Competition Commission of India
Suo-Moto Case No. 02/2010 (In Re: ... vs . on 10 January, 2012
COMP_E_T|T|ON CQMMISSION OF INDIA Dated: ...t 9. : .7.--.-- RTPE 05/2009 of MRTPC Suo-Moto Case No. 02/2010 In Re.: Domestic Air Lines Order under Section g7 of the Act
R. Prasad (dissenting) I have a different view on the subject and am therefore passing a separate order.
2. Before proceeding with the case it is necessary to examine the legal provisions of the Competition Act as explained by the Supreme Court in the case of CCI vs. SAIL Civil Appeal No. 7779 of 2010. In this case CCl had appealed to the Supreme Court against an order of the Competition Appellate Tribunal (COMPAT) -- wherein the issue was whether an appeal would lie to COMPAT against an order issued by the CCI, when there was no provision for appeal in the Competition Act, 2002 against such an order. While deciding the issue in favour of CCl that no appeal would lie against a direction issued under Section 26(1), the Supreme Court examined the provisions of Sections 19, 26 and 53A of the Act. The Supreme Court while dealing with the interpretation of the Act on Page 32 of the order held as follows:
On the contrary, the objective of the Act is more than clear that the legislature intended to provide a very limited right to appeal. The orders which can be appealed against have been specifically stipulated by unambiguously excluding the provisions which the legislature did not intend to make appealable under the provisions of the Acytwlhtvgis always expected to the court to apply plain rule of construction T statute which have been specificall,/S'/';t:j3:,:)i§'ii K7? "-
'\ ,1?'-
t' to read the words into the and on Page 35 as follows:--
A statute is stated to be the edict of Legislature. lt expresses the will of Legislature and the function of the Court is to interpret the document according to the intent of those who made it. it is a settled rule of construction of statute that the provisions should be interpreted by applying plain rule of construction. The Courts normally would not imply anything which is inconsistent with the words expressly used by the statute. In other words, the Court would keep in mind that its function is jus dicere, not jus dare. The right of appeal being creation of the statute and being a statutory right does not invite unnecessarily liberal or strict construction. The best norm would be to give literal construction keeping the legislative intent in mind.
and again on Page 37 as tollows:-
''...If the words of the Statute are clear and unambiguous, it is the plainest duty of the court to give effect to the natural meaning of the words used in the provision. The question of construction arises only in the event of an ambiguity or the plain meeting of the words used in the Statute would be self defeating. "
& "Where the words are clear and there is no obscurity, and there is no ambiguity and the intention of the legislature is clearly conveyed, there is no scope for Court to take upon itself the task of amending or altering the statutory provisions. "
In view of these propositions and as held by the Supreme Court in the SAIL case, appeal would lie only against orders passed under Section 26(2) and 26(6) of the Act. The Supreme Court also held that no appeal would lie against any directions for investigation and inquiry under Sections 26(1) and 26(7) of the Act [Page 39 of the order]. Similarly, no appeal would lie against an "den Section 26(8) of the Act. In the Act there is no section i.e. when ' j:,:;;':fiifitavention of the Act under Section 3 or Section 4 of the Act, thef/pregceefli is dropped by the X .
Commission. Something which is not proviidedilgyin,th_m_: jtgcarliniétffibe imported. The intention of the legislature is clear because the legislature has not provided an appeal under Section 53(1)(a) of the act against dropping of a case after the D.G. has found a contravention of the Act. The legislature never intended that after the [).G. found a contravention in a case, the case could be closed. Otherwise it would have provided for appeal under Section 53(1)(a) of the Act as it has provided for against orders under Section 26(2) and 26(6) of the Act.
3. The next issue is the procedure laid down in Section 26 of the Act. Though the Supreme Court in the SAIL case was not required to decide this issue, it gave an opinion on Section 26 which is reproduced as under (page 13 of the order):-
In terms of Section 26(3), the Director General is supposed to take up the investigation and submit the report in accordance with law and within the time stated by the Commission in the directive issued under Section 26(1). After the report is submitted, there is a requirement and in fact specific duty on the Commission to issue notice to the affected parties to reply with regard to the details of the information and the report submitted by the Director General and thereafter permits the parties to submit objections and suggestions to such documents. After consideration of objections and suggestions, if the Commission agrees with the recommendations of the Director General that there is no offence disclosed, it shall close the matter forthwith, communicating the said order to the person / authority as specified in terms of Section 26(6) of the Act. If there is contravention of any of the provision of the Act and in the opinion of the Commission, further inquiry is needed, then it shall conduct such further inquiry into the matter itself or direct the Director General to do so in accordance with the provisions of the Act.
In view of the reasoning given by the Supreme Court and the provisions of the Act, it is necessary to hear the parties, in cases where the D.G. has recommended closure, before closing the case under Section 26(6) of the Act. As this procedure has not been followed, l have no option but to disagre"eN_w_iAth the majority view. Another reason for disagreement with the majority is~1fh<at ) Is.-,,=:_ " ' \"h\as nowhere in his report 1 \ stated that there was no contravention of ;:t'h"e,g;A,c the DC. has found price parallelism.
I,.,.--«... ; ._
4. In one of the two cases, the parties were heard whereas in the other case this procedure has not been followed. But it may not be necessary to hear the parties again as the issue involved in both the cases is the same. The D.G. has not held that there is no contravention. The D.G. has held in both the reports that there is price parallelism. Finding of price parallelism has to be examined with reference to the provisions of this Act and economic analysis. But before this analysis it is necessary to examine the facts of the case.
5. The Commission ordered u/s 26(1) of the Competition Act, 2002 (transferred from MRTPC) in the matter of 'In respect of all Airlines Operators' on 1st July, 2010, calling upon the DG to investigate into the alleged Restrictive Trade Practice by way of forming a cartel wherein they have collectively increased their air fares and have withdrawn the promotional fares offered to the passengers.
6. The case was received in the Competition Commission of India by transfer on 432010 from the office of DG (l&R), Monopolies and Restrictive Trade Practices Commission u/s 66(6) of the Competition Act, 2002. The erstwhile MRTP Commission took cognizance of the matter on the basis of an article published in the Economic Times and Financial Express on 11-2-2009.
7. It has been reported in The Financial Express report that the full service carriers including Air India, Jet Airways and Kingfisher Airlines, over the last three days, have hiked the fares by 25% across the board which could be ranging from Rs.800 -- Rs.12.00/- depending upon the distance. The report also adds that the hike is expected to be replicated by low cost carriers like Spice Jet, Go Air and Indigo. The fact that the cost of aviation fuel prices have dropped bv 57°/o as compared to the prices prevailing in August, 2008 and that there has been an increase in the credit period for payment of aviation fuel to the Oil Companies from 60 days to 90 days has also been mentioned in the report. The views of the industry experts have also been mentioned therein stating that the full service carriers could not sustain the low fares schemes that were on sale in January, the leanmest month for air travel, and that the Airlines in a desperate move have increa,s_e?:l' improve their yields Der passenger. The experts have further ;;=t'he1:l~'ALtrend of low fares ".
,. , 2' .
continues even in the medium term, the sector will have to brace itself for huge losses in the ensuing financial year.
8. In the report of The Economic Times, it has been stated that just a few days after fare cuts on select sectors, domestic carriers have taken a U-turn. Airfares will now increase by Rs.1000-2000 per ticket with most carriers withdrawing their promotional offers of Fist and Rs.99/-- fares. And surprisingly, full service carriers Jet Airways, Kingfisher Airlines and Air India and Budget Airlines like Spice Jet, lndiqo and Go Air seem to have acted in unison. Views of the industry analysis and experts have been mentioned in the report stating that the sudden increase in fares by all carriers on the same day hints at a possible arrangement between them. The experts have further stated that the airlines generally first observe competitive fare structures and then take a call. But this time, the fares hike has happened simultaneously, which is surprising.
9. MRTP Commission directed the DG (l&R) vide order dated 13.4.2009 to carry out investigation and submit Preliminary investigation report. In pursuance of the said directions, DG (l&R) vide their notice dated 24.4.2009 had sought necessary information from the following 08 Airlines operators:
(i) National Aviation Company of India Ltd (NACIL)
(ii) Kingfisher Airlines Ltd.
(iii) Go Airlines (India) Ltd.
(iv) Paramount Airlines (Pvt.) Ltd.,
(v) Inter Globe Aviation Ltd.
(vi) Jet Airways (India) Pvt. Ltd.
(vii) Spice Jet Ltd.
(viii) Deccan Airlines Pvt. Ltd.
10. Replies from Air India (for NACIL), Kingfisher Airlines Ltd,. Inter Globe Aviation Ltd. Go Airlines (India) Ltd. Paramount Airlines (Pvt.) Ltd., were received wherein they had denied the allegations made againist.«t;\hé,.rfli{'ljn-jthe above said news item. No response was received from other Airways (India) Pvt. Ltd, Deccan Aviation Pvt. Ltd, and Spice Jet ;¢l"j_,td,;;-3»-At A ' H. ?age.;/this matter was transferred rw:' V41 .
to Competition Commission of India u/s 66 of Competition Act 2002 consequent upon the repeal of the MRTP Act.
Investigation by DG in the case (RTPE 05-2009) --
11. DG focused the investigation upon --
12. Whether the airlines have increased the airfares by 25%, ranging from Rs. 1000/- to Rs. 1500/- and in some cases by Rs. 2000/- and have withdrawn the promotional fares during the said period as a result of concerted action?
13. And if so, whether such formal / informal understanding among airlines service providers is in violation of Section 3 of the Competition Act 2002?
DG's findings of the Allegations for both RTPE 5/ 2009 and 2/2010 '14. DG has arrived at the market share of the seven important airlines. These are indigo (19.5%), Jet Airways (17.5%), Jet Lite (7.9%), Kingfisher (19.9%), Spice Jet (13.5%), Air India (Domestic)(14.9%) and Go Air (6.8%).
15. According to DG, for the purposes of analysis of allegation, the relevant market being taken here in terms of Section 2 and Section 19(5) read with 19(6) and 19 (7) is market for passenger carriage by all airlines operating in India. DG has also computed the HHI in the relevant market which is 1595.42.
16. DG has elaborated that airlines follow a dynamic fare regime. Each airline have fare buckets for every sector of operation, according to their dynamic fare policy. The airlines have a range of fares, which are placed in different fare buckets with inventory associated with each fare bucket. Airlines start selling from the lowest fare bucket and start moving upwards as a flight approaches its departure date, gradually selling higher prices inventory as demand picks up. The inventory is therefore prices in buckets which move as the load factor in'c.rea:§;g53...,]:he prices are determined based on past performance of the flights V ' M If 'fbuckets and so dynamically changing competitive and other operating en\vi,i?oni;i,r.i) f '1.
.. ..\_a 3,4'
17. So airlines have adopted a pricing system having several price bands (buckets). These price bands are from lower to higher side and reservations systems world over show tickets of lower value from lower price basket first. However the number of tickets in basket is not made known to consumers, agents or even DGCA etc. The number of seats allocated in a particular bucket can be changed in a short time.
18. DG has stated that computer reservation system is used for reservations of a particular airline and interfaces with a global distribution system (GDS) which g supports travel agencies and other distribution channels in making reservations for most major airlines in a single system. There were only three major GDS providers in the market place: Amadeus, Travelport, Sabre and shares and reservation systems like Navitaire hosts "ticket less airlines. An airline's inventory contains all flights with their available seats. While low cost carriers do not have interface with the Global Distribution Svstem, the full scale carriers have.
19. Common features for the low cost carriers (having no interface with the GDS) are --
o The inventory allocation or movement of seats in higher fare buckets can be done instantaneously.
o No requirement of filing with the Airline Tariff Publishing Company (ATPCo).
0 Generally LCCS do not participate in Global Distribution Systems.
0 LCCS not disclose the exact number of seats available in a price bucket to the web portal and travel agents and to the airline customers.
20. lndigo airline (Low cost carrier) employs the Navitaire Open skies reservation (RES) and this is a commonly used airline RES system by multiple LCCs around the world. Fares and allocations are entered directly into the system by each analyst. The system allows customers and travel agentsV\'tof:'tifi'al{e3'7fjoQl§ings directly through website and do not require to be filed V1'aflff.':;_l'?'ft.iblishing Company (ATPCo)- it W i
21. Spice Jet (Low cost carrier) has an internal IT system to handle pricing changes and do not need to communicate with any system/ vendor outside the airline. It also has a reservation system provided by M/s Navitaire of USA. The system takes schedules, fares and inventory as inputs and distributes the same to customers. The fare slabs are designed from lowest to highest and based on number of seats to be sold in a particular slab which is managed based on a variety of parameters like demand, competitive fares etc.
22. Go Air has Ftadixx as computer Reservation System and its tickets are not available on Global Distribution System like other LCCs. The dynamic pricing model adopted by Go Air takes into account the market changes, competition, demand, patters, seasonality events, weekends etc. to maximize revenue.
23. Kingfisher airline makes fare revisions by filing the fares with ATPCo (Airline Tariff Publishing Company) from where the fares get distributed to various Global Distribution systems (GDS) worldwide. Kingfisher Airlines reservation system is hosted by SABRE a global computer reservation system where all bookings made at Kingfishers Airlines offices/websites are priced and ticketed.
24. Jet AinNays and Jetlite have been using SABRE as its Computer Reservation system. All bookings for the Company are handled by SABRE reservations system and fares get distributed to various Global Distribution systems (GDS) worldwide. It gets traffic from international markets on domestic sectors through interline agreements and Code Share Arrangements with various airlines worldwide.
25. According to DG, the analysis of the price movement of airfares of each of these airlines shows that airfares did increase during the said period as compared to earlier, in particularly, huge increase in price was noticed in the last category and spot fares of these airlines. It has been stated by DG that airlines like Indigo have maintained that this year there was a huge dem,ands.aQd increase in demand and seasonality caused bookings in the higher the peak season. It has been admitted by the airlines that there, in if R e i:7r,r;.fa'i't'es during the Diwali.
7'' iris l:4)-
L .
126. DG has written that examination of data and replies furnished by the airlines shows that inventorv bucket svstem is not easilv discernible and could not be properly explained. It is not known how the bucket system is planned and how it is actually executed. Further it has been stated that allocation / re-allocation of seats between higher fare bucket and lower fare bucket is possible and is also sometimes done for commercial considerations. The airlines have neither been able to give details of number of seats which have been earmarked under each bucket nor shifting of the bucket fare from lower to higher fare. It is noticed that a passenger does not know as to why he is being asked to purchase a higher price bucket seat. Although fares and buckets are displayed on websites of airlines, in a particular bucket the number of seats allotted are not mentioned, and the consumers do not know since number of seats against each bucket. The enquiries made by the DG have revealed that the airlines can convert low bucket fare seats into high bucket fare seats by making changes in the inventory of buckets, if the demand is very high.
27. DG has further maintained that airlines do adiust the seats under different buckets and thus during the period of high demand and enhanced load factor, the seats in the lower fare buckets get reduced and the seats in the higher fare buckets category go up. The behavior of the airlines tend to be the same and there is a case of price parallelism in a sense that fares move upwards generally in case of all airlines during the peak demand time and to the close of departure. it is only the airline management which knows as to the number of seats allocated in each price band. This is not known to the customers as well as in full to the travel agents. Moreover, number of seas allocated in a particle bucket can be changed in a short time. Thus, by changing number of seats in a price band (buckets) a higher price can be charged when the demand increases. Therefore, no transparency is maintained in the system for disclosure of seats in a price band in the LCCs as well as FSCs working on GDS. Thus, the whole system is opaque.
28. According to DG, the basic pricing policy of all airlines is to benchmark its airfares against its competitors for a given sector which leads to a situation of price parity. It would thus be observed that the pril;:és'i7a:ni"ofnvgst full service carriers and no frills airlines are likely to be similar in in,efTri§§.s¢:§éil,v¢isgilégaorles for the same sector on 'a given date and time. The i'fa"ct'i't¥"fll:at5:3anf al . E , K s resp"onds"toi'the prices of its ' 'W; 'ilfgl l'_:A I' If! 9 competitors has been argued by almost all the airlines including NACIL may lead to some price parity. However, such price parity / parallelism alone may not be sufficient to establish that the airlines have acted in a collusive manner to fix the airfares in the stated period in question.
29. As stated above, the number of seats allotted against each bucket can also be increased or decreased. This requires changes in the computer systems also and last hour changes can always be done at least in case of LCCs since they do not have to resort to Global Distribution System and have a system of direct distribution. Even Full Service Carriers, who are on Global Distribution System, can take this decision anticipating the demand and make changes, although with some lead time.
30. DG has reported that in the course of investigation, based upon available materials and submissions of various respondents, no evidence of cartel like behavior among airlines on fixation of fares could be notices during the period of investigation. Generally during the peak demand time and high demand, fares of airlines move together. To this extent, there is a pattern of price parallelism in the conduct of all the airlines. However. evidence of organized. express or tacit collusion among the airline could not be found in the investigation. Regarding collusion DG has stated that there has been no direct information exchange. However, DGCA has facilitated indirect information exchange from December 2010 by asking all airlines to display the tariff sheet on websites.
31. It has been noted by DG that airlines have categorically denied that there was any reduction in number of flights during the period ofinvestigation."ll/loreover, i_t__l3_a_s_ been found that the airlines do adiust the seats under different buckets and thus during the period of high demand and enhanced load factor, the seats in the lower fare buckets get reduced and seats in the higher fare buckets category goes up. Therefore, during the peak demand period, the consumers are left with no choice but l:o purchase air tickets of higher bucket fare cate«gfc§E§7TT5'Ttte\ behavior of the airlines tend to be same and there is a case of priceq.e5a'}i%1ll'el:\Kll that fares move upwards generally in case of all airlines k time and to the r:lo.s.e.-of.the,A,depa.rture. i ~ E-7% S A M J
32. While concluding DG has mentioned that due to the practice adopted by the airlines, the consumers, particularly, who travel by booking tickets near the date of departure, feel as if they could not get the optimal choice. This practice of airlines also has the potential of the adiustment of the fares towards the higher cateqorv buckets from the lower category buckets after assessing load factor and high demand. The airlines may devise means to bridge the gap in information asymmetry, make the system transparent and more consumers' friendly to give consumers option to make optimal choice.
33. it is relevant to mention here that the said conduct of the airlines in their pricing of airfare under the different bucket fare system is separately under examination in a separate case No. 01/2011 so as to determine whether such practice is anti- competitive in violation of Section 3 of the Competition Act, 2002.
34. l, before, delving into the details of the price parallelism and determining whether the conduct of the airlines on the above two occasions had any element of concert or coordination among them; iterate the summary of what has been stated by airlines and DG as stated above.
35. Airlines follow dynamic fare pricing regime which entails allocation of the inventory of seats to several different fare bands (buckets). The same is followed by domestic airlines in India. Both FSCs as well as LCCs follow this pricing mechanism. Fares under this system increases as the date of departure comes closer. While the FSC's have an interface with the global distribution system (GDS) and supply tariff information to APTCo, the LCCs do not have such interface.
36. No limit has been prescribed by the DGCA or any other regulatory authority on the price cap or number of seats in each bucket. However, airlines report the fares to be charged one month in advance to the DGCA. The fares reported are in the form of ranges in particular baskets.
87. From the entire material given abov':é;VA'f .'clféia{% that issue before the "lafi_p'n"tiof section 3(3)(a) and / reefiifaéfft or practice among the Commission is to establish whether thererlwafs (I 1:
or 3(3)(b). This involves identification of 2 z).
airlines. Since the DG has already concluded at both the occasions the airlines were found to be involved in price parallelism / price parity and this is substantiated by the submissions as well as understanding of the movement of fares during both the periods under question. Thus the issue is only of determination / proving agreement or practice among the airlines. It is relevant to look at the definition of the agreement as per section 2(b) of the Act --
"agreement" includes any arrangement or understanding or action in concert,-
(i) whether or not, such arrangement, understanding or action is formal or in writing;
or
(ii) whether or not such arrangement, understanding or action is intended to be enforceable by legal proceedings;
38. The definition of the agreement as stated above is very wide since it includes any sort of arrangement, understanding or action in concert. Before elaborating the agreement it is required to evaluate the submissions of the airlines as well as the report of DG to see the correctness of the arguments. Para wise comment on arguments are in the subsequent paras.
39. I agree with the DG that airlines were found to be involved in moving seat inventory to the higher bands (buckets) during the relevant period. Since airlines have themselves admitted and DG has found that in the times of higher demand airlines shift the inventory from lower buckets to higher buckets. This is evidenced _t_>y the fact that the load factor increased during the period concerned as well as movement of seat inventory. The number of seats available in thehiqher price bands increased during the relevant periods.
40. All the airlines in their submissions have made it clear that presence of competitors and pricing or fares charged by the competitors are major factor in their pricing policy for the different sectors (pairs). Domestic airlines take clue from each other in setting their price or seat inventory lg» and accordingly shift the number of inventory from among varflb _rt~i\'t*i.4ularly lower to higher ' 2 \ ..
12buckets as and when situation demands so. That the airlines take clue from each other has been admitted by the airlines themselves in their submissions to the DG.
41. DG has in its own analysis given a detailed account of bucket system and how it operates. DG has noted that airlines fix various fare buckets ranging from very low reportedly if tickets are purchased in advance or high if purchased at the last minute and when the load is 100%. It is the inventory bucket system which is not easily discernible. if a consumer purchases air ticket just hours before travel, then he pays more as compared to the consumers who have purchased tickets in advance, in case of high demand situation. It is not known how the bucket system is planned and how it is actually executed. DG has mentioned the airlines' submission that allocation / re-allocation of seats between higher fare bucket and lower fare bucket is possible and is also sometimes done for commercial considerations. A consumer are not able to know how many seats are available in respective fare buckets by the airlines and whether consumer has made an optimal choice by purchasing a ticket out of a particular fare bucket. Accordingly. there is information asymmetry and non- transparency clue to non-display of number of seats available in a band (bucket). This non-display in turn increases the opactueness.
42. One of the LCC airline has mentioned that they source the data from a third party vendor. The data is a snap shot at a particular point in time of the "current" selling fare on a given future departure date and flight number. This data is a subset of the fare buckets and is generated by a web-robotic application based on pre- defined date / query for the next 25-30 future departure dates. DG has not made his observation as well as raised any issue on this. This information as stated by airlines is available to them for the next 25-30 days. Since. no investiqationrhas been pointed in this direction, one can easily assume that same can be made available for any number of future dates, if available on websites of the airlines. How this sharing of prices facilitate coordination among the airlines and what effect it has on reducing business uncertainty about the future demand and prices for the airline services. is dealt elsewhere. ~ ,
43. In another submission before the,/DG 1 r .
submitted that they track the Global dtslttbfitto if scale airlines (FSC) stemfthe prices offered by the "'\:}\/H' 1/ I other airlines. DG has not made any investigation in this regard. Like how the system could have been used by the airlines to coordinate and for reducing the uncertainty about the future business environment and situation. Similar issues were raised in the U.S. Department of Justice's investigation of the major U.S. airlines and the Airline Tariff Publishing Company (ATPCO), which is owned by the airlines and disseminate price change information to airline and travel agent computer systems. The investigation began in 1991, and the resulting case was settled with a consent decree in March 1994. ATPCO is a central clearinghouse for distribution of fare change information. Each day airlines send to ATPCO information on new fares to be added, old fares to be removed, and existing fares to be changed. At least once a day, ATPCO produces a compilation of all industry fare change information and sends that computer file, which includes thousands of fare changes, to a list of recipients. The complainant the U.S. Department of Justice, charged that the airlines, through ATPCO, had colluded to raise price and restrict competition in the airline industry. The Justice Department argued that the airlines had carried on detailed conversations and negotiations over prices through ATPCO. The DOJ argued that the announcement of fares that are to take effect at a later date allowed the airlines to negotiate over prices. While none of the announcements was binding, such "cheap talk" can still aid collusive behavior.
44. DG has identified the relevant market as the market for airlines services in India. The concept of the relevant market is fundamental to the issue of competition and its assessment in any sector. As defined in Section 2(t) of the Competition Act 2002, the relevant product market means a market comprising of all those products or services which are regarded as interchangeable or substitutable. The product in the context of the air transport sector is transit from one place to another i.e. between one city to another. Transit from one place to another can occur through direct as well as indirect flights. However for the time sensitive passenger, indirect routes may not be a substitute for direct routes. Also there will be preference for the time and date of travel. Keeping the above in mind, the relevant market is defined as the route between city pairs at a particular time§jorj:;A~a.. particular date.Accordinqlv, considering different competitive environ[rné7ri_tv;4%'::_' each of the citv pair connected by the airlines should be the Even the report titled "Competition Issues in the AilrlTlranls§port:'Siecttifiiiri awarded to ASCl also 4! '_ ,-3;
4 14identified city pair as the relevant market. It becomes more important when considering the US ATPCo case mentioned above, the DOJ's case also was based on patterns of multimarket coordination that it claimed to have identified.
45. Nearly all the airlines have stated that each city pair has its own pricing which is determined by lot of market factors. From this it can be inferred that DG has erred in determining the relevant market. The DG has thus left one area untouched which is determining the relevant market as city pair and then finding out if there is any multimarket coordination. in the absence of any investigation I am bound to conclude the presence of such multimarket coordination.
46. Accordingly, the HHl computed by the DG would not serve any purpose because each city pair would have different no. of operators and thus different HHI. HHI should have been computed city pair wise and then only a meaningful figure of concentration would come which have implications on making decision whether the market is conducive for a cartel like conduct or not. The study entitled "Competition Issues in the Domestic Segment of the Air Transport Sector" in India awarded to the Administrative Staff College of India computed HHl's for various city pairs and in many of them HHl's were at level which might have raised competition concerns.
47. None of the low cost carrier has an interface with the global distribution system (GDS) and do not supply tariff information to APTCo. LCCs have defense available to them of the cost involved for having interface with the GDS. However, equally important is the fact that airlines as prudent businessmen should have in these times of globalization considered as many avenues / sources / channels of income as possible, the reasons of not making availableuthe,information are not adequately justified by the airlines.
48. Airlines have submitted that they do not keep historical data with them and that can not be retrieved from their respective systems. This seems to be a ridiculous argument when the airlines themselves say that fa.res_,'Z'§?t5ififfs.,,,4'are decided on the basis of one of the important factor i.e. prices and its projection in to the future. If we believe the airlines r}ton~availability of the historical prices then their own submissiortthat psti;c'esf'arNe'*§sat on the basis of [historical data does not hold qood."l\/lore'ovAer,"eveTry tétktélslinto account the 15 past performance including past demand and price data. Thus this contention of the airlines is grossly untenable.
49. Airlines have given various submissions pertaining to the fixation of the tariff/ price of the tickets. However, the submissions of the airlines are contradictory to each other's as stated in the following lines. Some have said prices / fares depends upon the cost of the fuel. Others have said parallel pricing is due to similar cost conditions. One has iterated that the various cost elements like cost of fuel, PSR, ADF vary across airlines and even for a single airline, varv across different routes. It has been submitted that it is only the passenger demand and competition on the route which determines the price. It is interesting that neither anv airline has provided the data on cost of operation of a flight for any of the routes and they have not shown the method by which they have arrived at the cost of operations. Even the DG. has not examined this issue.
50. It would be better to look at the past few years' brief of the aviation industry, before 2008, the Indian Aviation sector had the presence of players like Air Deccan, Air Sahara, lndian Airlines, Spice Jet, Go Air, lndigo, Kingfisher, and Jet Air. Thereafter a phase of consolidation started. First it was the Air Sahara and Jet ainNays merger and then Kingfisher and Air Deccan. That phase saw the merger of Air Deccan with Kingfisher and Air Sahara with the Jet Airways. Air Deccan which was lndia's first low cost carrier made the prices of air tickets drastically down. After this consolidation, the prices of the air tickets went upwards and all this hue and cry of increase in fares started coming in the public domain. Perhaps the acguisition of the mavericks was the reason for such increases. The study entitled "Competition Issues in the Domestic Segment of the Air Transpon' Sec.tor.'.'..i.n lndiaawarded to the Administrative Staff College of India has written its apprehension on the situation after the consolidation about whether the operation of market forces are adequate to prevent the abuse of market Dower.The report also mentions that on some routes after merger of the airlines, HHI rose to the levels which a competition authority cannot ignore. N / Ki' S - -
51. From the information in the public domain, it Hnoticed that Ai_'rf'Deccan was the first airline which introduced the concept of dynamig\pri'ci_ng whi_tl:h...rneans,,selli,n,g at a
2.'; V .f' "uh _ .=~_,«I 16 higher price during high season (tourist season) and selling cheap during the off- seasons. Thereafter, every player in the Indian market switched to this dynamic price model. it was a new concept of pricing of air tickets that was first introduced in India. it is clear that before this concept, each of the airlines might be having different pricinq policies. Introduction of new concepts, schemes. techniques marks the presence of the competition wherein each competitor in order to meet competition try new schemes, policies etc. However, at present each of the players has in its submission qiven that they are using this dynamic price model for fare setting. it seems none wants to try any other pricing strategies. The information in the public domain is full of articles on how the airlines in the other jurisdictions are experimenting with different pricing strategies to meet the competition.
52. It becomes important to see whether the structure of the present airlines market is conducive for any sort of collusion among the airlines. The aviation industry in India, especially with regard to passenger airlines, follows a strictlyoligopoly-type structure with the characteristics such as (i) an industry dominated by a small number of large firms (ii) firms sell either identical or differentiated products (the only differentiation here being in service quality and frills offered), and (iii) the industry has significant barriers to entry (which holds true both with respect to regulations and huge capital investment required). One sees the following characteristics with respect to the Indian passenger airlines market --
(i) Fewer Firms and Higher Market Concentration
ii) High entry & exit barriers
iii) Homogeneity of products ( (
(iv) Similar production costs
(v) Market Transparency (
vi) Depressed Conditions or Low Innovation Rate
53. Lets relate these factors to the Aviation industry and see whether industry is really conducive to the cartel / price fixing type ofcfiognduct ----
a) Fewer Firms and Hfier Market,.§:'6'ni§éntfation-The industry has five big .9."
players barring some locally placft/d¥',?p'l 90% of the share is in 17 the hands of 5 firms for the industry as whole for all the city pairs. Thisfigure would vary when presented for individual city pairs.
b) fli_gh entrv & exit barriers :-Barriers to entry are important to the successful operation of a cartel. In the absence of barriers, an increase in price will attract new competitors into the market. The barriers to entry in this sector have been discussed elsewhere, however, the major barriers to entry are:
(i) High capital reguirements: According to the domestic air transport policy, the minimum fleet size for a scheduled operator is five (5) aircrafts and the minimum amount of the shareholders' funds is Rs.10 crores for aircraft of all-up-weight below 40,000 kgs and Rs.30 crores for aircraft of all up--weight exceeding 40,000 kgs.
§I3;§_:_ Grandfathering of slot rights poses a barrier to entry for a new player. There is currently no policy for allocation of slots to new players in the industry. However, we were given to understand that when a new slot is available the new player is given preference as far as possible.
(ii) Homoqeneitv of products :- It is easier for firms to collude where products are similar and where the main dimension of competition is price competition (competition is not multidimensional). Where goods are homogeneous the costs of collusion are reduced and the likelihood of successful collusion increased. The possibility for non-price competition through product differentiation is, of course, reduced.The service in the airline industry is of two types; Full Cost Carriers providing meals etc. on board -- Therefore have a higher price, Low Cost carriers which do not provide the meals- cheaper flight tickets. interestingly, notice that the two big players in the market -- Jet and Kingfisher -- both have a stake in low cost airline also. Therefore, the homogeneity of ownership may make collusion easier.
ta sjjqhe easier it is for 1 't iwhere costs are pric§s"'"t»lian other cartel fl' ' _ . . .. ,.'\.'V. V' .3' _.- f-.''/ K. ' -- 3'
- _ , , ». , ..«' ~.._h_', , .. five"
18members. In the absence of any investigations, I rely on the submissions made by the airlines that cost of operations of all the airlines are same.
e) Market Transparency:-All players in the market can monitor the price of the ticket on offer, hence any under-cutting is easy to detect. This reinforces the fact that collusion or cartelization is easy in this market.
54. The major issue to prove is an agreement among the airlines. The Competition Act, 2002 defines the word agreement as -
"Agreement" includes any arrangement or understanding or action in concert,-
(i) whether or not, such arrangement, understanding or action is formal or in writing; or
(ii) whether or not such arrangement, understanding or action is intended to be enforceable by legal proceedings;
55. Thus the word agreement encompasses all sorts of coordination among the parties / competitors. This coordination among the competitors is visible in the market place by observing or noticing the parallel conduct of the parties / competitors either in terms of price movements or market sharing, fixation of terms and conditions etc. Here the airlines have themselves admitted that there may be parallel movement in the prices in view of several factors. Moreover DG also has concluded price parallelism / price parity from the behavior of the airlines and from the simultaneous movement of prices of.dif.f,e,rent airlines. Since DG did not find any evidence of tacit as well as explicit agreement, it ruled out the possibility of any sort of coordination among the airlines and also denied any tacit collusion / understanding / arrangement among airlines. However, there is no doubt about the parallel prices.
56. Parallel conduct is either a proof of concgd presumption of collusion. Parallel conduct"
concerted behaviour is the only plausibfllew I bvelhauviour, or at least creates a a proof of collusion if » The_';deicision of the European .. F"//« 19 Court of Justice in this regard becomes relevant to quote. in Imperial Chemical Industries Ltd. (lCl) v Commission. ECJ stated that, although parallel behavior may not by itself be identified with concerted practice, parallel conduct may amount to strong evidence of concerted behaviour if it leads to conditions of competition which do not lead to the normal conditions of the market, having regard to the nature of the products, the size and number of the undertaking, and the volume of the said market. in Morton Salt Co. vs US 235 F 2d. 573, 577 (1956), it was recognized that "such behavior is another item to be weighed, and generally to be weighed heavily, in the determination".
57. Parallelism is not to be proved as it is already established. Accordingly what is required now is to establish connecting evidences / explanations / factors which when read together along with parallelism constitute a mechanism among the airlines to exchange the information among them. These factors / explanations / evidences are itself available in the submissions of the airlines as well as information available in public domain.
58. One of the airlines has made a submission that it sources the data regarding future selling prices of various other airlines for each of their flights on a given future departure date with flight number. Since the airlines have buckets the price of which according to them increases as departure date comes close, with this information in hand they can easily track month-wise information which probable corresponds to the buckets and accordingly align the prices / fares. There is no investigation in this regard carried out be the Director General to see how this platform of a third party is used by other airlines.
59. Airlines have also submitted that they track the changes in the fares made by the airlines in the Global Distribution System. There is no investigation in this regard carried out be the Director General to see how this platform of Global Distribution System is used by airlines. In US the airlines were found to be involved in the practice of using the platform of ATPCO for the information exchange and to secure the coordination of others over a price chang /
60. I believe in the absence of investigaitilh _'_ ' ' jtregagrdfthat this information is available to all the airlines. This is 7a sorti§5fi'.t3rice'li'nf€3frrtj;€i'tion'exchange which puts 20 the consumers at large in disadvantageous position. This information is more readily available to the airlines than the consumers. This coupled with the fact that airlines can easily make changes in the fare structure in computerized reservation with such information in hand reduces the consumer welfare. Had this been equally easily available and in equally comparable manner to both airlines as well as consumers, the consumers would have made more informed choice. This has the effect of reducing their uncertainty as to each other's future action, or diminishing their incentive to deviate from a coordinated strategy.
61. The consumer/ passenger is put to more disadvantageous position in the light of the additional fact that airlines do not disclose the inventory position as well as inventory in a particular bucket. This fact has been admitted by DG as well as Director General of Civil Aviation that there is information asymmetry in the market and the system is opaque and hard to discern.
62. The airlines are operating on various city pairs and each city pair has different competitive environment. The airlines operating on different city pairs might coordinate to keep the prices artificially high. As mentioned in the above paras DG has not conducted any investigation on whether there can be any sort of multi- market coordination among the airlines. In the absence of investigation, i presume coordination among them. This coordination has the effect of maintaining the prices higher level and against the consumer welfare.
63. From the information available in public domain, it has been observed that the starting fares in a bucket for any airline for a particular city pair is more or less same. if one start books a ticket for any distant date say for 15"' October 2012 Airlines on the one hand say they have different cost structures but ihaveisimiilar demand conditions for the industry as whole. But here also these demand conditions vary for different city pairs regarding which no investigation carried out by DG. The price parity for such a distant future is impossible unless there is any sort of arrangement to align the fares. Even if it is based on their own__APidng'egpgendent estimations, then their own submission that they do not keep histo§;ig{'aM'l:: be untrue. For once, even if we assume independent estimag ,. price, the immediate question will be whether the airlines are usingith hisgtoriical data as well as the same econometric tools / techniques of estimations, this is hard to believe. Moreover, DG has not delved in to this aspect.
64. Another factor that points towards some sort of coordination among the airlines ts the similarity in the Air Turbine Fuel Charge. According to airlines this Air Fuel Charge is the function of the distance as well as consumption of fuel by the aircraft. From the information in public domain ,this Air Fuel Charge levied by the airlines are found to be same for number of routes and same among the airlines for those routes. This aspect has not been touched upon bv the DG. Moreover the report of Administrative Staff Colleges of India also acknowledged the fact that the charges are same among the airlines.
65. industry has an association FlA-Federation of Indian Airlines, of which all major domestic airlines are members. Number of times there are articles in the newspaper media reports in the public domain when the executives announce that fares should be hiked and that there was no good reason for a price war. In one such article, one of the top executive of LCC suggested to his peers that airfares be hiked since nobody was gaining from such a fare war. DG has not properly related all the aspect in the disintegrated manner since the issues are all intertwined. Though in such an announcement no particular fare was specified, but when we relate aspect like sharing of prices together on a common electronic platform with the announcements, it does have such meaning. The DG has not investigated such announcement during the period after the date of notification of Section 3 and 4 i.e. 20.05.2009.
66. The issue in this case is of price parallelism. The Director General found that all the airlines were resorting to parallel pricing. in every business parallel pricing is necessary because when a person starts a business his price of goods based on the price of the similar goods in the market. Subsequently all the operators in the market realize that they must have a similar price in order not to lose business and this brings an equilibrium in the market. If one of the operators thinks of having a different price and reduces the price then his sales increases at the cost of the other operators of similar goods in the market. In the Vcron_s_§,qt»i_en_Ee the other operators in ,, the market would also reduce the price to bring"-' of the first operator.
Sometimes in order to teach a lesson to rr,:l'é 3wht§'i.lollvered the price, the it .3 fr 8 other operators reduce the prices to s'uchlan.3eXte V hat :%'t"_lteg_.§'riris"t operator" starts 22 incurring a loss and may ultimately go out of business. The economists on the basis of the game theory define price parallelism as a case of tacit collusion. But the different courts in the US and Europe have held that price parallelism itself is not violative of the Competition Act. I need not elaborate those decisions as they are many. The courts have held in the US and Europe that some plus factors with price parallelism are necessary to establish concerted price parallelism. Concerted price parallelism envisages a meeting of minds and without the meeting of minds there cannot be a violation of the Competition Act. But then the laws in the US and Europe are different from the Indian competition law. According to the legal provisions existing in the Sherman Act or Articles 81 and 82 of the European Commission there has to be an action in concert which has to be established by the competition authorities before any violation of the Competition Act can be found in those regimes. But concept of parallelism by itself does not show any conspiracy and it also cannot be classified as an agreement by any stretch of imagination. What is necessary is that it has to be established that competitors had knowledge of identical prices and that they had decided to whether to fix identical prices. But it is quite possible that parallel decisions may be independent and not inter dependent. The leading case on this issue is of American Tobacco Company vs. US where the prices of the products were identical for nearly 12 years. All three producers of cigarettes of then increased their prices at the same time by similar amounts. There was also material to show that they offered similar discounts to different distributors at different points of time. Though the companies took the plea that the price was fixed independently of each other, the other material showed that there were same plus factors. Such behaviour is possible when there is oligopoly situation where the sellers are very few. In the modern world the different operators in the market do not sit down together in a smoke-filled room and come to an agreement that they will fix the price. in the modern times with the coming up of the computers decision of the 'fixing prices is immediate and information is also available through various modes and means and a meeting together is not at all necessary. Considering this fact decision taken can reflect anti-competitive behaviour without an actual agreement taking place. Persistent price stability in the case#,ef.._ff etal excess capacity indicates confidence on the part of each seller that"lE§;n5'pe'titorsi;,will hold the price line. Therefore what transpires is a tacit agreemfntfan . . :,»'2i'., pfriceéfixing is involved 7,) .
23"'nlo"'acccrual of benefits to consume'rs':;in"'and"
such an agreement is unlawful. In such cases meeting of minds cannot be established but what can be established is an agreement of parallel pricing.
67. The Indian competition law is different. Under Section 3(3) of the Competition Act 2002 and an agreement or a or a decision taken by an association or a practice followed are treated as agreement though they are different items provided the conditions in clauses (a) to (cl) of Section 3(3) are satisfied. In the European law or the American law such a situation does not exist. Under those jurisdictions the competition authorities have to establish that there was a conspiracy. Under the Indian Competition Act a fiction has been created and according to this fiction if the conditions in clauses (a) to (cl) are established then an enterprise has to discharge the onus that it had not resorted of price-fixing. The first fact is that practice has been put on par with agreement and the second aspect in this case is that the onus has been shifted on the enterprise to establish that it had not resorted to price-fixing. The onus is rebuttable and when it is discharged, the enterprise can get out of clutches of Section 3(3) of the Act.
68. In this particular case the parallel behaviour over a long period of time by the different airlines amounts to a practice carried out by them. Practice has been defined under Section 2(m) of the Act in an inclusive manner and a defined as folIows:-
"Practice" includes any practice relating to the carrying on of any trade by a person or an enterprise.
69. Therefore price parallelism can be regarded as a practice in the case of airlines. It was for the airlines to establish by bringing material on record that they have not indulged in price-fixing. This onus has not been discharged by the airlines either before the DG or before the Commission. There is no material to hold that there has been conspiracy or there was an agreement but for the practice carried out it is necessary for the parties to establish that they have not indulged in price-fixing. Under the provisions of Section 3(3) of the.'€3o'rti"pjetti;ttgn Act it is not necessary for the Commission to establish adverse (AAEC). But if the factors mentioned in Section 19(3) are looked 'I y price parallelism there is ac¢:tt*>cnsumiers'are putito a. loss.
"' r ':-- .9 *1,-,;r' 24 Even there is no improvement in production or distribution of goods or the provision of services. Therefore the provisions of Section 19(3) are applicable.
70. Price parallelism in this case is established. It has to be examined whether there are plus factors in this case to establish contravention of the Act. The business model is that each airline has a number of 'fare buckets. The number of seats in each bucket is known only to the airline and not to the consumer. When the demand increases at the time of festive season or when there is a strike and the supply of airline seats decreases, the airlines move the seats from the lower buckets to the buckets having higher prices. Thus the consumer has to buy tickets which are costlier. Further some of the LCCs are now flying to foreign countries but they have not opted for Global Distribution System (GDS). The reason for not having interface with GDS has not been given by the airlines. Further, all forecasting of seasonal fares is based on historical data but before the D.G. the airlines have stated that they do not maintain historical data of fares. Such an explanation is not acceptable. The airlines have not given any data on the cost of operations and their relationship with the prices of tickets. Further, it is seen that the Air Fuel surcharge levied by the airlines are found to be same for number of routes and same among the airlines for those routes. This clearly shows information sharing among the airlines and may be a reason for parallel pricing. Another important fact to be noted is that when Air India did not fall in line with the other airlines as far as pricing of tickets was concerned, all these airlines complained to the Ministry of Civil Aviation that Air India was resorting to predatory pricing. This also shown a concerted action on part of the airlines. In fact there is no reason as to why all these airlines are having the same price for the same city pair. In fact, there is no evidence to support an alternative explanation.
Thus, the plus factors exist in this case which shows a concerted behaviour and leads to price parallelism.
71. To sum up, under the Indian Competition Law, it is not necessary for practice to have a meeting of minds. The onus w it 25 price parallelism have therefore contravened the provisions of Section 3(3)(a) of the Competition Act, 2002.
72. Therefore under the provisions of Section 27 of the Act the follo wing directions are issued to these airlines.
la)
(b) (C)
(d)
73. The Secretary is directed to issue a copy of the order They should cease and desist from price fixing through price parallelism and price fixing The number of seats in each price bucket should be indicated to DGCA and should also be indicated on their website.
They may consider introduction of the bid process in the purchase of their air tickets.
A proper rationale for fuel surcharge should be followed. A penalty of Rs.5 crores each is levied in the case of Indigo, Jet Airways, Kingfisher, Spice Jet and Go Air. The penalty is much below 10% of the turnover of each of the airlines.
to each airline as discussed above.
sell "
(R. Prasad) Member, CCI 'A _ . C?/~\l=~lLl¢\LiT .;;--;?3l'S/3lSl'ANT DlRl§CTOR gl. ,,l\~,g_.§_0:tn;§e"tition Commission of indie ' ' " New Delhi rm... an .1...»-