Portovesme argued that the preferential tariff addressed a market failure, namely the failure of the recently liberalised Sardinian electricity market to deliver competitive prices owing to the market power of the incumbent operators. Portovesme claimed that ENEL and Endesa (whose power generation assets in Sardinia were subsequently sold to E.ON) formed a duopoly, and artificially kept electricity prices in Sardinia in line with average Italian prices, whereas in fact Sardinian prices should be much lower, especially for large electricity consumers, given the overcapacity in electricity generation, which ought to encourage generators to sell at competitive prices.

Portovesme contended that the preferential price enjoyed corresponded to the price the company would have obtained under normal market conditions, i.e. if the electricity market functioned correctly.

According to Portovesme, the compensatory payments received from the Equalisation Fund (amounting to EUR 37 per MWh) brought the price paid by Portovesme into line with power prices paid by its competitors, and did not distort competition, since all EU zinc producers paid equivalent prices, and the measure merely offset the competitive disadvantage suffered by the company as a result of Sardinia’s structural problems (which included the absence of natural gas).

Metals like zinc, aluminium and lead were commodities for which there was one world price, arrived at on the London Metal Exchange. Variations in local production costs did not translate into differences in world prices. In particular, Portovesme’s zinc production was so small (at less than 1 % of world production) that it could not affect the world price.

Without the tariff Portovesme would have to close its plant, but the closure would not benefit other EU zinc producers, who would not be in a position to increase output, and the gap would be filled by imports from third countries. Portovesme concluded that the preferential tariff did not affect competition or trade in the EU.

Moreover, according to Portovesme, the compensatory payments from the Equalisation Fund did not involve State resources, since the Equalisation Fund merely redistributed private funds which came from electricity end-users.

Portovesme noted that power off-take by energy-intensive industries was important to ensure security and continuity of supply in Sardinia, and to avoid inefficiencies in power generation which would lead to higher electricity costs for end-users.

Finally, the company points out that, if it was obliged to close down its plant, the consequences for Sardinia’s production base would be very serious, as 790 direct jobs and 500 indirect jobs would be lost.

In addition to the general problems affecting European alumina refineries, Eurallumina also pointed to specific Sardinian problems that prevented local industries from reaping the benefits of electricity market liberalisation, notably the market power of incumbent electricity operators (see Portovesme’s observations in recital 55 above). Eurallumina also contended that the absence of natural gas was a specific handicap for the production of alumina in Sardinia, as the large quantities of steam required for the process had to be obtained using more expensive heavy oil.

According to Eurallumina, after liberalisation the market power of electricity generators in Sardinia led to an increase in electricity prices: without the tariff, Eurallumina would have had to pay EUR 65,95 per MWh for its electricity supplies, as against the EUR 44 per MWh it used to pay before liberalisation. The 2004 tariff brought the price down to EUR 35 per MWh.

Eurallumina concluded that the tariff introduced by the 2004 Decree was necessary pending the development of the new infrastructure that would remove the main factors of competitive disadvantage affecting Sardinian industry.

A competitor of Portovesme — the same company that had originally brought the case to the Commission’s attention — wrote to the Commission again after the opening of the formal investigation. It submitted an analysis of the 2004 preferential tariff in which it concluded that the tariff constituted unlawful State aid. The disputed tariff scheme was fundamentally different from the Alumix tariff granted to Alcoa: it was impossible to take the view that it did not constitute State aid. The aid must be held incompatible with the common market, as it could not be approved either as regional operating aid or as restructuring aid.

Solvay Chimica Italia and Solvay Chimica Bussi, who operate in the chlorine-soda industry, sent observations on the 2004 tariffs in July 2005, after the expiry of the one-month deadline for submitting comments in response to the decision to open the formal investigation. Despite the delay, these companies asked to be considered interested parties. In their observations they submitted that both the 2004 and the 2005 tariffs constituted unlawful and incompatible State aid.

Portovesme considered that the 2005 tariffs did not distort competition, that they were necessary to ensure the survival of the industries concerned, and that they could be considered compatible as regional aid. The company’s arguments are similar to those put forward by Italy, to which reference may be made (see recitals 87 to 104).

Eurallumina likewise considered that the 2005 tariffs did not distort competition, were necessary to ensure the survival of the industries concerned, and could be considered compatible as regional aid. The company’s arguments are similar to those put forward by Italy, to which reference may be made.

The two competitors of Portovesme and Syndial that had submitted observations in Case C 38/B/04 submitted observations in this case too. Both of them contended that the disputed measure constituted operating aid which should be considered incompatible with the internal market.

In 2005, according to the data provided by Italy, the electricity price negotiated by Syndial with its supplier exceeded EUR 60 per MWh, whereas the tariff would bring it down to EUR 27 per MWh. The rationale of the two measures was that the State would reimburse to the beneficiaries the extra costs they had sustained in order to purchase electricity in Sardinia compared to average prices paid by their EU competitors.

This argument also supposes that where a market was not functioning properly a Member State would be justified in setting prices that simulated conditions of effective competition (see Italy’s comment in recital 72 above).

Further, if this submission were to be accepted, there would be no State aid where a Member State granted a subsidy to bridge the gap between a price freely negotiated between two market operators and the theoretical price at which agreement might have been arrived at in conditions of full competition. This would defeat the fundamental purpose of State aid control.

Italy contends that these tariffs are identical to the Alumix tariff, which the Commission considered not to constitute aid (see recital 74 above), and should likewise be considered free of aid.

In the present case, the market operator principle does not apply. The position would have to be assessed in the light of the market operator test if the State sold electricity to the companies concerned, either directly or through a State-owned company, as it did in Alumix. But here the electricity is supplied by private companies. There is therefore no need, as there was in Alumix, to examine whether the State’s behaviour has been comparable to that of a rational market operator.

Since the preferential electricity tariffs in Italy are granted exclusively to undertakings operating existing facilities located in Sardinia and belonging to a closed list of industries identified in the legislative and regulatory instruments that form the legal basis of the measures, the advantage they confer is a selective one.

As explained in recital 28, the compensatory payments at issue here are financed by means of a parafiscal charge collected by the Equalisation Fund via the A4 component of the electricity tariff. This charge is obligatory, being imposed by means of decisions of the AEEG that implement national legislation. The Equalisation Fund is a public body established by law which carries out its functions on the basis of precise instructions laid down in the decisions of the AEEG.

In the PreussenElektra case, the Court held that an obligation imposed on private electricity supply undertakings to purchase electricity produced from renewable energy sources at minimum prices higher than the real economic value of that type of electricity did not constitute State aid, because the measure did not involve any direct or indirect transfer of State resources.

The Commission points out that in PreussenElektra the resources required to finance the measure were provided by the electricity suppliers directly to the producers of renewable energies, without transiting through any public body. In that system, the sums to be transferred could never be at the disposal of the Member State’s authorities. In the case at hand, however, the monies transit through a public body, the Equalisation Fund, before being channelled to the final beneficiary. Thus PreussenElektra addresses different factual circumstances, and is not relevant to the present case.

Italy contends that the Equalisation Fund does not have any discretion in the disbursement of the funds, which are earmarked for the financing of the tariffs and never come within ‘the perimeter of public finance’. Thus the State cannot freely dispose of the funds, and consequently the funds do not constitute State resources.

Before considering the role played by the Equalisation Fund, the Commission has ascertained whether any of the other tests enumerated in Pearle are satisfied. It is clear that here, by contrast with the Pearle case, the disputed tariffs were established at the initiative of the State, not at the initiative of an economic sector. In Pearle, moreover, the beneficiaries of the measure were also the sole contributors of the resources used, so that the intervention of the public body did not tend to create an advantage which would constitute an additional burden on the State. In the case at hand, the beneficiaries do not bear the financial burden of the levy, which rests solely on electricity consumers. Thus the Pearle judgment cannot validly be relied upon irrespective of whether merit can be found in the argument that the Equalisation Fund is a mere accounting intermediary.

However, the analysis cannot be limited to the powers of the Equalisation Fund in its capacity as a public body. What has to be ascertained, rather, is the more general question whether the State, either directly or through any other body designated by it, can exercise control over the funds used to finance the tariff. The same test would have to be applied if the Equalisation Fund were a private body.

The similarities with the case at hand are evident. The surcharge on the electricity price used to finance the tariffs at issue here is imposed, as in Essent, by law. The Equalisation Fund plays the same role as SEP, as it centralises and manages the proceeds of the parafiscal charge, and is subject to the same constraints, as it cannot use the proceeds from the charge for purposes other than those provided by law (the financing of preferential tariffs). The State is in a position to control and orient the use of the resources: the Equalisation Fund discharges its accounting functions on precise instructions from the AEEG, which acts within its statutory regulatory powers and/or in implementation of national legislation. Thus the resources handled by the Equalisation Fund remain under public control at all times.

Italy’s Supreme Court of Cassation had already ruled that the legal personality of the Equalisation Fund was not separate from that of the Italian State, and that the monies transferred to the Equalisation Fund should be considered State property, even if they derived from private entities and were intended for private undertakings (judgment No 11632/03 of 3 April 2003). In the Iride case, the applicants, Iride SpA and Iride Energia SpA, brought an action in the Court of First Instance challenging a decision whereby the Commission had categorised as State resources the sums administered by the Equalisation Fund in the A6 account. The arguments put forward by the applicants were very similar. They disputed the substance of the Court of Cassation’s ruling, contending that the role of the Equalisation Fund was merely that of an accounting intermediary between private citizens subject to a financial obligation and the beneficiaries of the sums concerned, a role which did not allow the Equalisation Fund to make use of the monies deposited even for a short period of time. The applicants also claimed that the case-law developed in PreussenElektra was applicable.

That conclusion concerned the Equalisation Fund’s A6 account, used to finance the stranded costs of Italy’s electricity sector. It can, however, logically be extended to the A4 component, which finances the disputed tariffs. The judgment of the Court of Cassation was based on an analysis of the legal personality of the Equalisation Fund, and its finding with respect to State property therefore applies to all of the monies deposited with the Equalisation Fund. The same is true of the finding by the Court of First Instance that the State can control the resources administered by the Equalisation Fund. There is no difference between the A6 account and the A4 component other than the purpose to which the resources are put (covering stranded costs in the case of the A6 account, and financing preferential tariffs in the case of the A4 component). The sums transferred to the Sardinian beneficiaries from the A4 component must therefore also be categorised as State resources.

As regards the effect of the measures on trade between Member States and the ensuing distortion of competition, it is undisputed that the markets concerned (lead, zinc, silver, aluminium and chlorine/soda) are fully open to competition. This by itself is enough to establish that the tariffs have an effect on intra-Union trade.

Nevertheless, the Commission has considered the argument advanced by the beneficiaries and by Italy that in the metals industry (the argument has not been put forward expressly in respect of chlorine) the tariffs have no effect on trade and do not distort competition because there are no actual trade flows between Member States; no such trade flows are likely to develop, and in view of the particular features of the industry the tariffs do not damage the beneficiaries’ competitors in the Union.

Moreover, the pattern of decreasing EU production, increasing imports from non-Union countries, and limited or no trade flows between Member States is not unusual: indeed it is typical of industries which are experiencing structural difficulties or which are subject to high competitive pressure. Such industries are particularly sensitive to measures taken by Member States to improve the competitive position of their domestic industries.

In any event, the complaints received from the beneficiaries’ competitors provide an indication that the disputed tariffs do raise competition problems and affect trade within the Union.

The Commission concludes that the disputed 2004 and 2005 tariffs constitute State aid within the meaning of Article 87(1) TFEU and can be authorised only if they qualify for one of the exemptions provided for in the Treaty. This conclusion is consistent with that arrived at in the Alcoa case, which concerned virtually identical tariffs.

The Italian electricity market is generally highly concentrated, although less so in the north of the country. The dominant operator in all zones is the former monopolist ENEL, except in Sardinia, where it holds a duopoly with E.ON. ENEL has considerable market power and has been found by the Italian competition authority to have abused this power in several instances. Electricity prices in Italy are generally high, owing to a generation mix based largely on fossil fuels (essentially gas), the absence of nuclear capacity, and congestion on the interconnectors with the rest of Europe.

Prices paid in Sardinia by the beneficiaries of the tariffs in 2004/2005, on the basis of bilateral contracts, ranged from EUR 60 to EUR 65 per MWh. It has not been possible to compare these prices to those recorded on the Italian mainland, as the Commission does not possess the relevant data (which Italy has claimed it is unable to provide — see recital 92 above). However, based on the above analysis of general price trends, which may be broadly representative of price trends in bilateral contracts, it seems possible to presume that, in the period under consideration, energy-intensive undertakings in Sardinia paid a higher price than on the Italian mainland.

In conclusion, the electricity market in Sardinia exhibits a combination of problems, including high prices, a high degree of market concentration, dominant operators’ market power, excess generation capacity in the high-cost segment, relative inefficiency of obsolete generating plants, lack of access to natural gas infrastructure, and insufficient interconnection. These problems are likely to translate into higher electricity prices for end-users, including the most energy-intensive ones.

But even if one were to suppose that there was indeed a regional handicap, the criteria laid down by the Regional Aid Guidelines would still have to be fulfilled. The Commission has accordingly assessed whether these tests are satisfied.

Aid must sustainably contribute to regional development and be proportional to the handicaps it seeks to alleviate. Operating aid in an assisted region cannot be authorised by reason of difficulties experienced by one industry, but must be shown to make a lasting contribution to regional development. In the case at hand, this is unlikely. Even if one were to accept that the maintenance on the island of some industrial production facilities would contribute to employment and to the maintenance of a manufacturing base on the island, the effects would not be lasting. If high electricity prices are a structural problem in Sardinia, as Italy claims, the difficulties experienced by energy-intensive industries will resurface as soon as the tariffs are removed.

The Commission also points out that the existence of a State subsidy aimed at reducing electricity costs for large users does not encourage electricity suppliers to bring prices down in order to avoid losing their largest customers, nor does it prevent a worsening of their cost structures. Instead, the subsidy tends to heighten the incentive for suppliers to exploit their market power. Therefore, even if were true, given the situation of overcapacity, that energy-intensive companies would normally be able to obtain a competitive price were it not for the market power of the electricity suppliers (who, according to Italy and the beneficiaries, have an interest in keeping prices high), the Commission considers that the preferential tariff is not the appropriate instrument to curtail such market power.

The subsidy granted to the beneficiaries of the tariffs is not calculated in such a way as to offset the regional disparity (i.e. the differential between electricity prices in mainland Italy and prices in Sardinia for the same category of customer), but rather aims to align Sardinian prices on those paid by the beneficiaries’ competitors in the EU. The prices set for the Sardinian beneficiaries are particularly low, and certainly below prices on mainland Italy, which are among the highest in the EU. Therefore, the tariffs are not proportional to the regional handicap they allegedly seek to alleviate.

The very method chosen to calculate the amount of the subsidy suggests that the measure pursues an industrial policy objective rather than a regional objective. Indeed, the industrial policy dimension of the tariffs has been underlined repeatedly by Italy in its submissions.

Regional operating aid must be progressively reduced (point 4.17 of the Regional Aid Guidelines). Since the updating mechanisms for the 2004 and the 2005 tariffs are different, they have been assessed separately.

As regards the 2005 tariffs, The Commission has assessed the indexation mechanism introduced by Article 11(13) of Law No 80/2005, as interpreted by the AEEG in decision 217/05 (see recital 33). According to the implementing rules laid down by the AEEG, the tariff is to be increased yearly by a percentage reflecting trends in electricity prices in the EU, but the increase cannot exceed 4 %. According to Italy, this mechanism ensures progressive reduction.

The Commission takes the view, however, that the fact that the nominal tariff is subject to a rather limited annual increase does not mean that the net amount of aid — as measured by the compensatory payments — will decrease over time.

Italy further contends that, in reality, bilateral contracts are less prone to price increases than prices on power exchanges, and that therefore the 2005 tariffs are de facto degressive. The Commission points out that the indexation method submitted for assessment is based on variations in prices on power exchanges, and that Italy has not submitted data on the evolution of bilateral contract prices. Thus Italy’s statement is not sufficiently substantiated.

In fact, the limited information at the Commission’s disposal suggests the opposite, i.e. that bilateral contract prices increased sharply in Sardinia (see, for example, Eurallumina’s statement that in 2005/2006 it paid a price of EUR 80 per MWh on the free market, up from EUR 65 per MWh in 2004). In these circumstances, had the tariffs been implemented, the compensatory payments would have increased over time.

Thus neither the 2004 tariffs nor the 2005 tariffs fulfil the degressivity requirement.

The Commission concludes that the tariffs at issue cannot be considered compatible as regional aid under the 1998 Regional Aid Guidelines. Since Sardinia ceased to be assisted pursuant to Article 107(3)(a) after December 2006, it is not necessary to examine the compatibility of the aid in the light of the Regional Aid Guidelines for 2007–2013.