Kupeli & Ors V Kibris Turk Hava Yollari Sirketi (t/a Cyprus Turkish Airlines) & Ors

The Defendant’s case has been, consistently, that there was no contract or obligation to provide tickets or seat reservations to those who did not have them, and that the cancellation of tickets or reservations could be explained in each case by requests made by or on behalf of the individual Claimant for cancellation. But the Claimants did not accept this case, and looked for a reason to explain why the Claimants had not got on the Defendant’s flights. The Claimants concluded that one possibility was that the Defendant’s flights were fully or even overbooked: that was why it had not been possible to obtain a reservation with the Defendant, and why the Defendant cancelled reservations which had been made (or so the argument went).

Mr Armutlu, solicitor for the Claimants, set about finding out as much as he could about the numbers of passengers carried by the Defendant, and how the reservations system worked. He gathered evidence from two main sources. The first was the Court of Audit of TRNC; the second was from the Defendant itself.

On 23 January 2013, the Court of Audit of the TRNC produced a report on the Defendant’s claims for payment for carrying CTA passengers in reliance on the Protocol of 9 February 2011. The report was prepared by Mr Ece Celikdal, the senior auditor, and Mr Mahmut Denizcioglu, assistant to the chief auditor. It contained 23 attachments which ran to 237 pages. The report appears to have been requested by the President of the Court of Audit to review an expert report dated 1 December 2011 prepared at the request of the Council of Ministers by two “experts”, Mr Cavusoglu (Chairman of the former Board of Directors of CTA) and Mr Mehmet Darbaz (previously a director of CTA) in relation to the Defendant’s claim for carrying CTA passengers.

The Court of Audit report refers to information and evidence provided by the appropriate authorities, without listing precisely what material it considered in arriving at its conclusions. However, Mr Ersoy told me that the Defendant’s finance department produced invoices for all the claims made to the TRNC. I assume that the report writers had access to those invoices, as the starting point for their analysis. Indeed, they refer to invoices at paragraph 3.1.2 of the report, under three headings: scheduled flights, additional flights (Turkey and UK), and charter flights, which can only sensibly be understood as a reference to the Defendant’s invoices. At paragraph 3.1.4 of the report, the writers refer to a large number of files of documents provided to them concerning the claims in each of these categories, which I assume must have come from the Defendant. There are a number of references in the report to documents provided by Atlasjet, which would make sense given the subject matter of the report.

So far as flights to and from the UK were concerned, the Court of Audit concluded at paragraph 3.3.2 that the Defendant had claimed to have flown more passengers to and from the UK than it had in fact carried (the claim was for 46,086 passengers, whereas the Court of Audit thought the correct figure based on the information before it was 42,851); it had also claimed for empty seats on its flights out of and into the UK (a total of 41,649 empty seats) which was a figure not agreed by the Court of Audit who thought the correct figure was much lower.

I note that the Court of Audit concluded that claims for empty seats within Turkey (including to and from Lefkosa, also known as Nicosia, the capital of the TRNC, and denoted for aviation purposes by the letters ECN) were only due for passengers who were ticketed one-way only (paragraph 3.3.1). If that is correct – and I make no finding on whether it is or not - it means that the Defendant would not be reimbursed for empty seats on its flights to and from Lefkosa from mainland Turkey, if passengers with return tickets to the UK had disembarked in mainland Turkey (at Ankara, known for aviation purposes as ESB, or Antalya, known for aviation purposes as AYT) before the aeroplane carried on to Lefkosa.

In several places, the Court of Audit recorded that information had not been provided by the Defendant, or that the information provided by the Defendant was inconsistent with information provided by third parties such as the border control agency, which the report-writers had also obtained.

Mr Ersoy told me that the Defendant was challenging the conclusions of the Court of Audit by litigation which is ongoing in the TRNC. Mr Teker told me that the Defendant no longer has copies of the information which was provided to the Court of Audit. I find that surprising if the Court of Audit report is subject to ongoing litigation in the TRNC. Be that as it may, the documents provided to the Court of Audit by the Defendant were not produced to this Court.

It seems that the Court of Audit compiled some spreadsheets for its own purposes, and appended those to its report. We were unable to establish at trial where the data contained in those spreadsheets had come from. When those spreadsheets, some of which contained lists of passenger names by flight, were put to Mr Ersoy and Mr Teker, each distanced himself from them saying that they were not the Defendant’s documents. The Claimants, by their Counsel Mr Bradley, placed reliance on those spreadsheets as showing passenger numbers significantly in excess of the aircraft capacity, and so (it was said) they demonstrated systemic overbooking. In submissions, the Defendant’s Counsel, Mr Adkin QC, appearing with Miss Holcombe, pressed the point that the provenance of the data in those spreadsheets was not clear, and in consequence, the data itself was unreliable; he invited the Court to be circumspect about the Court of Audit documents and what they might show.

I conclude that the Court of Audit documents raise as many doubts as they resolve. I am unable to reconcile the Court of Audit spreadsheets with other evidence produced in the case (which I shall come to) and I conclude that the Court of Audit spreadsheets are not a reliable record of the passengers booked on any given flight at the point of departure of the flight. They are certainly not a reliable record of the passengers who actually flew on any given flight, given that the names of some of the Claimants (for example, Mr Ozkes) are included in the spreadsheets, and we know that those individuals did not fly – hence they bring these claims. In short, I am not quite sure what these spreadsheets show and how they came to exist. I conclude that I cannot infer from them, as I am invited to, that there was systematic overbooking of the Defendant’s flights.

The Court of Audit report and the documents produced by the Court of Audit do, however, hold some significance for the resolution of these claims. The Court of Audit was not satisfied that it had received robust data from the Defendant. It did the best with what it had. I infer that its documents reflect that “best effort”. I have found myself to be in a similar position and I too have doubts about the robustness of the data I have received from the Defendant and the reliability of the Defendant’s records, for reasons I will now outline.

In its sixth supplemental disclosure list dated 8January 2016, the Defendant included what were referred to as “Passenger Manifests” and “Passenger Lists”. On 19^th January 2016, the Claimants’ solicitor wrote to the Defendant’s solicitor referring to the sixth supplemental list and requesting sight of what were referred to as the “Passenger Named Lists” (“PNLs”) and the manifestos (referred to as the “Actual Passenger Lists” or “APLs”) for UK outbound flights. This request was met with a consent order for further disclosure by the Defendant, dated 20 January 2016, extending to PNLs and APLs. It is important to clarify the terms used: the documents being sought at this point were the Defendant’s internal documents, generated by its own systems; at the time of the request, these were the only documents recording passenger numbers of which the Claimants or their legal team were aware.

In purported compliance with the Consent Order, the Defendant served its 8^th supplemental disclosure list and accompanying documents. This revealed few further documents. Dissatisfied with the disclosure made, and still wishing to have sight of all the PNLs and APLs for the Defendant’s flights from and to the UK, Mr Armutlu issued an application for further disclosure (and other things) dated 10 February 2016. I heard that application on the first day of trial, when Mr Bradley expanded the application to encompass a request for a larger search for further documents; the reason for the application and its expansion was the Claimants’ strong suspicion, based on the PNLs and APLs belatedly disclosed, that the Defendant had been systematically overbooking flights, and then cancelling passengers’ reservations unilaterally (I was taken to a number of examples on the documents where this was said to have occurred). Mr Adkin resisted the application, on a number of bases, including the submission, on instructions, that the Defendant had already disclosed all relevant documents in its custody or control, after conducting proper searches, and there was nothing more to come. I refused the Claimants’ application. My main reason was that to allow this application would have required the trial to be adjourned, which was disproportionate and contrary to the interests of justice; but I also said that I was not satisfied that there was further relevant information or documentation which was in the Defendant’s custody or control, which had not been disclosed. We got on with the trial. The issue of passenger numbers (booked and actual) remained live, and was to be explored with the Defendant’s witnesses.

When Mr Ersoy came to give evidence, he was asked about overbooking. The following interchanges were significant:

It was clear after Mr Ersoy had given evidence that there was more work to be done by the Defendant to clarify the numbers of CTA passengers it had carried from the UK, and to produce the further documents (namely the log books) to which Mr Ersoy had referred. As it was, there was to be a 2 day pause in the trial (for unconnected reasons) and this provided an opportunity to gather this further evidence. When the trial resumed, the Defendant produced:

Mr Teker explained the new material produced by Teker 3 and reflected in the Teker spreadsheet, alongside evidence he had given in his earlier statements about the Defendant’s reservation system. The following is a summary:

The PNLs produced by the Defendant are of no utility at all in establishing passenger numbers in the period before 5 July 2010, when the “first come first served” system was in operation. That is because, as Mr Teker explained, the people who simply showed up at the airport would not have been included on the PNL. Even after 5 July 2010, the numbers on the PNLs remained significantly divergent from actual passenger numbers. I conclude that the PNLs are wholly unreliable as a guide to passenger numbers.

If the system for counting passengers who boarded the flight was robust and assiduously implemented, the numbers shown on the log books, the customs forms and the Stansted manifests should all have tallied in recording the precise number of passengers on board each flight. Unfortunately, they did not do so. I accept that the differences between these documents were in most cases small, and only occurred for some flights. But they exist.

The discrepancies in the various documents produced by the Defendant lead me to doubt the integrity of the Defendant’s systems for recording passenger numbers. I conclude that the Defendant’s documents are, in a general sense, rather unreliable.  That said, only the Defendant is in a position to produce any data at all as to the numbers of passengers booked on flights (as opposed to the actual passenger numbers, which the customs documents and Stansted manifests record), and I can only look to the PNLs to assess the numbers of passengers booked.

On the central issue, my conclusion is that some flights were overbooked. That is the inference I draw from the fact that negative figures appear in the empty seats column. This must mean that there were more passengers booked on the flight, through the Defendant’s own systems, than capacity to carry them. Specific explanations were offered for some of those overbookings: Mr Teker thought that the overbooking on 31 July 2010 could be explained by a late change of equipment which had the effect of reducing capacity, for example. But Mr Teker struggled to explain other negative numbers; he speculated that the difference might be attributable to the delay between data transferring from the reservations system to the inventory system (from which the Teker spreadsheet PNL figures had been sourced). But even accepting that there was such a delay, that still means that at a given point in time, the Defendant’s systems recorded more passengers being booked on a flight than there was capacity to carry them and so this is not a coherent explanation. Although those overbookings are in quite small numbers (mostly single figures), their existence leads me to doubt Mr Teker’s and Mr Ersoy’s assertions in evidence that flights were never overbooked and could not be overbooked on the system: they plainly could be and at times were.

As to actual passenger numbers: on the vast majority of flights, there were in fact empty seats, but only very few on the flights departing the UK after 6 July 2010 until early August 2010 when a different pattern with more empty seats can be seen to emerge. On the flights on which the Category 1 Claimants were booked, there were in the end empty seats, in sufficient numbers to have accommodated each of the Category 1 Claimants with their families if they had wanted to travel on those flights. However, it is not true that all flights had empty seats; some flew full to capacity, and I reject Mr Ersoy’s evidence that all flights flew with empty seats.

I have not attempted to reconcile the Teker 3 “actual passenger” or “empty seat” figures with the Court of Audit figures for the claims made by the Defendant. Mr Armutlu has tried to perform that exercise, or something like it, by comparing Court of Audit figures for ticketed passengers with the aggregate number of passengers booked on each flight (the Court of Audit figures plus the Defendant’s figures for “own sales”), compared with the capacity of each flight, as shown in Teker 3. Mr Armutlu says that this demonstrates that the flights were persistently and significantly overbooked. But Mr Armutlu’s submission is dependent on the integrity of the data produced to and ultimately by the Court of Audit. That is a matter which the Defendant disputes and which I have been unable to resolve. Further, the Court of Audit passenger figures are seriously discrepant with the log books, customs forms and Stansted manifests. What is clear is that if the Defendant has claimed payment from the TRNC for passenger numbers as shown in the Court of Audit documents, then Mr Teker’s recent research and analysis suggests that the Defendant has overclaimed payment, and the actual passenger figures were much lower.

In summary, I find no evidence of systematic overbooking of flights although some flights were overbooked. But the various documents disclosed by the Defendant which purport to record passenger numbers disclose many inconsistencies. This leads me to conclude that those documents, some or all, are unreliable in material respects, and that this is, I conclude, indicative of a wider problem with the Defendant’s systems for recording passenger numbers and details.

In his first witness statement, Mr Teker described the system for making reservations. He said that when a flight reservation was made, a Passenger Name Record (“PNR”) was generated by the Defendant’s system. The PNLs listed those passengers with confirmed reservations on the flight (each of whom had a PNR, it follows). He said, amongst other things:

I was left with the impression that Mr Teker went into the reservation system and changed entries on that system pretty much at will. Mr Teker reacted to circumstances as they arose, and took what he perceived to be corrective action by cancelling or changing flight reservations where he thought that was appropriate. It was not his invariable practice to check with the passenger or the travel agent before making changes. Because of the defective notification system, any such changes were not notified to the passenger or the agent and in truth remained “internal” to the Defendant. To be fair to Mr Teker, he had been extremely hard pressed in the weeks after the Defendant started carrying CTA passengers and I accept that he was just trying to do his best to cope with the many requests for replacement flights which he tried to process. But the system was disorganised, and mistakes got made. I reject Mr Teker’s evidence that he would never cancel a confirmed reservation without an express written request from the passenger or agent: there are examples within the documents of him doing precisely that. More generally, I conclude that he reacted to events, sometimes in a rather haphazard way, and I can place little faith in what he now says would have been his “usual practice”.

Before considering the nature of the Defendant’s relationship with each category of Claimant, it is necessary to clear the ground of some of the larger assertions advanced at various stages in this litigation as to the basis of the Defendant’s liability. First, it was argued that the Defendant entered into a partnership with CTA and thereby assumed all CTA’s liabilities. This argument was advanced relying on certain press reports referring to a “partnership” between the Defendant and CTA, and assertions by some Claimants that they had been told by their travel agents of this partnership. It is clear that there was no partnership or joint undertaking between the Defendant and CTA. I accept Mr Ersoy’s evidence on this point. I conclude that the press reports of a partnership, and suggestions by travel agents that there was a partnership in existence, were simply wrong. It is possible that CTA encouraged the perception of some sort of partnership by its language of “cooperation scheme” on its website on 21 June 2010; but if so, again, that was just a misperception. The Defendant was at all times a legal entity separate from CTA. It never assumed CTA’s liabilities.

The second argument, advanced at an earlier stage of the litigation but abandoned by the time of trial, was that the Claimants derived rights against the Defendant from the Protocols. This was not correct: under the Protocols, the Defendant was contractually bound to the government of the TRNC to provide replacement flights to the CTA passengers, and the Claimants had no rights as third parties.

Thirdly, the Claimants argued that they had a bipartite contract with the Defendant, consisting of the Defendant’s promise to fly the Claimants on its flights to and from Turkey. It is that assertion which is at the heart of this case, and to which I now turn. The analysis varies according to the circumstances of each category of claimants. I shall consider the contractual position for each category generically, before turning to the individual cases.

The Category 1 Claimants all had confirmed reservations to travel on the Defendant’s flights. Some of them had tickets reflecting those reservations; others did not, but had reservations (at one stage) on the Defendant’s systems. The issue with this category is not whether these Claimants had a contract with the Defendant – it is accepted in principle that they did – but rather, (i) when that contract came into existence; and (ii) whether, on the facts of each individual Category 1 case, the cancellation of the flight reservation was a breach of contract by the Defendant.

On issue (i), the timing of the contract, Mr Teker gave important evidence which was not challenged. He described the booking procedure. When a flight reservation was made, a Passenger Name Record (“PNR”) was generated by the Defendant’s system. Mr Teker said this:

“once a customer has a PNR and a confirmed booking that person will be entitled to board the relevant Atlasjet flight”.

He thereby acknowledged that the entitlement to board was consequent on a passenger having a confirmed booking.

Mr Teker confirmed that

“the KTHY customers were permitted to cancel their re-booked flights with Atlasjet in order to change the time, date or destination”.

He thereby acknowledged that CTA passengers who were rebooked on the Defendant’s flights were permitted to alter a booking without penalty (subject, of course, to availability on the chosen replacement flight(s)), and there was flexibility on their side. However, Mr Teker also maintained that he would not cancel or amend any booking without having received a specific request to that effect from the agent or passenger; I have addressed that evidence above (and rejected it as evidence of Mr Teker’s practice, in fact) but I accept that evidence as representing the position as it ought to have been, and as was understood by both parties as part of the agreement, namely that once a booking was made, the Defendant had no unilateral right to change it (possibly, unless there was some unavoidable force majeure or crisis – this caveat was not the subject of argument and not relevant on the facts of the individual cases before me).

I conclude that a contract came into existence between the Defendant and the CTA passenger with a confirmed booking on one of the Defendant’s flights at the time that the booking was confirmed, not before, and not after. It was at that point that the Defendant accepted the passenger’s request for a replacement flight, the Defendant became obliged to provide a seat on a specific flight to a particular destination, and the CTA passenger surrendered his or her rights to claim against CTA. The surrender of that right was the consideration for the Defendant’s contractual commitment to carry that passenger on one of its flights; no money changed hands when the contract between the CTA passenger and the Defendant came into existence, a peculiarity explained by the context of these operations.

Mr Adkin sought to persuade me that no contract came into existence until a ticket was issued, and to that end he relied on Dillon v Baltic Shipping Co (the “Mikhael Lermontov”) [1991] 2 Lloyds LR 155, a decision of the Court of Appeal of New South Wales (subsequently reversed on other grounds). In that case, the booking form (for a cruise) expressly stipulated that the contract only came into existence on issue of a ticket and subject to the terms and conditions contained on that ticket; the Court of Appeal could find no justification for disregarding the stipulation in the booking form and confirmed that the contract came into existence when the ticket was issued to Mrs Dillon (a conclusion which led to Mrs Dillon succeeding in her action against the cruise company). The Court of Appeal acknowledged that this was a conclusion different from the “conventional analysis” where the contract comes into existence only after the ticket has been issued (by way of offer) and the recipient has had an opportunity to consider and accept the terms of the offer; it was also a conclusion different from those cases where the contract comes into existence when the booking form was issued or the balance of fare was paid (the Court cited Oceanic Sun Line Special Shipping Co Inc v Fay (1988) 165 CLR 197 as an example of the latter). In my judgment, the Mikhael Lermontov is authority for the proposition that the timing of any contract for carriage is fact-specific, and depends on the particular circumstances of the particular case.

In this case, Mr Teker’s evidence was that his practice was:

“I would send the Atlasjet ticket, or more often an email confirming the new PNR and the flight details, to the travel agent.”

It is clear that tickets were not invariably issued to passengers with confirmed bookings, and that the issue of the ticket did not assume any particular importance for either party. I conclude that the issue of the ticket was not, on an objective analysis, the point at which the contract came into existence. The critical moment, when the CTA passenger’s status changed from being a “hopeful applicant” to being a passenger with a contractual right to be carried on the Defendant’s flights, was when the booking was confirmed on the Defendant’s system. This was an electronic “event” not necessarily accompanied by the issue of any separate document of entitlement in the passenger’s name. Although emails should have been sent according to Mr Teker’s practice, this was not invariably done (there are examples in the papers where no email was sent, or so it appears), and when it was done, it was done at a time convenient to Mr Teker, in order to confirm the reservation which had already been made; the emails were not themselves contractually significant.

I conclude that once a CTA passenger had a confirmed booking on the Defendant’s system, that passenger had a contract enforceable against the Defendant. My analysis parts company with Mr Adkin’s analysis only to the extent that I identify the confirmation of the booking, rather than the ticket itself, as being the relevant moment when the contract came into existence. Each of the Category 1 Claimants had a reserved booking and (I conclude) they had a contract with the Defendant. I shall deal with the remaining issue for Category 1 Claimants, namely whether there was a breach of contract by the Defendant, below.

The Claimants in Category 2 all turned up at the airport during the period governed by Regulation 1 (23 June 2010 to 4 July 2010), only to find they were unable to board a CTA flight. They contend that this amounted to a breach of contract by the Defendant. The Claimants’ argument goes thus: the Defendant asked a number of travel agencies to “spread the word” that passengers with CTA tickets should go to the airport where they would be given replacement flights by the Defendant; in delivering that advice to the Claimants, the travel agencies were acting for and on behalf of the Defendant; that advice amounted to a unilateral offer by the Defendant to CTA passengers generally (from which class the Claimants are all drawn), communicated to them via the agents acting on the Defendant’s behalf; that offer was open for acceptance by the Claimants’ performance in attending the airport with their CTA tickets to hand, without more; when the Claimants did attend, as advised, they accepted the Defendant’s offer and the Defendant was in breach of contract in failing to provide air transport for them.

There are two problems with this argument. The first and fatal problem is that there was no “unilateral offer”, made by the Defendant with the serious intention that if accepted (by the act of going to the airport) it would give rise to a contract binding on the Defendant to fly the CTA passenger to a destination in Turkey or TRNC. Chitty on Contracts, 3d Ed, defines an offer is “an expression of willingness to contract on specified terms made with the intention that it is to become binding as soon as it is accepted by the person to whom it is addressed” (para 2-003). It is difficult to see what the specified terms of this offer might have been: the Defendant was putting on ad hoc flights at this early stage, so there was no particular flight on offer to any CTA passenger; the Defendant was not flying to all destinations previously served by CTA, so that any given CTA passenger might find that the destination for which he or she was ticketed on CTA was simply not served by the Defendant, or if it was, that the flights to that destination were full and only a different destination was on offer; further, the Defendant was trying to clear a backlog of CTA passengers which had built up since CTA stopped flying on 21 June 2010 and any offer to provide replacement flights had to be understood as being subject to space being available. None of these caveats needed to be stated: they were or should have been obvious to any reasonable observer.

Regulation 1 was not made with the intention that the invitation to CTA passengers to come to the airport should give rise to a contract binding on the Defendant to fly that passenger to the destination of choice on the date of choice, the moment that the CTA passenger got on a bus or tube to come to the airport: it would be beyond common sense for any entity to offer to contract on that basis. It was merely CTA’s advice, communicated via the travel agents, about how to get on a flight if a suitable flight was in fact available for that CTA passenger and if the CTA passenger wanted to take that chance.

The second problem is that in disseminating this advice, the travel agents were not acting as the Defendant’s agents or intermediaries. They were not in a position to bind the Defendant contractually, even if the “advice” could be seen as an offer with contractual intent. The Category 2 Claimants’ cases proceeded on the basis of various documents which were said to give authority, actual or ostensible, to the travel agents to bind the Defendant contractually. These documents included:

None of these documents or statements amount to an appointment of the travel agents as the Defendant’s agent, capable of binding the Defendant contractually. Nor do any of these statements come close to suggesting that such an agency arrangement is already in place so as to clothe the travel agents with ostensible authority. Therefore, the Claimants’ agency argument fails.

But in any event, I fail to see how the agency argument assists the Category 2 Claimants: if the Defendant was making an offer by Regulation 1, which offer was capable of acceptance by the CTA passengers by their performance in turning up at the airport, it would not matter whether the CTA passengers learned of that offer via their travel agent, heard about it directly from the Defendant, or heard about it in some other way (eg a newspaper report). The analysis would surely be that it was an offer to the class of CTA passengers who had been failed by CTA generally, and was open to any member of that class to accept by performance. The method of communication of that offer is irrelevant. (The flipside must be that if it was not an offer of this type, then it did not become one simply by repetition to CTA passengers via intermediaries or agents.)

In summary, there was no contract between the Category 2 Claimants and the Defendant.

I wish to make two further comments in relation to the claimants who fall within Category 2: first, it is quite clear that Regulation 1 caused a number of CTA passengers a great deal of upset and disappointment. Regulation 1 was obviously problematic, because it depended on the CTA passengers actually turning up at the airport, ready to fly. There is plenty of evidence in press reports and the Claimants’ witness statements to demonstrate that many CTA passengers failed to get on flights and were sent home from the airport. This situation was chaotic and unsatisfactory.

Secondly, it is also clear that the Defendant did carry a number of passengers on its flights in this early period. Those passengers were booked onto the Defendant’s flights in one of two ways: by turning up at the airport, and finding (happily) that there was availability on a flight; or by obtaining a confirmed reservation by email contact with the Defendant in advance of going to the airport. In either case, a contract between the Defendant and the particular passenger(s) came into existence, at the point that their booking on a particular flight was confirmed by the Defendant. Mr Adkin is right to contend that at that point, and not before, the CTA passenger surrendered his or her right to claim against CTA for CTA’s failure to provide transport services on the specified CTA flight, and that surrender was the consideration by the passenger for the new contract with the Defendant. The disappointed Category 2 Claimants therefore still have a right to claim against CTA, although this may be of little comfort to them.

The Category 3 Claimants were assured that they would get flights with the Defendant, but the flights never materialised. Save for an aborted booking in the case of Mr Yuzidik, addressed below, the Defendant has been unable to find a confirmed booking for any of the Category 3 test Claimants on its systems.

It follows from my analysis of Category 1, that unless the Category 3 Claimants can show that they did in fact have a confirmed booking (in which case they would become Category 1 Claimants), they did not have any contract with the Defendant.

The contractual analysis for the Category 3 test Claimants is little different from Category 2: there was no confirmed booking, and thus no contract.

The Category 3 test Claimants are doubtless disappointed that they did not get replacement flights from the Defendant. But the Defendant had no contractual liability to those Claimants to provide replacement flights. They too retain their right to claim against CTA.

There is in the end a single feature which divides all the Claimants into two (not three) categories: whether they did or did not have a confirmed booking with the Defendant. Those with a confirmed booking have established a contractual relationship with the Defendant (Category 1); those without have not (Categories 2 and 3). Against that background, the CARI Button can be seen merely as a means of access to the Defendant’s system, and it is not necessary to examine that facility further or decide which agents had access to it and which did not.

I will consider the Category 1 Claimants first in order: only they can establish any contract with the Defendant, and so in their cases it is necessary to determine on the facts why the confirmed bookings did not materialise, and whether the Defendant is in breach of contract. Given that there is no contract in relation to Categories 2 and 3, the cases falling in those categories can be considered more briefly, because those claims must fail.

Mrs Atesogullari purchased four tickets for herself and her family to fly to and from Adana on 13 July, returning in September 2010, on CTA. She purchased her tickets from Istanbul Travel. When she heard that CTA had stopped flying, she went to Istanbul Travel and was told that the Defendant would re-arrange the flights, but these would not be to Adana but to Ankara or Antalya, and she would have to make her own arrangements for onward travel. She agreed. On a later visit to Istanbul Travel, she was told that the Defendant had emailed with availability for four people on two different flights from Stansted to Ankara return, one departing on 14 July 2010 and returning on 1 September 2010, the other departing on 16 July 2010 and returning on 4 September 2010. Mrs Atesogullari accepted the first option, departing on 14 July 2010. The travel agent wrote the times of the new flights on her CTA tickets.

Mrs Atesogullari had not returned to see her family in Turkey for 10 years and was very pleased to be going home. She was taking her three children with her. Two of those children have been raised by her although she is not their natural mother; she wanted to take them to visit their natural mother’s grave in Turkey. That was the reason for the visit.

On 14 July 2010 Mrs Atesogullari travelled to Stansted by taxi with her three children. She presented her CTA tickets with the annotations handwritten on those tickets by the travel agent. She was told that there was no reservation. Her English is not good and her 15 year old son translated for her.

She was then asked some other questions, and was taken to a room for questioning by security officers who were suspicious that she had come to the airport without a proper ticket and without a reservation. Her children became very upset when she was taken into the interview room. She asked for an interpreter (and got one) because her son was no longer with her. The interviewers released her once, they said, they had spoken to Istanbul Travel who confirmed that she was due to travel that day.

Mrs Atesogullari then purchased three tickets to travel with Turkish Airlines, at a cost of £1,700. She left her eldest daughter behind because she could not afford a ticket for her.

The documents produced by the Defendant show that on 9 July 2010 at 09.44, Mrs Atesogullari and three others in her party were booked on flights to and from Ankara departing on 14 July 2010, returning on 13 August 2010 (the mismatch in the return date was never explained: Mrs Atesogullari thought she was coming back on 1 September 2010, as noted by hand on her CTA ticket). There are two entries in the Defendant’s reservation system to evidence the booking, both created by Mr Teker: one indicates a “pre-booking”; the other indicates a “Confirmation from booking” and says “PNR created”. A third entry at the same time shows “20 EUR Service Fee collected from each Passenger” (although there was no evidence to suggest that in fact any money was collected from Mrs Atesogullari at this point or at any point). Other documents produced by the Defendant show Mrs Atesugallari’s booking accompanied by a specific reference number associated with each passenger in her party under the heading “Bilet No”.

From these documents, I conclude that Mr Teker made a confirmed booking for Mrs Atesogullari and her family to travel to Ankara on 14 July 2010.

That confirmed booking was cancelled at 18.08 on 12 July 2010 by Mr Teker (shown on the relevant document as the “user”), with the name “Sevgi Saritas” noted in the “RCVD” column. The system appears to have generated an SMS message to “11”, which (as we know) was a fake number. It therefore seems that the cancellation was never, in fact, notified to Istanbul Travel or to Mrs Atesogullari.

When Mr Teker came to investigate why this booking had been cancelled, he concluded that Sevgi Saritas (the name of a travel agent working for Istanbul Travel) had requested that cancellation, but was unable to find any email requesting that cancellation. On 10 July 2010, Mr Teker made a second booking (which was similarly confirmed and subject to a note about a service fee collection) for Mrs Atesogullari and her family to travel to Lefkosa (ECN) via Ankara on 16 July 2010 and returning on 4 September 2010. This too is evidenced by separate documents showing a unique reference number in the column headed “Bilet No”. (Some of the documents produced by the Defendant show that the return leg was in fact booked for 4 August 2010 via Antalya rather than Ankara: these are mismatches which were not explained.) That confirmed booking was cancelled at 18.10 on 12 July 2010 by Mr Teker (shown on the relevant document as the “user”), with the name “Sevgi Saritas” again noted in the “RCVD” column. The system appears to have generated an SMS message to “11”, which (again, as we know) was a fake number. It therefore seems that this cancellation was never, in fact, notified to Istanbul Travel or to Mrs Atesogullari either.

As at 10 July 2010, I conclude that Mrs Atesogullari and her family had confirmed bookings on two return flights from Stansted, one on 14 July 2010 and one on 16 July 2010. The Defendant was not entitled to cancel or alter either one of these confirmed bookings without express authority from Mrs Atesogullari.

The central issue of fact in Mrs Atesogullari’s case is why the confirmed bookings were cancelled. Mr Bradley maintains that it was a unilateral cancellation in breach of contract by the Defendant; Mr Adkin contends that it was a cancellation at the request of Mrs Atesogullari or her agent.

I accept Mrs Atesogullari’s evidence that she did not want or ask for either booking to be cancelled (she did not even know about the second booking for 16 July). She had good reasons for wanting to travel to Turkey, she could not afford to lose the four flights, even if the replacements were not for her destination of choice, and she would not have gone to the airport with her children if she had cancelled the flights for that day.

Ms Saritas was not called to give evidence, but I daresay she would not remember Mrs Atesogullari or the details of this transaction at this distance in time. I daresay she would also tell me that her usual practice would have been to cancel flights only when specifically asked to by her client. I was shown an email from Ms Saritas dated 10 July 2010 requesting flights for 16 July 2010. It may be this email which prompted Mr Teker to make the second booking for 16 July 2010. But this email did not request cancellation of the 14 July 2010 reservations; nor, evidently, did Mr Teker understand it to be a cancellation request because he did not cancel the 14 July 2010 booking on 10 July 2010 when he made the second reservation. He left both sets of bookings in place.

In his statement Mr Teker said he could remember Sevgi Saritas requesting this cancellation. But when he came to give evidence he confirmed that he had no such recollection. He was just making assumptions based on what was recorded in the “RCVD” column.

On the balance of probabilities, I conclude that there never was a request from Ms Saritas to cancel the 14 July 2010 booking. There is just no coherent reason why Ms Saritas would make such a request when that is not what her client wanted. Further, there is no evidence of such a request being made – no email has been recovered requesting cancellation (even though one other email from Ms Saritas, not requesting cancellation, has been recovered); and Mr Teker’s annotation in the RCVD column does not refer to any email in terms.

I am asked to prefer Mr Teker’s evidence as to his usual practice, first that he would never cancel without an express request from the passenger or their agent in writing, and secondly that by writing “Sevgi Saritas” in the RCVD column Mr Teker was recording the fact of a specific request from that agent. I reject Mr Teker’s evidence on this matter. I have already indicated my conclusion that I can have little faith in Mr Teker’s claimed usual practice (see above). Further, the words entered in the RCVD column are inconclusive and could just mean that Ms Saritas was noted to be the travel agent handling this passenger’s case, or that she should be notified of the cancellation (although I cannot see that she was ever notified, in fact); I am not persuaded that her name signifies that any request had in fact come from Ms Saritas.

I conclude that Mr Teker cancelled the reservations without instructions to do so. I am unable to say why he did that. But I have noted elsewhere the Defendant’s and Mr Teker’s rather chaotic management of CTA passengers. Perhaps Mr Teker, pressed as he was to find spaces on flights in this busy period in the middle of July, realised on the evening of 12 July 2010 that Mrs Atesogullari was holding two sets of bookings when she could not possibly use both, so he decided to cancel them both in order to be able to offer those seats to others; perhaps he just made a mistake: either explanation would fit the facts as they have emerged.

My conclusion on the narrow issue of fact in Mrs Atesogullari’s case is that Mr Teker unilaterally cancelled Mrs Atesogullari’s confirmed booking, without any authority from her or her agent to do so. The Defendant, by Mr Teker, its employee, thereby breached its contract with Mrs Atesogullari.

Mrs Atesogullari’s claim against the Defendant succeeds.

Mr Ozkes purchased twelve tickets from CTA for himself, his family and the Ceritli family to travel on 12 July 2010 from Stansted to Gaziantep in Eastern Turkey, returning from Gaziantep to Stansted on 30 August 2010. He paid £2,429.40. He purchased these tickets from two agencies, Sun El Travel and Istanbul Travel.

When he heard in early July that there was a problem with CTA, he tried to call the agents but the lines were busy. He went to Sun El Travel, which agency referred him to Istanbul Travel. At Istanbul Travel he found a large crowd of people, who were angry. He finally spoke to an agent at Istanbul Travel who told him that the Defendant would arrange replacement tickets for a new destination, because the Defendant did not fly to Gaziantep; the agent told Mr Ozkes that his party would have to fly to Lefkosa (ECN). The Defendant would then fly the party to Ankara (ESB), but from Ankara Mr Ozkes and his party would have to make their own way to Gaziantep. Mr Ozkes accepted this.

Mr Ozkes went back to Sun El Travel a day or two later, because he had still not received any tickets from Atlasjet. The agent called Atlasjet in Mr Ozkes’ presence, and reported back that Atlasjet’s representative had said that the Ozkes party was booked to fly the next day, and that the tickets would be available later that day. When Mr Ozkes returned later that day to collect the tickets, he found that the tickets were in fact for travel to Lefkosa on two different days, 14 July 2010 and 16 July 2010. The agent said that it was too late to try to change the tickets to allow the party to fly together on one day. Mr Ozkes accepted the tickets. These “tickets” were in the form of two printed pages, listing the passengers and showing the flights outbound on 14 and 16 July 2010 respectively, inbound on 2 and 4 September 2010 respectively. The page showing the 14 July 2010 flight has a handwritten annotation which appears to show the time of an onward flight from Lefkosa to Adana (ADA).

He went to Stansted with his tickets the next day, 14 July 2010, with six members of his party who were intending to fly on that day. At the check-in desk, he was informed that there was no reservation on the system and that he would have to go back to his travel agent who should call Atlasjet. His party were refused boarding. The check-in desk staff could not give him any information about the position for the remaining six members of his party who held tickets for 16 July 2010. He says he saw others being turned away from the flight too.

Mr Ozkes went back to Sun El. He asked for a refund of the cost of his CTA tickets but was refused and told that he had to fly with Atlasjet. The agent said he could not guarantee when he would get another flight with Atlasjet because many people were trying to fly. He asked the agent whether the tickets on 16 July 2010 were valid, and was told that those tickets could not be guaranteed and that the passengers should simply go to the airport on 16 July 2010. On the advice of the agent, Mr Ozkes booked six tickets and reserved another six tickets with Turkish Airlines for 21 and 22 July 2010 to and from Adana, which is not far from Gaziantep. He paid cash on 14 July 2010 for six of those tickets; and held the remaining reservations in case the rest of the party was unable to fly with Atlasjet on 16 July 2010.

Mr Ozkes went to the airport again with six passengers on 16 July 2010, with his tickets for the flight on that date. He was again told at the check-in desk that there was no reservation in his name and that he should go back to his agent and re-book the tickets. The six passengers were refused boarding.

Mr Ozkes went back to the agent and paid for the remaining Turkish Airlines tickets. As a result, he paid an additional £4,832.60 to fly the whole party of twelve to and from Adana on 21 and 22 July 2010 returning on 11 or 30 September 2010.

The Defendant’s case is that Sun El Travel requested replacement flights for the six flights booked through it, for 16 July 2010 (outbound) and 2 September 2010 (inbound). I have seen that email request which was timed at 14.25 on 10 July 2010. That request was verified by the Defendant as relating to CTA passengers (marked by a tick on the hard copy of the email), but the return date was manually overwritten with 4 September. On 10 July 2010, timed at 15.02, Mr Keksin, employed in the Defendant’s Route Management Department, booked five or six flights (not clear from the Atlasjet documents) for 16 July returning on 4 September 2010, from Stansted to Lefkosa via Antalya.

Mr Teker cancelled that reservation on 11 July 2010. He wrote in the free text box of the reservation system “HATALI TARIH” which means wrong date. There is no email chain or other evidence that the cancellation was notified to the agent or to the passengers. The only evidence of any notification is the entry in the reservation system “RCVD” column that an SMS message was sent to “11”. But as Mr Teker confirmed in evidence, this number was a fake number.

On 19 July 2010, Mr Teker created two new bookings (one for five members of the Ozkes family, one for a single member of the family – together comprising the party that had booked with Sun El Travel) to fly one way from Lefkosa to Stansted via Antalya on 4 September 2010. There is no evidence that this booking was notified to Mr Ozkes or his agent.

So far as the group booked with Istanbul Travel is concerned, Mr Keksin booked replacement flights for 14 July 2010 to travel to Lefkosa via Ankara, returning on 2 September 2010. Mr Teker cancelled these on 14 July 2010, writing in the free text “RCVD” column “Istanbul Travel”. There is no evidence that this cancellation was notified to anyone, beyond the SMS message to “11”.

My conclusions are as follows. I accept Mr Ozkes’ evidence that he accepted the flights that he was offered, even though they were not on the dates that he had originally wanted, even though the party was split into two sub-groups, and even though the flights were not to the destination desired. He had paid a lot of money for the group to travel, and on any view, getting to Gaziantep from wherever he was taken by Atlasjet would have been cheaper than having to buy the tickets to travel from London, in high season, afresh. Mr Ozkes struck me as a truthful man, and I reject the Defendant’s case that he is lying about his acceptance of those tickets or in any other way putting forward a false version of events. I can see no reason why Mr Ozkes would attend the airport on two occasions with his family, if in fact he knew that the flights had been cancelled at his request. I find I can place very little reliance on Mr Teker’s version of events or the documents advanced by the Defendant as evidencing that version of events, and I reject Mr Teker’s evidence to the extent that it is inconsistent with Mr Ozkes’ evidence. I conclude that Mr Teker cancelled Mr Ozkes’ confirmed bookings without any request to do so, and in breach of contract.

There are three final points to make in relation to Mr Ozkes’ case. First, it is not clear to me why Mr Teker bothered to make a one-way booking from Lefkosa to Stansted on 19 July 2010, for 4 September 2010. The Ozkes party had no outbound flight at that time with the Defendant; and there was no reason at that stage for Mr Teker to believe that he or they wanted the return leg of the flights which had been cancelled by him a few days earlier. Having made this booking, Mr Teker did not apparently try to communicate it to the agents or the passengers (no evidence to that effect has been produced by anyone) - Mr Ozkes was by that date in Turkey, having flown there at considerable cost with Turkish Airlines. The fact that the Ozkes family were recorded as “no shows” on 4 September 2010 in Lefkosa is to misdescribe the position: the Ozkes group had no idea they were even booked on those flights.

Secondly, it is unclear whether the Defendant has claimed for Mr Ozkes’ bookings from the government of TRNC. If it has, this would be incorrect. I say this because Mr Ozkes and his family are listed as having travelled with the Defendant on 14 July 2010 in the documents which have been obtained from the Court of Audit (see above). I am unclear about how Mr Ozkes’ name came to be included within the list, or what that signifies. It is clear that Mr Ozkes did not fly with the Defendant. Further, if any claim has been made for the “no show” for the return leg on 4 September 2010, then for different reasons this is wrong given that Mr Ozkes did not even know about the reservation on that date.

Thirdly, Mr Ozkes says that he was compelled to fly to Lefkosa (via Ankara), and that he was assured that from Lefkosa he would be flown to mainland Turkey (Adana) from where he could make his way to Gaziantep. There may be a perfectly benign explanation for the complexity of the proposed route (for example, that flights direct to Adana were by this time full). But there is some evidence to suggest that Atlasjet was not permitting its passengers to disembark in mainland Turkey, namely, an internal email exchange referring to the Defendant’s “loss of revenue” resulting from passengers getting off in mainland Turkey and leaving empty seats to Lefkosa. I am not able to make any finding, in this case, as to why it was proposed that Mr Ozkes should fly to Lefkosa via Ankara then to board another plane to Adana. I simply record that point of confusion here.

Mr Ozkes succeeds in his claim for breach of contract.

Mrs Akguc booked to fly with her two children from Stansted to Gaziantep on 26 July 2010 returning on 7 September 2010 on CTA. She purchased the tickets through Tulip Holidays, since 2 July 2010 (and at all times material for this claim) an authorised agent of the Defendant.

When she heard that CTA was in trouble, she contacted Tulip Holidays. She was initially told simply to go to the airport, advice which she did not accept; she subsequently heard that the Defendant would issue replacement tickets and asked Tulip Holidays to arrange that for her. She understood that she would not be able to fly to Gaziantep but was willing to fly to Ankara.

On 14 July 2010, Tulip Holidays booked flights for the three passengers to Ankara for 18 August 2010 to return on 17 September 2010. Mrs Akguc noticed the different dates and explained to the agent (from whom she collected the tickets on 14 July 2010) that the return date was no good because she had to be back with her children in time for the start of the school term in September.

Mrs Akguc says that she received a phone call from the agency the day before her flight saying that the flights had been cancelled and that she should not go to the airport as planned. The agent said that a new ticket would be arranged, but she said that she did not want a new ticket but wanted a refund, which the agent said was not possible. She did not travel to Turkey that summer and remains out of pocket.

The Defendant’s evidence demonstrates that her flights were cancelled on 16 July 2010 by Mr Salih Ince (shown in the user column of the screen shot), an agent working for Tulip Holidays. Mr Bradley argues that because Tulip Holidays is an authorised agent for the Defendant, the Defendant bears legal responsibility for Tulip Holiday’s errors (assuming for argument’s sake that this is one). Mr Adkin argues that the agency agreement does not have that effect, and extends only to providing Tulip Holidays with access to the Defendant’s reservations system; it does not have the effect of putting the Defendant in Tulip Holiday’s shoes for all purposes.

Mr Adkin is surely right. The Defendant has not assumed responsibility for all of Tulip Holiday’s liabilities. Tulip Holidays acts as an agent for the Defendant for certain defined purposes, but otherwise remains an independent entity, with separate liability.

In this case if the flights were wrongly cancelled, I conclude that on the balance of probability the mistake was made by Mr Ince, an employee of Tulip Holidays. That is because, first, he is recorded as the “user” on the Defendant’s reservations system at the time that the cancellation was entered (by contrast with other individual cases where the user was Mr Teker or someone employed by the Defendant). But secondly, there is a coherent explanation to be inferred for Mr Ince’s actions in cancelling Mrs Akguc’s flights: Mrs Akguc had to be back in England for her children to start school in early September 2010 and I am sure she made that position plain to Mr Ince (or whoever she dealt with at Tulip Holidays – she thought it was the owner, Cadri, now sadly deceased); the probable explanation for the cancellation must be that although Mr Ince could get the Akguc family out to Ankara in August, he could not get them back before the school term started, because by the middle of July when he was looking for flights, all the Defendant’s return flights, arriving back at Stansted before school term started, were full. That explanation is consistent with the fact that Mrs Akguc was given tickets for a return on 17 September 2010 in the first place (from this I infer that there was nothing available any closer to her original flight return date on 7 September 2010). I conclude that the flights were cancelled because they were unsuitable and suitable replacements could not be found: this was to act in accordance with Mrs Akguc’s instructions, even if this position was not fully explained to her at the time.

Mrs Akguc’s claim against the Defendant must fail.

Mrs Harman purchased three tickets through Diplomat Travel to fly with CTA from Stansted to Gaziantep on 29 June 2010, returning on 21 September 2010. When she heard that CTA was in trouble, she tried to contact Diplomat Travel but the lines were busy. So she went to the office instead. She found a crowd of people outside the agency all waiting for news about CTA. Someone from the agency then made an announcement that the Defendant had signed a “partnership agreement” with CTA and the Defendant had taken over all CTA’s flights. That person said that the CTA tickets would not be changed and the passengers should simply go to the airport at the time stated on the CTA tickets and they would be able to fly with the Defendant. Mrs Harman went to the airport on 29 June 2010 with her children and her CTA tickets. She went to the CTA desk which had a sign showing the Defendant’s name on it. The staff member at that desk told her that there were no flights and all the CTA flights were cancelled. Mrs Harman went back to Diplomat Travel. An agent there told her that the Defendant was in process of rescheduling additional flights to take CTA’s passengers and that she should wait for further information. She never got any more information; no one called.

The Defendant relies on Mr Beyzade, the owner of Diplomat Travel, who denies that there was ever a crowd outside his travel agency, or that his staff would ever advise a client simply to go to the airport without a valid ticket to get on a flight. But secondly, the Defendant stresses that Diplomat Travel was not one of its agents and so was not in a position to represent the Defendant in any dealings with the public, and whatever may or may not have been said by staff at Diplomat Travel could not bind the Defendant contractually. Further, the Defendant says that what Mrs Harman says she was told by the staff at the airport was substantially correct: there were no flights to Gaziantep and all CTA’s flights had been cancelled. In fact, the Defendant operated only one flight on 29 June 2010 from Stansted: it flew full to capacity to Antalya. The Defendant has no record of any reservation by Mrs Harman on its system, and it is common ground that the Defendant never issued her with any ticket to travel.

I accept that Mrs Harman was told, by someone at Diplomat Travel, that the Defendant was now in a partnership with CTA, because that was the commonly understood position which had been widely reported; however, it was an inaccurate understanding, because the Defendant never did enter into any contract with CTA; so this was misinformation. I also accept that Mrs Harman was told, by someone at Diplomat Travel, that she should go to the airport at the time shown on her CTA ticket, because this accords with the advice being given widely by CTA and the Defendant to the CTA agents, which they were asked to disseminate to passengers. I make no finding on the unimportant point about whether there was a crowd outside Diplomat Travel.

Mrs Harman’s claim must fail for reasons I have already outlined above: she had no confirmed reservation and thus no contract with the Defendant. In any event, Diplomat Travel was not the Defendant’s agent, and did not have the authority (ostensible or actual) to bind the Defendant contractually, even if the advice to turn up to the airport did amount to a unilateral offer (contrary to my earlier findings).

Mrs Harman’s case exposes clearly the weakness of the Category 2 cases. Mrs Harman’s CTA tickets were to a destination that the Defendant did not even fly to. So the Defendant was never going to be in a position to fulfil CTA’s contractual obligation to Mrs Harman. No fresh contract could ever have been formed unless and until the Defendant made a specific offer to fly Mrs Harman, on a specific date, to a different destination altogether, which offer was accepted by Mrs Harman. That moment never came to pass.

I am sorry for Mrs Harman. She should not have been told to go to the airport. There was no prospect she would be able to board a plane to Gaziantep. But she had no contract with the Defendant, and her claim against the Defendant must fail.

Mrs Manto purchased tickets from Holiday Centre to travel from Stansted to Gaziantep on CTA on 24 June 2010 and back on 5 August 2010. When CTA’s licence was suspended, she contacted Holiday Centre and was told that she should attend the airport on the same date and at the same time as her ticket, and show her CTA tickets to the Defendant who would fly her to a different destination, probably Antalya. Mrs Manto says she went to Stansted with her husband on 24 June 2010 and was told that there was no flight. There were many others at the airport, she said, who had also been turned away. She returned to Holiday Centre who managed to get her a single return flight, but she was forced to buy a new ticket for the outbound leg of the journey and she had insufficient funds to buy a new ticket for her husband who stayed behind in the UK.

The Defendant has no reservation recorded for Mrs Manto and no ticket to travel out to Turkey was ever issued to her by the Defendant. The Defendant did not at any stage operate flights to Gaziantep.

No contract ever came into existence between Mrs Manto and the Defendant. What Mrs Manto was told at the airport was correct: there was no flight that day to Gaziantep. The advice passed on by Holiday Centre – that she should attend the airport – was not a unilateral offer capable of acceptance by the fact of her attendance at the airport. And in any event Holiday Centre was not an authorised agent of the Defendant capable of binding the Defendant contractually.

Mrs Manto’s claim against the Defendant must fail.

Mr Bozdere did not pursue his claim. I dismiss it.

Mr Caglar booked flights on CTA through Istanbul Travel. He was due to fly out to Gaziantep with three members of his family. When he heard about CTA’s flights being suspended he went to Istanbul Travel and was initially told simply to go to the airport with his original tickets, where he would be flown by the Defendant. He later heard that the Defendant was issuing replacement tickets, so he went back to Istanbul Travel. He was told that he could not fly to Gaziantep but that he could fly to Ankara. In his witness statement, Mr Caglar said that the agent offered flights on 4 August 2010 to Ankara, which Mr Caglar accepted; he was told to come in shortly before the departure date to collect the tickets. He went to Istanbul Travel on 1 August 2010 to collect his tickets but was then told that the flights were cancelled. After some chasing of Istanbul Travel, with no response, he decided to purchase new tickets for his party with Turkish Airlines.

In his oral evidence, Mr Caglar suggested that the agent at Istanbul Travel had asked him whether he was available to fly on 4 August 2010, to which he had answered yes; and that when he went back on 1 August 2010 the agent had said that there were “no flights” rather than that the booked flights had been cancelled.

The Defendant has no record of any reservation ever being made in Mr Caglar’s name.

I conclude that there was never any contract between Mr Caglar and the Defendant. Istanbul Travel may have thought that it would be possible to get him tickets to travel on 4 August 2010, but in the event they could not secure a reservation on that date and no booking was ever made. Mr Caglar did not have a contract with the Defendant and his claim must fail.

Mr Caglar felt that he had been treated like a fool by the Defendant, by CTA and by the government of TRNC. He had been given false reassurances that the Defendant would be able to fly him to Turkey, he was kept hanging on and waiting, and in the end he had no choice but to buy new tickets at considerable cost. Mr Caglar was put under great strain by the events of the summer of 2010. A clearer explanation by CTA and the Defendant, earlier in time, might have avoided that. But in the end, Mr Caglar lost money because CTA went into administration, not because the Defendant was in breach of any contract with him.

Mr Yuzudik purchased tickets through Happy Days Travel for him and his family to fly on CTA to Adana on 13 July returning on 1 September 2010. When Mr Yuzudik heard of CTA’s difficulties, he contacted Happy Days Travel, who told him that the flights could be rearranged with the Defendant. The agent at Happy Days Travel said that it would not be possible to fly to Adana, and offered Antalya or Ankara as alternatives. Mr Yuzudik said that Antalya was acceptable, and provided his CTA tickets to the agent. On 8 July 2010, the agent informed him that he was booked to Lefkosa via Antalya; Mr Yuzudik said that he did not want to fly to Lefkosa, but wanted to fly to Antalya. The agent then said that his flight was cancelled because there were no seats available to Antalya. Mr Yuzudik did not believe what the agent was telling him, because he understood that he had seats on the flight to Lefkosa via Antalya already (although now cancelled). Mr Yuzudik then purchased tickets with Turkish Airlines (outbound 14 July, inbound 4 September).

The Defendant has disclosed a document from Happy Days Travel relating to several passengers including the Yuzudik family, dated 6 July 2010. Happy Days requested replacement tickets for the Yuzudik family to Antalya on 13 July 2010. It appears from the Defendant’s reservation system that Mr Yuzudik and his family were booked, the following day (7 July 2010 at 11.29), onto a flight leaving Stansted on 14 July 2010 to Lefkosa via Ankara. The agent wrote to Mr Teker at 12.19pm on 7 July 2010 saying that the Yuzudik family “is not stn-ecn” (ie is not travelling from Stansted to Lefkosa), but wanted to go to Antalya, and asked for the records to be corrected as follows: “it will be 13 jul stn-ayt” (ie Stansted to Antalya). The Defendant’s reservation records show that a second reservation was made at 19.19 on 7 July 2010 for the Yuzidik family to travel from Stansted to Antalya on 31 July 2010 returning on 5 September 2010. Mr Teker replied to Happy Days the following morning (8 July 2010) at 6.12 am, saying that the

“quota of our flight number STN-AYT dated 13 July is unfortunately full. In this case we have cancelled this PNR. Please determine the dates which are convenient for passengers from the website www.atlastjet.com and make your request accordingly”.

This is a confusing sequence of events. But it is clear, at least, that the original booking made by Mr Teker for the Yuzudik family went to the wrong destination (Lefkosa), and was cancelled for that reason. Mr Yuzudik was misinformed by the travel agent that that flight was via Antalya; it was in fact via Ankara (ESB). If Mr Yuzudik had appreciated that, I am sure he would have agreed to the cancellation of a booking which was, in truth, made in error and unwanted.

There is no evidence that the second booking for 31 July / 5 September 2010 was communicated to Happy Days or to Mr Yuzudik, but plainly this outward bound date was unsuitable for Mr Yuzudik and his family who wished to be away from mid-July; if asked, no doubt he would have confirmed that the outbound date was unsuitable and would have requested that this booking too should be cancelled.

I conclude that there never was a contract to fly Mr Yuzudik and his family to Antalya on or about 13 or 14 July. There was no breach of contract in cancelling the reservation for 14 July 2010 to Lefkosa. Mr Yuzudik’s claim must fail.

Mr Sendur had booked flights through Alternative Travel to fly to Lefkosa via Bodrum on 26 June 2010 with CTA. When he heard that CTA was no longer flying, he called Alternative Travel who told him that the Defendant was arranging additional flights for CTA passengers. The day before he was due to travel on his original CTA tickets, the agency called him and told him that the flight for 26 June 2010 was cancelled and that he would need to wait for the next announcement from the Defendant. Mr Sendur said he could not wait. He demanded his money back but was told he should wait for the next flight, but they did not know when that would be and he would have to wait. He decided to purchase new tickets from British Airways. The evidence shows that on 25 June 2010 he purchased two tickets to Cyprus departing on 29 June 2010.

Mr Sendur had no reservation with the Defendant.

There was no contract with the Defendant. His claim against the Defendant must fail.

Mrs Korkmaz booked five tickets to fly to Izmir (and herself to and from Bodrum) from Stansted. Her own ticket departed on 29 July and returned on 9 September 2010; tickets for others in her party departed on 3 August and returned on 18 August 2010. Those flights were booked on CTA, through CTA’s offices in Green Lanes in North London. She paid £1,441.55 in total. She was intending to travel with her family (including grandchildren); hotel accommodation in Izmir was already booked and paid for. When she heard from friends in July that CTA’s flights were suspended, she went to CTA’s offices and found them closed. She contacted First Club Holidays for help. The person she spoke to at First Club Holidays said that CTA’s flights had been suspended but that the Defendant would take on CTA’s commitments. The particular individual, an employee of First Club Holidays, indicated that she could issue tickets to fly with the Defendant, but these would be to a different destination and were subject to availability, so would possibly be on a different date. Mrs Korkmaz gave the agent the details of her tickets; the agent then said that flights were available to Antalya, for Mrs Korkmaz to fly outbound on 29 July and return on 9 September 2010, and for her family to travel out on 3 August and return on 17 August 2010. Mrs Korkmaz accepted this destination and these dates. When she went to the travel agent to collect her tickets, she was told that her reservation was cancelled.

In her witness statement, Mrs Korkmaz said that “I re-booked our flights for 29 July 2010…” but when she came to give evidence, she said this (this is an agreed note of this passage): “These people told us they would arrange booking for us which would coincide with our hotel booking dates, they would fly us to Antalya and from there we would go to Ismir by ourselves. They would look for tickets but afterwards when I attended they said they had not been able to reserve a ticket for us”.

I conclude that First Club Holidays never did “rebook” tickets for Mrs Korkmaz. The agent undertook to look for tickets for her; but no tickets to Antalya were available on the dates that she wanted, or anything close to those dates. There was no availability.

This is consistent with Mrs Korkmaz’s oral evidence, and with the fact that the Defendant has no record at all of Mrs Korkmaz’s booking on its reservation system (what it has produced in her name bears no relation to the dates outlined above, and must relate to another passenger with a similar name). If the agent told Mrs Korkmaz that her booking had been cancelled, that was imprecise and inaccurate language, because there had been no booking in her name in the first place.

Mrs Korkmaz had no contract with the Defendant. Her claim must fail.

There is no dispute that the measure of loss for breach of contract, if established (as it has been in the case of two of the Category 1 Claimants) is the cost of purchasing a replacement ticket with a carrier other than the Defendant if that was done, otherwise the wasted cost of the CTA ticket. The figures involved are small. If there is any dispute on quantum (I do not understand there to be), then the parties are at liberty to make further submissions in writing.

Further to their claims for breach of contract, the Claimants claim compensation under the Denied Boarding Regulations (the “EU Regulations”), implemented into domestic law by means of the Civil Aviation (Denied Boarding, Compensation and Assistance) Regulations 2005 (SI 2005/975). Article 3 of the EU Regulations provides as follows:

“1.

This Regulation shall apply:

(a)

to passengers departing from an airport located in the territory of a Member State to which the Treaty applies;

(b)

to passengers departing from an airport located in a third country to an airport situated in the territory of a Member State to which the Treaty applies, unless they received benefits or compensation and were given assistance in that third country, if the operating air carrier of the flight concerned is a Community carrier.

2.

Paragraph 1 shall apply on the condition that passengers:

(a)

have a confirmed reservation on the flight concerned and, except in the case of cancellation referred to in Article 5, present themselves for check-in,

- as stipulated and at the time indicated in advance and in writing (including by electronic means) by the air carrier, the tour operator or an authorised travel agent, or, if no time is indicated,

- not later than 45 minutes before the published departure time; or

(b)

have been transferred by an air carrier or tour operator from the flight for which they held a reservation to another flight, irrespective of the reason.

3.

This Regulation shall not apply to passengers travelling free of charge or at a reduced fare not available directly or indirectly to the public. However, it shall apply to passengers having tickets issued under a frequent flyer programme or other commercial programme by an air carrier or tour operator.

…”

It was common ground that the EU Regulations could only apply to outbound flights from the UK to Turkey or Northern Cyprus, and not to the return leg (see Article 3.1).

There is a fundamental problem with the claims based on these Regulations. They could only apply in the context of this action if two conditions precedent were fulfilled: first, the passengers must have a confirmed reservation, and secondly, the passengers must present themselves for check-in (Article 3.2(a)). On the facts presented by these Claimants, none of the Claimants who presented themselves for check-in (within Categories 1 or 2) could at that point be said to “have” a confirmed reservation. The Category 2 Claimants never did have a confirmed reservation so they plainly do not meet the requirement. The Category 1 Claimants “had” confirmed reservations, but those reservations were cancelled days earlier, and did not exist at the time those passengers presented at the airport, at which point they did not “have” reservations. The article must sensibly be read to require the confirmed reservation to exist at the time of presentation at the airport, because if that was not so, any passenger who at any prior stage “had” a reservation – even if that reservation had been cancelled or changed - could simply turn up at the airport and demand compensation when he or she was refused boarding, which would be absurd. Further, the recitals to the EU Regulations make clear that this provision is a protection to passengers who are denied boarding “against their will” (see for example recital (9)), which assumes that those passengers would otherwise be in a position to board, which is surely predicated on their “having” a confirmed reservation. That was not the case for the Claimants, none of whom had confirmed reservations at the point that they got to the check-in desk.

I prefer that analysis and reason for concluding the EU Regulations are inapplicable to the Defendant’s analysis that the EU Regulations are inapplicable as a result of Article 3.3. Although it is right to say that the Claimants were all travelling (or would all have travelled) free of charge, it is also right to say that these passengers were flying under a “commercial programme” operated by the Defendant, namely the programme envisaged by the Protocols.

If I had concluded that the EU Regulations did apply, I would have deducted any compensation due under them from the damages due for breach of contract as I am invited to do by Article 12(1). Any other conclusion would risk over compensation, which is at odds with the rationale of that article which is to “compensate passengers for the entirety of their damage” (see Case C-83/10 Sousa Rodriguez v Air France [2012] Bus LR 1596 at [38]). For that reason, the EU Regulations probably do not make any difference to the outcome of these cases, even if they are applicable.

When Master McCloud listed this case for trial, her and the parties’ focus was on the list of legal issues which were annexed to her order for directions. That list comprised three groups of issues: (i) issues concerning the Protocol and the extent to which the Claimants could rely on the Protocol, (ii) whether CTA acted as agent for the Defendant and if so, the nature and scope of CTA’s public announcements; and (iii) the scope and effect of the EU Regulations.

In relation to (i), it was common ground that the Claimants could not derive any rights from the Protocol. For the record, the following issues were listed under the first heading:

“(1)

The Protocol between Atlasjet and the Government of the Turkish Republic of Northern Cyprus

1)

Can the English Court recognise this document, given that one ‘party’ to the agreement is a minister of the Government of the Turkish Republic of Northern Cyprus signing on behalf of that Government?

2)

If so, what is the governing law of the Protocol?

3)

Can the Protocol be enforced as against Atlasjet by reason of the Contracts (Rights of Third Parties) Act 1999?

4)

If so, what is the identifiable class and/or the particular description of third parties under the Protocol within the meaning of S.1(3) of the Contracts (Rights of Third Parties) Act 1999?

5)

If so, what, on their true construction, were the enforceable obligations of Atlasjet to this class/description of third parties under the Protocol?

6)

Was the Protocol in any event abandoned by Atlasjet and the TRNC and, if so, when?

7)

Insofar as Atlasjet can be said to have assumed obligations to the Claimants pursuant to the Protocol, prima facie enforceable at the suit of those Claimants, what is the effect on those obligations of the Protocol dated 9^th February 2011?”

None of these issues were aired at trial, which I take to signify the Claimants’ abandonment of these points, and acceptance of the Defendant’s case that the Claimants could not rely on the Protocols. Given the Claimants’ concession, there is no need to consider the issues individually.

In relation to (ii), the following issues were listed under the second heading:

“(2)

Agency of CTA.

8)

Did CTA have actual authority from Atlasjet to act as its agent?

9)

If so, what was the scope of that authority?

10)

In light of the answers to the above, what is the status, if any, of the notice published by CTA on its website as referred to in paragraph 8 of the Particulars of Claim vis-à-vis Atlasjet?

11)

Were the passengers’ contracts of carriage with CTA frustrated by the suspension of CTA’s operations on 21^st June 2010?”

None of these issues were aired at trial, which I take to signify the Claimants’ acceptance that the Defendant was not CTA’s agent. Given the Claimants’ concession, there is no need to consider the issues individually.

In relation to (iii), the following issues were listed under the third heading:

“(3)

Regulation EC261/2004 issues common to some or all of the Claimants

12)

Was Atlasjet the operating carrier of the Flights listed in Schedule 1 to the Particulars of Claim, or any of them?

13)

If the relevant flights were not the flights listed in Schedule 1, but were some other flights (ie ‘rearranged’ flights), does EC261/2004 apply given the provisions of Article 3 thereof?

14)

In either case, does EC261/2004 in any event apply to:

i)

Flights departing Turkey

ii)

Flights departing the Turkish Republic of Northern Cyprus?

15)

Were the Flights in Schedule 1 cancelled and, if so, on what date?

16)

Does Article 12 of EC261/2004 operate as a set off to damages claims?”

The Claimants did advance a generic claim under the EU Regulations, which I have addressed above; of the specific issues listed above (12-16), only issue 16 remained in issue before me, and my conclusion on that is set out above. Otherwise, the Claimants abandoned these points and there is no need to deal with them individually.

The Category 1 Claimants succeed in their claims, to the extent that they can establish that their confirmed bookings were cancelled by the Defendant without authorisation. This is established for Mrs Atesogullari and Mr Ozkes, but not for Mrs Akguc. Noting the recital to Master McCloud’s order, the parties will wish to consider where that leaves the remainder of the Category 1 claims. The Category 2 and 3 Claimants fail, because in their cases there was no contract with the Defendant; all claims within those categories must fail for that reason.

I thank all Counsel for their assistance in this case and in particular for their mastery of the extensive documents.