The IFS paper purports to calculate the revenue consequences of a change in beer duty. Those calculations are flawed for four reasons:

—  They assume that the sole objective of any tax is to maximise revenue from that tax;

—  They exclude the effect of duty changes on illegal cross-channel imports;

—  They exclude the effect of duty changes on direct and other Government revenues;

—  The conclusions were derived from the FES data, a small sample which is known to conflict with nationally collected data because it excludes over a third of all beer drinking.

1.  ILLEGAL CROSS-CHANNEL IMPORTS

  The IFS study excludes the effect of duty changes on illegal cross channel imports. But illegal imports are crucial to any calculations of the impact of duty changes, because:

  1.1  Illegal imports are a significant proportion of total beer consumption. According to Customs evidence, the level of illegal imports exceeds legal imports by a factor of x3.

  1.2  Illegal imports are particularly price-sensitive for the reasons outlined below. Indeed, it is arguable that a duty cut of a few pence could eliminate beer bootlegging altogether.

2.  PRICE-SENSITIVITY OF ILLEGAL IMPORTS

  2.1  Bootleggers exploit the duty differential between the UK (currently 33p per pint of 5 per cent abv beer) and France (around 5p/pint). So each 1p increase in UK beer duty increases the bootleggers'/gross margin by, at minimum, 3 per cent.

  But the proportional effect on bootleggers' net profits is far greater, as bootleggers' margins have to cover their costs of sales and distribution.

  2.3  Sales costs create the incentive for bootleggers' customers to buy illegal beer rather than beer from legitimate sources. This incentive might take the form of a discount, of a door-to-door delivery service, or of persuasive threats—whatever way, a cost is involved. We would have thought that sales costs of at least 10p/pint would be necessary to persuade customers to buy beer illegally, given the wide availability of competitively priced beer from legitimate UK distributors.

  2.4  Distribution costs will include:

—  shipping British beer across the channel (a high proportion of bootleggers' loads are now UK brands);

—  warehousing the beer in Calais;

—  shipping it back to Dover;

—  an additional cost to represent the risk of occasional confiscations of vans and loads by Customs;

—  and onward distribution from Dover throughout the UK, on average over 100 miles.

  We would be surprised if these distribution costs did not exceed 10p per pint.

  2.5  We consequently estimate the net margin remaining to the bootlegger as under:

—  duty difference (France v UK) 28p per pint

—  sales costs (say) 10p per pint

—  distribution costs (at least) 10p per pint

—  Net margin (at most) 8p per pint.

  This is illustrated by the graph of duty rates included in CAMRA's evidence and reproduced in below.

3.  RECENT BOOTLEGGING TRENDS

  3.1  The calculations above suggest that every penny on duty alters bootleggers' net profits by over 10 per cent. This is confirmed by recent trends in bootlegging.

  3.2  Over the last two years bootleggers have suffered from far better Customs enforcement, far higher penalties, greatly increased competition from UK supermarkets (who started selling cheap French beer a year ago), and higher ferry prices (since the ending of duty-free allowances in July 1999).

  3.3  Despite all this, beer bootlegging volumes have increased. It is difficult to suggest any economic reason for this except for the 1p/pint increases of January 1998 and January 1999, which together increased bootleggers' net profits by some 25 per cent.

  3.4  Indeed, the four 1p increases in beer duty since the introduction of the Single Market in 1993 have had the unfortunate effect of doubling the profits that bootleggers would have made had duty rates been left unchanged. Had duty rates been reduced by 4p/pint instead of being increased by that amount, then little if any net margin would now be left to bootleggers, and bootlegging of beer might well have been substantially diminished.

4.  NET REVENUE CONSEQUENCE OF A DUTY REDUCTION

  4.1  We have demonstrated that a duty cut of around 8p/pint should be sufficient to make beer bootlegging uneconomic. The elimination of beer bootlegging would increase UK beer duty revenue by around 3.5 per cent (this is Customs's estimate of the proportion of illegal imports to total UK sales).

  4.2  According to the IFS calculations, an 8p duty cut would reduce beer duty revenue on UK sales by more than the 3.5 per cent benefit from the elimination of beer bootlegging. But the elimination of beer bootlegging would have other important consequences. For example, considerable Customs resources would be freed if beer bootlegging became uneconomic. The transfer of these resources to detection of other smuggling such as tobacco should increase revenues from other duties.

5.  THE EFFECT OF DUTY CHANGES ON DIRECT AND OTHER GOVERNMENT REVENUES

  5.1  The IFS report considers only the effect of a change in duty levels on indirect taxation. But changes in beer duty also affect Government receipts from direct taxation. This is because lower duty would increase beer sales: greater beer sales would create greater employment (pubs are very labour-intensive and employ some 400,000 people): and that extra employment would generate NI and income tax, while reducing unemployment support.

  5.2  The IFS does not deny that such effects occur, but they have suggested that such effects are likely to be smaller than the effect on indirect taxation. They have produced no calculations to back up this assertion.

  5.3  The only calculations that we are aware of that attempt to assess the effects of direct as well as indirect taxation are the studies undertaken in the mid 1990s by the Henley Centre, and by Oxford Economic Forecasting using the Treasury model of the economy. Both studies showed that the effect of a duty cut on direct tax revenues would be large, possibly exceeding the effect on indirect taxation.

  5.4  If the IFS wishes to contribute to this debate, it would be interesting for them to repeat this exercise and see if they obtain the same result. Until they have done so, it is unhelpful and somewhat presumptive for the IFS to suppose that their "gut feel" is more accurate than the detailed work undertaken by Henley Centre and OEF.

6.  THE USE OF FES DATA

  6.1  The IFS conclusions are derived from data from the Family Expenditure Survey (FES). This is a very small sample that is highly unrepresentative.

The FES acknowledges that it excludes a considerable proportion of beer consumption, as the following statements by the FES indicate:

—  1993  ". . . the estimated average expenditure . . . in the FES on beer is about two-thirds of corresponding estimates from statistics produced by HM Customs & Excise"

—  1998  ". . . the estimated average of all households on beer is rather over half of corresponding estimates produced by HM Customes & Excise"

  6.2  The IFS defends their use of FES data on the grounds that, although unrepresentative, the extent by which the FES under-records beer consumption remains constant. But the above reports show this is not true even over a timescale as short as five years, let alone the 15-years compared by the IFS.

  6.3  The IFS concludes from this unrepresentative data that "real spending on beer has remained fairly constant, with an average real increase . . . of less than half a per cent each year between 1979 and 1996". This is not only wrong, but wrong by a large margin. The Consumer Expenditure survey shows at 20 per cent decrease in real spend on beer over this period.

  6.4  Based on this highly flawed data, the IFS proceeds to derive figures for elasticity of beer consumption which are equally unreliable. The figures they derive not only differ from those generally used within the industry, but also from those used by HM Customs and Excise in the SPIT study—which IFS themselves advised on.

  6.5  The IFS admits that their estimates of price elasticity in this latest report have wide confidence intervals. In fact these confidence levels are so wide that many of their "conclusions"—eg that beer is less price elastic than spirits—have no statistical significance, and are probably simply untrue.

  6.6  In conclusion, the IFS report has done much to muddy the waters—unnecessarily—and nothing to clarify the real figures. It is disappointing that an institution of their standing should have produced such a poorly researched report, and we urge that its conclusions are treated with considerable caution.

30 November 1999