin does not fall below its marginal cost. In doing so it will take into account, among other things:
    1. Combining equations 3.8b and 3.10b of chapter 3 of the study produces the following expression determining the margin:

      2. [a.(1-r).(x-1) + c + t]/[x.(1-r)]

      where c is the marginal cost expressed as a percentage of revenue, t is the rate of tax and levy on turnover, r is the rate of gross profit tax (including levy if appropriate), x can range between 1 and 2 and indicates the degree of monopoly, and a is a constant number.

    2. Given values for the other variables one can derive a value for the constant (a) that produces the 22.5% margin. Changing the rate of tax/levy (t) from 9% to 1% and the rate of tax (r) from zero to 15% results in a fall in the margin to only about 20%.