Some further thoughts on Sardinia Ferry Fares or how succeed in establishing a concerted practice

This post also deals with the decision of the Italian Competition Authority (ICA) in Sardinia Ferry Fares (Case I743). More precisely, it dwells on the approach taken by the ICA to discharge its evidentiary burden to establish a concerted practice. In short, following the receipt of many reports from passengers complaining about increasing fares charged for the maritime links between Sardinia and Continental Italy, by a decision made on 11 May 2011 the ICA opened an Article 101 TFEU investigation against a number of ferry operators: Moby, Grandi Navi Veloci (GNV) and its parent companies Marinvest and Investitori Associati, SNAV and parent company, again, Marinvest, and Forship trading with the Sardinia Ferries brand. The ICA enquiry focuses on the fares charged for the links between the Sardinian ports of Golfo Aranci/Olbia and Porto Torres and the ports of Civitavecchia, Livorno and Genoa/Vado Ligure on Continental Italy.

The ICA closed the investigation with the decision made on 11 June 2011. It found that Moby, SNAV, GNV and Marinvest to have breached Article 101 TFEU as they collectively agreed to rise the 2011 Summer season fares for routes connecting the Sardinian ports of Golfo Aranci/Olbia and Porto Torres with the ports of Civitavecchia, Livorno and Genoa/Vado Ligure. The ICA qualified the conducts carried out by Moby, SNAV, GNV and Marinvest as a single complex anticompetitive agreement falling within the concept of concerted practice. Indeed, consistently with its decisional practice, the ICA inferred the existence of the concerted practice from the parallelism in the market behavior of the parties and two other relevant elements such as the absence of credible alternative explanations for such parallelism and qualified contacts between parties. This post reviews how in the reasoning of the ICA the above elements have been assessed, with a particular on the nature of qualified contacts between the parties the ICA took as evidence of the price-fixing arrangement.

A) Market parallelism

This element was found by the ICA in the parallel pricing polices followed by the parties. In that regard, the ICA conducted a time comparison test, under which it contrasted the fares charged by the parties for the 2006-2010 period with those applied for the 2011 Summer season under investigation. Interestingly, since the parties supplied highly differentiated products, the ICA took the view that it was not feasible to identify the fares effectively charged on average by each operator. Then, for the purpose of the time comparison test the ICA used a proxy for the fares effectively charged by the parties that it was found in the single average revenue per passenger. The results of the test indicated that the parties aligned their pricing policies for the 2011 Summer season fares with substantial fare increases, on average of 42%, for the Civitavecchia-Olbia and Genoa-Olbia routes and 50% for the Genoa-Porto Torres route.

Interestingly, the ICA used this proxy also in CIN/Tirrenia (Cases C11613 and C11613B) in the context of merger review. The ICA had found that the CIN acquisition of Tirrenia to restrain competition on the market for ferry services between Sardinia and Continental Italy. Therefore, it cleared the merger by imposing on CIN and its majority shareholder, Moby, a set of remedies, which, among other things, comprised freezing fares obligations. In that regard, the parties committed to apply 2012 Summer season fares to the Genoa-Porto Torres, Genoa-Olbia and  Civitavecchia-Olbia routes that were such to not increase the average per unit profit that Moby had achieved in the 2009 Summer season. Believing that the parties failed to comply with the above obligations, the ICA opened a monitoring proceedings against CIN and Moby. To ascertain whether the fares charged by the merging parties complied with the freezing fares remedies, the ICA used the concept of the single average revenue as a proxy for the fares. The data submitted by Moby and CIN indicated that single average revenues increased by a percentage much higher than the increases in fuel prices. According to the ICA, the parties’ rising fares could not be explained with higher fuel costs. Therefore, it took the view that the parties were likely to have failed to correctly implement the freezing fare remedy and it opened a monitoring procedure.

B) Absence of alternative explication

To prove that their parallel pricing policies were not the result of unlawful pricing coordination the parties put forward a number of market conditions. They contended that these conditions constitute a plausible explanation for their pricing policies other than an anticompetitive price-fixing arrangement as alleged by the ICA. The ICA carefully studied all these conditions. Drawing on the economic analysis of relevant markets, it eventually took the view that none of them could be constituted a credible alternative explanation heavily. The market conditions submitted by the parties and considered by the ICA were as follows: i) elasticity of demands; ii) degree of market transparency; iii) growing fuel costs; and iv) recovering losses.

i) Elasticity of demand

The parties contended that the demand for ferry link services was inelastic. Then, had the parties lowered fares, they would not have attracted more passengers. The ICA rejected this view. Evidentiary gathered through the investigations showed that, following the fare increases implemented by the parties, passengers travelled with the competitors that applied lower charges or they did not travel at all. Then, the ICA concluded that customers were sensitive to price increases and, logically, the decision of the parties to rise fares was not rational.

ii) Market transparency

The widespread use of revenue management systems by ferry operators since 2002 led to a high degree of transparency in market. Therefore, the parties submitted the argument that such level of transparency explained their similar pricing policies. However, this argument was dismissed by the ICA. It considered that at least until 2009 the fares were decreasing, notwithstanding reliance on revenue management systems. Hereby, the high market transparency could not be taken as a credible reason for the price-fixing practices at hand.

iii) Fuel costs

Moby and GNV claimed that they had to increases fares to compensate the losses generated by increases in fuel prices. Also this argument was considered by the ICA as ill-founded and then  dismissed. In fact, the ICA pointed out that fuel expenditure accounted at maximum for 45% of the total costs of the parties. Nevertheless, the fares increased by a measure much higher than fuel prices. Thereby, the ICA reached the conclusion that the fare increases were not proportionate to the fuel price increases and for this reason the growing fuel costs could not constitute a valid explanation for the parallel pricing

iv) Recovering losses

Finally, Moby and GNV argued that the contested price increases were necessary to recover from the financial losses they were suffering from during the previous seasons. Neither this argument was successfully. The ICA pointed to the different financial conditions of the parties. Whilst GNV operated all its routes at substantial loss, Moby profitably run the Genoa-Porto Torres and Genoa-Olbia routes, while being slightly on red ink for the Civitavecchia-Olbia route. Therefore, considering the differences in the parties’ profitability, the ICA ruled that Moby had no incentive to align its pricing policies to that of GNV which was had to recover substantial losses.

C) Qualified contacts

For the European Commission as well as for the ICA trying to prove an anticompetitive agreement finding a qualified contact between the parties can be a turning point, as it reverses the evidential burden of proof. A causal link between the contact and the anticompetitive market conducts is then presumed and it is upon to the parties to submit evidence to rebut such presumption. In Sardinia Ferry Fares the ICA succeeded in finding two qualified contacts linking the parties: i) the CIN joint-venture set up to participate in a competitive tender procedure and ii) two commercial agreements entered into by Moby and GNV.

i) The CIN joint-venture

Moby, GNV and SNAV set up the CIN joint-venture to bid for another major ferry operator, Tirrenia. Tirrenia was an ailing publicly-owned undertaking to be divested by the Italian Government that launched a tender procedure to select to whom sell it. The ICA took the view that the member of CIN when drafting their bid for Tirrenia exchanged sensitive information concerning their commercial policies. This was because the future profitability of Tirrenia also depended on the commercial policies of the parties would. Interestingly, the ICA recognized that the contacts among Moby, GNV and SNAV within the CIN vehicle did not have an anti-competitive object per se. Yet, the ICA considered such contacts as indicia of the parties talking about their future commercial policies; hereby they constitute evidence of the concerted practice. However, as has been correctly said here, quite paradoxically the parties have been encouraged by the Italian Government to form a joint-venture to bid for Tirrenia to save it from a likely bankruptcy. However weird this ruling may be, it must be said that it appeared to be consistent with the ICA decisional practice in assessing evidence of concerted practice. A similar ruling, indeed, can be found in Prezzi del latte per l’infanzia. In this case, the Italian Government invited the manufacturers of baby milk to reduce their retail prices. A trade association regrouping the milk producers called for a meeting to discuss how to coordinate their pricing policies in response to the governmental invitation. Strikingly, the ICA found a qualified contact among the parties to a long-running price-fixing arrangement.

ii) The commercial agreements concluded by Moby and GNV

The ICA looked at the code share agreement for the Civitavecchia-Olbia route and the commercial agreement for the Genoa-Porto Torres route signed by Moby with GNV. Under the agreement for the Civitavecchia-Olbia, the revenues generated by the sales of tickets were allocated to Moby and GNV on the basis of pre-agreed criteria. Such criteria did not factor in how many tickets were sold by each of the parties. The effect of the agreement was that the Civitavecchia-Olbia route was jointly run by Moby with GNV. Consequently, the parties had no incentive to compete on prices.

The commercial agreement for the Genoa-Porto Torres allowed GNV to accept reservation and sell tickets for the Moby vessels operating this route for the June-December 2011 period. Since both Moby and GNV operated this link, passengers that were unable to find a place with GNV would buy tickets from Moby, instead from the other competitors Saremar and Tirrenia also operating this link.

Also those commercial agreements did not have an anticompetitive object per se. Notwithstanding that, in the ICA view they qualified for relevant contacts between the parties. As a result, the ICA viewed them as evidence of the contested price-fixing agreement. In addition, the ICA found the agreements to facilitate the parties’ anticompetitive practices as they were instrumental in stabilizing the price-fixing agreement by frustrating the efforts of the competitor Saremar to gain clientele.

Incidentally, it may be worth noting that the two above commercial agreements were also examined by the ICA in the in CIN/Tirrenia merger case. Believing that the agreements were liable to strengthen the market power of the merged entity, the ICA imposed on Moby and CIN the obligation to terminate them by the deadline of 28 June 2012. The ICA dealt with these agreements also in the above mentioned CIN/Tirrenia monitoring proceedings as it believed the merging parties to not have comply with the termination obligation. More specifically, as for the Civitavecchia-Olbia agreement, Moby contended that the tickets sold after that deadline had been bought by clients coming from the GNV web site on the basis of a minilink agreement which was different from a commercialization agreement. The ICA briefly discussed the legal nature of the minilink agreement. It pointed out that under the minilink agreement the only obligation entered by GNV  was to direct potential customers from its site to that of Moby. Moby, on other hand, directly sold and issued the tickets to customers. Regardless of the obligation entered into by the parties to the minilink agreement, what the ICA stressed was that the termination obligation in the CIN/Tirrenia decision applied to any commercialization agreement. Therefore, also minilink agreements fell within the scope of application of the conditional authorization of the merger, with the result that Moby had to terminate the minilink agreement they had signed with GNV before the deadline.

In conclusion, the finding of a price-fixing agreement in Sardinia Ferry Fares was based by the ICA on indirect evidence, such as a thorough economic analyses of the relevant markets to dismiss the parties’ contention that such features of the markets were conducive to parallel commercial conducts. More importantly, the ICA also relied on the qualified contacts between the parties. Not only such contacts were lawful but, as in case of the CIN joint venture, they were even encouraged by the Italian Government. And, as has been explained, such strict stance in assessing the evidential relevance of contacts between the parties is consistently with the ICA decisional practice.