Important legal notice

61985C0433

European Court reports 1987 Page 03521 Swedish special edition Page 00157 Finnish special edition Page 00157

Opinion of the Advocate-General

++++

Mr President, Members of the Court, 1 . The Tribunal de grande instance ( Court of First Instance ), Mulhouse, has submitted the following question to the Court : "Does Article 95 of the Treaty of Rome forbid the imposition, on motor vehicles exceeding a certain power rating for tax purposes, of a differential tax which increases exponentially according to that rating, where the rating is arrived at by means of a formula which has the effect of subjecting to the said exponential increase any vehicle of a given cylinder capacity and such vehicles are manufactured only in certain other countries, in particular those of the Community, and not in France?" 2 . The question has been submitted to the Court in connection with proceedings brought by a private individual, Mr Feldain, the owner of a foreign car, against the French tax authorities . Mr Feldain considers that the system of annual taxation applied to luxury motor vehicles under Article 18 of Law No 85-695 of 11 July 1985, which was adopted following the Court' s judgment of 9 May 1985 in Case 112/84 Humblot (( 1985 )) ECR 1367, at p . 1375, is still contrary to Article 95 of the Treaty . 3 . Details of the system previously applied and of the new legislation are set out in the Report for the Hearing and need not therefore be repeated here . 4 . The Court has had occasion in numerous judgments to define the criteria for the application of Article 95 . One of the most detailed judgments is that of 27 February 1980 in Case 169/78 Commission v Italy (( 1980 )) ECR 385 in which the Court stated in particular that "Article 95 must guarantee the complete neutrality of internal taxation as regards competition between domestic products and imported products" ( paragraph 4 ). 5 . In a judgment delivered on 15 March 1983 in Case 319/81 Commission v Italy (( 1983 )) ECR 601, paragraph 21, the Court added that the Member States are empowered "to adopt, whilst observing the relevant directives, a higher rate of VAT on luxury products as opposed to domestic or imported products not having that quality, provided, however, that the criteria chosen to determine which category of product is to be more heavily taxed are not discriminatory as against imported products similar to or in competition with domestic products in the manner contemplated by the second paragraph of Article 95 ". 6 . More particularly with regard to the taxation of cars, the Court further stated very recently ( judgment of 16 December 1986 in Case 200/85 Commission v Italy (( 1986 )) ECR 3953, at p . 3969 ) that "in order to decide whether contested differential tax arrangements are compatible with Article 95 of the Treaty it must be considered whether the differentiation is based on an objective criterion, whether it is directly or indirectly discriminatory and whether it is of such a nature as to protect domestic products from competing products imported from other Member States" ( paragraph 9 ). In the same judgment the Court stated ( at paragraphs 16 and 20 ) that the competitive relationship which must be taken into account in determining whether or not there is a protective effect must extend to all cars, both those with diesel engines and those with petrol engines . 7 . In the present case, it is clear that the Humblot judgment of 9 May 1985 ( supra ) is of the greatest importance . In that case the Court stated that "as Community law stands at present the Member States are at liberty to subject products such as cars to a system of road tax which increases progressively in amount depending on an objective criterion, such as the power rating for tax purposes, which may be determined in various ways ". 8 . "Such a system of domestic taxation is, however, compatible with Article 95 only in so far as it is free from any discriminatory or protective effect ". 9 . The Court then went on to say that the application to all cars of a power rating for tax purposes in excess of 16 CV of a special tax which is five times the tax in the band immediately below "entails a much larger increase in taxation than passing from one category of car to another in a system of progressive taxation embodying balanced differentials like the system on which the differential tax is based ". 10 . It is therefore necessary to consider in the first place whether the new method of taxation now constitutes, as a whole, a "system of progressive taxation embodying balanced differentials ". 11 . I - The Tribunal de grande instance, Mulhouse, asks whether the new system is compatible with Article 95 because, for vehicles exceeding a given power rating for tax purposes, the amount of the tax increases exponentially according to that rating . 12 . In that connection it should be pointed out in the first place that although it is true that the curve representing the different levels of taxes does in fact from a certain point display the characteristics of an exponential curve, that is due to the combined impact of all the factors involved . 13 . The various coefficients were determined not on an "exponential" basis but according to a "geometric progression ": each level of taxation is arrived at by multiplying the lower level by a given factor ( called the "common ratio" of the progression ). In this case, that factor is not constant . It varies between 1.2 and 2.4 for the categories up to 16 CV inclusive, and is 1.5 for the categories above 16 CV . Since almost without exception the multiplication factor used above 16 CV is lower than those used below that threshold ( the factor used for the transition from the 8 and 9 CV tax band to the 10 and 11 CV tax band is only 1.2 ) it is appropriate to conclude that the French rules are not open to criticism from that point of view . 14 . However, the compatibility of a tax system with Article 95 does not depend only on the rate of progression but also on the way in which the basis of assessment is defined and on the other detailed rules for the application of the tax ( 1 ) and in particular, in the case of a geometric progression, upon the number of terms . This prompts me to consider the sequential relationship between the various tax levels under the system in question . 15 . There are nine tax bands, as follows : up to 4 CV : 2 actual power ratings for tax purposes, ( 2 ) from 5 to 7 CV : 3 ratings, 8 and 9 CV : 2 ratings, 10 and 11 CV : 2 ratings, 12 to 16 CV : 5 ratings, 17 and 18 CV : 2 ratings, 19 and 20 CV : 2 ratings, 21 and 22 CV : 2 ratings, 23 CV and above : 9 actual ratings . ( 3 ) 16 . The immediately striking feature of that sequence is the variation in the number of power ratings for tax purposes contained in the various tax bands . 17 . The inclusion of five power ratings in the 12 to 16 CV tax band, in particular, is rather surprising . According to the information provided by the French authorities in reply to a question put by the Court, 74 car models come within that tax band . 18 . The next three bands are each limited to two power ratings . The 17 and 18 CV band covers only two models of cars, the 19 and 20 CV band ten models and the 21 and 22 CV band three models . 19 . The French Government has not explained the reasons for that particular tax structure and furthermore I have not succeeded by my own efforts in identifying any objective and neutral criterion underlying that structure . 20 . In the first place the method chosen has the effect of ensuring that 14, 15 and 16 CV vehicles, which one would expect to find in a separate tax band, are not taxed more heavily than 12 or 13 CV vehicles . Accordingly, no French vehicle is subject to a tax exceeding FF*1*729 ( the figure applicable in the Paris region in 1984 ). 21 . By contrast, the creation above the 16 CV threshold of "slim" bands embracing only two power ratings, combined with the geometric progression which characterizes the system as a whole, leads to a very steep increase in the amounts of tax payable in the categories in which the vehicles sold are exclusively of foreign manufacture ( FF 2*594, FF*3*882, FF*5*832 and FF*8*758 ). 22 . Whereas the four steps between a 12 CV car and a 16 CV car do not give rise to any increase in tax, the same four steps between a 16 CV car and a 20 CV car lead to difference in taxation of 224 %. 23 . It is clear that such a progression is not based on "balanced tax differentials" in the sense in which the Court used that expression in Humblot . The effects of the exponential nature of the system are mitigated with respect to French cars at the top of the range, thanks to the creation of the "wide" tax band mentioned earlier, whereas they are accentuated by the return to "narrow" bands once the stage is reached where only foreign vehicles are affected . The only possible conclusion is that the method of classifying cars is not based on an objective and neutral criterion as regards competition between French and foreign vehicles . 24 . Must it be concluded from this that cars of a power rating for tax purposes in excess of 16 CV ( or whose cylinder capacity exceeds 3 litres ), which all happen to be cars of foreign manufacture, should escape any taxation or that they should be taxed at the rate laid down for the 12 to 16 CV band? I do not think so . In my opinion those cars should, until such time as the whole system is overhauled, be taxed as if there were further bands containing five power ratings ( 17 to 21 CV, and so forth ). 25 . II - In the second part of its question, the Tribunal de grande instance, Mulhouse, goes on to ask whether or not the formula according to which the power rating for tax purposes is calculated is compatible with Article 95 of the Treaty . 26 . In the Humblot judgment the Court conceded that the power rating for tax purposes could be regarded as an objective criterion on the basis of which the road tax could be increased progressively . 27 . The Court stated, however, that "the power rating for tax purposes ... may be determined in various ways ". It did not therefore express any view regarding calculation formulae of the type at present used in France . The French formula was laid down in an administrative circular of 23 December 1977 and is as follows : 1.48 P = m ( 0.0458 C ) K In that formula, P stands for the power rating for tax purposes, m is 1 for petrol and 0.7 for diesel fuel, C is the cylinder capacity of the engine expressed in cubic centimetres . K is a parameter representing the transmission of motion and is obtained by calculating "the weighted arithmetic mean of the speeds expressed in kilometres per hour which would in theory be attained by the vehicle at an engine speed of 1*000 revolutions per minute for the various ratios of the forward gears of the gearbox ". 28 . It seems to me that the objective and neutral character of such a formula would be beyond doubt if it were applied outright to all vehicles without distinction, whether they were manufactured in France or abroad . The actual position, however, is that for vehicles whose power exceeds 100 kW, the value of K is subject to a ceiling of 21, even if an objective valuation produces a higher figure . As we shall see in due course, very few French cars are affected by this rule . 29 . Since K is the denominator, the fact that it is subject to a ceiling entails an artificial increase of the power rating for tax purposes ( P ) whenever the real value of K exceeds 21 kilometres per hour . ( a)*The effect of the imposition of a ceiling on the factor K 30 . In its reply to the second question put by the Court, the Agent for the French Government purported to justify the limitation of the factor K to 21 by the following argument : "In view of the need not to afford any abnormal advantage to certain sports models or 'top-of-the-range' models, whose real power and performance were very high and which might have appeared to be relatively economical in terms of energy consumption when driven so that their full potential was not exploited, the value of the parameter K was limited to 21 km/h for private vehicles whose real power exceeded 100*kW ." 31 . In the light of the explanations provided at the hearing it appears that that statement should be understood as follows : 32 . As the cylinder capacity ( and therefore the power ) of an engine*(4 increases, the less the engine speed needs to be geared down . On the basis of the normal method of calculating the power rating for tax purposes in France ( where the factor K is not subject to a ceiling ) many engines with a large cylinder capacity would, because of their high gear ratios, have to be placed in relatively low tax categories . Vehicles which have such gear ratios are relatively economical as regards fuel consumption when they do not exceed 120 to 130 km/h . 33 . However, vehicles of that kind, with a large cylinder capacity, are capable of top speeds of 180 to 250 km/h . When they use their capacity to the full, their fuel consumption is much higher ( since air resistance, which must be overcome, increases in proportion to speed ). In order to ensure that such vehicles, which are likely in practice to consume a great deal of fuel, are not unjustly favoured, it is necessary, beyond a certain threshold, to disregard the gear ratios . That result can be arrived at by freezing the coefficient K at 21 km/h for vehicles whose real engine power exceeds 100*kW . 34 . The logic of that reasoning does not seem to me to be open to criticism . It would be paradoxical if, solely as a result of applying a reduction ratio of K = 30, for example, a vehicle with a cylinder capacity of 3*590 cc ( Jaguar XJ6, 6 cylinders, rated at 21 CV for tax purposes ) were to be placed in the same tax category as a vehicle with a cylinder capacity of 2*500 cc ( Citroën CX 25, 4 cylinders, rated at 13 CV for tax purposes ). 35 . It is also important to note that almost all foreign vehicles with a cylinder capacity of less than 3 litres which are virtually identical to French top-of-the-range vehicles as regards space and comfort ( Citroën CX, Peugeot 505, Renault 25 ) are placed in tax bands no higher than 16 CV ( the Alfa 6, all Audis, 5.0 Series BMWs, the BMW 628 Csi and the BMW 730, the Fiat Croma, the Ford Scorpio, the Lancia Thema, the Maserati Biturbo, the Mercedes 260*E, 260*SE, 300 E/SEL with manual gearbox, the Rover 820, the Saab 9000, the Volvo 740 and 760 and so forth ). 36 . On the other hand, the manner in which the power rating for tax purposes is calculated in France is open to the following criticisms : 37 . The official aim of introducing the factor K in the formula for calculating the power rating for tax purposes is to favour vehicles which consume relatively little fuel . As we have seen, vehicles with a large cylinder capacity but high gear ratios use large amounts of fuel only at speeds in excess of 130 km/h . In France it is in any event prohibited to drive any faster than 130 km/h . 38 . In the second place, it cannot be denied that vehicles with a cylinder capacity of less than 3 litres can also have a heavy fuel consumption . That is true of what are known as sports models with particularly high engine speeds and high compression ratios and relatively low gear ratios . They too can achieve speeds in excess of 200 km/h . 39 . Such vehicles may consume more at 180 km/h than a vehicle with a larger cylinder capacity but higher gear ratios . However, sports models of that type are not penalized in the same way as vehicles with a large cylinder capacity . 40 . Moreover, the French Government has not put forward any objective criterion to justify imposing a ceiling for the factor K of 21 rather than 20 or 22 km/h . 41 . For his part, Mr Feldain, as indicated in the Report for the Hearing, has calculated that if the parameter K is limited to 21 km/h, any vehicle with a cylinder capacity exceeding 3*109.7 cc will automatically be given a power rating for tax purposes exceeding 16 CV regardless of its fuel consumption characteristics . Only vehicles manufactured abroad have a cylinder capacity exceeding 3*109.7 cc . 42 . However, I do not find that argument very convincing since the 17 CV rating has to start somewhere and I do not see why the threshold should not be 3*109.7 cc . If the ceiling for K had been 20 km/h, the threshold for the 17 CV rating would have been 2*961.75 cc . That would not have caused any French vehicle with a rating of 16 CV to be given a rating of 17 CV since no French car has a cylinder capacity in excess of 2*664 cc . ( Even the Peugeot 604, which is no longer manufactured, had a capacity of only 2*849 cc ). 43 . If the ceiling for the factor K had been 22 km/h, the threshold would have been 3*257.92 cc . Only one other foreign vehicle ( the Porsche 911 ) would possibly have been able to "drop" into the 16 CV rating, but in order to be certain it would be necessary to know the factor K for that vehicle . 44 . But there is a last argument which, in my opinion, is decisive . It is the fact that the result of the calculation 0.0458 x C continues K to be raised to the power of 1.48 even where the factor K is "frozen" at 21 . The result of this is to accentuate the impact of the higher cylinder capacity of vehicles of more than 3*109.7 cc without any offsetting effect, big or small, resulting from the inclusion at the same time in the calculation of the higher gear ratios which are usually a feature of such vehicles . 45 . I said just now that it was not reasonable for a car of 3*590 cc to be rated at 13 CV . Conversely, it seems to me doubtful whether there is any justification for giving it a rating of 21 CV . 46 . It is difficult to understand why, starting at the threshold where the real value of K ceases to be applied, the authorities do not simply adopt a method of classification based on cylinder capacity alone . 47 . I should not like to say whether it would be sufficient, by way of alternative, to reduce the exponent of 1.48 for cars whose coefficient K is affected by the ceiling . ( The mere fact of reducing it from 1.48 to 1.44, for example, would in itself have the effect of reducing the power rating for tax purposes of such cars ). What is certain, in any event, is that the "freezing" of one of the components of a mathematical formula should logically lead to a review of the other components of the formula, to ensure that their influence does not suddenly become disproportionate, particularly where international competition is involved and the change in the calculation method affects mainly cars manufactured in other Member States . 48 . That is the result in the present case . In response to a question asked at the hearing, the French authorities have forwarded to the Court a list of six models of French cars whose power rating for tax purposes has been changed as a result of limitation of the factor K . However, it is a list of cars "type-approved since 1 January 1978 ". Three of the models are either no longer manufactured or else were given a different rating only because of their automatic three-ratio gearbox . The three remaining models are sports type cars, of which relatively few are produced, and only two moved to a higher tax band . 49 . It thus appears that the disputed regulations affect French cars only marginally . 50 . Unfortunately, a complete list of the foreign cars which are given a higher rating as a result of this calculation method is not available . But it must be a long list since, according to the specialized press, German manufacturers sell around 56 models whose power exceeds 100 kW, British manufacturers about 20 and Italian manufacturers about 34, whilst it seems that French manufacturers produce only 8 . 51 . It is therefore appropriate to conclude that the method of calculating the power rating for tax purposes of cars with a power output of more than 100 kW is not neutral from the point of view of competition and is of such a nature as indirectly to protect cars of that power output manufactured in France . ( b ) Inclusion in the calculation of the various ratios of automatic gearboxes 52 . Before 15 April 1983 the factor K for vehicles with a four-ratio automatic gearbox took account of all four ratios, according to the following formula : K = k1 + k2 + k3 + k4 53 . Since the issue of a circular dated 15 April 1983, the average of only three ratios has been taken into account instead of four, according to one or other of the following formulae : K = k1 + k2 + k4 if k4 is equal to or less than 1.4 k3 or else K = k1 + k2 + 1.4 k3 if k4 is equal to or exceeds 1.4 k3 54 . The change in the regulations thus means that the factor K has, since 1983, been lower than its previous value : consequently, the power rating for tax purposes of a vehicle equipped with a four-ratio automatic gearbox is increased . 55 . However, four-ratio automatic gearboxes are more economical as regards fuel consumption than three-ratio gearboxes . Legislation whose purpose is to favour the most economical vehicles should therefore give vehicles equipped with such gearboxes a lower rating, not a higher one . 56 . It appears, however, that in recent years four-ratio automatic gearboxes have also been introduced in vehicles manufactured in France ( see the French Government' s reply to the third question put by the Court ). 57 . Thus, all automatic Peugeot and Citroën BX cars are equipped with four-ratio gearboxes . The calculation rules referred to earlier apply to those models in the same way as to models manufactured abroad . 58 . In those circumstances, the new method of calculating the parameter K for vehicles fitted with a four-ratio automatic gearbox cannot, in my opinion, be regarded as discriminatory or protective . Conclusion 59 . For all the foregoing reasons, I suggest that the Court should reply as follows to the question submitted to it : "Article 95 of the Treaty establishing the European Economic Community must be interpreted as meaning that a Member State may not impose upon cars exceeding a certain power rating for tax purposes, all of which are imported, in particular from other Member States, a differential tax which increases much more steeply than the differential tax applied to cars which do not exceed that power rating for tax purposes, and such a power rating for tax purposes must not itself be calculated in such a way as to arrive at an excessively high rating for certain categories of cars most of which are imported, in particular from other Member States ." (*) Translated from the French ( 1 ) See the judgment of 27 February 1980, supra, paragraph 7 . ( 2 ) The car known as the "2 CV" is now rated as 3 CV for tax purposes . ( 3 ) It seems that cars exist only with the following ratings : 23, 26, 27, 31, 32, 34, 39, 40 and 54 CV . ( 4 ) I consider that for the purposes of this comparison it is necessary to disregard "turbo" and "16 valve" engines which enable cars of a middle-range cylinder capacity to achieve a real power output exceeding that of vehicles with much higher cylinder capacity . It seems to me that as a result of these new developments most of the Member States will be obliged at some stage to review their car taxation systems .