Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

A structure conduct performance assessment of alternative Canada-United States Air Services Agreements Roberts, Tony Selwyn 1994

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1994-0294.pdf [ 4.87MB ]
JSON: 831-1.0087527.json
JSON-LD: 831-1.0087527-ld.json
RDF/XML (Pretty): 831-1.0087527-rdf.xml
RDF/JSON: 831-1.0087527-rdf.json
Turtle: 831-1.0087527-turtle.txt
N-Triples: 831-1.0087527-rdf-ntriples.txt
Original Record: 831-1.0087527-source.json
Full Text

Full Text

A STRUCTURE-CONDUCT-PERFORMANCE ASSESSMENT OF ALTERNATIVECANADA-UNITED STATES AIR SERVICES AGREEMENTSbyTONY SELWYN ROBERTSB.A., The University of British Columbia, 1988A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIREMENTS FOR THE DEGREE OFMASTER OF SCIENCE (BUSINESS ADMINISTRATION)inTHE FACULTY OF GRADUATE STUDIES(Department of Commerce and Business Administration)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIAApril 1994Copyright, Tony Selwyn Roberts, 1994In presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of ,41ø i4/,VWThe University of British ColumbiaVancouver, CanadaDate Q9 ,9Pi/4 f94DE-6 (2/88)ABSTRACTCanada and the United States have the largest, bilateral traderelationship of any two nations. Fittingly, they also exchange thelargest volume of international air travellers of any pair ofcountries. The terms under which Canada-United States airtransportation are provided are set forth in the Canada-United StatesBilateral Air Services Agreement. The current Agreement was foundedupon the consumer demands and industry operating practices thatprevailed in 1966. Although the Agreement was substantially modifiedin 1974, the essence of the regime has been rendered obsolete by thedevelopments of transborder airline market characteristics. Canada andthe United States have recognised that a new bilateral air servicesagreement is a necessity.Three general strategies have been proposed as the bases for a newregime: the specified rights option, the open border option, and thecabotage rights option. Specified rights is the genre of the currentregime: all routes having entry strictly controlled. The open borderoption would entail complete freedom for either country’s carriers tocontest all transborder routes. Cabotage rights allow carriers tocontest any market within or between the two countries.A new agreement has yet to be achieved. The delay in finding anacceptable scheme has been the difficulty in meeting both major*i 1*objectives for the new policy: efficiency and equity.This report examines the alternative schemes for a new Canada-UnitedStates air services regime. The structure-conduct-performance paradigmof industrial analysis is utilised to evaluate the nature of thedistribution of benefits that would arise following the adoption of thevarious alternatives.The report concludes that the adoption of a phased-in, open borderregime would best meet the twin objectives of efficiency enhancementand equity of opportunity.*iii*TABLE OF CONTENTSAbstract iiTable of Contents ivList of Tables viAcknowledgement viiChapter One The Purpose of Analysing AlternativeProposals for a New Canada-United States AirTransport Regime 1The Structure-Conduct-Performance Paradigm 7Organisation of the Study 9Chapter Two Characteristics of International AirTransport: Background 14Foreign Ownership of Airlines 30Alliances Between Canadian and United StatesAirlines 52Chapter Three Introduction to the Structure-Conduct-Performance Paradigm 64The Fundamental Principles of the Structure-Conduct-Performance Paradigm 65Chapter Four The Structure of the Air TransportationIndustry 96Evidence on the Conduct and Performance ofAir Carriers 134Conclusions on the Relationship Between theStructure of the Air Transport Industry,and Airline Conduct and Performance 158Chapter Five General Alternative Strategies forTransborder Airline Services Reform 167Literature Survey 170The Specified Rights Option: Revising theStatus Quo 186The Open Border Option 216The Cabotage Option 249Strategies Utilising Phase-Ins of LiberalRegimes 293Summary of Analyses of the Proposed Models 308*1 v*Chapter Six The Need for a New Structure for CanadaUnited States Air Transport Services 317The Three Solutions Under Consideration 318Conclusions 322Bibliography 327Appendix I. Calculation of Average Number of CompetitorsPer Route in United States Airline Markets,1988 335LIST OF TABLESTable 1. Foreign Airline Ownership of U.S. Airlines 36Table 2. Extent of Leased Aircraft in U.S.Carrier Fleets 103Table 3. Effective Tax Rates 133Table 4. Comparison of Total Factor Productivities 176Table 5. Comparison of Unit Costs by Region 177Table 6. Sumary of Effects of Open Border Regime onMarket Categories 242Table 7. Revenues and Costs of Canadian and U.S. AirCarriers 274*v1*ACKNOWLEDGEMENTMany individuaTh have assisted me in the preparation of this thesis.wish to acknowledge their assistance, which has come in many forms.For their academic assistance, I wish to thank Mike Tretheway, TaeOum, Tom Ross, Jim Brander, Trevor Heaver, James Higginson, BillWaters, and Steve Wexier.For administrative assistance, I wish to thank Teresa Cheung and PatShanahan.For support in many ways, shapes, and forms, qualitatively andquantitatively above and beyond comparison, I wish to thank Anne,Cohn, and Ian Morse, and Raphael Perez.Last, and far from least, I wish to thank Virginia McKenzie,for her undying faith and enthusiasm.This thesis is dedicated to Chuimny*v ii *CHAPTER ONE.INTRODUCTION1. The Purpose of Analysing Alternative Proposals for aNew Canada-United States Air Transport Regime1.1 Statement of ObjectivesThe purpose of this report is to evaluate the alternative approachesthat have been proposed for a new Canada-United States airtransportation agreement. The current agreement, which provides theregulatory structure for transborder airline services, dates back to1974.1,2 Although there have been several amendments to thisagreement, it has remained substantially the same for close to twentyyears.3 Over this period, may facets of the air transportationindustry have changed. These changes have brought about the need for anew air transport services agreement between Canada and the UnitedStates.This study intends to evaluate the proposed alternative schema for anew Canada-United States air services agreement, and to determine whatthe most desirable such alternative is. There are two criteria forrating these alternatives: it should improve the efficiency with which1transborder airline services are provided; it should ensure thatopportunities for gainful enterprise and employment are retained forinterested parties of both Canadian and U.S. nationality. Inparticular, we will devote a significant proportion of our analyses tothe consideration of Canada-United States carrier alliances. Muchattention will be given to how these relationships have influenced, andcan be expected to influence, the structure, conduct, and performanceof Canada-United States airline markets.1.2 Important Changes in Industry Operating PracticesNecessitating Policy ReformSince 1974, the air transportation industries of Canada and theUnited States have undergone substantial transformations. Economicderegulation has occurred in both countries, allowing market forces togovern the pattern of, and prices paid for, airline services. Aircarriers, spurred-on by greater freedom to enter and exit from routes,have developed hub-and-spoke route networks.4 These operationspromoted the development of a new generation of airliners: mediumsized, twin-engined types, such as the Boeing 767. Also during thisperiod, innovations in information systems technologies led to theevolution of yield management techniques. Inter-airline alliances havealso emerged as important aspects of industrial structure/conduct. In2effect, the way airlines do business has changed dramatically, whereasthe regulatory structure in which they operate has not kept pace withthese changes. Unfortunately, the achievement of a new agreement hasbeen delayed due to concerns over a number of key issues.13 Equity and Efficiency: Two Important Considerations1.3.1 Equity ObjectivesWhile Canada and the United States have recognised the necessity ofachieving a new air transportation agreement, contentious issues haveprevented the realisation of this aim. Equity considerations have beeninimical to the conclusion of a new agreement. Equity objectives areroutinely pursued in bilateral air transport agreements; in this case,equity refers to equality of opportunity for interests in both Canadaand the United States to benefit from engaging in transborderoperations .In the Canada-United States setting, particular equity concerns stemfrom the difference in the magnitudes of the respective countries’airlines. For example, it would be a relatively simple matter for aU.S. carrier, having 300 or more aircraft, to allocate a portion of itsfleet to new transborder services. In contrast, the Canadian carriers,3with barely 100 aircraft, would find it more difficult to dedicatecapacity to any new services permitted by a new agreement.There are also operational and geographical characteristics of theU.S. industry that give U.S. carriers decided advantages over theirCanadian counterparts. In particular, the well-developed hub-and-spokenetworks of U.S. carriers allow them to realise superior productiveefficiencies relative to Canadian carriers. Also, the bulk of Canada’sairline traffic travels east-west between points within 200 miles ofthe Canada-United States boundary. This geographic reality presentsthe possibility for U.S. airlines to provide de facto domestic servicesto Canadian consumers. The Canadian carriers are not in a position tooffer reciprocal services to U.S. consumers. A new Canada-UnitedStates air transportation agreement must provide an equity ofopportunity for Canadian and U.S. interests alike. The determinationof an agreement that would allow Canadian carriers to compete on anequal footing with U.S. firms has eluded negotiators thus far.1.3.2 Efficiency ObjectivesThe motivation for concluding a new air services bilateral agreementhas come from two main sources. First, the domestic deregulations ofair transportation in both Canada and the United States has yielded4improved efficiencies in the provision of airline services.6’7 Second,the liberalisation of bilateral agreements between other countries hasimproved the value consumers receive for air transport expenditures.8Such advantages should be available to North American air travellers.A new bilateral air services agreement should provide a setting inwhich the same types of efficiency gains that have been achieved indomestic markets may be realised in the transborder sector. Thesegains would be in the form of productive efficiencies or allocativeefficiencies: productive efficiency referring to the minimisation ofthe cost of providing a given level of output; allocative efficiencyreferring to the extent to which the allocation of resources achievesthe optimum distribution of benefits between different economicagentsVarious authors have studied the expected efficiency effects to beassociated with alternative schemes of bilateral air servicesliberalisation.1° Their conclusions invariably point to theadvantages, in terms of economic efficiency, that can be expected toaccompany the adoption of liberalised bilateral regimes. Moreover,there is extensive evidence, from the experiences of both Canada andthe United States, following those countries’ moves to domesticderegulation, that allowing the market to dictate the allocation ofairline resources leads to more efficient service provision.” Similar5performance improvements can be anticipated for transborder airlineservices under liberalisation.The challenge for analysts and negotiators is to determine the mostdesirable scheme, given the prospects for efficiency gains, and theconstraints of equity considerations.1.4 Outline of Alternative StrategiesSeveral schemes have been proposed for a new Canada-United States airtransport regime. There have been three general strategies suggested:specified service rights, an “open border,” and varying degrees of“cabotage rights.” The existing agreement allows services to beconducted on the basis of route specific rights, allocated to either orboth country. In turn, the respective regulatory authorities allocatethe routes to a particular carrier. An extension of this scheme,wherein additional routes would be specified, has been suggested as thebasis for a new agreement.The “open border” concept would permit any Canadian or United Statesairline to serve any cross-border city-pair. This proposal has beenadvocated as a means of introducing new services to currently unservedor poorly served markets. It should increase the competitiveness of6transborder airline markets. The drawback to this scheme is that itshould be anticipated that, in the absence of any restrictions, itwould allow U.S. carriers to further dominate the transborder sector.The ‘cabotage” option would provide the opportunity for carriers toserve the domestic markets of the other country. For example, AmericanAirlines might be permitted to fly Calgary-Toronto. This proposal isalso intended to foster additional competition in airline markets.There is concern that, unfettered, this format would also prove ananathema to the viability of the Canadian air carriers.2. The Structure-Conduct-Performance Paradigm2.1 Foundations of the ModelTo analyse the different proposals under consideration, this studyemploys the structure-conduct-performance paradigm. The structureconduct-performance model was pioneered, during the 1930s, by Edward S.Mason of Harvard University.’2 The premise of the model is that thestructure of an industry dictates the conduct of the participatingfirms, which in turn determines the performance of the industry.Dimensions of performance are the technical and allocative efficienciesof the industry, and the distributive equity associated therewith.137While the original formulation of the model supposed a unidirectionalflow of causality, from structure, to conduct, to performance, thecontemporary view is that causality may flow in both directions.’4Thus, structure determines conduct, conduct determines performance,performance determines both structure and conduct, and so forth.2.2 Suitability of the Structure-Conduct-Performance Modelfor Application to Air TransportationThe structure-conduct-performance model was chosen as the means forevaluation of policy alternatives for several reasons. It is lesscomplicated than game theoretic approaches. It is not susceptible tosurvey difficulties, and generalisation problems, as is the case studyapproach. 15Second, the basis of the paradigm in neoclassical economic theory hasled to the wide-spread acceptance of structure-conduct-performancewithin the field of industrial economics. In turn, its popularity hasled to considerable advancement of the methods associated with themodel.Finally, numerous studies of air transportation have been done using8the structure-conduct-performance methodology. Brander and Zhang(1990,1993), Borenstein (1989), Gillen, Oum, and Tretheway (1985),Morrison and Winston (1986), and Oum, Stanbury, and Tretheway (1991)are examples of analyses which employed the structure-conduct-performance paradigm. Our study will make use of the methodology andconclusions of these foregoing reports to determining the desirableformat for a new Canada-United States bilateral air services agreement.3. Organisation of the Study3.1 General MethodologyThis study examines the alternatives for a new Canada-United Statesair services bilateral agreement. In order evaluate thesealternatives, we will utilise the structure-conduct-performanceparadigm. This model provides a framework for analysis of theeconomics of air transportation. We utilise existing evaluations ofthe performance of pre- and post-deregulation air transport markets inthe United States and Canada in order to gain insight as to what theliberalisation of transborder markets would result in. The basis forthis approach is that the fundamental principles governing the conductand performance of airline markets should not differ with changes inlocation, but rather should vary primarily with respect to the9structure of the market. The performance of these markets will be thebasis on which comparisons between alternative Canada-United Statesregimes will be based.Of particular interest will be the factors that have caused theperformance of deregulated airline markets to deviate from competitiveexpectations. Specifically, situations in which the structure of amarket or markets has afforded a firm or group of firms the ability toexercise market dominance and/or capture supra-normal returns will beexamined. Where this has occurred, we will also examine the nature ofthe conduct that has arisen to determine whether interventionist policycould be expected to achieve more desirable performance or not.After developing a solid repertoire of structure-conduct-performanceindications from analogous air transport markets, the methodology willbe applied to the three alternative forms for a new Canada-UnitedStates air transport regime under consideration. These alternativeswill be assessed, and a recommendation made as to what the mostadvantageous policy would be. The basis for recommendation will bethat the policy should maximise industrial efficiency, whilst achievingthe bilateral equity aims that can be expected to remain a tenet ofinternational air transport relations.103.2 Chapter by Chapter Layout of the StudyChapter II begins by examining the general context in which NorthAmerican air carriers operate. This includes an examination of thehistorical background of international air transport agreements, andparticularly Canada-United States agreements. We examine the alliancesbetween Canadian and U.S. air carriers. We are interested inanalysing these relationships in order to both understand the carriers’responses to existing policy, and to anticipate what forms of conductmay arise from the bilateral policies under contemplation.Associated with the issue of airline alliances is the question offoreign ownership of airlines. We therefore explore the regulationssurrounding foreign investment in airline firms. We consider whetherthese long-standing regulations are desirable given today’sinternational economic relations.Further, in keeping with the emerging nature of internationaleconomics, we consider the implications that trading blocks may have onthe face of international aviation. Therefore, Chapter II also surveysthe air transportation policy of the European Community in order togain insight into what the trend in international air transportrelations may be.11Chapter III is the section of this study that develops the structure-conduct-performance paradigm. It is then given application to theairline industries of Canada and the United States in Chapter IV. Ofparticular interest is the relationship between regulatory structureand industrial performance, and between market concentration andindustrial performance.Chapter IV contains an extensive survey of the literature regardingstructure-conduct-performance as applied to North American airlinemarkets.Having developed the structure-conduct-performance model, we thenturn to an analysis of the alternative policies for a new Canada-UnitedStates bilateral air services agreement, in Chapter V. We will analysethe expected welfare effects of the adoption of the alternativestrategies.A summary and our conclusions are presented in Chapter VI.12Footnotes to Chapter One1. Chesen, p.72. Transport [1991c], pp.3,43. Transport [1991c1, p.84. Morrison and Winston [1986], pp.5-85. Chesen, p.416. Morrison and Winston [1986], p.727. Oum [1991], pp.15-188. Tretheway and Windle, p.189. Brander, p.1610. For examples, Dresner and Tretheway (1990), Dresner andWindle (1992), Gillen, Hansen, and Ramos (1990).11. For example, Borenstein (1989), Caves, Christensen,Tretheway, and Windle (1987), Graham, Kaplan, and Sibley (1983),and Oum, Stanbury, and Tretheway (1991).12. Reid, p.1113. Scherer, p.40014. Reid, pp.14,26,2715. Reid, pp.37,38,45,4613CHAPTER TWO.CHARACTERISTICS OF INTERNATIONAL AIR TRANSPORT1. Background1.1 The Foundations of BilateralismThere are many forms of law that apply to air transport activities.Bilateral agreements are one form of regulation in aviation. They areimportant because countries have found it difficult (if not impossible)to conclude multilateral agreements on air transportation. However,there are some multilateral conventions that exist in air law thatinfluence the form that bilateral agreements take.“The Treaty of Paris” in 1919 gave the first legal recognition to therights of sovereignty of countries over that airspace overlying theirterritories. This sovereignty extends to all activities conducted in acountry’s airspace.In 1944, the “Chicago Convention” was concluded which established theregime of economic and legal interrelations with respect tointernational airline transportation. The Chicago Convention wasoriginally signed by 50 countries. It has since been adopted by over14150 states.One of the important principles of the Convention is that allcountries should be able to participate in the pursuit of airtransportation on an equitable basis. The notion of an equitable basisrefers to the distribution of economic benefits that should accrue tothe interests of the parties involved in an air transportrelationship. 1There tends to be somewhat of a conflict between this principle andArticle 1 of the Convention, which restates the principle of nationalsovereignty as originally described in the Treaty of Paris. Theconflict stems from the desire of countries to promote their owncarriers’ economic fortunes, versus the principle of equity thatunderlies the entirety of the Convention. As discussed later in thischapter, this has given rise to the tendency toward bilateralism in airtransport agreements.The Convention established a list of “freedoms” pertaining to airtransportation. Some of these freedoms will be referred to later inthe text, and should be defined.151.2 Definitions of TerminologyFirst Freedom: the right of an airline to overflya foreign country for commercial purposes.Second Freedom: the rightof an airline to make stopsof a technical naturein a foreign country whileenroute toanother foreign country.(Technical stops are those made forthe purposes of maintenance, particularly for re-fuelling.)Third Freedom: the right of an airline to carrypassengers from its homecountry to another country.Fourth Freedom: the rightof an airline to carrypassengers from another country to its home country.Fifth Freedom: the rightof an airline to carrypassengers from a point not within its home countryto a pointwithin another country that is also not its home country.Eighth Freedom: the rightof an airline to carrypassengers between two points in a country that isnot its home country. Eighth freedom rights are more commonlyreferred to as “cabotage.”16Consecutive Cabotage: the operation of a flight betweentwo points in a foreign country where that segment is thecontinuation of a flight that originated in another country,and onto which the carrier has the right to board passengers.Stand Alone Cabotage: the operation of a flight betweentwo points in a foreign country, that is not the continuationof a service originating in another country.1.3 The Bermuda Agreements and Canada-United StatesBilateral Agreements1.3.1 The Bermuda AgreementsIn 1946, the United States and the United Kingdom concluded theBermuda I Agreement. Bermuda I established the general format forbilateral air transport agreements to come. The focus of Bermuda I wason routes and capacity. Article 4 of the Agreement gives recognitionof the principles of fairness and equality of opportunity that bothparties should enjoy in the provision of air transport services betweenthe two states’ territories.2 Bermuda I featured covenants regardingThird, Fourth, and Fifth Freedom services. An agreement was reached on17the pricing of services: prices were to be set by negotiation, with theability of either country to veto fares that it consideredunacceptable.The later Bermuda II Agreement (1977) introduced significant changesfrom Bermuda I. Bermuda II featured less flexibility in allowing airroutes, virtually abolished Fifth Freedom services, and introducedcapacity controls. These moves were initiated by the United Kingdom inresponse to the competitive threat posed by the growing number ofAmerican air carriers. The British were concerned that the liberalnature of Bermuda I would prove an anathema to their airline industryin the face of increasing US competition.31.3.2 Canada-United States Bilateral AgreementsThe first Canada-United States air services bilateral agreement wassigned in 1929. It allowed any Canadian or U.S. carrier to flytransborder routes. The second Canada-United States agreement wassigned in 1939. In contrast to the earlier bilateral, the 1939agreement severely restricted access to transborder routes, with fourcity-pairs being granted to each country. In 1945, the two countriessigned yet another bilateral. This time, there was an asymmetric awardof route rights: Canada received but 7 of 20 routes specified in the18agreement.4The first “modern” bilateral between Canada and the United States wasconcluded in 1949. It substantially conformed to the Bermuda I format,with pricing and capacity provisions contained within the agreement.5It also included Fifth Freedom rights for Canadian carriers, forflights from Canada, through the United States to the south Pacific andthe Caribbean.In 1966, the basic Canada-United States bilateral that prevails tothis day was negotiated. Capacity restrictions were absent from theagreement; these had never been a contentious issue between Canada andthe United States.6 Market entry was based on “designation”: eachgovernment selected which airline (or airlines) under its jurisdictioncould serve a particular market. The determination of prices was by“double approval.” Double approval means that an airline may establisha price on a route which will prevail as long as it is accepted by bothgovernments. The 1966 agreement gave United States air carriers therights to serve 27 city pairs. Canadian carriers were granted rightsto serve 16 city pairs. One Fifth Freedom right was granted to eachcountry.1974 saw the agreement revised to its present form. The 1974 versionof the Canada-US bilateral specifies that US carriers are permitted to19serve 92 routes; Canadian carriers are allowed to serve 71 routes. Atotal of 129 transborder routes can potentially be served; 87 of theseroutes were being operated as of November 15, 1990. Significantrevisions to the Canada-US bilateral were made in 1974. The amendmentsand additions included the following:-an increase in the number of designated routes;-an agreement covering the conduct of non-scheduled, orcharter, services;-the Pre-clearance Agreement.Pre-clearance refers to the practice of establishing customsfacilities at the point of departure. The Pre-clearance Agreementallowed Canadian carriers to establish 13 pre-clearance facilities inthe United States, while US carriers were given the rights to create 11such facilities in Canada. In actual practice, only US carriers haveinstituted pre-clearance operations. They exist at six Canadianairports: Vancouver, Calgary, Edmonton, Winnipeg, Toronto, andMontreal 8While attempts have been made to negotiate a substantially improvedCanada-US bilateral, there has been only limited progress.The deregulation of air transportation in both the United States and20Canada led to the growth of regional or commuter airlines. Thesecarriers use relatively small, turboprop aircraft on stage-lengths ofgenerally under 600 miles. They are usually affiliated with a largeair carrier, acting as a feeder service for traffic destined to, oroutbound from, smaller centers that cannot justify economical jetservices. Their emergence promoted the need to modify the Canada-USBilateral.In 1984, the two countries negotiated a liberalised agreementcovering the realm of transborder services conducted by regional aircarriers: the Regional, Local, and Commuter Services Agreement. Thisagreement permitted the growth of short-haul, transborder services byregional carriers. The authority to operate short-haul, transborderservices under this Agreement is automatically granted, provided thatthe following conditions are met:1. The aircraft operated have a maximum passengercapacity of 60 or less, and a payload capacity of no more than18,000 pounds;2. The city pairs have not been designated in the 1966Agreement;213. One of the cities must have a population of less than 500,000if in Canada, or 1,000,000 if in the United States;4. The stage lengths must not exceed 400 miles if on routes inthat area bounded by Thunder Bay and Quebec City, otherwisethey must not exceed 600 miles. Routes between Alaska andpoints in Canada are not restricted with respect to stagelengths .Also in 1984, the two countries reached the Experimental TransborderAir Services Agreement. This agreement was aimed at promoting thegrowth of transborder services through the use of under-utilisedairports. The airports designated for concentration under this schemewere Montreal-Mirabel in Canada, and San Jose, in the United States.Features of this plan were limitless carrier participation, in terms ofthe number of carriers and the amount of capacity they could offer, andliberal pricing provisions.’0 Being experimental in nature, the ETASagreement had a fixed term, and expired without being renewed.Services to Mirabel were instituted, but discontinued. Services to SanJose were initiated, and have outlived the ETAS program: Vancouver-SanJose rights were grandfathered into the Canada-United States bilateralat the conclusion of ETAS.221.4 Working Toward a New Transborder Regulatory Regime:Considerations, Complications, and Influences1.4.1 Structural FactorsRecent rounds of bilateral negotiation have been fruitless due to theinability of the two sides to reach a compromise on the issue how toachieve an equitable balance of opportunities for airlines of bothcountries in the provision of transborder services. A key issue ofcontention was that of cabotage- -or “Eighth Freedom” rights as definedin the Chicago Convention. Canadian negotiators advocated a cabotagerights regime for North America. The United States took the positionthat it did not wish to grant cabotage rights to Canadian carriers.This was due to their fear that it would establish a precedent thatcould ultimately jeopardise the viability of American carriers.Ironically, the American concern had little to do with Canadianprospects for seizing control of the US market, but rather that, underthe terms of the Chicago Convention, the US would be legally bound toconfer the same rights to non-Canadian, foreign carriers.The United States preferred the concept of an “open border” regime.This would allow airlines of either country to offer direct servicesbetween any city in Canada and any city in the United States. TheCanadian side felt that this was unacceptable since it would give the23larger, more efficient American carriers the ability to not onlycapture more of the transborder traffic (in which US airlines alreadyenjoyed a substantial advantage in terms of market share), but alsothat the Americans would establish “mirror image” routes to thedetriment of Canadian carriers.“Mirror image” routes are services that parallel the Canadiantranscontinental routes. Under this scheme, Canadian domestic trafficis carried from a Canadian city, to a US hub, then back into Canada.The Canadian concern was founded on the premise that this practicewould give the Americans de facto cabotage in the most lucrative ofCanadian markets: the east-west mainlines.In considering the relative positions of the two nations innegotiating toward a new agreement, two facts should be kept in mind:1. Air Canada and Canadian Airlines International combined arenot as big as US Air (the fifth-largest carrier in the UnitedStates);2. Transborder travel accounts for close to 20 per cent of allair traffic in Canada, whereas it accounts for only about 2per cent of the traffic in the United States.1124It is worth noting the difference in perception that is evident inboth countries regarding the status of the other in terms of marketpresence or market influence. When American politicians discuss theactivities of foreign air carriers operating into and out of the UnitedStates, the companies that are invariably mentioned are BritishAirways, KLM, Lufthansa, and JAL.12 In actual fact, the foreigncarrier conducting the most aircraft movements into and out of theUnited States is Air Canada; Canadian Airlines International ranksthird on the list, being slightly behind British Airways.In terms of actual passengers travelling between the United Statesand foreign countries, routes linking the United States and Canadaaccount for the second largest quantum: 14.9 per cent of totalinternational passenger movements. Only Germany, with 17.7 per cent oftotal international passenger movements, was a more significantorigin/destination than Canada (these figures were for 1987).13In 1989, US carriers captured 62 per cent of transborder travellersutilising scheduled airline services. 82 per cent of transbordertravellers made their journeys using scheduled services. Canadiancarriers capitalised on the charter market, enjoying a 96 per centshare of the non-scheduled market.’4 In general, it is Canadians whodo the majority of transborder travelling. In 1988, Canadiantravellers accounted for close to 60 per cent of total transborder25passengers.15It has been widely recognised that there are fundamental realities ofdemographics and geography that will determine the feasible solutionsto the problem. The Canadian market is located within a relativelynarrow, 150-mile-wide strip along the Canada-US boundary. Canada hasthree urban centres having metropolitan populations of over 1 million;the United States has 39 such centers. The climatic regime of Canadais such that Canadians wish to travel to US destinations, particularlyin the winter months, to enjoy their holiday times. In contrast, theattractiveness of Canada as a holiday destination for Americans hasbeen relatively poor, especially in comparison to Europe. The UnitedStates is the more powerful of the two economies, and should continueto be so ad infinitum. It therefore has business centres that attracttravellers from Canadian business interests.The essential reason for the success of US carriers in the scheduledservices market is in their provision of beyond-the-gateway services:US air carriers are able to offer Canadian travellers services to morepoints in the United States by virtue of the intra-United Statesnetworks they have established. A Canadian passenger seeking transportto, for example, Tucson is confronted with the choice of travelling ona Canadian carrier to, say, Las Vegas, and then having to transferonto a regional carrier to get from Las Vegas to Tucson.26Alternatively, he can fly on a U.S. airline all the way from Canada toTucson, enjoying the advantages of single-line service.1.4.2 Efficiency Consequences of LiberalisationMuch research has been done in recent years tooutcome would be of opening the Canada-UScompetition. These inquiries have suggestedefficiency gains from a more liberal regime.efficiency gains varies according to the type ofinvestigate what themarket to transborderthat there would beThe extent of theagreement instituted.The range of alternative models for a new bilateral goes from a newset of route allocations--essentially following the status quo, to acompletely open skies policy--full cabotage rights for carriers of bothcountries. The outcomes of these alternative approaches have beenanalysed. by Chesen and Associates (1989), Gillen, Hansen, and Ramos(1990), Korenic (1990), Oum (1990), and The Ministerial Task Force onInternational Air Policy (1990). Without exception, these studies haveconcluded that there should be improvements in efficiency associatedwith the liberalisation of airline services in North America.In general, the beneficiaries of liberalisation will be consumers.Some airlines may benefit from liberalisation schemes by being able to27increase their traffic densities due to improving the structure oftheir route networks. Airlines enjoying economic rents under thecurrent regime should see these rents disappear (or at least be greatlyreduced) under the contemplated changes.Unfortunately for Canadian carriers, in addition to the advantages ofgeography and demographics noted above, their U.S. counterparts enjoysuperior factor productivity, and unit cost advantages.16 Therefore,those in the Canadian airline industry have urged the CanadianGovernment to consider their well-being in crafting a new bilateralagreement. Pursuant to this, it has been suggested that there shouldbe a phased approach to implementation of the opening of North Americanairline competition. The phased approach is aimed at “leveling theplaying field” so that Canadian carriers can compete on a equal basiswith U.S. carriers.There are questions that arise with regard to this notion ofequitable opportunity for Canadian companies:1. How to determine what scheme would afford Canadiancompanies the measure of protection that would allow them toget onto an even footing with American competitors;282. What time frame should be provided for thisobjective to be realised;3. Will this approach be meaningful in the long term.In order to frame an advantageous new agreement, consideration mustbe made of several factors influencing the air transport industries ofCanada and the United States. These considerations include: anunderstanding of the ramifications that domestic deregulation has hadon the nature of airline operations in both countries; the trend towardglobal airline networks through inter-airline alliances; the associatedissue of foreign ownership (which may inhibit the achievement of anideal alliance strategy); the direction that other countries are takingwith respect to bilateralism and multilateralism.We now turn our attention to a consideration of the relationshipsthat exist between Canadian and U.S. air carriers with respect tomarketing and operational alliances. Their current manifestations areevidence of the strategic conduct of the carriers. Understanding theirfoundations should therefore be instructive as to showing how the aircarriers would react under alternative formulations of a new bilateralagreement.292. Foreign Ownership of Airlines2.1 General History of Controls on the Foreign Ownership ofAirlines2.1.1 General Basis for Ownership ControlsAs noted in the preceding discussion of the bases of internationalair transport relations, airlines have historically had a distinctnationality. Such international operation rights as may be exercisedby an airline are conferred on the basis of the carrier’s nationality.Nationality generally has been determined on the basis of thecitizenship of those who substantially own and effectively control theairline.17In general, countries have not permitted foreign nationals to own anymore than minimal holdings in national air carriers. This practicearose out of concerns from both military and economic perspectives.Nations have viewed air transport as a strategic industry.In recent years the policy has endured because of the importance thatair transportation has assumed in national economies in terms ofcommunication, tourism, and as a realm of high technology employment.Moreover, operating a national airline is a source of pride for most30countries. It is for these reasons that so many of the world’s aircarriers are owned either partially or entirely by nationalgovernments.Canada and the United States are unusual in this regard. The U.S.airline industry has, since its inception, been operated by privatecapital. Canada has had strictly privately-owned airlines since 1989--when Air Canada was fully privatised.2.1.2 The Genesis of Foreign Ownership ControlsThe sovereignty of countries over their airspace, and over theoperation of aircraft therein, was defined by the Treaty of Paris.This led to national identities being formed with regard to airtransport operations. Since a nation had the authority to dictate whatoperations would be permitted in its airspace, governments sought toensure that their own nationals would, as much as they were able (whichvaried greatly from country to country), to reap the rewards of theiraviation enterprises. Moreover, state aircraft were employed topromote government policy and to ensure security.With the end of World War II approaching, the allied powers, andneutral and non-aligned nations met at the 1944 Chicago Convention to31determine a legal regime that would facilitate the pursuit ofcommercial aviation in the post-war era. Some discussion of provisionsconcerning ownership and control of airlines occurred at Chicago. TheUnited States advocated that such restrictions be placed oninternationally operating airlines. Opposition, largely from LatinAmerican countries, led to no formal agreement being reached as to astandard for these matters at Chicago.The 1947 Bermuda Agreement between the United Kingdom and the UnitedStates established the first formal conditions regarding nationality ofownership with respect to air carriers. The purpose of therequirements for national ownership and control over air carriers wasto prevent enemy nations and nationals from being able to overfly andenter the territories of the signatories. This was the beginning ofthe Cold War era, and the specter of espionage was looming over theminds of British and American negotiators. The potential forsurveillance, reconnaissance, and infiltration posed by commercialaviation operations was intended to be minimised by the introduction ofthe foreign ownership restrictions included in Article 6 of the BermudaAgreement. 1832The substance of Article 6 is as follows:Each Contracting Party reserves the right to withhold or revokethe exercise of the rights specified in.. .this Agreement by acarrier designated by the other Contracting Party in the eventthat it is not satisfied that substantial ownership and effectivecontrol of such criers are vested in nationals of eitherContracting Party...There is no explicit indication that this article was intended topertain to economic considerations. It has been widely interpreted,however, as so-doing. This interpretation formed a precedent for laterbilaterals.As the Bermuda Agreement has served as the model upon which mostbilateral agreements have been based, countries have embraced theinterpretation of the Article as a form of economic protection in airtransport relations. The result has been the continuing limitation onforeign ownership of airlines. Carriers have had to rely on domesticsources of capital to establish and expand their businesses. Thispractice is in marked contrast to the internationalisation that hasbeen a feature of corporate endeavour in other industries.332.2 Foreign Ownership Controls in Canada and the UnitedStates2.2.1 Canadian Foreign Ownership ControlsAs mentioned previously, Canada and the United States permit anairline to have no more than 25 percent of total voting equity to beheld by non-citizens. The reasons for this limit are slightlydifferent in both countries.The motivation for the equity limit in Canada is one that has beenpervasive in Canadian economic thinking since Confederation: Canadaseeks to protect its economic sovereignty through legislation. Thedevelopment of air transportation legislation in Canada began with thepassage of the Trans-Canada Airlines Act of 1937. This Act created themandate for the growth of a national air carrier. Thus, Trans-CanadaAirlines (TCA) was formed.The purpose of TCA was to provide a Canadian trans-continentalairline service. Ottawa was concerned that if it did not act towardthis end, that pressure to permit United States carriers to providesuch services would prove to be indefatigable. Again, the aim was toensure Canadian control over transportation services in the country.342.2.2 Foreign Ownership Controls in the United StatesThe US experience with the genesis of air transportation was inmarked contrast to that of Canada. From the outset, US firmsestablished airline services on a private basis. US firms were notonly able to establish viable service within the United States, butsome US carriers were first to introduce air transportation withinforeign countries--Pan American Airlines being the best example ofthis.Though the United States Government has never owned airlines, it didparticipate in fostering the growth of the air transportation industryby giving mail contracts to early air carriers. In later years,recognising the strategic importance of aviation, Washington took stepsto ensure that the US industry could provide the US military withaircraft which could be called into national service to transport menand materiel in times of conflict. The current manifestation of thisscheme is the Civil Reserve Air Fleet (CRAF).35Table 1.FOREIGN AIRLINE OWNERSHIP OF US AIRLINESForeign airline % ownership US carrierAir Canada 24.0 ContinentalSwissair 5.0 DeltaSingapore 5.0 DeltaAnsett 20.0 America WestJAL 20.0 HawaiianKLM 49.0 NorthwestSOURCES: Dempsey, p.294; Erickson, p.35.The case of the relationship between Northwest Airlines and KLM isparticularly noteworthy. KLM has a 49 per cent holding in NorthwestAirlines. However, this 49 per cent of equity is not all votingequity; it is voting equity that is restricted to 25 per cent of thetotal. The issue here centres on relaxation of this restrictionbecause if substantial airline alliances are to be formed, the partiesdesire that they be given some measure of control over each other’soperations. This control is yielded by having a position that allowsfor their wishes to be directly conveyed to management; this is whatvoting shares facilitate. It is unlikely that, in most cases, airlineswould wish to invest in companies where their desires could not bedirectly expressed to management.Given a highly regulated airline industry, and the presence of thethreat (whether perceived or actual) posed by the Communist Bloc, the36restrictions on foreign ownership made some sense. Prior toderegulation, the airlines were granted route authority, and applied tothe applicable regulatory agency in order to enter new markets and toestablish prices in markets being served. The determination ofappropriate prices was the responsibility of the regulatory agency--theCivil Aeronautics Board in the United States; the Canadian TransportCommission in Canada. Because these agencies approved fares on a cost-plus basis, any cost increases on the part of the carriers wereultimately incident on the air traveller. This practice was one of thebiggest reasons why advocates of deregulation eventually weresuccessful. It was deemed in the best interests of consumers thatpricing in air transportation be left to the forces of the marketplace.Under the regulated price regime, airline profitshigher than they have been under deregulation. Thismonopoly rents stemming from entry restriction.opportunity for strategic allocation of fixed costsinterests of the company, the ownership of airlineswould have been politically unworkable.also tended to bewas the result ofCoupled with theto suit the bestby non-citizens372.3 Foreign Ownership Limitations Reconsidered2.3.1 The Case For Reducing and/or Removing ForeignOwnership LimitationsIn this era of deregulation and global airline alliances, therestrictions on foreign ownership of airlines should be re-assessed.Gone are the route awards and concomitant price regulation. Also, theexistence of a relatively clearly defined enemy to free enterprisecapitalism has disappeared as of 1991. If deregulation has beensuccessful in providing better services to consumers, capital should begiven the same sort of freedom to move within the airline industry.There are two primary reasons why airlines would like to see therestrictions on foreign capital eliminated. The first is that therewould be a greater pool of capital on which the air carriers coulddraw. If domestic investors interested in airline equity have beenexhausted, then foreigners wishing to acquire airline holdings wouldincrease the supply of available funds. This would have severaladvantageous effects on airline finances.A larger pool of funds would reduce the amount of airline financethat would have to come through debt rather than equity. Currently,many major carriers in the US and both major airlines in Canada have38high debt-to-equity ratios. This exposes the carriers to the whims ofinterest rate changes to a frightening extent. A debt-financed carrierhas less latitude regarding its disposition of cash than does anequity-financed company.Second, and most importantly, the relaxation of foreign ownershiplimitations would permit airlines to enter into more “strong alliance”relationships. The current scheme only allows carriers to engage in“weak alliances”: they may exchange equity, but only to an extent thatdoes not effectively merge the companies into a single entity. Thecurrent regulations prevent non-citizens of either country from havinga controlling interest in an airline, regardless of the proportion ofequity that is held. There tends to be a direct relationship betweenthe “strength” of the inter-airline relationship and itseffectiveness.20 The most effective alliances are those where there issubstantial equity exchange between the carriers. Foreign ownershiprestrictions limit the extent to which equity may be exchanged, andtherefore preclude the optimal coordination of operations betweencarriers.Another reason that foreign ownership restrictions should be changedis that they are being somewhat circumvented through the use of debtissues. While foreign concerns are prevented from owning more than 25per cent of voting stock in Canadian or American air carriers, there39are no limitations on the amount of debt holdings that a foreign partymay have in an airline in either country. By acquiring creditorstatus, a foreign interest may be able to exercise some measure ofcontrol over the decisions of the airline. Since this practice iscontrary to the intent of the law regarding foreign control overdomestic carriers, some action should be taken to ameliorate thissituation.Two courses of action appear obvious to rectify the problem of theinfluence of foreign creditors. The first would be to createregulations to deal with the amount of foreign debt that can be assumedby the carriers. This solution, however, would really not serve theinterests of the carriers as it would mean that their pool of capitalwould be that much smaller, hence the interest rates that they shouldexpect to pay for loans would be that much higher.The second remedy would be to increase the amount of equity holdingsthat foreign interests are permitted to hold in the airlines.The point may be raised that liberalising foreign ownershipregulation may only serve to put the Canadian industry into foreignhands, while not otherwise being beneficial to Canadians. However,given the current financial status of Canada’s two major airlines, itis likely that inaction will leave Canada with an effective air40transport monopoly. This situation will clearly not be advantageousfor Canadian consumers--unless one takes the position that it ispreferable to endure monopoly for the sake of maintaining Canadianeconomic sovereignty. It should be stressed that the aim ofliberalisation is not to give carte blanche to foreign investment, butrather to allow for a greater percentage of foreign capital toparticipate in the Canadian industry.Air Canada and Canadian Airlines International have apparentlyengaged in destructive competition. A superior cash position hasafforded Air Canada the opportunity to attempt to drive Canadian out ofbusiness. If Canadian had been earlier allowed to conclude its equitydeal with AMR Corporation, in which the latter proposed purchased a 49per cent share in the company, Air Canada would not have been able touse this strategy successfully due to the substantial financialresources of the American company. This would have obviated the needfor the two Canadian carriers to resume “normal competition.” While inthe short term this would lead to higher prices for Canadiantravellers, in the long run prices would be kept lower sincecompetition would be maintained thereby.Another area of concern is that the increase in foreign influencemight lead to a decrease in employment within the airline industry inCanada. Again, the liberalisation being contemplated here is not to41the extent that foreign parties would be given majority control.Furthermore, Canadian employment and immigration law would remain inforce, thus limiting the access of foreign labour to airlineemployments in Canada.2.3.2 Arguments Against Increased Foreign OwnershipNo course of public policy is without its drawbacks. With regard tothe increasing of the proportion of equity that foreign investors canhold in domestic airlines, there are issues that arise against pursuingthis policy.There will be at least a perceived loss of control over the nation’sair transportation system with an increase in foreign ownership. Thefear that arises is that courses of action may be taken by non-nationalinvestors that will be beneficial to the investors and/or thecorporation which will be adverse to the country. These adversitiesare usually manifested in the following forms:1. Restrictions on trade imposed by the parent country.The regulations of the investors’ home government may preventthe subsidiary from capitalising on certain opportunitiesconsidered not in the best interests of the home country. An42example of this might be that the prohibition on United Statesairlines from serving Cuba would be applied to a Canadiancarrier that was majority owned by a US firm;2. A purely domestic focus. The subsidiary will notcompete against the parent company in international markets.The problem with this is that the opportunities for the firm toexport its services will be curtailed, thereby eliminating asource of export earnings that would have helped the country’sbalance of payments position. In the airline industry, this wouldmean that the foreign-controlled subsidiary would only carrydomestic traffic, serving international travellers to the parentcompany for flights to points abroad.3. Lack of research and development. This is a concernmore related to manufacturing industries. The problem thatmight surface in the airline industry would likely be in the areaof computer reservations systems. However, given the strangleholdthat the largest airlines have on this technology, this should beof little concern.4. Utilisation of foreign management. Foreign-controlled firmstend to populate the top management positions of thesubsidiary with nationals of the parent country. This43denies domestic citizens the opportunity to advance to the upperechelons of the organisation. While this is likely to be aneventuality in the airline business, it should be noted that thereis already an exchange of top executives between the United Statesand Canada: Hollis Harris, a US citizen, is the president and CEOof Air Canada, for example.As with any public policy, there will be costs that accompanybenefits. Given the above described drawbacks that can be expected toarise from the liberalisation of foreign investment in air carriers,what must be determined is the net benefits to be obtained from such aliberalisation.2.3.3 Foreign Investment Liberalisation and Bilateral AirServices AgreementsThere is another realm of potential difficulty associated with theeasing of foreign equity restrictions on air carriers. This is in thearea of the bilateral agreements between countries for airlineservices. Bilateral air services agreements can be expected tocontinue to be the framework upon which international airtransportation is conducted. Changing the regulations on foreignownership may compromise the workability of these agreements.44The basis of the bilateral air services agreements is that thesignatories’ air carriers will be domestically owned and controlled.The agreements stipulate the conditions of services that will be madeavailable, and, in general, which firms may offer these services. Theyare premised upon the legal notion of equity: the benefits accruing toeach party should be comparable. Historically, since the airlines ofeach nation have been domestically owned and controlled, thebenefits arising out of the agreement could be clearly visible betweenthe parties. With the relaxation of ownership restrictions, theallocation of benefits becomes much more difficult to discern.Critics of liberalising airline ownership regulation are concernedthat such relaxation would necessitate a change in the whole way thatair transportation agreements are achieved. Professor Paul Dempsey, anair transportation law expert, comments that“...foreign ownership jeopardizes the integrity of bilateral airtransport negotiations between the United States and foreigngovernments. International routes are traded by nations on abilateral basis. usually with candid input from their carriers.Multiple2llegiances may well jeopardize the integrity of thatprocess”The adoption of more liberal foreign ownership policies may result inturmoil in international air services relations, as the bilateralismthat has prevailed since World War II cannot be sustained if one of its45fundamental premises would be violated. In this case, the premise isthat airlines have national identities. With free investment in aircarriers, the national identities of these firms may become obscured,and the regime of air transport agreements will have to be reformed toreflect the resultant absence of carrier nationalities. It may bedifficult to reconcile the wishes of multinational airline enterpriseswith those of the government-owned airlines that persist as being thecoimnon form of carrier; it may be difficult for countries that wishto pursue liberal ownership regimes and those that desire to maintainnational control over their carriers to reach agreement over theprotocol that will prevail in international aviation.2.4 Recent Developments Toward a More Liberal ForeignOwnership Regime2.4.1 The Scandinavian Air Services CaseAirlines have been owned and controlled on a multinational basis forsome years. Scandinavian Airlines System (SAS) has been a successfulexample of multinational airline enterprise. Incorporated in 1947, SAShas conducted international airline services on behalf of Sweden,Norway and Denmark. SAS has had to contend with on-going squabblesrelated to contentions of favouritism related to company appointments46and the distribution of facilities and contracts among the nationalsand interests of the three countries respectively. Being administeredby the three governments, SAS has had to allocate such benefits betweenDanes, Norwegians, and Swedes on political bases rather than onstrictly economic terms. This has resulted in a certain degree ofinefficiency for the airline. Nonetheless, SAS has survived, withcredit to its longevity being given to the commonality of culturebetween the three constituent nations allowing them to overcomeproblems that have arisen.222.4.2 The “Clinger Bill”In February of 1993, Congressman William F. Clinger Jr. ofPennsylvania introduced a Bill into The House of Representatives of theCongress of the United States that intends to increase the proportionof voting equity in United States air carriers that can be owned byother than US citizens. This Bill is officially known as H.R. 926 butis commonly referred to as the “Clinger Bill.” As of the time of thiswriting, the Bill has not been enacted into law.The Clinger Bill, if adopted, would allow for foreign interests tohold up to 49 per cent of the voting equity in an airline company. Therequirement that the president, chairman of the board of directors, the47chief operating officer, and two thirds or more of the board ofdirectors be United States citizens that exists under current US lawwould remain in effect. An escape clause is contained within the Billin that the acquisition of such equity must be considered to be in thepublic interest.There are specified considerations regarding the determination of thepublic interest. These considerations include the financial conditionof the carrier in question, and the effect of the purchase on service,competition, and on employment. The most important considerations arewhether the purchaser’s home country would permit United Statescitizens to acquire similar holdings in airlines of its register, andwhether a purchasing airline was owned, controlled, or subsidised by aforeign government. The Bill provides that the Secretary ofTransportation can disallow a purchase if he believes that thesecriteria are not met by the interested purchaser. Furthermore, thePresident can disapprove a purchase if he finds that it is not in theinterests of the United States on the basis of national defenseconsiderations.The motivation for the “Clinger Bill” was provided by United Statesairlines, who have sought a relaxation of foreign ownership guidelinesso as to access a larger pool of investment capital. This has been ofparticular interest to smaller American carriers, who believe that such48reforms would afford them the needed capital with which to expand theiroperations. It is also seen as a vehicle through which US carriers canbetter integrate their operations with partners in other countries.2.5 Attitudes Toward Foreign Ownership Held by SelectedInterested PartiesVarious interested groups made submissions to the Ministerial TaskForce on International Air Policy. These submissions often includedopinions regarding the desirability of increasing the proportion offoreign ownership permitted for Canadian air carriers. Air Canada,Canadian Airlines International, Canadian charter carriers, Canadiantour operators, and several public organisations (including Canadianprovincial governments) supported the relaxation of foreign investmentlimitations. It should be noted that the text of the Task Force Reportdoes not clarify whether these groups support greater foreign ownershipin terms of voting equity per se.23 There is, however, clearindication in the Report that most groups wished to see control overCanadian carriers retained by Canadians.The position of commuter and smaller scheduled carriers in Canadawith regard to foreign ownership liberalisation was less clear in termsof their submissions to the Task Force. These groups indicated that49they wished to continue the practice of code sharing. This would allowthem to continue to provide connecting services to the majorairlines.24 The most secure and profitable feeder-major relationshipson a domestic basis have involved equity exchanges between the carriersconcerned. By implication, commuter and other smaller carriers shouldbe expected to support the liberalisation of foreign ownershipregulations so as to enter into more secure relationships with majorairlines.Other groups that approached the Task Force expressed opposition tothe relaxation of foreign ownership restrictions. U.S. airlines wereseen to have been divided on the issue of foreign ownership. Some U.S.firms expressed concerns that national security and labourconsiderations might be compromised by the allowance of greater foreignownership.25 Labour unions indicated that they wished to see theretention of existing ownership and control guidelines. They wereanxious that any form of international liberalisation should notcompromise the viability of the Canadian airline industry.26As the presumption of nationality of air carriers is a feature of thebilateralism paradigm, any alteration of the norms with respect tonationality of carriers will affect the nature of international airtransport relations. The adoption of less restrictive ownershipregulations would impact international aviation agreements. The50practice of international airline alliances may be effected by changesto ownership restrictions and nationality requirements. A relaxationof ownership requirements may encourage the development of moreextensive international airline alliances--both qualitatively andquantitatively.513. Alliances Between Canadian and United States Air Carriers3.1 Airline Alliances in GeneralIn the 1980s, the expansion of air travel led companies in variouscountries to opt for forming alliances with foreign carriers in orderto provide their customers with greater geographic coverage. Airlinealliances can take four general forms: marketing and serviceagreements, pooling arrangements and joint ventures, equity exchanges,and outright mergers. These arrangements are emerging as importantmeans for airlines to secure their futures. They have been importantas a means of getting carriers into markets that they had little or nopresence in without having to establish ground and flight operations inthe particular markets. Alliances allow airlines to offer customers amore comprehensive network of services.Marketing and service agreements are the simplest form of alliance.The common manifestation of these agreements are the use of code-sharing. Code-sharing involves the use of multiple flight designatorsfor the same f light.27 For example, Air New Zealand operates a servicefrom Honolulu to New Zealand, the flight designator code of which isNZ 53. Canadian Airlines International also provides service on thisroute; its flight designation is CP 1037. In fact, the only aircraftactually flying the route is an Air New Zealand DC-b or B747. Both52Canadian and Air New Zealand offer customers seats on this aircraft,but under different flight numbers.The purpose of code-sharing is to keep customer loyalty throughoffering what is apparently on-line service. This provides thetraveller with the convenience of single-airline service; there is noneed for the passenger to collect his baggage in Honolulu fromCanadian, and check it in with Air New Zealand. It diminishes thechance that the customer will miss his connecting flight as the Air NewZealand flight will be held pending the arrival of the CanadianAirlines flight. This makes for better service between Canada and NewZealand. Hence the airline that does not fly its own services to aparticular destination is nonetheless be able to bring that travellerto a partner’s gate and vice-versa. This increases the attractivenessof the allied carriers to the consumer, who incurs lower informationcosts in securing passage, and who receives a higher quality of serviceduring the journey. Code-sharing also has the advantage that it allowsthe partner carriers to dodge the penalty of computer reservationssystem (CRS) bias: code-shared flights are listed on CRSs as thoughthey were single-line services.28A more sophisticated form of airline cooperation is the poolingarrangement. The scheme sees two or more airlines coordinate theirflight operations, and share revenues therefrom according to some fixed53formula. If the airlines also share costs, the arrangement is termed ajoint venture.29Pooling arrangements and joint ventures are superior to marketing andservice agreements in that the relationship between the carriers ismore firmly defined through extensive contractual obligations. Thecoordination of the participating airlines’ schedules on a system-widebasis distinguishes the pooling arrangement or joint venture fromsimpler marketing arrangements. While under more rudimentaryagreements there would be some consideration given (as indicated above)to making sure that connecting passengers make their flights, poolingarrangements and joint ventures involve planning services on a network-wide basis so as to maximise service quality and revenues, whilstminimising service costs.Canadian airline firms are currently involved in pooling arrangementswith foreign air carriers. Canadian and United States carriers havenot entered into pooling arrangements for North American services.30A still stronger form of alliance is the equity exchange.3 This forminvolves the acquisition of voting shares by an airline in anotherairline. Frequently, there is a two-way exchange of shares. Theholding of such securities makes for a semi-permanent relationshipbetween the carriers. It also reduces the risk associated with54exposing details of the firms’ financial, operational, and technicaldata to each other as the possibility of defection is lower. Theexchange of equity increases the stakes that the partners hold in eachother’s fortunes.32 In turn, this promotes the depth of cooperationbetween the carriers with respect to honoring their constractualobligations for the delivery of services to the consumer.33Finally, mergers are the strongest form of airline alliances. Themerger is simply the complete combination of two or more companies intoone entity. The advantage of the merger is that it brings all of theassets of the merging companies under the control of one managementgroup. Thereby, the merged entity is able to coordinate flightschedules, fare structures, discount and seat-management strategies,and take advatage of the general benefits of larger corporate size asit affects capital acquisition, and the breadth of market-presence.While the merger is the strongest form of corporate alliance ingeneral, in international air transport a problem arises in thatcountries reserve the rights to operate domestic, and Third and FourthFreedom airline services to companies that are native to the country.The definition of a national carrier may vary from country to country,but largely ownership of the enterprise determines nationality. If themerger was contrived by the purchase of a national carrier by a foreigncarrier, the ultimate nationality of the resulting entity would be that55of the purchasing carrier. This would preclude the resulting companyfrom operating the aforementioned services with respect to the acquiredcompany’s country without some change to the air law of that country.3.2 Barriers to Airline AlliancesCurrently, there are limited alliances between Canadian and US aircarriers. These are confined to marketing and service agreements, and,to a lesser but growing extent, some degree of equity exchanges. Therestrictions on foreign ownership of airlines in both countries haveprevented substantial equity arrangements from being contrived betweencarriers from opposite sides of the border.3.2.1 Regulatory BarriersIn Canada, the legal regime pertaining to all transportation modes isthe National Transportation Act (1987). Under the terms of the Act(NTA), a Canadian transportation enterprise must be owned by a Canadiancitizen to be eligible to receive an Operating Certificate in Canada.The Operating Certificate grants the holder the legal right to conducta transportation business in Canada.56A stipulation of the National Transportation Act requires that anairline be owned not more than 25 percent by foreign nationals. Therequirement has precluded the entrance by Canadian carriers intosubstantial equity exchanges with foreign companies.United States law is similar in substance to that in Canada. UnderSection 101 of the U.S. Federal Aviation Act, a United States airlinemay not be more than 25 percent owned by non-US citizens; this is partof the definition of “fitness” in the “fit, willing, and able criteria”used to determine whether an applicant for a licence will have onegranted. Again, this restriction applies to the voting stock of theairline. Foreigners may hold more than 25 per cent of total equity,but the effective control of the airline must remain in the hands ofAmerican citizens. This includes the provision that the president ofthe company, and more than half of the board of directors must beUnited States citizens.3.2.2 Corporate Charters as Alliance BarriersBeyond the restrictions imposed by federal legislations, there may beother legal requirements that would prevent an increase in foreignparticipation in domestic airlines. For example, Canadian AirlinesInternational has a charter condition that limits any one shareholder57to holding not more than 10 percent of the voting stock in theairline’s parent company, PWA Corporation.343.2.3 Competition LegislationBeyond those regulations that are concerned directly with the conductof air carriers, the firms are subject to the guidelines of generalcompetition legislation. In both Canada and the United States,competition legislation exists to protect consumers from the effects ofinter-firm conduct that is intended to enhance the profitability ormarket power of companies that are arrived at by means consideredoutside the realm of fair business practices. In the United States,this body of regulation is known as Anti-Trust Legislation; in Canada,it is termed Competition Legislation.Competition legislation is particularly concerned with businessactivities that are considered to reduce competition. Forms of anticompetitive conduct that are applicable to the air transportationsetting include predatory pricing, refusals to deal, price fixing, andmergers between firms that are intended to lessen competition. Withrespect to airline alliances, the questions of price fixing and mergersare likely to arise.58Where two firms dominate a market, they can be tempted to colludewith respect to pricing, setting fares at above competitive levels inorder to either eliminate smaller competitors from the market, or forthe purpose of rent extraction, or for both reasons. Where these firmsare alliance partners, they will be able to execute such collusion morereadily; where they are strongly allied through equity exchanges theycan do so with little or no possibility of defection.3.2.4 Two Cases in Canada-United States Airline AlliancesRecently, these restrictions on foreign ownership have beenchallenged--on both sides of the border. In Canada, AMR Corporation ofTexas, the parent company of American Airlines, sought to acquire a 49percent interest in Canadian Airlines International. The deal wouldhave given AMR effective control of Canadian. In order to close thisdeal, the provision of the National Transportation Act with respect toforeign ownership limitations would have to have been waived by CabinetOrder.One of the conditions of the deal was that Canadian would withdrawfrom the Gemini reservations system. However, Canadian was undercontractual obligation to remain in the Gemini system untilGemini was 33.3 per cent owned by Air Canada, which was determined not59to allow Canadian to abrogate its obligations to Gemini. Ultimately,Canadian was permitted by the Canadian Competition Tribunal, to exitfrom the Gemini system.The aforementioned chain of events brought Canadian to the verge offinancial collapse. The solution was to have the Canadian Governmentbail out the troubled carrier: Canadian Airlines received $50 millionin loan guarantees from Ottawa.36 Further loan guarantees wereprovided by the provincial governments of Alberta and British Columbia.The federal government justified this action on the premise that, hadit not done so, Canadian would have failed and Canadian consumers wouldhave been left with a virtual airline monopoly in the guise of AirCanada.In December of 1992, Canadian announced that it had reached anagreement with AMR which would see the latter acquire a 33.3 per centeconomic interest in the airline. This holding would give AMR a 25 percent voting share in the company as the result of its investment of$246 million. This arrangement avoided the problem of the foreignownership limitations of the NTA. However, the problem of the Geminisituation persisted until the spring of 1994 before finally beingovercome.Canadian’s rival has also been active in trying to form alliances60with US air carriers. Air Canada purchased a substantial interest inContinental Airlines of Texas in November of 1992. The deal saw AirCanada acquire a 27.5 per cent share of Continental. Because of therestrictions on foreign ownership, Air Canada netted only 24 per centof voting stock in the arrangement.With the alliance with Continental, Air Canada achieved twocompetitive advantages. Continental has a fleet of 330 aircraft-- incomparison to Air Canada’s livery of 120. The deal therefore gives AirCanada some measure of influence over these aircraft. In addition,Continental has major hub operations at Newark, Cleveland, Houston,Denver, and Honolulu. With coordination of schedules, this alliancegives Air Canada the ability to send the traveller on-line from anyCanadian city to virtually any United States destination. Moreover,Air Canada obtained influence over U.S. markets, and in access to U.S.airport assets without the need for dealing with nationalitycomplications and/or securing operating eigth freedom rights.In both the case of Canadian Airlines and AMR Corporation, and of AirCanada and Continental, there was the desire to have a more substantialequity exchange between the companies. Because of the limitations onforeign ownership in both the United States and Canada, a larger equityposition was not attainable for either Air Canada or AMR Corporation.61Airline alliances have been shown to increase both the value andconvenience of air travel for consumers. It has been asserted thatwhile economies of scale are absent from airline operations in terms ofcosts, they are present in terms of these demand-side forces. Thefreedom to form alliances is therefore important for air carriers inthat they enhance the competitiveness of the individual firms throughbetter acconmiodating consumer demands. With deregulation on each sideof the border, mergers have led to an increase in the concentrationmeasures of the industries in both Canada and the United States.However, it has been demonstrated that, although industrialconcentration has increased, the performance of the industries hastended toward more competitive outcomes in terms of fare levels andservice quality. The practice of alliance formation has therefore beenbeneficial for the participating airlines, and for the consumers oftheir services. It is likely that a change in foreign investmentpolicy would extend these benefits, on a continental basis.62Footnotes to Chapter Two1. Diederiks-Verschoor, p.112. Diederiks-Verschoor, pp.40,413. Diederiks-Verschoor, p.414. Transport [1991c1, p.25. Transport [1991c], p.36. Chesen, p.247. Transport [1991c], p.1098. Transport [1991c1, p.279. Korenic, p.1310. Korenic, p.1311. Korenic, p.1912. [louse of Representatives, p.3513. Korenic, p.2014. Transport [1991cj, p.3715. Oum [19901, p.1716. Windle, p.3717. Balfour, p.25618. Dierikx, p.11919. Dierikx, p.12120. Patterson and Tretheway, p.521. House of Representatives, p.29822. Gidwitz, p.1023. Transport [1991a], pp.80-10124. Transport [1991a], p.8625. Transport [1991aj, p.9226. Transport [1991aj, pp.95,9627. Hadrovic, p.19328. Hadrovic, p.20929. Diederiks-Verschoor, p.1730. Transport [1991bj, p.4731. Patterson and Tretheway, p.832. Feldman [1989], p.3033. Patterson and Tretheway, p.534. Patterson and Tretheway, p.635. Wings [#4,1993], p.836. Wings [#4,1993], p.1037. Wings [#4,1993], p.863CHAPTER THREE.MODELLING THE INDUSTRY: THE STRUCTURE-CONDUCT-PERFORMANCE PARADIGM1. IntroductionTo systematically analyse an industry, a framework is needed thatwill allow the interrelationships between causes and effects to beidentified and evaluated. A model provides the means by which suchanalysis can take place.The structure-conduct-performance paradigm is a model that has beenwidely used in the economic analysis of industrial structure. Themodel was pioneered by Edward S. Mason of Harvard University during the1930s. It has been further refined by the efforts of Joe S. 8am,Richard Caves, F.M. Scherer and others. It has gained widespreadacceptance with policy analysts, largely due to its basis onneoclassical microeconomic theory, particularly with regard to itsemphasis on Pareto optimality, to rate policy alternatives.1The model will be described in this section, and will subsequently beused to analyse the character of the airline industry in general, andthe nature of existing and contemplated transborder air servicesbetween Canada and the United States in particular.64The basic premise of the structure-conduct-performance paradigm isthat industries have a characteristic structure, determined bytechnical and regulatory realities, that leads participating firms tofollow certain modes of conduct, which in turn result in performancecharacteristics. The model seeks to describe an industry in terms ofits structure, conduct, and performance, and considers the linkagesbetween these modules. Its purpose is to both identify and describethe causation of existing phenomena, and to provide a logical means ofprediction of what results can be expected from alterations to any ofthe modula of the industry.2. The Fundamental Principles of the Structure-Conduct-Performance Paradigm2.1 Structure2.1.1 General Elements of StructureThe structure of an industry refers to the productive activities offirms that participate in an industry, and the distribution of outputbetween these firms. There are five major determinants of thestructure of an industry: technology, the horizontal and vertical65integration of firms, the concentration of the firms, barriers to entryinto the industry, and governmental regulation.2’3 There are alsolinkages between these determinants.Technology refers to the physical characteristics of the processes bywhich firms offer their goods or services to the consumer. Itencompasses both the realms of production and distribution. The natureof the productive and distributive processes determines the firms’ coststructures. Cost structure in turn dictates the number of firms thatmay successfully participate in the industry.4Important concepts related to the issue of technology are economiesof scale, economies of scope, and economies of density. Economies ofscale arise when per-unit costs decrease as total output increases.Economies of scope exist where a single firm, producing different typesof outputs, can do so at lower cost than could several firms, eachspecialising in a particular type of product.5 Economies of densityare related to network operations. Density economies exist where per-unit costs decrease as more units are moved over a given network.6The extent of inter-firm integration is a significant aspect ofindustrial structure.7 There are two types of such integrations:horizontal and vertical. Horizontal integration involves the partialor complete merger of two firms in the same industry. For example, the66merger of two airline companies is a horizontal integration. Verticalintegration occurs when a partial or complete merger is concludedbetween firms that have a supplier-customer relationship. An exampleof this type of integration would be the purchase by an airline (inthis case, the customer) of an airport terminal facility (in this case,the supplier). The usual objectives of integrations are to captureeconomies of scale, scope, or density, or to create barriers toentry.8’9Barriers to entry have a large influence on industrial structure.1°Entry barriers may be ex ante: for example, where the productiontechnology favours natural monopoly. Entry barriers may be ex post:where the incumbent firms in the industry engage in conduct aimed atdetering entry.If entry barriers are significant, the market may be effectivelyclosed to new competition from would-be entrants. This can permitexisting firms to continue to earn supra-normal profits, to thedetriment of the consumer. Alternatively, entry barriers may also giverise to the inefficient production processes. Firms producing behindentry barriers may be sufficiently protected from market disciplinethat their production processes incur higher-than-necessary costs.This is termed “X-inefficiency,” and represents a loss to society inthat these resources would have been more greatly valued in other67uses.Finally, government regulation can have a decided impact on thestructure of an industry. Regulation is generally divided into twocategories: technical and economic. Technical regulations areprimarily aimed at ensuring the physical safety of workers andconsumers. This form of regulation is applied to most if not allindustries, regardless of market structure. Economic regulation isgenerally applied to industries in which some form of market failureexists.’2 Industries characterised by monopolies, monopolisticcompetition, and dominant firm situations are example cases. Theobjective of economic regulation is to move the industry toward a moresocially desirable level of production, output, and distribution.Economic regulation is usually concerned with the specific areas ofindustrial activity: pricing, entry and exit, mergers and acquisitions,collusion, advertising, and research and development. It is economicregulation that is of relevance to the discussion of the effect ofgovernment regulation within the context of the structure-conductperformance paradigm.2.1.2 Five General Types of Market StructuresIn a purely competitive market structure, the firm faces a perfectly68elastic demand curve for its output. In short, the firm may sell allthat it can produce, at a price that is determined solely by playersoutside of the firm: consumers and other firms. Purely competitiveindustries are typically those having low fixed costs both in absoluteterms and in terms of their relation to marginal costs. Thistechnology gives rise to ease of entry (due to low absolute fixedcosts), and to the absence of economies of scale (due to therelationship between fixed and marginal costs). Firms are able toeasily enter the industry to capture what profits are available.Equilibrium is achieved when the firms participating are all earning anormal return on investment. Any further entry would result in allfirms earning sub-normal returns.13A natural monopoly situation arises when it is possible for only onefirm to supply the market economically. Natural monopolies arecharacterised by having a high proportion of fixed costs relative tomarginal costs. The minimum efficient scale of production is such thatonly one firm can reach this scale, given the existing demand for thegood or service. Entry by other firms would result in a duplication offixed costs, and therefore a rise in the per unit cost to a level abovethat which is socially optimal. Therefore, it is socially beneficialto have only one producer, thus the term natural monopoly. 14There are two general types of intermediate market structures: the69oligopoly and monopolistic competition. Oligopoly arises in industrieswhere there are relatively significant entry barriers, such as highfixed costs. In these industries, average costs are not such that anatural monopoly should prevail, as a small number of firms are able toproduce at the minimum efficient scale. Instead, a limited number ofsuppliers can economically participate, selling substitutable,differentiated products to consumers. Because of the small number ofsuppliers, the oligopolists are aware that their production quantityand pricing decisions will influence the conduct of rival firms.’5Monopolistic competition is the fourth form of industrial structure.It is somewhat similar to oligopoly, yet differs in that the firms inthis structure do not base their quantity and price decisions on theexpected reactions of competitors. Instead, they presume thatcompetitors will not alter their behaviour in response to the firm’sconduct. Products are considered to be close substitutes for eachother. There are usually relatively lower proportions of fixed costsin these industries compared to those in oligopolies, and the number ofparticipating firms is likely to be higher. Firms will face downwardsloping demand curves since other firms’ products are easilysubstituted for their own.’6The final form of industry structure is the contestable market. Acontestable market is one in which competitive outcomes are achieved70through the threat of competition rather than the presence of actualcompetition. While a lone firm may supply the market, it is limited inthe extent to which it can extract economic rents because of the threatof entry by other firms. Thus, even in the absence of actualcompetition, competitive pricing and output are realised in contestablemarkets.17Contestability is founded upon the premise that the cost of entrywill determine the behaviours of both incumbent firms and potentialentrants. Entry barriers are any costs that would affect an entrantbut which incumbent firms can avoid. The degree of contestability of amarket varies inversely with the extent of entry barriers facingpotential competitors. A perfectly contestable market is one in whichthere are no entry barriers. The perfectly contestable market mimicsthe market performance of competition, since, even if there is a lonefirm supplying the market, the ever-present threat of entry disciplinesthat firm’s conduct.182.2 Conduct2.2.1 Basic ElementsThe conduct aspect of the paradigm concerns the behaviour of firms in71their production of goods or services. Conduct refers to the decisionsfirms make with regard to the scale and methods of production, pricingpolicy, distribution methods, geographic coverage, advertising andresearch activities, and other behaviours.’9 It encompasses thestrategies that firms pursue with respect to agreements andrelationships with other firms, and research and development. Conductis considered primarily influenced by the structure of the industry.There may also be performance-related influences on conduct.2°Industrial structure forms the boundaries within which the firm’sdecision-makers must operate. For example, if the firm isparticipating in a purely competitive market structure, the managementof the firm may not presume to be able to establish the price of thefirm’s output(s); the marketplace has absolute control over the pricethat the firm may receive. The degree of latitude that the individualfirm has in self-determination of its conduct tends to vary directlywith the extent of concentration of the industry: if sufficientlyconcentrated, competitors may react directly to changes in the firm’sconduct. If the firm anticipates such reaction, its conduct will bethereby influenced.2’As basic decisions, firms must decide what type of goods or servicesthey will offer to the marketplace, whatgeographic areas they willoperate in, and what technology to employ in the production process.72In choosing to enter or remain in the business of providing a good orservice, a firm must consider the probability of its successfulparticipation in that pursuit. This decision is influenced by theexisting firms in the industry, and by the characteristics of thedemand for the product being consumed.Existing or incumbent firms may have certain advantages that poseentry barriers to would-be competitors. In addition to the naturalbarrier to entry that may be posed by economies of scale, incumbentfirms might have other advantages such as patents, name recognition,superior relationships with suppliers (often yielding cost advantages),or “regulatory capture,” in which incumbent firms have the ability toinfluence government regulators to disallow entry.22There are characteristics of market demand that also influence thestructure of an industry. The most important of these characteristicsare buyers’ price elasticities of demand, the substitutability of otherproducts, and the rate of growth of demand.23The own-price elasticity of demand for a good or service may beparticularly significant in dictating conduct. Own-price elasticitydetermines the price-cost margin or profit-revenue ratio:24P - MC = 1P e73As the elasticity of demand (e) increases, the price-cost margin(P -MC) decreases. This limits the extent to which firms may affectprices in the pursuit of profits. In turn, this influences the conductof firms within an industry.2.2.2 Objectives of FirmsThe objectives of the firm ultimately determine the firm’s conduct.In general, there are three objectives that are considered the drivingthe forces behind the conduct of enterprises: profit maximisation,revenue maximisation, and growth maximisation.25Classical economic theory presumes that profit maximisation is theaim of the activities of the firm.26 Under this credo, the managementof the firm is primarily concerned with maximising the return to theshareholders’ investment. The firm should be able to enhance thewealth of shareholders to an extent superior to the returns that theinvestors would be able to earn by choosing alternative investments.Those firms having the best profit performance should attract moreinvestment capital, and thereby be able to sustain their operations andbe able to grow in their industry. Rewards accrue to the firms whichare able to use their assets most efficiently, as they will earn for74their shareholders the most advantageous rates of return. In turn,this will draw assets into the most efficient companies, giving themost productive uses of society’s scarce resources.The idea of the firm as a profit maximiser has its foundationin classical economics. An essential premise of the classical schoolof economic thought is that the firm would be participating in a purelycompetitive market structure. A feature of this structure is that allfirms are considered to be price takers, selling undifferentiatedproducts, to consumers whose buying decisions are motivated by theirdesire to maximise utility through making purchases based on rationalevaluations of their choices in a marketplace having perfectinformation regarding all available alternatives. The emergence of themonopoly and oligopoly forms of market structure necessitated a changein the theory of the firm. In these market structures, the firm hasthe ability to effect prices, differentiate its products, influenceconsumer tastes, and possibly bar entry by new firms. Moreover, theemergence of the large corporation, particularly in the realm of theaforementioned market structures, as the common form of enterprise sawthe separation of ownership and control of the firm. With theincreased dispersion of shareholdings in large corporations, themanagement of such companies often became somewhat less inclined topursue the classical goal of profit maximisation, and instead turned tothe objectives of revenue (or sales) maximisation, and growth75maximisat ion.27’8The objectives of revenue maximisation and growth maximisation leadto conduct that is notably different from that which results from pureprofit maximising strategies. Firms pursuing revenue or growthmaximising goals will tend to sacrifice short-term returns forshareholders in anticipation of longer-term rewards for the firm, or,according to more cynical analysts, for the short-term rewards of thefirm’s executives, whose remuneration and/or non-monetary utility isoften directly related to total revenue or the rate of growth.29’3°The less cynical explanation for revenue or growth maximisation is thatoligopolists recognise that there is often a direct relationshipbetween market share and long-term viability. The pursuit of revenueand growth maximisation strategies is usually associated with firmsoperating in oligopolistic market structures.It has been demonstrated that a revenue maximiser will produce ahigher level of output than will a profit maximiser.3’ The purpose ofthis higher level of production is to yield greater revenues that canbe used to expand capacity, which in turn is expected to lead togreater returns in the long run. The fundamental goal is to maximisethe net present value of the firm’s profits in the long term.Therefore, evidence that an oligopoly is pursuing objectives other thanprofit maximisation is that the industry is characterised by having76current sub-normal returns and higher-than-expected output. Thus,conduct is shaped by the objectives of the firm.In order to achieve their objectives, firms have a variety of actionsthat they may take. In general, these actions can be categorised aspricing behaviors and non-pricing behaviours. Pricing behaviour refersto the setting of prices for the firm’s outputs; it is generallymeaningful in the context of market structures in which the firm is aprice setter. Non-pricing behaviour refers to actions taken by thefirm in other aspects of marketing such as product differentiation,selection of geographic areas in which to compete, and advertising. Italso encompasses the realm of inter-f irm agreements related to theforegoing marketing activities.2.2.3 PricingPricing behaviour is particularly important in the analysis ofindustrial conduct since it affects the quantity demanded in themarketplace, and the distribution of surpluses between consumers andproducers. The difference between price and marginal cost is ofparticular interest because this affects the returns to, and hence theviability of, participating firms; these, in turn, may influence thenature of the structure of the industry. Important aspects of pricingbehaviour are the nature of the demand curve confronting the firm, the77expectations that the firm has with respect to the responses ofcompetitors to its price changes, and the extent to which the firm canpractice price discrimination.32The conduct of firms in the pricing arena is heavily influenced bythe structure of the industry in which the firm is operating. Theability of firms to establish prices varies along the spectrum ofmarket structures. Firms participating in purely competitive marketstructures, wherein a large number of competitors each supply a smallproportion of the total output offered to consumers, will havegenerally little, if any, latitude in establishing the price of theirproduct(s). At the other extreme, the firm that is the lone seller,the monopolist, will have absolute control over what price prevails inthe marketplace. Intermediate market structures exist wherein firmsmay have some degree of control over the prices of their output. Theprinicipal determinants of the extent of this control are thesubstitutability of the products, and, most importantly, the firm’sexpectations regarding how its competitors will react to changes in theprice(s) of the firm’s product(s); the number of competitors isconsidered to be of little importance.33 Where a firm participates ina purely competitive market structure, it may sell all of its output ata market-determined price; its demand curve is perfectly elastic. Thefirm is termed a “price taker” because it cannot influence the goingprice by altering its supply of product to the marketplace. The firm78could choose to sell at a lower-than-market price, but rational firmswill not do so because they could earn larger revenues for the sameincurred cost by selling at the market price. Conversely, they couldraise the price of their goods or services, but since other firmssupply perfect substitutes, any such price increase will result in zerosales for the price-raising firm.The monopolist will, if free to do so, price so as to be able to sellthat level of output that will maximise profit. It accomplishes thisby setting its price such that, the difference between total revenueand total cost is maximised. This price is associated with the levelof output at which marginal revenue (MR) equals marginal cost (MC).Two problems may arise in monopoly pricing. If the industry is suchthat average costs are declining over all possible quantities ofoutput, it will not be possible for the monopolist to price at MR = MC,since this would mean that total revenue (TR) would be less than totalcost (TC). This situation occurs in cases where market demand is suchthat minimum efficient scale cannot be reached, even by themonopolist.34 The other problem in monopoly pricing is that in seekingto maximise profits, the monopolist will constrain output to maintainthe profit maximising price. There is a social welfare loss in thissituation as there are consumers who would be willing to pay themarginal cost of additional output, yet who are unable to acquire the79good or service due to the behaviour of the monopolist. To addressthese difficulties, some form of second-best pricing may be sought bythe monopolist (in the case of the declining cost situation), or beimposed by regulators (in the attempt to reduce welfare loss in thesecond scenario).Firms operating in intermediate market structures are faced withrelatively more complicated pricing decisions than are pure competitorsor monopolists. The complication arises in these industries since thefirm, in establishing its price(s), recognises that its competitorswill react, altering their prices accordingly. If a monopolisticcompetitor or oligopolist decides to reduce its price, for example, itshould enjoy short term increases in sales. However, its rivals,experiencing a loss of sales as the result, are certain to take similaraction once they become aware of the first firm’s actions. Where“price wars” ensue, firms may be financially damaged.Price wars can mean downward price spirals that may see prices fallto where P < MC, and firms experience negative economic returns. Whileprice wars may be a boon to consumers in the short run, over the longerterm, sustained losses may drive more marginal firms out of theindustry.35 If the number of failing firms is substantial, thesurviving firm or firms may, once the shakedown period is over, emergewith sufficient market power to charge price premia due to decreased80competition. The deliberate pursuit of this strategy is known aspredatory pricing, and is specifically prohibited by legislation inmany jurisdictions, including Canada and the United States.36In contrast to the problem posed by price wars, oligopolists mayagree to raise prices/restrict output as a group, thereby to capturethe benefits normally associated with monopolists. Such pricingagreements are known as collusive price agreements. If oligopolistscan collude successfully, they can sustain monopoly prices, and willmaximise the total revenues of the firms as a group. The same socialwelfare loss that is associated with monopoly will result. Suchcollusive agreements necessitate voluntary restraints by the individualfirms on the amount of output that each offers to the marketplace. Thegroup of firms engaging in such behaviour is known as a cartel.37Cartels tend toward instability. Cartels suffer from the unavoidablepresence of temptation for individual firms to cheat on the cartel byoffering more output to the marketplace than has been allocated to itby the agreement. Alternatively, the defector may cheat by pricecutting: offering the good or service in question to marginalcustomers, who are not consuming at the cartel price. Such cheatingbehaviour is known as defection. The defector is motivated in thisdirection because it recognises that it can earn greater surplus byselling more than its agreed upon share at the cartel price. Since all81cartel members face this temptation, it is almost inevitable that onewill take this course of action, not necessarily because it wishes tobreak the cartel, but because it believes that others are planning to,and it wishes to not be left “holding the bag,” so to speak.38A final aspect of pricing behaviour that is significant to the firmis price discrimination. Price discrimination refers to the practiceof charging different consumers different prices for the same product,based on differential valuations placed on the product by differentgroups of consumers.39 Price discrimination shifts the distribution ofsurplus away from the consumer and toward the producer. Surplus is thedifference between what it costs to produce the good or service, andwhat the consumer would be willing to pay for it.4°The most advantageous form of price discrimination, from the firm’spoint of view, is that in which each individual consumer is chargedexactly what he is willing to pay for the product in question. In thisway, the firm is able to capture the entirety of the availablesurplus.41 While ultimately desirable, the application of perfectprice discrimination is complicated by the requirement that the firm beaware of each consumer’s valuation of the product. It is more commonlypracticable for firms to only be able to price discriminate on a groupby-group basis.82In actual practice, the firm accomplishes price discrimination byidentifying groups of consumers having common traits in terms of theirwillingness to pay for the firm’s product. Successful pricediscrimination has two requirements: the groups must have differentprice elasticities of demand for the product; there must be a meansof preventing the resale of the product by members of the group beingcharged the lower price to people in the group being charged the higherprice.422.2.4 Non-Price BehaviourNon-pricing behaviour refers to actions taken by the firm in otheraspects of marketing such as product differentiation, selection ofgeographic areas in which to compete, and advertising. It alsoencompasses the realm of inter-firm agreements related to the foregoingmarketing activities.Firms also pursue activities in the non-price realm in order tofurther their interests. These activities can be classified into threegeneral areas, in keeping with the standard concepts of marketing.These areas are product, place, and promotion. The product conceptfocuses on the qualitative attributes of the good or service beingoffered, which extends to innovation and invention. Research and83development activities come under the purview of the productclassification. The place aspect of conduct refers to where the firmdecides to offer its goods or services. Place covers both thegeographic component of where the product is offered, and the terms ofconditions for sale of the goods or services. Finally, promotionencompasses those activities that are aimed at building and sustainingconsumers’ awareness of, and loyalty to, the firm’s products. Theprevalent aspect of promotion is the firm’s advertising behaviour.The quality of the firm’s product(s) is important to the fortunes ofthe firm in the marketplace. Where a firm faces competition, anydeficiency that is observed in its products will make the firm losemarket share to its competitors. Once a poor perception is created inthe minds of consumers regarding a firm’s goods or services, it can bedifficult to win back customers. This forces firms to be vigilant withregard to the satisfaction that the consumer receives from consumptionof its products. If competitors are observed to be offering bettergoods or services, the firm must be prepared to respond in kind, orendure losses of sales. The on-going attempt to sustain market share,and/or to win customers away from competitors influences the firm’sconduct.Conduct relating to changes in the qualitative attributes ofa firm’s products depends upon two primary factors: the sensitivity of84consumers to the qualitative aspects of the product(s), and theexpected reaction of competitors to such changes. Where consumers areexpected to react quickly to qualitative improvements to a product,there is, by implication, relatively high demand elasticity facing thefirm. If the product can be improved, the firm may enjoy increasedsales as the result of qualitative improvement. The question arises asto the extent of consumers’ reactions to the quality improvement.Where consumers are more sensitive to quality, the magnitude of theelasticity can be expected to be relatively great in contrast tosituations where consumers exhibit what may be termed “threshold”behaviour. Threshold behaviour means that the consumer will only reactto relatively significant changes in product quality.43 The firm canbe expected to incur additional costs in improving product quality,thus raising the marginal cost of the output. Therefore, the marginalrevenue, determined by the elasticity of demand for the product, mustbe weighed against the addition to marginal cost associated with theproposed qualitative improvement, in order to determine thedesirability of effecting such improvement to product quality. Theoptimum expenditure on qualitative improvements to the firm’sproduct(s) is identified by that expenditure that equates the marginalcost of the change(s) to the marginal revenue that is expected toresult from providing the improved product(s). For the firm that isstarting from a condition of equilibrium (ie. its current behaviour issuch that MR = MC for the product line), the change in the product must85incur no greater change in MC than the anticipated increase in MR thatwill result from offering a superior product.The place aspect of marketing concerns the firm’s behaviour withrespect to where and how it offers its goods or services to theconsumer. There are two dimensions to the where aspect of placebehaviour: there is a geographic concept and a demographic concept. Interms of geography, the firm may choose to offer its products on a widedistribution basis, or may elect to concentrate on a specific area.Similarly, the firm may choose to offer products that have a broadappeal, or may concentrate on groups of consumers that have particularneeds or preferences. The determinants of the firm’s place strategyare the opportunities presented by the marketplace, the financialresources of the firm, and the firm’s objectives.The how aspect of place behaviour concerns the methods ofdistribution that the company uses to get its products to the consumer.A firm may act as its own agent, selling directly to the consumer, ormay employ the services of other parties to do so. In some cases, theother parties are companies that are at the same stage in theproduction process with whom the firm may enter into agreements aimedat achieving synergies. In other cases, the other parties are playersat other levels in the distribution chain.86Finally, there is the non-price behaviour of firms in terms ofpromotion. Promotion is aimed at building and sustaining customerloyalty to the firm. The chief means of promotion is advertising.Advertising performs two essential functions: it informs and persuades.Informational advertising is that which relates to consumers theattributes of the firm’s products. This form of advertising helps theconsumer to identify those products that will meet their needs fromamongst the many available options in the marketplace. Persuasiveadvertising attempts to convince the consumer that there are benefitsto be obtained from the product that are outside of the strictlyperformance realm.It should be noted that the relationship between the cost ofpursuing any non-price activity and the benefits that are expected fromso-doing (as described in the section concerning product qualitydifferentiation), hold true for the other realms of non-pricebehaviour. The ultimate consideration is whether the activity willcontribute more to the firm than the cost the firm incurs in followingthe particular behaviour.2.3 Performance2.3.1 Elements of Performance87In evaluating the performance of an industry, there are two majorperspectives that should be considered: the allocative efficiency ofthe industry, and the technical or productive efficiency of theindustry. Four additional considerations under the subject ofperformance are the technical progress of the industry, the quality ofthe products being offered, and the opportunities for employment andinvestment within the industry.44 The performance of an industry israted in terms of its success in meeting society’s objectives withrespect to of the above-described dimensions.2.3.2 Allocative EfficiencyAllocative efficiency refers to the extent to which society’s scarceresources are appropriately distributed in order to meet the needs ofconsumers. The fundamental premise of allocative efficiency is thatproduct prices should directly reflect the real costs required to bringthe goods or services to the marketplace. Such prices reflect thevalue that society places on the resources being devoted to theproduction of the products, so that consumers can, by their purchasingdecisions, signal producers what goods and services are desirable to88produce. Prices should include a normal rate of return on the capitalresources that have been devoted to the industry, thereby signalinginvestors what industries should be expected to grow, hence willbenefit from the addition of capital. To the extent that pricesdeviate from reflecting actual resource costs, society’s scarceresources will be misdirected, thus allocative efficiency will be sub-optimalThe important concept regarding allocation is that of Paretoefficiency. An allocation is considered Pareto efficient if oneparty’s well-being can only be increased at the expense of anotherparty. In other words, no one may be made better off without someoneelse being made worse off. Output restrictions in an industry wherethere is a degree of monopoly power have the effect of maintainingprice above marginal cost. This means that there are some consumerswho would be willing to pay the cost of additional output, but who areunable to acquire the product because of the strategy of theproducer(s). This represents a deadweight loss to society: neither theproducer nor the consumers is capturing the benefits that additionaloutput would provide. If output was expanded to meet the demand ofthese additional consumers, with the condition that the producer(s)could retain their level of profitability, no one would be made worseoff, while the added output would make these marginal consumers betteroff. Therefore, since for monopolies, P > MC, Pareto efficiency is not89satisfied.46Another dimension of allocative efficiency concerns the allocation ofsociety’s scarce resources to alternative production activities.Ideally, resources should be so allocated so that the value that theconsumer places on the product (P) should match that which is placed onalternative uses for its constituent resources (MC). In competitiveindustries, this criterion is automatically satisfied, since price isequivalent to marginal cost: P = MR = MC. However, in monopolisticmarket structures, where P > MC, allocative efficiency problems arise:price does not reflect the true opportunity cost of the resourcesdevoted to the production of the good or service; fewer resources aredevoted to this production than are warranted by the pattern ofconsumer demand.472.3.3 Productive EfficiencyThe performance of an industry in terms of its productive efficiencywill be influenced by structure and conduct. Productive efficiencyrequires that the production process be such that no alternativeprocess exists that would provide the same level of output at lowertotal cost.48 Competitive industries necessitate that participantfirms employ the most efficient production process available; those90that do not soon find that their marginal cost of production exceedsthe industry average. By definition, such a firm will not becompetitive in the market, since the efficient firms’ marginal costswill determine the market price. Conversely, where the firm has somedegree of market power, the force of competition as a promoter ofproductive efficiency may be reduced or, in the extreme, wholly absent.Two cost increasing factors may come into play.49First, the firm will not be motivated to make the most efficient useof its factors of production. Managers will not have competitivepressures upon them to optimise the conversion of resources intofinished products. The resulting excess of costs over what could beotherwise achieved in an efficient production process is termed “Xinefficiency. ,,50The second form of productive inefficiency associated withmonopolistic industries is that of rent seeking. Rent seeking involvesthe devotion of resources into pursuits aimed strictly at strengtheningthe market power of the firm.51 Manifestations of this behaviour aremany: product differentiations having no qualitative basis, but ratherserving to make entry more difficult for potential competitors;investments in capacity expansion that are not needed to meet peakdemand variations, but rather that serve to discourage entry byallowing immediate output increases in the face of incipient91competition. Regardless of the vehicle used to accomplish astrengthening of the firm’s market power, the effect is to increase thetotal cost of producing any level of output. In turn, this increasesthe marginal cost of the product, leading to increased prices,resulting in yet further reductions of the quantity supplied to themarketplace. Rent seeking both transfers surplus from consumers toproducers, and contributes to the quantum of deadweight social loss.522.3.4 Other Aspects of PerformanceOther performance considerations are notable. Industries should beevaluated on the basis of the progressiveness of the production anddistribution processes they employ. An industry that is performingwell in other respects cannot be considered a good performer if it isnot taking advantage of the most effective means of bringing itsproducts to the consumer; such an industry would not be optimising theuse of scarce resources. Important dimensions of industrialprogressiveness are the extent to which producers innovate to increaseproductivity, and the extent to which they introduce new products tothe marketplace.53Industries may have their performance with regard to product quality.It should be such as to provide the consumer with the maximum value for92his expenditure. An industry producing poor quality products canhardly be considered a good performer. Net value benefits areassociated with product improvements that are non-linearly related toproduct price. When product quality increases, this represents anincrease in the utility that the consumer enjoys from the consumptionof the product. Coupled with a less-than-linear price increase, theconsumer’s utility is further increased due to the income effect: hismoney income can purchase a greater level of satisfaction thanpreviously available.Conversely, the industry can have negative effects on society’sutility. An industry producing “bads” as well as goods and serviceswill adversely effect social welfare. Pollution in its many forms isthe most evident example of an industry’s poor performance outside ofthe realm of product characteristics. Where an industry imposes costsupon society that are not explicitly recognised as part of the moneycost of production, negative externalities are said to have beencreated. Negative externalities have the effect of reducing theutility of affected parties. In considering the performance of anindustry, its generation of negative externalities must be taken intoaccount.Finally, an industry should provide opportunities for interestedinvestors and labour to participate therein. Industries that are93Footnotes to Chapter Three1. Reid, p.112. Needham, pp.1,23. Reid, p.124. Scherer, pp.88,895. Bonsor, p.226. Giflen [1985], p.1077. Scherer, p.708. Needham, p.1989. Scherer, pp.69,7010. Reid, p.1511. Scherer, p.40512. Brander, p.2813. Scherer, pp.12,1314. Scherer, pp.79,519-52115. Brander, pp.82,8315. Needham, pp.56,5717. Brander, p.8018. Morrison and Winston [1987], pp.53,5419. Needham, p.120. Reid, pp.26,2721. Needham, pp.121,12222. Brander, p.23223. Scherer, p.524. Waterson, p.325. Needham, p.326. Gaibraith p.8527. Gaibraith, pp. 40-41, 81-8228. Thurow, pp.32-3929. Gaibraith, pp. 79,8530. Needham, p.531. Needham, pp.10-1432. Needham, pp.55,64-6633. Needham, p.6134. Needham, p.25135. Scherer, p.20036. Brander, p.20337. Scherer, p.15838. Waterson, pp.47-5239. Needhani, pp.68,6940. Sassone and Schaffer, pp.73-8041. Brander, p.20542. Kraft, p.11743. Needham, p.9344. Reid, p.1345. Brander, pp.16-lB94Footnotes to Chapter Three (cont.)46. Needham, pp.237-23947. Brander, pp.77,7848. Needham, p.23149. Scherer, p.40550. Waterson, p.1351. Brander, p.4952. Brander, p.5053. Scherer, pp.4,595CHAPTER FOUR.STRUCTURE-CONDUCT-PERFORMANCE APPLIED TO THE AIR TRANSPORT INDUSTRIESOF CANADA AND THE UNITED STATES1. The Structure of the Air Transportation Industry1.1 Technological Structure of the Industry1.1.1 Flight EquipmentAirlines are in the business of transporting passengers and cargobetween different points on the earth’s surface. Airlines utilisedifferent types of aeroplanes in order to accomplish such movements.While airlines may differ in the scope, and scale, of the services theyprovide, the technological nature of their operations is greatlysimilar.The dynamics of aircraft operations are such that there are economiesof distance for airline f lights.1 These economies exist becauseaircraft have two operating costs that apply to every flight: the costof getting the aircraft from the ground to cruising altitude; the costof cruising at that altitude from the point at which it is attained tothe destination. Cruise altitude is that altitude at which theaeroplane operates most efficiently in terms of its fuel consumption,96speed, and range. The greatest expenditure of energy, hence fuel, ismade to accelerate the aircraft to takeoff speed, thence to climb tothe cruising altitude. Therefore, it is most economical to have anaircraft cruise as far as possible once the given cruising altitude isachieved. These fundamental operational characteristics are coanon toall different classes of aeroplanes.Aeroplanes may be classified according to two dimensions: theirmaximum gross take-off weight (GTOW), and their form of powerplant.Three classifications exist regarding GTOW: light aircraft have GTOW of12,500 lbs. or less; medium aircraft have GTOW of between 12,500 and300,000 lbs.; heavy aircraft have GTOW of more than 300,000 lbs.2Three general types of aircraft propulsion systems also exist: pistonengines with propellors, turbine engines with propellors, and pure jetengines. (Since we are concerned in this paper with airlineoperations, the following discussion will relate only to thoseaeroplane classes used in such operations.)Airlines generally utilise medium and heavy aeroplanes in providingtheir services. An exception is the use of light aeroplanes by feederairlines. Since the 1960s, airlines have used turbine-powered aircrafttypes almost exclusively. Turbine powered aeroplanes include both purejets, and turbo-props. The suitability of a class of aeroplane for agiven operation depends on the stage-length of the flight and the97volume of traffic that is expected.For long stage-lengths, all airlines use pure jet aircraft types.Pure jet types include turbojets and turbofans. Turbojets are theoriginal jet engine form. They provide power by ingesting andcompressing air at the front of the engine, using this air to oxidisekerosine fuel, and exhausting the resultant gases rearward to produceforward thrust. Turbofan engines are similar to turbojets, but withthe addition of a ducted fan that propels a sizeable amount of noncombusted air rearward utilising the power generated by the turbojetcomponent of the engine. Turbofans are more powerful, more fuelefficient, and quieter than turbojets. As the result, all newairliners have these engines as they afford airlines cost savings andfewer operational restrictions related to noise externalities.Turboprop systems employ a small turbojet engine, whose power istransmitted, via a drive-train, to turn a propellor. Turboprops areused to power aeroplanes at the smaller end of the medium size-class ofaircraft. Turboprops have the advantage of being more fuel efficientthan pure jets for operations conducted at lower altitudes. Turbopropaircraft types also require far less runway length for take-off andlanding. They have therefore have been the mainstay of airlineoperations on relatively short stage lengths (those of 600 miles orless), and/or on routes serving airports having relatively short98runways.Assembling a proper fleet of aeroplanes is important to the short-and long-term costs of the airline. The airline must have the rightnumber and types of aircraft to meet the needs of both its existingroute structure, and the anticipated opportunities that will bepresented in the future. There are three general methods of aircraftacquisition: outright purchase, capitalisation, and leasing.The purchasing of aircraft has strategic implications for the firm.The acquisition of aircraft from manufacturers involves lengthy leadtimes; as much as two years may elapse between the ordering of anairliner and its delivery. The price of modern airliners is extremelyhigh: the market price of aircraft in the A310/B767 size range isupwards of $50 million.3 The nature of the relations between airlinesand airframe manufacturers is such that large purchase orders fromloyal customers make the manufacturers inclined to offer their productsat a lower per unit price in contrast to smaller orders from lessfrequent customers.Outright purchase was more common in the earlier days of airlineoperations. The astronomical cost of contemporary airliners almostcompletely precludes one-time cash payments, as an airline would needto have cash reserves on the order of scores of millions of dollars99aircraft purchased. The only way in which modern airlines makeoutright purchases of flight equipment is through the equityacquisition of a competitor or subsidiary. Other methods of assetfinancing are more likely.The most common method of financing is through what is known as anequipment trust. The equipment trust is a form of chattel mortgage,usually held by large banks or insurance companies. Sometimesequipment trusts are provided by the manufacturers themselves- -a meansof inducing sales of their product line.4 Manufacturers may financeaeroplane purchases by accepting payments during the course of theproduction period. This payment scheme is known as the progresspayments method. Installments are made at different stages of theproduction process. The stages might be as follows: on signature ofthe contract; upon fuselage completion; on completion of the inner andouter main planes; at receipt and mounting of the engines; oncompletion of the control surfaces; on final completion. A simplerversion sees the carrier make payments at six month intervals.Besides the issue of payment schemes, the nature of the contractualagreement between the airline and the airframe manufacturer is asalient fleet planning consideration. A carrier wants to negotiatecontracts with manufacturers which allow for conditional changes suchas inflation or failure of the product to meet performance100specifications. For this reason, the inclusion of options along withfirm purchases has come into vogue in the industry.Options spread the risk of aircraft purchase over airlines,manufacturers, and leasing organisations. Airlines benefit from thesearrangements due to relief from capital payments for aircraft that haveturned out to be unusable in the short term. They also have theadvantage of being able to acquire marginal aircraft, that have turnedout to be needed, in less time than if they had to place firm orders.Options allow a carrier to be flexible and conservative at the samet i me.Leasing of aircraft is an alternative means of acquisition. Leaseaircraft are available from various organisations includingmanufacturers, airlines, and leasing companies. Leasing isadvantageous in that it is likely that an aircraft can be obtained froma lessor more quickly than from a manufacturer. This allows for fasterfleet reconfiguration than could be accomplished through purchasing.Aircraft may be acquired through capital leases or operating leases.Capital leases are essentially a form of borrowing to purchase theaircraft. As such, the lessee assumes the risks associated withownership. Conversely, operating leases are analogous to the long-termrental of flight equipment. Most North American airlines have optedfor operating leases.101Due to the volatility of the demand for airline services, and theenormous cost of purchasing flight equipment, the riskiness of aircraftownership is transferred from the airline to the leasing companythrough the leasing on an operating lease basis. The disadvantages ofoperating leases are their cost premia (charged to reflect the riskbeing assumed by the lessor), the possibility that, in times of generalexpansion of airline activity, there may not be sufficient leaseaircraft available to meet all airlines’ needs, and that the aircraft’ssalvage value cannot be recovered by the airline.5 That lease paymentsare higher than would be the capitalisation payments for owned aircraftmeans that the leasing airline must be extra vigilant regarding itscash flows during periods of recession.Leasing of flight equipment has grown in popularity. There arenumerous advantages to leasing: it eliminates the need to borrow; itprovides the lessor and lessee an investment tax credit; it allows theairline to avoid hefty progress payments (which can be as much as 20per cent of the purchase price); it acts as a hedge against inflation;it reduces the risk of technological obsolescence. Leasing is mostlikely for companies having higher earnings since they stand to benefitthe most from the investment tax shields. The aircraft are usuallyleased over a term of from 4 to 18 years, with 10 to 12 year leasesbeing most common.6102Leasing has become so popular that some airlines own but a smallfraction of the aircraft that they operate. Table 2 shows the fleetsof major United States carriers as they were composed on December 31,1990. Leased aircraft accounted for 51 per cent of the total flightequipment in their use at that time.Table 2.EXTENT OF LEASED AIRCRAFT IN U.S. CARRIER FLEETSCompany Total A/C Leased A/C % of TotalAmerica West 92 70 76American 552 315 57Continental 340 245 62Delta 444 186 42Northwest 342 130 38Southwest 106 42 40TWA 207 135 65United 462 203 44USAir 456 209 46TOTAL 3,001 1,535 51Source: Air Transport World, July 1991, p. Computer Reservations SystemsAside from aircraft, the most important technology internal to theairlines is that of the computer reservations system (CRS). Computerreservations systems are the means by which airlines distribute theirproducts to the consumer. They also permit the airline to optimise theallocation of seats on board a particular flight so as to maximise103revenue. Airlines may have their own CRS, or they may utilise the CRSof another airline for a fee. Airlines sell tickets directly throughthe CRS as operated by the airline’s own personnel, or customers canobtain tickets from a travel agent that has access to the CRS via acomputer terminal on the travel agency’s premises.1.1.3 Air Transport InfrastructureThere are two infrastructural technologies that are importantdeterminants of airline conduct which are external to the carriers:airports and air navigation systems. Airports and airways areinfrastructures provided usually by governments.Airport technology has significant influence over the conduct ofairline operations. Important aspects of airport technology areairfield capacity and terminal facilities. The most importanttechnological factor in airport operations is the capacity of theairfield. Airfield capacity is usually expressed as the number ofaircraft movements that the airport’s runways can accommodate per hour.An aircraft movement is a take-off or landing conducted by an aircraft.An airfield is considered to be at its optimum capacity when the numberof aircraft movements is such that the average departure delay peraircraft is four minutes.7 At many airports, airfield capacity is104insufficient to accommodate all of the flights that airlines would liketo operate at a given time. Therefore, a means of capacity rationingmust be employed in response to the air traffic congestion that wouldoccur if carriers were free to schedule as they wished. There are twooptions for capacity rationing that can be used to limit the congestionproblem: peak-load pricing and capacity allocation. Canadian airportsuse a form of peak-load pricing; American airports usually employ theallocation method.There are three methods that may be practiced for capacityallocation. The first method is allocation by airport committees.These committees are composed of representatives from the airportoperator(s) and the air carriers. The airport operators establish thenumber of aircraft movements that will be permitted, then therepresentatives from the individual carriers decide on the allocationof traffic between the airlines. The efficiency of this method iscontingent upon two factors. The first, that the carriers involvedwill recognise the need to find allocations that are acceptable to allparticipants. Implicit in such an allocation is the individualcarrier’s awareness of the valuation of each time slot to its ownoperations and to the operations of the other companies. By trading-off relatively lower-valued slots, the airline hopes to be able toobtain slots that it values more highly from other carriers.8105The second method is allocation by lottery. This scheme sees slotsrandomly assigned to carriers through a lottery event, after which timethe carriers are free to exchange their allotted rights in an openmarket. The rights remain in force until the time of the next lottery,when a completely new round of lottery allocation and re-allocationthrough market exchange takes place. The advantage of the lotterysystem is that the valuation of the slot times is made explicit throughthe amounts that carriers actually are observed to pay to acquiredesired times. In turn, this allows the airport operator to determinethe optimum number of slots to offer, thus more closely matchingcapacity to actual demand.The third possibility for slot allocation is the auction method.tinder this plan, the slots are offered at auction to the carrier thatis willing to pay the most for the particular slot time. The auctionscheme is similar to the lottery method in that the allocation ofrights remains in effect until the next auction occurs, and in that thecompanies are free to exchange rights after the initial distributionhas taken place.The alternative to the capacity allocation methods is the use ofpeak-load pricing. The basis of peak-load pricing is that the userwill be charged a different amount for the use of the facilitydepending on the time, day, and season that the slot time is associated106with. Peak-load pricing is intended to keep the demand for theairport’s capacity at acceptable levels by ensuring that only thoseoperations that truly value the slot time are able to get access to it.The price of a particular slot is determined through an iterativeprocess, homing in on the market value of the slot. Peak-load pricingis more efficient than direct capacity allocation methods.9There are several advantages to the use of peak-load pricing: usersthat value the slot times at least as much as the price can always getaccess at that time; the airport operator does not have to guess at theappropriate amount of capacity to ration; strategic behaviour iseliminated with respect to access rights; the revenues from the sale ofslots accrues exclusively to the airport operator, allowing the airportto better cover operating costs.A physical solution to the capacity problem would be to buildadditional capacity at existing airports, or to construct wholly newairfields. In a simple world, construction of new facilities would beundertaken when the net present value of the additional capacity met orexceeded the cost of construction. Aside from the problem of thesubstantial funding required for such projects, the decision as towhether or not to build (or expand) capacity is complicated by the factthat the required land for construction or expansion must be available,the total social costs of the airport plan must be considered, and the107possibility of strategic behaviour on the part of interested partiesmust be taken into account.A study by the United States General Accounting Office concluded thatmost large- and medium-sized airports in that country have sufficientland available on which airport capacity expansion could beundertaken.’° In Canada, the congestion-prone airports at Toronto andVancouver also have available land for capacity expansion.11 A largerobstacle to airport capacity expansion is posed by interest groupopposition.Airport planners in both Canada and the United States haveencountered strong opposition to expansion plans. This oppositioncomes largely from community groups who are concerned about theaddition to the noise pollution (and, to a lesser extent, road trafficcongestion) that will accompany increased flight operations, and fromenvironmentalists who are concerned about negative effects on wetlandareas, water courses, and urban atmospheric conditions. The true costof airport operations should include the social cost borne by thirdparties in terms of noise, road congestion, and other negativeexternalities. Determination of these costs complicates the evaluationof the true net present value of airport projects, and the studies thatare required to estimate these costs slows the process of airportexpansion. Nonetheless, there is legitimacy to these considerations,108so they must be factored into project evaluations.A further complication to the determination of the welfare effects ofairport projects is the strategic behaviour of incumbent airlines.Since these companies are often closely involved with airport operatorsand planners they can influence these groups through both direct andindirect means.The technology of airways and approach aids influences the operationsof air carriers. Airways are the highways of the air. They arenavigation tracks, provided by high-frequency (HF) and very-high-frequency (VHF) radio signals, that guide an aircraft from point topoint during the en route phase of the flight. There is a finitecapacity to the airways system due to the need to maintain separationbetween aircraft, and due to the workload on air traffic controllers.Therefore, aircraft operating on highly travelled routes may experiencedelays due to the holding procedures that are necessary to maintainaircraft separation.Where there are insufficient terminal facilities airlines may also beprevented from offering services. The biggest problem related toterminal facilities is that of gate unavailability. To load and offload passengers, an airline must have access to gate facilities.Unavailability of gates will not permit service to be offered. In the109United States, the structure of the organisations responsible forairport facilities allocation is such that, as described previously,incumbent airlines have influence over the dispensation of gate andservice counter facilities. Once again, the behaviour of incumbentcarriers with respect to control over facilities allocation may prevententrant carriers from accessing an airport. The nature of airportadministrative structures in the United States makes this an issue ofparticular concern in that country.1.2 The Supply of, and Demand for, Airline ServicesThe market for airline services shapes the conduct of the aircarriers. On the supply side, the industry can be analysed from twoperspectives: at the macro level, or at the micro level. The macrolevel involves looking at the industry on a national (or international)basis. The micro level involves looking at specific airline markets,or city pairs. Decisions taken at the micro level tend to determinethe structure of the industry at the macro level.1.2.1 The Supply of Air Transport ServicesDealing first with the supply of airline services, the output of an110airline is measured in terms of the number of available-seat-miles (forpassenger operations), or the number of available-ton-miles (for cargooperations). The supply of available-seat- or ton-miles offered by acarrier in a particular market determines the cost of operation in thatmarket.The choice of a particular aeroplane type to use in a specific marketdepends on the distance between the origin and destination airports,and the expected volume of traffic. In general, a large aircraft willhave a lower cost-per-seat-mile-produced than will a smaller aircraft.This is due to the distribution of fixed or constant costs versusvariable costs. The costs of providing service on a particular flightmay be divided into flight costs and terminal costs.Fixed flight costs include the cost of the flight crew, maintenance(which is directly related to the amount of time the aircraft flies),and the fuel that is required to get the aeroplane, at operationalweight, from the flight origin to the destination.12 The variablecosts of a flight operation include the costs of the fuel required forthe payload (passengers and/or freight), and the cost of meals for thepassengers.Terminal costs are those expenses incurred to get passengers andcargo onto and off of the aircraft. They include ticket processing (at111the airport counter), baggage handling, ground crew expenses, andaircraft servicing in terms of the cleaning of the aircraft interior.The nature of operational costs makes it more economical for airlinesto service a given market using a larger aircraft. This is because, ofthe above-described costs, only the flight costs vary with aeroplanetype utilised, for a given volume of traffic. For a flight of givendistance, the same number of people (or cargo) may be transported morecheaply using larger aircraft types because of the enhanced labourproductivity associated with larger aircraft. The size of the flightcrew varies less than proportionately with aircraft size, therefore thelabour expense associated with a smaller airliner versus a larger oneis largely similar. Since a larger aircraft has more seats to spreadthe labour expense over, it will usually have a lower per-seat labourcost. As the labour component accounts for a large proportion of totalflight cost, savings in this area translate into lower total costs peravailable-seat-mile.The other cost factor militating in favour of the utilisationof larger aircraft is the relationship between fuel requirements andaircraft size. In general, the larger the aircraft, the lower the perseat-mile cost of production. A complication to the costing of aparticular airline service is posed by the need to allocate the nonjoint coniion costs and overhead costs associated with air112transportation. Since modern aircraft carry both passengers and cargoon most flights, some allocation of the fixed flight costs must be madebetween different classes of passengers and freight. The nature of allallocation methods for such common costs is that arbitrariness isinevitable. The same holds true for the allocation of overhead costs.Airline overhead costs include those related to administration,advertising, and ownership charges for flight and ground equipment. Incosting any particular flight, the question of how to properly allocatethese costs has no definitive answer.Turning now to the macro perspective on the supply of airlineservices, the determinants of the cost level of the airline’s overalloperations warrant discussion. Of particular interest in this regardare the possibilities of various economies of airline structure.Inevitably, the question of whether or not there are economies of scalein the provision of airline services arises as the starting point insuch discussions.Economies of scale would exist in air transportation if it could beshown that the cost of supplying an available-seat-or ton-mile ofcapacity decreases with the increase in the number of points served inthe airline’s route network, or the available capacity offered by theairline.13’4 Many analysts have considered the question of scaleeconomies in air transport; they invariably conclude that constant113returns to scale exist in air transportation.15 This implies that anyairline should be able to serve a particular market equally as cost-efficiently as any of its rivals, regardless of the size of its overallnetwork.Research into airline economics has shown that there are othereconomies that exist in airline operations: economies of scope andeconomies of density. Economies of scope are present in airtransportation because of the contributions to total revenue thatresult from carrying different types of payloads on a given flight.Passengers on the same aircraft may come from different consumptiongroups, having different willingnesses to pay for the same flight. Anairline offering differentiated services on the same aeroplane canserve these different consumer groups more cheaply than couldmultiple airlines specialising in each type of service. Moreover, thecarriage of air freight by what are primarily passenger airlinesresults in scope economies because both passenger and freight revenuescombine to cover flight costs. In summary, airlines can offeravailable-seat- or ton-miles more cheaply by performing services ofdifferent qualitative characteristics on the same flight.If per-unit costs are found to decrease with an increase in theamount of capacity offered within a given network, economies of densityare said to exist. Economies of density arise because the fixed costs114of ground operations and overhead expenses associated with servicing agiven number of cities in an airline network can be spread over moretraffic units as the volume of traffic increases. Economies of densityhave been identified as having sizeable magnitude, and are found topersist up to relatively large traffic volumes.’6An important cost consideration rooted in the technological nature ofthe provision of air transport services is the absolute perishabilityof the airline product. Once the airliner leaves the departure gate,any un-sold capacity on the flight is permanently lost to the airline.Airlines can therefore not afford to be flying unused capacity aroundif they are to operate at minimum cost. Efficiency dictates that theairline take two measures to avoid this form of inefficiency: toschedule aircraft of appropriate capacity for the route segment; tooptimise the load factor for the particular flight. Capacity is statedin terms of the available-seat- or ton-miles offered in a market. Loadfactor is the ratio of revenue-passenger-miles or revenue-ton-miles toavailable-seat-miles or available-ton-miles. In order to determine theappropriate aircraft type and load factor for a particular service, theairline must consider the demand characteristics of the market to beserved.1151.2.2 The Demand for Air Transport ServicesThe demand for airline services is shaped by influences both internaland external to the firms, and by influences on both industry-wide andmarket-specific bases. The fundamental nature of the demand forairline services is that it is a derived demand. People undertake airtravel in order to get to where they want, or need, to be; cargo goesby air for essentially the same reasons. Air transportation ischaracterised as being the fastest, and one of the most reliable, meansof transport. Passengers and shippers desiring these qualitiestherefore select air transportation.The pricing of airline services affects the quantity of thoseservices that will be demanded. Airline services are normal goods inthat the quantity of services demanded is negatively related to theprice charged for the service. In Canada and the United States,airlines are free to establish any prices they wish for their services--with the exception of services in remote areas, for which fares may bedisallowed. 17Airlines can also influence the level of demand for their services.They can accomplish this by changing the qualitative nature of theservices they offer. Qualitative dimensions include the flightfrequency, flight scheduling, and in-flight service. Of these, flight116frequency and flight scheduling are most important. Travellers andshippers choose air transportation largely because of their valuationof time. While the speed of aircraft clearly is superior to that ofany other form of conveyance, what really matters to the customer isthe total time that must be devoted to transportation. Totaltransportation time includes time spent en route and at the terminal,but also includes schedule delay time. Schedule delay time is thedifference between the time that the passenger wishes to commence thejourney and the time that the flight is actually scheduled to departat. By scheduling flights at more opportune times for the customer,the airline decreases the schedule delay time for the consumer,prompting an increase in demand for its services. A means of reducingschedule delay time is increasing the frequency of flights. When thereare more departure times to choose from, the customer will select thatflight that minimises his schedule delay time. Increasing flightfrequency may also divert traffic from other modes (particularlyautomobiles) for relatively short distance trips, thereby increasingthe demand for air services.Other factors influencing the demand for a carrier’s servicesthat are controllable by the firm are in-flight amenities, on-timeperformance, related travel services (eg. rental cars, hotelarrangements), and corporate image (eg. safety record). Inter-airlinealliances also affect demand. Such arrangements offer the traveller a117service that is substantially similar to on-line travel. They increasethe number of points effectively served by the carrier, while reducingthe terminal time component, and the probability of baggage loss,associated with inter-line travel.While in the previous section it was asserted that economies of scaleare absent from airline operations in terms of cost savings, it isgenerally conceded that there are demand-related benefits accruing toairlines that have large route networks. Consumers recognise that thelarge carrier can offer services to a greater number of destinationsthan can a smaller rival. There is a greater probability that thelarge carrier can accommodate the customer’s travel plans without theneed for inter-lining than can a smaller airline. There is also areduction in the information costs required to secure passage since alarge-network airline can be expected to be able to meet the customer’sneeds more readily.18 Finally, a large carrier’s frequent flier programis more attractive to the consumer than is a smaller airline’s. Thesedemand-related characteristics confer advantage to large carriers.There are significant determinants of demand that are outside of theability of the firm to control. Foremost among these exogenousinfluences is the consumer’s disposable income. Air travel issubstantially a luxury good: increases (decreases) in consumer incomeresult in greater-than-proportional increases (decreases) in the amount118consumed. Oum and Gillen (1983) found that the income elasticity ofdemand for airline services was in the range of 1.6 to 2.5.19Therefore, airlines face dramatic shifts in demand in response to thechanges in consumers’ disposable incomes.More predictable sources of demand variation in air transportationare hourly, daily, and seasonal variations. Hourly variation occursbecause travellers wish to depart from, or arrive at, locations atparticular times of the day. Because of the synchronisation of many(primarily economic) activities, the demand for flights is not constantover the day. Hourly demand tends to peak at between 7 and 9 a.m., andthen again between 4 and 6 p.m. in major markets having substantialvolumes of business travellers.20 Daily variation refers to the shiftsof demand that are related to the day of the week. Seasonal variationrefers to the shifts of demand associated with different times of theyear. Seasonal variation arises largely because of the changes ofdestinations for vacation travellers: people like to journey to warmdestinations in the winter, and to temperate climes in the summer.Other sources of demand shifts are noteworthy. Publicity surroundingairline accidents may negatively affect the level of demand for airlineservices. Such incidents are likely to affect the general level ofdemand for air travel. In similar fashion, episodes of terrorisminvolving airline flights will deter people from flying, particularly119on routes where such situations are most likely. Geographical factorsmay also impact the demand for air transportation on specific routes.Increases (decreases) in business opportunities at a given locationwill likely lead to increases (decreases) in the demand for travel tothat location. Changes in demographics will also influence the levelof demand for services to/from particular points. Moreover, increasesin population, whether of a city, region, or country, are likely toincrease the level of demand for air transport services involving thatarea.Naturally, for an individual firm, a significant external influenceon demand is posed by the actions of rival firms. Price cuts, andservice improvements in terms of amenities, on-time performance, flightfrequency, and scheduling will all have adverse effects on the quantitydemanded of the firm’s services.Finally, the level of demand will be affected by the availability ofsubstitute means of transportation. Short-range air services areparticularly vulnerable to substitution, as alternate transport modessuch as automobiles, trains, and ships can potentially provide servicesthat are comparable when considered in terms of the total-traveltime/price trade-off.Demand elasticities influence the structure of the air transport120industry. Own-price, cross-price, and income elasticities are allimportant factors determining the both the overall, and market-specific, structures of the industry.Own-price elasticity is a measure of the responsiveness of quantitydemanded to a change in the product’s price. Research examining theelasticity of demand for air transportation has shown that air fareshave a decided impact on the volume of demand. Oum and Gillen (1986)determined that, in Canada, a 10 per cent drop in price would increasedemand by 11 to 13 per cent.2’ This implies a price elasticity ofdemand for air transport services of 1.1 to 1.3. Moreover, Oum,Waters, and Yong (1992) assembled the results of 13 different studiesof air transport demand elasticities, and state that the majority ofsuch estimates are in the range of 0.4 to 2.0 (in absolute value).22The width of the foregoing interval is accounted for by differences inelasticity between different groups of consumers of airline services.Air travellers can be generally categorised into two groups: businesstravellers and vacation travellers. Business travellers are expectedto be less price-sensitive than vacation travellers because they areundertaking non-discretionary trips, and because the person travellingdoes not pay for his own ticket. Oum, Waters, and Yong report commonelasticities for business travellers of between 0.65 and 1.15, with themajority being somewhat less than unity.23 Conversely, vacationers’121elasticities of demand are substantially higher. This group oftravellers is making discretionary voyages, and will fly only if theprice of air travel fits their budget. Most estimates of demandelasticity for this group are greater than unity: in the range of 0.40to 4.60, with those above 1.4 being common.24Turning to the matter of income elasticity of demand, it has beenaffirmed that there is considerable volatility in the relationshipbetween income and the quantity of air transportation consumed. Oumand Gillen (1983) estimate that income elasticity of demand for airtransportation is in the range of 1.6 to 2.5. In general, businesstravellers are seen to have lower responsiveness to changes in incomethan are vacation travellers. Oum and Gillen estimate the incomeelasticity for the former group to be 1.5, as opposed to 2.1 for thelatter.25Finally, the cross-price elasticity of demand for an airline’sservices should be considered on two dimensions: the effect of a changein the price of some competing mode’s services; the effect, for anindividual firm, of a change in a competitor’s services prices. Wherecompeting modes are concerned, cross-price elasticity becomes an issuein short stage-length markets. The consumer may be confronted with achoice between air travel and other modes only to the extent that airtransport’s qualities of travel time and cost are comparable to that of122alternative modes.Where the responsiveness of demand facing the individual firm isaffected by changes in the prices of competitors’ services there can bedifferential elasticity effects expected depending on the direction ofprice changes. Since airlines provide basically indifferentiableproducts, there should be substitution of one’s services for another ifprice differentials appear on the same route.Airlines should face kinked demand curves for any given market. Ifan airline raises its fares in a market, other companies should not beexpected to follow suit, as they could capture more traffic bymaintaining their existing fares. Conversely, if an airline drops itsfares, it should expect that its rivals will do the same. Thus, demandfor the individual airline’s services in a particular market shouldfeature relative inelasticity with regard to price decreases, andrelative elasticity with regard to price increases.1.3 The Reulatory Structure of the IndustryA significant determinant of the structure of the airline industry isthe character of governmental regulation concerning the industry.Canada and the United States are somewhat unique countries in that123their air transport industries feature economic deregulation. Since1978 in the United States, and 1984 in Canada, the relevant governmentshave had a hands-off policy with regard to the economic affairs of aircarriers.1.3.1 The Era of RegulationPrior to deregulation, the economic affairs of the airlines wereclosely monitored and largely controlled by government agencies. Theseagencies were the Civil Aeronautics Board (CAB) in the United States,and the Air Transport Committee (ATC) in Canada. The CAB and ATCregulated which carriers could serve which routes, and determined thefare structure that would prevail in each market. Additionally, theregulators could award subsidies, and had to approve any mergers and/orinter-firm agreements.26 Flight frequencies were not directlycontrolled by the CAB. In Canada, the ATC could specify not only thefrequency of flights to be provided, but also the type(s) of aircraftthat were to be operated, the number of intermediate stops that had tobe performed, and the types of traffic that could be carried.27Entry and exit from markets was directly controlled during theregulation era. The authority to operate services on a route wasgranted to a carrier on the basis of “public convenience and necessity124(PCN).” The onus was on the entrant carrier to show that thetravelling public would benefit from the inception of its services.Incumbent carriers, who had substantial influence with the regulatoryagencies, could often block such entry. Airline firms were generallyunable to construct efficient route networks because of entry and exitrestrictions. They were also often required to provide uneconomicservices on low-volume routes on the basis of cross-subsidy from moreprofitable markets. Inefficient resource allocation was inimical tothis structure of air transport services.Under this regulatory regime, the responsible government agenciesdetermined fare structures. These were based on the cost structures ofthe participating airlines in a given market. Qualitative aspects ofservices were also dictated by regulatory agencies. This led to highprices for the consumer because of the absence of competitivediscipline on fare levels, and the fact that airlines could pass costincreases on to the ticket buyer quite readily. Major beneficiaries ofthis structure were airline owners (especially those with holdings inairlines having better-than-average efficiency), and airline labour(whose wage demands could be accommodated by passing the resulting costincreases to the consumer).28The regulation of prices prompted airlines to engage in servicecompetition. Services that were provided are considered to have been125of a higher quality, in terms of in-flight service, direct flights, andaircraft types, than was actually valued by consumers at large. Thishigh level of service quality increased the cost of providing airtransportation. Since regulated prices were based on the average costof service provision in the particular market, unnecessarily highservice quality reinforced the high price for air travel. Thisdiscouraged travellers who would have been prepared to pay to fly at aprice dictated by a lower level of service quality; regulated highprices prevented these types of services from being offered.29Travellers also had to endure a substantial degree of inter-lining:flying on more than one carrier from point of departure to destination.Inter-lining was necessary because often the airline that served thedeparture airport did not have authority to carry the passenger to theultimate destination. The air traveller had to contend with circuitousroutings, terminal delays due to the need to collect his baggage fromone airline and transfer it to another, and schedule delays due to thelack of coordination of flight times between carriers. Coupled withthe monopolistic nature of air fares, these time costs presented thepassenger with a significant overall cost of air travel.The foregoing factors influenced the concentration of the airlineindustries in both Canada and the United States. In Canada, theindustry was dominated by the crown carrier, Air Canada. The other126major Canadian carrier was Canadian Pacific Air Lines (CP Air). Thesetwo companies performed nation-wide and international services. Therewere five regional air carriers: Pacific Western Airlines, Transair,Nordair, Quebecair, and Eastern Provincial Airways. Though there was areasonable number of carriers, the allocation of routes was such thatcompetition at the individual route level was minimal; the averagenumber of competitors per route in Canada was 1.3.30 The regulatorystructure divided the country into regions, each being served by thetwo national carriers (Air Canada and CP Air), and (usually) one of theregional carriers. The regional carriers acted as feeder services forthe two trunk airlines. There were no substantial equity positionsheld by any of the above-listed airlines in each other.In the United States, there were many large airlines prior toderegulation: at the inception of deregulation, in 1978, there were 23large carriers providing scheduled passenger services using jetaircraft in the US.311.3.2 Economic Deregulation of Air TransportationIn 1978, the United States Congress passed the Airline DeregulationAct. This Act provided the framework for the removal of governmentalcontrols over the economic affairs of US air carriers. Impetus for127deregulation was provided by research regarding both the efficiencyconsequences of the existing regulatory structure versus what could beanticipated under deregulation, and by direct comparisons of thefederally regulated air carriers’ performances with the performances ofintra-state carriers who were not subject to the federal regulatoryregime.The essential feature of the Act was the elimination of the controlsover prices and entry that were exercised by the CAB. The airlineswere to be able to establish fares at their own discretion, withrestrictions being limited to those of market forces. Airlines couldoffer services in any market they so desired, providing that they metthe “fit, willing, and able” (FWA) criteria. “Fitness” required thatthe airline be owned by US citizens. “Willingness” meant simply thatthe firm have the desire to provide such services. “Ability” referredto the airline’s possession of the technical expertise to maintain asafe operation, and that the company had the proper insurance asrequired by federal regulations concerning air transportation.Canadian deregulation was officially ushered in with the NationalTransportation Act (NTA) of 1988. De facto deregulation of theCanadian industry was introduced prior to this time. In 1979, inresponse to the deregulation of the United States industry, theCanadian government made the first substantial move away from close128control over the economic conduct of the carriers. In that year,restrictions on the ability of CP Air to compete with Air Canada in thelucrative transcontinental markets were removed. Charter carrierWardair was given freedom to offer advance booking charters (ABCs) ondomestic routes. Prior to this time, Wardair had only been able tooffer ABC services on international routings.Between 1979 and 1984, Ottawa introduced several general reforms topolicy regarding charter services. These reforms made such servicesmore readily available to consumers by reducing the restrictionssurrounding access to charter fares. ABCs had originated as discountfares available only to “affinity groups”: groups of people having someidentifiable relationship. Eligibility for ABCs also was premised onbooking seats far in advance of the departure time, without exception.Ultimately, policy was reformed to the point where ABCs were availableto individuals (with no group affiliation) and a portion of seats on acharter flight could be purchased without any advance requirement.Effective deregulation occurred in Canada in 1984. The “New CanadianAir Policy” was announced on May 19 of that year. It was not a pieceof legislation, but rather a policy statement. It made use of thepowers of the Cabinet to make orders allowing conduct outside of theletter of administrative law. The Policy gave the airlines the freedomto offer fare discounts at will and removed the mandatory restrictions129surrounding eligibility for these fares. Entry and exit from routeswas left to the discretion of the carriers, with some advance notice ofservice termination being required. Capacity and aircraft typerestrictions were eliminated, as was the Regional Carrier Policy. TheNew Canadian Air Policy based the granting of operating licences on thecriteria of “fitness, willingness, and ability” rather than publicconvenience and necessity. These reforms were primarily applicable tooperations in southern Canada; those in northern Canada remainedclosely monitored and controlled.Official deregulation in Canada came with the NTA, which came intoforce on January 1, 1988. As in the United States, governmentalcontrol over the normal economic affairs of air carriers was terminatedunder the NTA. Canada too opted for FWA criteria for the granting ofoperating certificates for air carriers. The pricing freedomsintroduced under the New Canadian Air Policy are legislated under theNTA. There are neither restrictions on capacity nor on flightequipment to be operated. Entry and exit has been decontrolled, exceptfor services wholly within, or to or from, northern Canada.It must be noted that the deregulation of air transport in bothCanada and the United States only pertains to domestic services.International services, including those connecting points in Canadawith points in the United States, continue to be regulated in terms of130access, capacity, and fares.1.3.3 The Existing Structure of the Airline Industry inCanada and the United StatesIn Canada, the airline industry has evolved into a duopoly marketstructure. Two major airlines, Air Canada and Canadian AirlinesInternational, provide international, and nation-wide domesticservices. Their affiliations with regional or commuter carriers are soextensive that 92.8 per cent of all revenue-passenger-miles consumed inCanadian domestic markets were performed by these two organisations in1992.32There are four substantial charter carriers in Canada: Air Transat,Canada 3000, First Air, and Royal Airlines. Revenue-passenger-milesaccounted for by these carriers represented the majority of theremaining 7.2 per cent of the total domestic output of Canadian aircarriers in 1992.In the United States, the industry has come to be dominated by fivemega-carriers: American Airlines, Delta Airlines, Northwest Airlines,US Air, and United Airlines. There are three additional carriers thatcontinue to provide significant services, but have been in such131precarious financial positions that their survival, even in the mediumterm, is in doubt: these are Continental Airlines, Pan American WorldAirways, and Trans World Airlines (TWA).Like their Canadian counterparts, the major American carriers havecreated extensive networks involving feeder carriers.TaxationA final form of regulatory structure that influences airlineeconomics is the regime of taxation that the firms are subject to.Countries have differing tax regimes confronting their carriers.Canada and the United States impose taxes on air carriers in such a waythat the effective cost of doing business in the two countries differssignificantly.McKenzie, Mintz, and Scharf (1992) concluded that the Canadian aircarriers are decidedly disadvantaged versus their American counterpartsby the tax regime in Canada.34 Their analysis showed that effectivetax rates on fuel and capital are lower in the United States than inCanada. Canadian labour inputs had a somewhat lower effective tax rateas compared to the United States. These effective tax rates appear inTable 3. The differential in taxes on aviation fuel is particularly132dramatic.Table 3.EFFECTIVE TAX RATESInput Canada United StatesFuel 18.1% 5.6%Labour 5.6% 9.2%Capital 22.6% 19.5%Source: Directions, pp.1659,1663It is interesting to note that the differential in effective labourtax rates between the two countries is 3.6 per cent. This is the exactpercentage that Windle (1991) found was the labour input cost advantageto Canadian firms in comparison to US carriers.35The effective tax rates affect the cost level of any quantity ofoutput. Where taxes are higher, the cost of producing a given outputis similarly higher. Since Canadian companies on balance face greatereffective tax rates, their input costs are higher than those of U.S.carriers. These taxes have created a cost structure that hasnecessarily contributed to the higher average fare levels experiencedin Canada relative to the United States.1332. Evidence on the Conduct and Performance of Air Carriers2.1 The Pricing of Airline ServicesThe fare structures of airlines in the deregulated era are aimed atprofit maximisation through the use of price discrimination. Being atliberty to establish their own prices, airlines have sought tostructure fares so as to tap into the differential surpluses ofdifferent groups of consumers. An airline’s primary pricing goal is toset prices so as to sell as many full fare seats as possible perflight. The remaining seats are then sold to more price-sensitivetravellers, to maximise per flight profit.36The passenger that must undertake a non-discretionary journey will berelatively price-insensitive: he will fly, even when fares arerelatively high. Business travellers are the major group of passengersthat exhibit low price elasticities of demand. Three factors help tomake business travellers price-insensitive: the need to travel oftenbecomes evident very close to the desired departure time; the companypays the cost of the traveller’s ticket; the cost to the company of theticket is less than the stated price because of the tax-deductibilityof travel expenses.37As the result of the characteristics of business travellers’ demand,134airlines price services aimed at this market segment at relatively highlevels. In return for the price premium, the business traveller isable to book a seat with little advance notice, is able to change histravel plans reasonably readily and without penalties, and, usually,can travel at hours that allow him to maximise his time productivity.With regard to the leisure traveller, airlines recognise that pricewill significantly impact the volume of such traffic. Therefore,carriers offer a wide variety of fares intended to extract the maximumof consumer surplus from this group, while maintaining demand so as tooptimise per-flight profits. Discount fares are the means by which theairlines entice leisure travellers to opt for their services. On anygiven flight, there may be many different fares being charged for thewhat is apparently the same quality of service; the passengertravelling on a discount fare may be paying but 20 per cent of theprice paid by the full fare customer.38 Airlines face considerablecommon and joint common costs: different classes of passengers are allflying on the same aircraft, and using the same terminal facilities.Given the relatively good ability of airlines to practice pricediscrimination, coupled with the considerable extent of non-separablecosts in air transportation, air fares can be most appropriatelycharacterised as value-of-service prices.39The charging of differential fares results in different revenues per135passenger-mile. Revenue-per-passenger-mile is referred to as yield.To optimise flight revenues, the firm must allocate seats to thedifferent passenger categories in a systematic, efficient manner,making its allocation decisions on the basis of the different passengergroups’ demand elasticities. This practice is termed yield management.The successful implementation of yield management rests on threebases: that demand for a particular flight is predictable; that thediscount fares will induce demand for the service; that diversion ofpassengers having a higher willingness-to-pay for a seat on the flightcan be substantially prevented. The emergence of computer reservationssystems has allowed airlines to develop sophisticated yield managementactivities. They provide the means of storing historical dataregarding demand for individual flights and fare offerings.With the possession of a sufficient number of observations regardinghistorical demand, the airline may accurately predict the number offull fare passengers to expect for a given flight. This allows thefirm to determine the number of seats to allocate to full faretravellers, leaving the remainder of capacity available for discounttravellers. The airline takes into consideration the variability offull fare demand when allocating capacity. Since full fare travel isusually booked relatively near to the departure time, the properallocation of such capacity is both complicated and crucial:136complicated because discount travellers book in advance, so are firstto make reservations, hence affirming claim to capacity; crucialbecause it is full fare customers that offer the highest yield, thusmaking the greatest contribution to flight revenue. The proportion offull fare travellers that are unable to obtain seats is referred to asthe spill rate. An important yield management objective is theminimisation of spill rate.The airlines have discovered that the magnitude of the discount, andthe severity of the restrictionsdiscount are significant determinantsof induced and diverted air travel over alternativealso make full fare travellers moreTo prevent this, the carriers erectFences are restrictions regardingseveral forms: the time in advancemust be purchased; the minimumreturnat thereturnsurrounding eligibility for theof the quantum, and proportions,Deep discounting induces people tomodes. However, sizeable discountslikely to opt for discount travel.fences” around the discount tickets, which may takeof the flight date that the ticketinterval between the departure andSaturday nightchanges in theDeregulation, in both Canada and the United States, has had a decidedimpact on average air fares. Ironically, fares have generallytimes; a requirement that the traveller spend adestination; penalties surrounding the making ofdate or time, or with regard to trip cancellation.137increased for short-distance flights--those of up to 1,500 kilometers.Morrison (1992) has shown that, in Canada, increases in these fareshave been in the range of-0.5 to 25 per cent in real terms. Forflights in the 1,500 to 2,500 kilometer range, real fares have beenlargely unaffected by deregulation. Long distance flights have seenfares decrease by between 0 and 25 per cent.4° These results have beeninfluenced by economies of distance and the distortion of fares thatexisted under regulation. The non-linear decay of yield with flightdistance has a technological basis, so has not been affected byregulatory changes. Conversely, short-distance air fares were undervalued during regulation. Deregulation led airlines to increase faresfor these services, to reflect the true cost of their provision.Oum, Stanbury, and Tretheway (1991) examined the trend in average airfares for the United States and Canada over the period 1977 to 1989.This time frame included that from deregulation in the United States,to one year after official Canadian deregulation. Their study showedthat, in constant dollar terms, fares in the United States decreased atthe rate of 8 per cent over the period 1977 to 1984, while Canadianfares, which remained regulated, decreased at the rate of 4 per cent.1984 saw the substantial deregulation of Canadian air transport underthe “New Canadian Air Policy.” From 1984 to 1989, Canadian air faresdecreased at a 10 per cent annual rate, whereas US fares declined at 15per cent.4’138Oum, Stanbury, and Tretheway (1991) also examined the differential inthe magnitude and utilisation of fare discounts on domestic serviceswithin the two countries. US companies have offered progressivelylarger average discounts over the period from 1980 to 1989. Theaverage discount in United States markets in 1989 was 63 per cent. Theaverage discount in Canada has also progressively increased, with the1988 average being 45 per cent.42The proportion of passengers travelling on discounted fares has beendifferent in Canada and the United States. In the United States, theproportion of total travellers flying on discounted fares has leveledoff at approximately 90 per cent. In Canada, some 60 per cent ofpassengers travel on discounted fares.43A less glowing report on the pricing behaviour of United Statesairlines is presented by Dempsey (1990). Dempsey examined real yieldsover the period from 1967 to 1988. He found that real yield decreasedduring these years, with the post-deregulation rate of decrease beinglarger than that during regulation. Dempsey calculated 1967-1977 realyields to have decreased at the rate of 1.7 per cent; 1978-1988 realyields were determined to have declined at the rate of 2.4 per cent.However, Dempsey shows that, when real yields are adjusted to accountfor changes in real fuel prices, the post-deregulation performance is139not as good as it first appears. The fuel-price adjusted rates ofdecrease were 2.7 per cent for the pre-deregulation period as opposedto 2.0 per cent since deregulation.44 Critics of Dempsey’s approachpoint out that the author’s selection of 1978 is inappropriate as thestarting year for the deregulated era in the Unites States, as fareswere liberalised beginning in 1977, so 1976 should be considered thewatershed year.Generally, analysts have agreed that deregulation has brought loweraverage prices for air travel.45 However, as discussed previously, thechange in price levels has not been unidirectional from market tomarket. Aside from the technological basis, in terms of economies ofdistance, for price increases in some markets, the question ofconcentration arises with respect to pricing. The effect of the numberof carriers in a market, and the distribution of traffic amongst thecarriers, must be considered with respect to their influence on airfares.Borenstein (1989) examined the effect on market concentration on farelevels in city-pairs having a significant proportion of total trafficbeing carried by a single airline. The airline in each case operated amajor hub at either the origin or destination end of the route.Borenstein compared the yields for the dominant carrier in each case,to those for other carriers serving the routes. He found that there140was a direct relationship between an airline’s share of traffic on aroute and at an endpoint airport, and the magnitude of the pricepremium that it was able to charge. He estimated that a 1 per centincrease in route traffic share led a carrier to raise price by 0.03 to0.22 per cent.46 He also calculated that, where the carrier isdominant at both ends of a route, it may sustain prices that average 6per cent higher than those of competing airlines.47Morrison and Winston (1989), in analysing the effects of the numberof competitors on a route, found that the withdrawal of a company froma market could have a substantial effect on the average price chargedin that market. They calculated that fares increased from between 2and 32 per cent when a competitor ceased operations on a route. Theyfound that there was a negative relationship between the increase inthe average fare and the number of competitors remaining in the market.Naturally, the greatest fare increases were experienced where thecessation of service by a firm resulted in a monopoly on the route.48Morrison (1992) examined the trend in Canadian airline pricing overthe years 1983 to 1990. Morrison concluded that there was nodiscernable effect of route concentration on Canadian airline prices.He attributes this to the contestability of Canadian markets.49 Thecontestability hypothesis suggests that an operating airline cannotcharge supra-normal prices without inviting entry by competitors141because of the near-risklessness of entry into a particular airlinemarket.The difference in the Canadian versus the US experience with respectto the effect of concentration on price levels is likely to be rootedin airport access. The control that US carriers often have over hubairports imposes significant costs on competitors wishing to introduceor increase their services on routes involving the hub airport. Theexistence of such significant costs violates the riskiess entry premiseof the contestability hypothesis. The outcome is that incumbent firmsreap some degree of monopoly rents. The nature of the demands oncapacity at Canadian airports, and of the capacity allocation mechanismused in Canada, have doubtlessly prevented the charging of price premiaby Canadian carriers.Comparing Canadian air fares to those in the United States, it isapparent that Canadian prices are somewhat higher than those south ofthe border. Morrison (1992) found that Canadian fares in 1990 were, onaverage, 7 per cent higher than US fares.5° He points out, however,that there is some difficulty in making direct comparisons betweenprices in the two countries because of the way in which fare data arereported. Canadian services, involving intermediate stops, have thefares disaggregated between flight legs; US fare data report the totalfare from the origin to the destination.51 This practice is cited as a142reason why short-haul fares in Canada appear to be lower than those inthe United States; the distance taper effect is obscured by theapportionment of long-haul fares by individual flight legs.2.2 Airline Marketing PracticesConduct relating to the development and applications of computer-reservations-systems (CRSs) has had a significant impact on theperformance of airlines. It is the CRS that allows the air carrier toimplement yield management and the accompanying price discriminationthat exists in the provision of airline services. Airlines that ownCRSs have garnered competitive advantages over non-CRS-owning carriers.These advantages stem from numerous sources: the hierarchy of flightinformation presentation on the systems; the fees that are charged tonon-owners for “hosting” their flight information on the CR5; therewards to travel agents for utilising the systems; the control overbookings in geographic areas where a particular CRS is dominant.One of the first anti-competitive behaviours that was noticed withregard to CRSs was the practice of listing the owning airline’s flightsfirst on the display screen. The effect of this practice was toincrease the likelihood that the travel agent would book the travelleron the vending-airline’s flights, rather than those of a hosted143airline, even though the latter might actually offer the consumerbetter utility. This practice is referred to as CRS bias. In theUnited States, legislation has been enacted to force CRSs to listflight information according to a specified hierarchy. The hierarchyused could be either actual flight time, or total elapsed time. InCanada, the solution had been the amalgamation of the reservationssystems of the two major carriers into one system.The CRS-owning carriers receive substantial revenue from hostingother carriers’ services on their reservations systems. Fortunereported that, in 1985, American Airlines’ revenue from operation ofits “Sabre” CRS was $336 million; operating profits associated withthis revenue were $143 million.52 This represents a rate of return onCRS operations of nearly 50 per cent. In contrast, the overall profitmargin of airlines has hovered in the single digits--when not in therealm of losses.In the United States, the two largest CRSs are operated by AmericanAirlines and United Airlines. The United States Department ofTransportation found that these two systems accounted for 75 per centof ticket sales in the US. The Department determined that the feescharged to hosted airlines were roughly double the cost of providingthe services.53 Therefore, conduct relating to CRSs has transferredprofits from non-CRS-owning airlines to those having these systems.144There are advantages accruing to the CRS vendor from its relationshipwith the travel agent. Travel agent commission overrides (TACOs) andthe “halo effect” are two forms of vendor-agent conduct that are viewedas having anti-competitive results. TACOs see the commission paid tothe travel agent increase non-linearly with the number of tickets sold.This gives the agent the incentive to sell as many seats on thevendor’s flights as possible, so as to maximise the average commissionthe agent receives. The “halo effect” refers to the practice offavouring the vending carrier, and is related to the TACO and the factthat the vendor supplies a more detailed and up-to-date package ofinformation relating to its own flights compared to those of hostedairlines.54In US markets associated with major hub airports, the concentrationof ticket sales through the dominant carrier’s CRS is often much higherthan the national average. For example, Sabre processed 88 per cent ofall tickets sold in the Dallas-Fort Worth market in 1985. This levelof concentration of CRS patronage in hub airport markets is consideredto present a significant barrier to entry.56 In turn, this reinforcesthe ability of the dominant carrier to charge price premia in theconcentrated market.In Canada, in 1989, 82 per cent of all air segments booked were done145so using the Gemini Reservations System.57 At that time, Gemini wasowned by Air Canada, Canadian Airlines International, and Covia (theCRS of United Airlines); each partner owned one third of the company.While the partnership of the country’s two major carriers in a singleCRS may have been beneficial to consumers in that the sort of CRSrelated problems that were outlined above for the United States areabsent from Canada, there remained the problem of the anti-competitiveeffects that Gemini posed for other Canadian air carriers. Inresponse, Canada’s Competition Tribunal established rules of conductfor Gemini to follow, so that the CRS would fairly represent the flightinformation of its owners and of other carriers.An outgrowth of the introduction of the CRS was the development offrequent-flier-programs (FFPs). The principle of the FFP is that theconsumer can be enticed to travel almost exclusively on one airline byoffering him free travel or seat up-grades based upon the number ofmiles he has flown with the company.The nature of these programs confers advantages to carriers havinglarge networks: the traveller can more readily accumulate travel“points if an airline can fly him to a wider variety of destinations;the accumulated bonus miles can be use for trips to more and varieddestinations.58 Therefore, the customer has an incentive to patronisethe FFP of the airline that operates the greatest number of flights,146and to the greatest number of destinations, from his “home” airport.Since network size confers an advantage in terms of theattractiveness of a FFP, smaller airlines are put at a distinctdisadvantage with respect to the demand for their services. Tretheway(1989) noted that this may force the smaller company to offer pointsmore readily: the customer of the smaller carrier will be givenequivalent reward after having flown less distance than with the largerairline. This adversely affects the cost structure of the smallercarrier, since it must endure both the opportunity, and out-of-pocket,costs of awards of free travel more frequently than its largerrivalsFFPs have been particularly effective at capturing businesstravellers’ loyalties. There is a principal-agent effect in thepurchasing of business air travel: the company buys the tickets, yet itis the employee that accumulates the travel points. The company isable to offer the travel advantages of the FFP, without incurring taxexpenses for itself or the employee. The United States GeneralAccounting Office (GAO) queried travel agents about the extent to whichthe pursuit of FFP point accumulation dictated their customers’selection of flights. The GAO found that 81 per cent of the travelagents they surveyed believed that, more than half of the time, theircustomers chose flights with the aim of accumulating travel points on147FFPs.6°The principal-agent problem can be cited as a cause of a non-optimalamount of air travel being consumed. The pursuit of free leisuretravel may encourage unnecessary business trips, trips involving point-accumulating, circuitous routings, or travel on carriers that do notoffer the least expensive fares for a given origin/destinationcombination. The result is an excessive level of airline servicesbeing consumed, with an associated reduction in social welfare.6’To overcome the advantages that large airlines have with respect toFFPs, coalitions of some smaller carriers have agreed to honour pointsearned on each others’ FFPs in the hope of luring the point-consciousconsumer away from the larger carriers. More significantly, FFPdisadvantages are considered to be a leading motivation for theconsolidation of smaller airlines into larger networks.62 This type ofconsolidation is not based on any cost-savings, so does not improve theproductive efficiency of the industry.Another effect of the growth of FFPs is the implications that freetickets have on the availability of seats for paying passengers.Social welfare considerations dictate that the paying-passenger shouldhave precedence over those travelling on points when it comes tosecuring a seat on a given flight. However, the nature of the booking148process does not discriminate between these types of customers. Theconsumer who is willing to pay the actual cost of the service will findthat he is unable to fly at his desired time because the seat thatwould have been available to him has gone to a “points” ticket-holder.Social welfare is thus adversely affected because capacity is allocatedon other than the basis of willingness to pay.2.3 Airline Conduct in Other Non-price AreasIn order to capture economies of both density and scope, airlineshave opted to construct hub-and-spoke route systems. These routestructures increase average traffic densities by consolidating trafficbound for various ultimate destinations onto flights through hubairports. Hub-and-spoke networks have had important effects on airlinecosts and on the demand for airline services.McShan and Windle (1989) found that, in the United States during theperiod 1977 to 1984, airline costs were reduced by 0.11 per cent forevery 1 per cent increase in the proportion of the airline’s departuresemanating from hub airports.63 As evidence of the recognition by thecarriers of the advantages of hubbing, McShan and Windle also observedthat the rate of increase in hubbing was 6.9 per cent per year sincederegulation; the pre-deregulation rate was 1.7 per cent per year.64149The scarcity of airport capacity in the United States, the mechanismsfor allocation of capacity amongst airport users, and the recognition,by the carriers, of the cost advantages of hubbing have promptedairlines to establish “fortress hubs” where possible. The fortresshub is one in which a single airline dominates the traffic into and outof that airport. By dominating an important airport with a hubbingoperation, an airline is able to capture a large proportion of thetraffic to and from that city. Moreover, it is able to (as discussedpreviously) extract economic rents from consumers. Where a strategichub is shared between carriers, it is said to be a “strong hub” foreach of these airlines.On the demand side, the practice of hubbing has been shown to haveincreased the attractiveness of air travel. Hubbing allows thecarriers to serve more destinations from any given point of originthrough the use of on-line connections. The increase in the trafficdensities enables the carriers to offer more frequent flights, therebyreducing the schedule delay time confronting the passenger.The practice of hubbing has had negative effects on passengerutility. Hubbing adds to the flight time of the average passenger: theextra distance that is associated with non-direct flights, the timethat is required to land at the hub, the terminal time to board the150connecting flight, and the time it takes that flight to take-off forthe next leg of the journey. These additions to total travel time maybe substantial where there is the need to travel through congested hubairports, wherein congestion-related delays are significant.Oum and Tretheway (1990) estimated that the disutility of making atransfer between flights was equivalent to a time delay of 45 minutes.Assuming a valuation of time of $30 per hour based on the estimates ofKanafani and Gobrial (1985), Oum and Tretheway assign the disutility ofhub connection to be worth $22.50 (in 1990 dollar terms).65 Thedisutility of the extra travel time associated with hubbing is expectedto vary directly with the passenger’s valuation of time.For the business traveller, who presumably has a high opportunitycost of time, the circuity of routing through hub-and-spoke routestructures represents a negative aspect of the conduct of the airlineindustry, US air carriers have not eliminated direct services,probably in recognition of the needs of the business traveller. McShanand Windle (1989) observed that the proportion of total departuresaccounted for by hub airports was 38.6 per cent in 1984.66 If astrictly hub-and-spoke network was being employed, departures from hubairports would account for 50 per cent of all departures; McShan andWindle did not observe this division of departures. The provision ofdirect services has continued in the United States. Aside from meeting151the needs of passengers having high time valuations, direct servicesare provided where there would be significant “backtracking” if theservice was conducted through the nearest hub airport, or where thevolume of traffic between two non-hub airports is sufficient thatdirect service is warranted. Huston and Butler (1988) showed thatfollowing the merger of TWA and Ozark Airlines, the new carrier wasable to offer each city in the newly created network an average of 28.7per cent more destinations than the two companies could before theiramalgamation.67 The merger afforded the new company a better routestructure in general, and specifically, better utilisation of its hubat St. Louis-where TWA and Ozark both had hubbing operations prior tothe merger. The merger did have the effect of raising averagerestricted fares by 19.2 per cent, and unrestricted fares by 5.7 percent, on the airline’s flights through the St. Louis hub.68 Moredramatically, average unrestricted fares rose by 13.1 per cent, whileaverage restricted fares increased by 45.9 per cent, for TWA-Ozark’sservices having St. Louis as an end-point.69Airlines in the United States recognised early on in the era ofderegulation that the ability to substantially control access toimportant airports afforded them significant advantages in terms ofcosts and competitiveness. The desirability of establishing a huboperation at an airport depends on the location of the airport vis-avis the overall network that the airline intends to operate, upon the152origin and destination traffic characteristics of the city that theairport serves, and upon the physical capacities of the airport. Theairlines that have been the most successful in the United States arethose that have been able to establish, and defend, fortress hubs atkey airfields.Each of the five largest US air carriers has a major hub operation atat least one airport. There is no location at which two of theseairlines have their most important hub. This is evidence that thecarriers realise that there must be domination at their central hub inorder for the network to be effective. Note also that there is noprimary hub that is located at an airport that is near the margins ofthe continental United States- -with the possible exception ofPittsburgh. Effective hub-and-spoke layouts require a somewhatcentralised main hub.An airfield capacity problem exists at four congested airports in theUnited States: O’Hare (Chicago), LaGuardia and John F. Kennedy (NewYork), and National (Washington). In 1969, the 11High Density Rule”came into effect at these airports. This policy required carriers tomake advance reservations for slot times. There are separate slotallocations for air carrier, commuter, and international services.7°Originally, the slot times were administered by airline153representatives. Though this system worked well, its format wasrendered inappropriate with the advent of deregulation. Withderegulation, a conflict of interest emerged regarding this slotallocations scheme in that established airlines on the allocativecommittees became responsible for conferring slots to new rivals.Since the new airlines had no representation on the committees, theyhad no negotiating strength with which to secure access to slot times.A 1985 amendment to the High Density Rule allowed airlines tobuy, sell, and lease slots. The objective of this amendment was tocreate a market for the airport slots. Existing slot allocations weregrandfathered, and the committees were disbanded. However, thestrategic nature of the possession of slot times resulted in thehoarding of slots by incumbent carriers. Sales of slots actuallydecreased since the inception of the program. Instead, slot-holdingairlines have preferred to lease their unused slots to other carriers.Slot leasing has increased since 1986. In many cases the slots areleased to the majors’ feeder carriers. This represents an underutilisation of the slots because they were intended for use by largeaircraft; feeder carriers generally use smaller, turboprop equipment,meaning less passenger movements per aircraft movement. Thus, a suboptimal number of passenger movements is realised at the slotcontro lied airports.154Code-sharing partners have also been favoured recipients of leasedslots.Incumbent carriers therefore continue to hold the rights to theoptimum slot times for services into these facilities. This affordsthese airlines on-going control over access to these importantairfields. Potential entrants find that they are effectively barredfrom operating into these airports because slots are simply notavailable on bases that would permit viable operations. Those slotsthat are available are at unfavourable times of the day, or areavailable only on a short-term lease basis. There is no value insecuring slot times in the middle of the night since most travellershave no desire to fly at these times; short term leases make investingin establishing operations at these airports very risky.Another dimension of the physical scarcity problem is that there areinsufficient terminal facilities to accommodate expanded operations atmany airports in the US. Terminal facilities refer to ticket counters,baggage check-in areas, passenger waiting rooms, baggage claim areas,and the actual gates for enplaning and deplaning.71 Gate availabilityis of particular importance to the issue of airport accessibility.Gates may be unavailable because they are not allocated to flights on155an as needed basis, but rather they are on leases to the airlines. TheGeneral Accounting Office (GAO) determined that close to 88 per cent ofall gates at large and medium-sized airports in the United States areunder lease. At large airports, the percentage of gates leased toairlines for their exclusive use is higher yet: 90 per cent arecontrolled on this basis.72 The durations of the leases vary, but somegates are leased for periods in excess of twenty years.The long term nature of a substantial proportion of gate leasespresents a substantial barrier to entry, particularly at largeairports. The vast majority of gate leases at these airports are forwhat are classed as long terms. Roughly 60 per cent of these leaseshave 10 years left on them, and another 35 per cent have more than 20years until expiration.73Since the provision of terminal facilities in the US is accomplishedthrough private financing, established carriers have made substantialinvestments to develop airport terminals. In addition to directinvestment in terminal facilities, incumbent carriers have long-termleases on more than 80 percent of available gate space.74 Therefore,these facilities are not going to be readily available to entrants tointroduce competitive services.The Canadian experience with airport domination has been decidedly156different from that in the United States. There are only two Canadianairports that experience capacity problems: Lester B. PearsonInternational in Toronto, and Vancouver International Airport.Terminal capacity allocation at these facilities is done by the AirportAuthority (in the case of Vancouver) or by Transport Canada (in thecase of Toronto). Transport Canada also allocates the use of gates atother major Canadian airports, and distributes landing and take-offslots at all Canadian airports.Canadian carriers have not really developed hub-and-spoke networks inCanada. The reason for the absence of such networks in Canada is thedemography of the country. The major air travel markets in Canadainvolve flights in close proximity to, and parallel with, the Canada-USborder. Such a physical structure gives rise to linear route networks,and only permits the use of directional hubs. It is for this reasonthat Toronto is the only real hub airport for the Canadian majors.What has developed in Canada is regional hubbing. Regional hubbingis the accumulation of traffic at regional airports outbound from, ordestined to, communities served by the regional, or feeder, airlines.The government provision of airport facilities coupled with thegeneral absence of congestion has not led to single airline dominanceof Canadian airline markets. Both of Canada’s major carriers provide157services in most markets in Canada. This is a marked contrast to thesituation in the United States, as outlined in the precedingparagraphs. While the Canadian industry is certainly dominated by AirCanada and Canadian Airlines International, neither company laysindividual claim to an overwhelming share of the revenue-passenger-miles on any significant domestic route. In 1989, of the 146 mosttravelled domestic city pairs, 77 per cent had at least two competitorsoffering services on the route.753. Conclusions on the Relationship Between the Structure ofthe Air Transport Industry, and Airline Conduct and Performance3.1 Pricing of Air Transport ServicesThe preceding paragraphs have outlined the price characteristics ofair transportation in North America. In general, air fares representgood value to the consumer. The practice of price discrimination hasallowed airlines to increase their per flight profitability. It hasafforded a higher level of output than would be possible in the absenceof price discrimination. Therefore, the pricing procedures of NorthAmerican air carriers should be seen as being beneficial in a socialwelfare sense.158While there has been concern that cartelisation has occurred in theindustry, there is evidence that this is not the case.76 Though therehas been evident collusion in airline pricing, the positive economicrents that are expected in a cartel situation have not materialised inair transportation at the aggregate level. In specific airlinemarkets, however, there rent accumulation has occurred. This has beengenerally associated with U.S. domestic routes where at least oneendpoint-airport is the strong or fortress hub of a major carrier.77Fares in Canada tend to be higher than those in the United States,over routes of equivalent distance. Morrison (1992) found that airlineyields in the fourth quarter of 1990 in Canada averaged 7 per centgreater than those in the U.S.78 It should be noted that internationalcomparisons of air fares can be misleading because of difficultiesregarding exchange rates and differences in the means by which data arereported between countries’ statistical agencies.79’8° Also,operational considerations may dictate differential fares betweencountries. Since Canadian carriers operate in a generally more hostileclimate than do U.S. operators, their costs tend to be somewhat greaterthan those of U.S. carriers.Canadian carriers cannot capture the economies of density that areavailable to their U.S. counterparts. This is due to the generallylower traffic volumes in Canadian markets. Moreover, since the pattern159of traffic is linear in Canada, the Canadian air transport market doesnot allow Canadian carriers to develop true hub-and-spoke routestructures, as have been created in the United States.The number of carriers participating in the industry on a nation-widebasis does not appear to influence the average price level. Though theairline industries of both Canada and the United States have becomemore concentrated since deregulation in the respective countries, theaverage real air fares in both countries have declined slightly overtime. What does affect average fares is the extent of competition inan individual city-pair market. The more concentrated the market, thegreater the average air fare. This has been most readily apparent inthe United States, where carriers have the potential ability to controlaccess to an airport through possession of slots and gates, and throughthe dominance of its CRS system with travel agents in the region wherethe airport is located.81’23.2 Quality of Airline ServicesThe qualitative performance of the industry has several dimensions,though the convenience and safety of the system are foremostconsiderations. Convenience measurements include schedule delay times,terminal delay times, and congestion delay times. The measure of160safety is the number of accidents and incidents per flying hour.The deregulation of the industries in both Canada and the UnitedStates has resulted in reduced schedule delay times. This is due tothe increase in flight frequency afforded by the creation of huboperations, and from the freedom of the carriers to allocate aircraftto routes on the basis of their technical merits. As we have seen, theeconomies of scope associated with hubbing allow for more flights perday to/from any given location. More departures imply less scheduledelay time, hence an increase in service quality.Unfortunately, the current hub networks also have increased theaverage traveller’s terminal and congestion delay times. Congestiondelays have appeared at several US airports, and at Toronto andVancouver in Canada. These delays are largely due to the timing offlights that is associated with hubbing operations. Morrison andWinston (1989) estimate that there would be a welfare gain of $11.0billion annually in the United States if more capacity was provided,and there was a more efficient means of capacity allocation employed.83In the realm of safety, the evidence shows that there has been asteady increase in the safety of air transport in Canada and the UnitedStates. There is a long-term declining trend in the number of fatalaccidents per 100,000 departures in both countries. This trend has not161been disrupted by deregulation.3.3 Barriers to EntryMeaningful entry barriers exist in both the Canadian and UnitedStates airline industries. There are some forms of barriers common toboth countries; other barriers are associated with only one of thecountries.Common barriers to entry are the advantages of large airlines withrespect to consumer demands, and the vertical integration of feedercarriers into the large carriers’ organisations. The large airline canget the passenger to more places, without the need for interlining.Moreover, the existence and nature of frequent flier programs endowsthe large carrier with advantages that small carriers or new carriersfind exceedingly difficult to match.The necessity of having feeder carriers to serve smaller communitiesalso confers advantage to the established airlines in both countries.Any new carrier would have to establish its own feeders, as those thatexisted in Canada and the United States at the times of theirrespective deregulations have been closely allied with, or purchasedoutright by, existing companies.162The United States industry has particular entry barriers not found inCanada. The most imposing such barrier is the control that incumbentairlines have over the allocation of airport facilities at keylocations. To have any hope of becoming a viable competitor, anentrant would have to access these airports. This may not be possiblein the medium term, as the slot rights, and more particularly gateleases, are held by incumbents for periods of up to twenty years.Furthermore, CRS penetration by dominant carriers at hub cities make itdifficult for an entrant to establish a presence in many markets. Thiseven holds true where the would-be entrant is an existing large carrierhaving some share of the markets involving a city where another carrieris dominant.The Canada-specific entry barrier is the relatively low volumes oftraffic in markets in that country. The existing large Canadiancarriers find it difficult to provide profitable services on domesticroutes in Canada. An entrant would be hard-pressed to capture anymarket share without losing money on its operations. This is becausethe entrant would have to serve the highest density markets to have anychance of wresting traffic from Air Canada and Canadian AirlinesInternational, yet would not have these carriers’ advantages of trafficaccumulation through their feeder operations. They also do not havethe FFP enticement that the large carriers have, and would be unlikely163to be able to construct such programs given the scale of theiroperations.164Footnotes to Chapter Four1. Tretheway [1991], p.12. A.I.P., RAC 4-13. Wings [#4/1992], p.144. O’Connor, p.725. O’Connor, p.746. Taneja, p.857. Directions, pp.1425,14288. Directions, pp.1476,14779. Directions, p.148210. G.A.O. [1990], p4411. Directions, pp.1447,146112. Operational weight is the weight of the aircraft itself,the flight crew, and the necessary fuel to get from origin todestination, including mandatory reserves of fuel: for flight toan alternate airport plus fuel sufficient to provide 45 minutes offlight for contingency purposes.13. Tretheway [1991], p.214. Bonsor, p.5415. For example, see Gillen [1985].16. Tretheway [1991], p.317. N.T.A. (I), p.11818. Oum [1991], p.1019. Tretheway [1991], p.820. Taneja, p.13121. Tretheway [1991], p.722. Oum [1992], p.14923. Oum [1992], p.14924. Oum [1992], p.14925. Tretheway [1991], p.826. Directions, p.117527. Transport [1981], p.2028. Taneja, p.12729. Directions, p.118030. Directions, p.115131. Jordan, p.31832. Statistics Canada, Catalogue 51-206 [1992], p.xxii33. Statistics Canada, Catalogue 51-206 [1992], p.xxii34. Directions, p.166335. Windle, p.4536. Kraft, pp.117,11837. Shaw, p.2438. Kraft, p.11539. Coyle, p.7440. Directions, p.115741. Oum [1991], p.1216542. Oum [19911, p.1343. Oum [1991], p.1344. Directions, p.118545. Directions, p.73846. Borenstein, p.35747. Borenstein, p.358 48. Directions, p.119549. Directions, p.117350. Directions, p.116151. Directions, pp.1160,116152. Bailey and Williams, p.18953. G.A.O. [1990], p.6354. Borenstein, p.34755. Bailey and Williams, p.18956. G.A.O. [1990], p.6357. Directions, p.24658. Taneja, p.14759. Tretheway [1989], p.19860. G.A.O. [1990], p.6261. Tretheway [1989], p.19862. Tretheway [1989], p.19963. McShan and Windle, p.22464. McShan and Windle, p.21665. Oum and Tretheway [1989], p.39066. McShan and Windle, p.21867. Huston and Butler, p.21368. Huston and Butler, p.21269. Huston and Butler, p.211,21270. G.A.O. [19901, p.2171. G.A.O. [1990], p.3272. G.A.O. [1990], p.3373. G.A.O. [19901, p.3474. House of Representatives, p.13475. Directions, p.75176. Brander and Zhang [1990], p.58077. Huston and Butler, p.21378. Directions, p.116179. Directions, pp.1160,116180. Jordan, p.32281. Bailey and Williams, pp.188-19082. Borenstein, p.36283. Directions, p.1188166CHAPTER FIVE.POLICY ALTERNATIVES AND ANALYSIS1. General Alternative Strategies for Transborder AirlineServices Reform1.1 GeneralThe goal of liberalisation of the Canada-United States air transportrelationship is to provide air travellers of both countries withimproved services at the least expensive prices. To accomplish this, anew regulatory regime should also encourage industry efficiency. Inkeeping with the equity principle of the Chicago Convention, industryparticipants in both Canada and the United States should receivebalanced benefits from the new regulatory regime.Three general concepts on which to construct the new bilateralagreement have been advocated:1. A new regime of specified transborder rights, in similar formas the existing bilateral;2. An open border concept, allowing unfettered point to pointservice across the Canada-United States border;1673. An open border accompanied by limited cabotage rights;1.1.1 Specified RightsThe most conservative approach would be to negotiate a new set ofspecific routes. The result would be a new bilateral, having the samesubstantial form as the existing agreement. This plan would increaseboth the number of city pairs served, and the amount of competition onexisting routes. It represents a furtherance of the managedcompetition that has prevailed in the transborder sector.1.1.2 Open BorderThe second concept, that of an “open border,” allows carriers ofeither country to serve any cross-border city pair. This policy isaimed at increasing the number of city pairs that would enjoy directservices, and would increase the amount of competition on existingroutes. Any Canadian or United States airline could serve anytransborder market that it wished. Pricing should have a doubledisapproval regimen. The open border plan introduces effective marketdiscipline to transborder routes.1681.1.3 CabotageThe third concept is a cabotage regime. This scheme would allowcarriers of one country to offer domestic services in the othercountry. Again, the objective is to increase competition. In thiscase, competition would be increased in both transborder and internalmarkets. Naturally, a cabotage regime would be combined with openborder privileges.Numerous analysts have researched the options available to Canada andthe United States with regard to what the structure that a newbilateral air services agreement should be. These studies haveconsidered several variations on the three general types of bilateralregimes, as described in the foregoing subsection. The findings ofthese inquiries will be discussed in the following section of thispaper. In addition, research regarding ancillary issues related toCanada-United States air services issues will be considered in thefollowing paragraphs. Such research includes studies of the pricing ofinternational air services under liberal versus closely-regulatedbilaterals (Dresner and Tretheway, 1990), the division of traffic amongcarriers under liberal bilateral regimes (Dresner and Windle, 1992),the relative efficiencies of international air carriers (Windle, 1990),169and the barriers to entry confronting carriers wishing to serve marketsin the United States (United States General Accounting Office, 1990).2. Literature Survey2.1 Dresner and Tretheway (1990)Dresner and Tretheway (1990) sought to establish the effect ofstructural variables on air fares. In particular, they analysed theeffect of the introduction of liberalised, bilateral agreements on airtransport prices in international air services markets. Their datawere for the years 1976-1981, a period in which the United Statesentered into a series of reformed bilaterals that intended to achievemore competitive pricing for airline services. The authors set out todetermine whether price competition had in fact been realised ininternational markets involving the United States through theliberal I sed approach.Dresner and Tretheway assumed that a carrier operating under aliberal bilateral agreement would engage in price competition. ThisBertrand-type behaviour would lead to low price equilibria) Incontrast, carriers operating under traditional, restrictive bilateralswere assumed to have colluded with regard to pricing, thereby producing170price levels above those where liberalised conditions prevail.2Dresner and Tretheway made the following assertions regarding theeffect of liberal bilaterals on prices:1. Liberal bilaterals had no significant effect on “full fares”:those usually associated with non-discretionary travelmarkets;32. Liberal bilaterals had significant effects on fare levels fordiscount travel in markets where competitive routes prevail;43. The average differential in discount fares as betweencompetitive and non-competitive routes was 34.7 per cent.5Dresner and Tretheway demonstrated that liberal bilaterals did havethe desired effect of reducing fare levels in comparison to traditionalbilaterals. They recognised that it was important to define whatconstituted a liberal bilateral. They used a broad definition ofliberal, which had the following characteristics:1. Freedom for carriers to determine capacity levels;2. Freedom from intervention by regulatory authorities with171respect to pricing;3. Freedom of other airlines to enter routes to which theagreement was applicable.While the authors have shown that liberalisation has positivelyaffected particular fare classes, the applicability of theirconclusions to Canada-United States markets will be limited. First,Dresner and Tretheway used routes of 4,000 kilometers or more in theiranalysis. The majority of Canada-United States transborder routes aremuch shorter than this, presenting the opportunity for intermodalcompetition. Second, the third characteristic of liberal (asdelineated above) requires that there be freedom of entry to the routesin question. For Canada-United States services, entry may betheoretically free, yet may be actually prevented by airline controlover airport facilities.A further complication to the application of the authors’ findings tothe Canada-United States situation is that Dresner and Trethewayexamined routes between countries having inherently high densities and(presumably considering the years involved) few attractive alternativeroutings (as the traveller would have had to make intermediate stopsvia third countries, and using inter-line services). For Canada-UnitedStates markets, most densities are relatively lower, and there are172numerous routing and carrier alternatives. (Note that “thinly”travelled routes were intentionally deleted from Dresner andTretheway’s data set.6)2.2 Dresner and Windle (1992)Dresner and Windle (1992) examined the impact on passenger volumesand U.S. carriers’ market shares that resulted from liberalisations ofbilateral air transport agreements. They studied the effects ofbilateral liberalisations between the United States and 51 othercountries. Their data were for the period from 1975 to 1987.Dresner and Windle categorised the countries into two agreementgroups: partially liberalised bilaterals, and fully liberal bilaterals.Fully liberalised bilaterals were those in which both prices andcapacity restrictions had been removed. Conversely, partiallyliberalised bilaterals were those retaining either price restrictionsor capacity restrictions, but not both.The authors found that there were significant effects from thepresence of fully liberalised bilaterals. Their model showed that suchagreements could be expected to cause an 11 per cent increase in therate of passenger growth. The model results indicated that partially173liberalised bilaterals would not lead to higher traffic growth rates.With respect to US carrier market shares, Dresner and Windle foundthat the liberal bilaterals led to non-significant changes. Thisconclusion was reached through the weighted-averaging of US carriermarket shares over more than 600 routes to the various countries in thestudy. It should be noted that there were particular situationswherein dramatic changes in US market share were evident (eg. theUnited States-Singapore markets). Dresner and Windle determined thatthe proportion of US citizens heavily influenced the selection of UScarriers.In summary, Dresner and Windle concluded that:1. Liberalised bilaterals positively affect the totalpassengers in a market;2. Liberalised bilaterals positively affect the rate of trafficgrowth;3. Partially liberalised agreements affect neither the totaltraffic level nor the rate of growth;1744. Neither type of liberalisation affects the share of U.S.carriers in aggregate markets.Dresner and Windle do not speculate as to the reason(s) for therobustness of the U.S. carriers’ market shares. It would beinstructive to know whether the consistency of these shares was theresult of nationalism on the part of US air travellers, or whetherthere were structural factors that militated in favour of US carriers.With regard to the latter, it is possible that there is an advantageconferred to US companies from the fact that they probably owned theCRS systems that customers utilised to book their flights through. Ifthis was indeed the explanation, it would have implications for Canada—U.S. services under a liberalised regime in that US CRS systems can beexpected to become more widely used in Canada with the anticipated exitof Canadian Airlines International from the Gemini system.2.3 Windle (1990)Windle (1990) measured the productivity and unit costs for U.S. andforeign air carriers. The aim of his research was to reveal thefactors that caused differences in efficiency between airlines invarious countries. In turn, these results were intended to givedirection as to what forms of international deregulation would be most175appropriate for future air policy.The metric of total factor productivity (TFP) indicates the ratio ofconversion of inputs into outputs. Windle grouped airlines by region,with Delta Airlines as the base case; Delta’s TFP had the value 1.000.Table 4.COMPARISON OF TOTAL FACTOR PRODUCTIVITIES BY REGIONRegion TFP % higher than USUS 1.113 n/aCanada 1.107 -0.6Europe 0.920 -19.0East Asia 1.297 15.3Others 0.688 -48.1SOURCE: Windle, p.37Three Canadian carriers (Air Canada, Canadian Pacific, and PacificWestern) appeared in the study. Their TFPs were dispersed over thesample: Canadian Pacific ranked 3rd; Air Canada ranked 13th; PacificWestern ranked 37th. The total sample size was 41. The results forTFP show that Canadian firms were slightly less efficient than theirU.S. counterparts.Unit costs were used to extend the analysis to consider what theairlines must pay for their inputs. This measure was used to accountfor efficiency differences that arise because of input price176differentials between countries. Retaining the previous groupings andbase case, Windle found the following results:Table 5.COMPARISON OF UNIT COSTS BY REGIONRegion Unit Costs % higher than USUS 0.910 n/aCanada 0.903 -0.9Europe 0.976 6.9East Asia 0.700 -26.3Others 0.911 0.1SOURCE: Windle, p.39Windle concluded that the unit cost differences were largelyexplained by labour prices and traffic densities. US firms enjoyedeconomies of density superior to those of Canadian carriers. Canadiancompanies had labour price advantages over their US counterparts.Windle states that Canadian companies had firm specific effects thatoffset US carriers’ traffic density advantages. However, he does notspeculate as to what these firm-specific effects might stem from. The“firm” variable used by Windle represented a carrier group average,having been was created from a cost-weighted average of that group’sindividual firm duimny variables.7 However, Gillen, Oum, and Tretheway(1985) concluded that the performance of models employing firm dummiesshould be considered suspect.8 That Windle analysed data only for oneyear (1983) leaves open the possibility that these firm specific177effects could actually reflect some anomalous influence, confined tothat year, and only affecting an individual firm or a specificcountry’s firms.Windle recommends that any policies aimed at increasing trafficdensities should be pursued since this promotes efficiency. Heidentifies several means by which traffic growth would occur, resultingin higher traffic densities:1. Induced traffic due to population growth and/or farereductions;2. Network re-configuration;3. Mergers;4. Attrition of weaker airlines, leaving remaining carriers withgreater traffic densities.In making these recommendations, Windle has neglected to consider theeffect of demand-side factors that influence the density of trafficrealised by carriers. These factors include those related to CRSownership, and frequent flier enticements. For example, the proclivityto select United States carriers on the part of U.S. citizens (as noted178by Dresner and Windle [1992]) probably bears a relation to the factthat these consumers obtained their tickets through U.S.-owned CRSsystems, and, moreover, was likely motivated in a substantial way bythe desire to accumulate frequent flier rewards. Encouraging theattrition of weaker airlines to increase traffic densities for theremaining carriers may be a means of increasing the returns to whatmany analysts have considered to be anti-competitive tools.92.4 United States General Accounting Office (1990)The United States General Accounting Office (GAO) (1990) examinedwhether the structure and conduct of the US airline industry hadresulted in conditions that made it difficult for new competitors toenter domestic markets in the United States. Their inquiry focused oncapacity constraints at US airports, and on the marketing practices ofcarriers, as possible barriers to entry.With regard to capacity constraints at airports, the GAO examinedseveral issues:1. whether the property rights associated with airport slots hadallowed incumbent carriers to block the entry of newcompetition;1792. whether the practice of leasing airport gates and otherphysical facilities represented a barrier to entry;3. whether there were limitations on airports to their futureexpansion (where physical facilities were concerned);4. whether noise control programs presented an entry barrier;5. the extent to which airline marketing practices constituted anentry barrier.Four key airports in the United States feature slot controls:Washington National, Chicago O’Hare, and John F. Kennedy and LaGuardiaairports in New York City. The Federal Aviation Administration (FAA)was responsible for administering the slot allocation program. 95 percent of the slots that existed at the outset of the program were“grandfathered” to the holding carriers. The remainder were madeavailable to entrant airlines through a lottery.10 After slots had beenacquired, they could be sold to other carriers. Although the FAA hadintroduced the “Buy/Sell Rule,” under which airlines could buy and selltheir slot rights, the GAO determined that this program had not met itsobjective of efficiently allocating slots. Incumbent airlines werefound to have been able to hoard slots; would-be entrants have been180excluded from slot-controlled airports.1’Airlines cannot offer services to airports without having gate andticketing facilities. The GAO examined the availability of suchfacilities at US airports and discovered that the control of incumbentcarriers over these facilities presents a sizeable entry barrier inmany markets. The essence of the problem is that incumbent airlineshold long term leases over these facilities. The GAO found that themajor airlines lease over 80 per cent of all gates at large-and medium-sized airports in the United States.12 Moreover, while these carriersallowed other airlines to utilise their gates to some extent, 76 percent of leased gates were used solely by the leasing carrier.13Considering that only 12 per cent of all gates at large- and medium-sized uS airports were not under lease control in 1990, the influencethat incumbent carriers had over market access was enormous.’4The most serious anti-competitive effect of airline control overairport facilities was found to be where long-term, exclusive leaseswere in force. The GAO noted that effective “use it or lose it” leaseprovisions reduced the adversity posed by lease controls.’5 They alsonoted that while it might be possible to develop a more equitablepolicy surrounding capacity allocation, the implementation of suchpolicies would be complicated by the fact that existing leases haveterms of up to 20 years, thereby delaying the time at which sweeping181reforms could be introduced.16The GAO concluded that airport noise restrictions did not generallypose a barrier to entry to most airline markets. The exception to thisfinding was that, where incumbent carriers had been granted exemptionsfrom noise restrictions, there could be an anti-competitive effect.’7With regard to the potential for anti-competitive ramifications ofairline marketing practices, the GAO considered the effect of frequentflier programs, CRS ownership, volume incentives for travel agents, andcode sharing.Frequent flier programs (FFPs) were determined to have influenced theselection of an airline by the consumer. The GAO found that 81 percent of business travellers selected carriers on the basis of expectedfrequent flier rewards more than half of the time they booked af light.18 The anti-competitive effects of FFPs were that theydiscouraged consumers from switching from well-established airlines tonew entrants, and that they served to further strengthen the ability ofan airport-dominant airline to capture the majority of trafficoriginating from the area surrounding that airport.’9The GAO concluded that the control over CRS systems gave thoseairlines so-doing advantages over carriers that had to purchase CRS182services. The outcome of the nature of the provision of CR5 serviceshas been to transfer millions of dollars from airlines purchasing CRSservice to airline who were CRS owners. The GAO determined that thefees that had been charged for these services were roughly double thecost of their provision.20 Non CRS-owning airlines were at a furtherdisadvantage in that travel agents were found to have booked adisproportionate number of flights on the CRS-owning carriers.2’ (Thisis the “halo effect,” as discussed in Chapter V, Section 3.2)A related marketing strategy that the GAO found to be anti-competitive was the practice of offering travel agents volumeincentives for booking with the airlines. These incentives involvedrewards to the agent in the form of free travel, commissions thatincreased non-linearly with the number of tickets sold, and other suchincentives. The outcome of these programs was that, to be competitive,all airlines had to offer similar rewards. This put smaller and/orincipient carriers at a disadvantage versus larger, well establishedairlines, who could spread the cost of such programs over a largertraffic base.22 The GAO reported that its survey of travel agentsrevealed that the objective of capturing these rewards influencedalmost two-thirds of travel agents to select a favoured carrier..23’4Finally, the GAO study considered the possible anti-competitiveeffects of code-sharing. They reported that the significance of the183anti-competitive effects of code-sharing were indeterminate. They did,however, find that agents had observed that, where their customersexpressed a preference, they preferred code-shared services to interline services by a margin of two to one.25 The GAO believed that code-sharing should not be prevented as it conferred benefits to thetravelling public in the form of better connecting times, moreconvenient gate locations, and less baggage-related problems.26In conclusion, the GAO determined that the greatest anti-competitiveforms of conduct in the United States airline industry were:1. The control that airlines had over the allocation of airportfacilities, especially at the four slot-controlled airports;2. The long-term leases that were associated with gates andterminal facilities. This was a particular problem wheremajority in interest clauses were associated with the leases,and where there was little or no room for physical expansion;3. Frequent flier programs, which the GAO claims affects theselection of airline by over half of the market;1844. The methods of CRS use and availability, which conferadvantages to owning airlines at the expense of non-owningcarriers;5. Travel agent incentive schemes.1853. The Specified Rights Option: Revising the Status Quo3.1 Policy Objectives Favouring the Adoption of SpecifiedRights OptionsSpecified rights alternatives would be selected where the objectiveof the regulators was to achieve a better balance of benefits betweenthe two countries’ carriers.Specified rights options are forms of managed competition. Theyallow regulators to directly influence the distribution of benefitsbetween both the two countries and the individual airlines. Thestructure of the existing bilateral has afforded the pre-eminence of UScompanies in scheduled services.27 With regard to the division oftransborder opportunities as between the Canadian carriers, Air Canadahas realised the majority of the benefits of the current bilateralregime. Their operating authority extended to 16 routes versus 5 forCanadian Airlines International as of December 1988.28 Air Canada isadvantaged by the current route allocation, beyond a mere numericalsuperiority: it operates on 8 of the 10 highest volume, transborderroutes; it has exclusivity on the Toronto-Los Angeles route (whichranks sixth in terms of volume); Canadian Airlines Internationaloperates on only one of the top ten routes.29’3° The disparities withregard to transborder opportunities between the countries and their186carriers could be directly addressed through the use of a new specifiedrights regime; a more equitable distribution of service provision couldbe achieved.Given the current disparity in the distribution of benefits betweenCanadian and US airlines, the majority of interested parties in Canadahave advocated that any new agreement give preferential treatment toCanadian carriers. The consensus is that any new bilateral agreementwith the United States must satisfy the goals of equity betweenbenefits and opportunities for both countries’ carriers, while at thesame time preserving the viability of the Canadian airlineindustry.3’,323.2 The Structural Implications of Specified Rights Options3.2.1 Alternative Means of Route AllocationThe specified rights option would retain the essential structure ofthe current transborder regime. The Ministerial Task Force onInternational Air Policy proposed several means by which a newspecified rights regime could be developed:1. Status Quo: Under this scheme, no new transborder routes would187be created. Existing routes would continue to be served bythe country and carrier currently providing the service.Routes that have been authorised but have not been actuallyoperated could see services initiated;332. Increased Double Tracking and/or Multiple Designation:Existing routes would have the number of countries authorisedto operate them expanded through additional double tracking.Existing routes would have the number of airlines authorisedto operate them expanded through multiple designation. Thiscould apply to both operating and non-operating existingroutes;343. Designation of Named Routes: New transborder routes would bespecified. The particular routes would be agreed upon by abi-national panel. They would then be allocated between thecountries through bilateral negotiations, and to theirindividual carriers, by the country concerned;354. Allocation of New Routes by Quantum: Under this scheme, eachcountry would be allocated a fixed number of new routes thatits carriers could operate. The actual city-pairs to beserved would be determined by the individual countries;361885. Round Robin Route Selection: Under this format, the countrieswould alternate turns in selecting routes that their carrierswished to operate. The selection process would stop when oneof the countries had no more routes that it wished toselect .‘3.2.2 Entry and ExitWhichever means of achieving a new set of specified routes is used,the resultant structure will be similar. The entry of carriers intothe specified routes will depend on the approval of the appropriateregulatory authority of the carrier’s home government: the Minister ofTransport in Canada; the Department of Transportation in the UnitedStates 38The exit of airlines from operating routes would require priornotification, as under the present agreement. A provision allowingairlines to buy and sell route rights might also be incorporated,however this would have to be confined to exchanges between airlines ofthe same country. (The Royal Commission on National PassengerTransportation advocated the freedom of Canadian air carriers to selldesignated routes to each other. )391893.2.3 Barriers to EntryIt will be necessary, particularly where services to presentlyunserved, hub airports in the United States are concerned, to makeairport access guarantees part of the new bilateral. These guaranteeswould encompass access to landing slots, gates, and other essentialterminal facilities.40’The utilisation of user charges as the means of airport capacityallocation in Canada do not pose a discriminatory barrier to U.S.carriers with respect to introducing new services to Canada. AlthoughPearson International Airport (Toronto) has a slot allocation system inplace that is administered by a committee of airport management,airline representatives, and the federal government, the general basisof allocation remains user charges rather than property rights.42Proposed changes to the user fee policy may affect the fortunes of U.S.carriers wishing to introduce new transborder services.The National Transportation Act Review Commission has recommendedthat Canada adopt a new procedure for slot allocation that “recognisesthe precedence of established airlines but at the same time makes newentries possible.”43 As plain user charges cannot confer favour toincumbent carriers, it appears that some measure of property rights arebeing considered for future Canadian airport policy. If applied only190to Canadian carriers, such policy would give these airlines parallelinfluence to that enjoyed by US carriers over airports in the UnitedStates. The wording of the recommendation is such that US carrierscould conceivably secure such rights at Canadian airports, as therecommendation refers only to “established airlines.” Of course, UScarriers are well established at many Canadian airports.The pursuit of a specified rights regime implies the continuation ofbarriers to entry at the individual route level. Although doubletracking and multiple designation would increase the potential if notactual number of airlines contesting any particular transborder route,the intent of specified rights is to ensure that there is equity ofopportunity between the two nations’ carriers. The equityconsideration can only be achieved through the constraint of the numberof carriers that have access to transborder routes. While theindustrial structure of the total transborder sector will likely see alarge number of carriers participating, individual routes will featurevery high concentration. (High market concentration will have priceeffects that will be discussed in the following Section.)The restriction on route entry by regulation will maintain very highconcentration for individual routes, thereby sustaining the ability ofcarriers to charge price premia.44 While the existence of alternativeroutings will moderate the magnitude of these premia for most markets,191the time cost of (what will probably be) circuitous routings willundermine the desirability of these alternatives. The agreement musttherefore continue to provide regulators with a means to review prices,to prevent the airlines from transferring surplus from transbordertravellers to themselves through unreasonable pricing.45 Doubleapproval pricing will therefore be a component of any new specificrights regime.3.2.4 Inter-Airline RelationshipsInter-LiningIt will often not be possible for the transborder traveller tocomplete a journey using the services of a single carrier. Therefore,inter-lining will be a necessity for many transborder trips. Whereinter-lining is possible through the use of the services of alliedcarriers, the consumer receives superior service quality, approximatingthat of single line service. Carlton, Landes, and Posner (1980)estimated that the benefits of single carrier service were between 7.0and 9.8 per cent of the total cost of air travel-the fare plus thevalue of the passenger’s time.46 The benefits of allied carrierservice can be expected to be within the range of 0 to 9.8 per cent, asall of the benefits of actual single carrier service may not be fully192captured by alliance arrangements.47 The value to Canadian consumersand Canadian carriers of alliances between Canadian and US carriers hasbeen recognised, and should be supported as part of a new specifiedrights agreement.48Foreign Ownership and Code SharingRelated to international carrier alliances are the issues of foreignownership and code sharing arrangements. It has been asserted that theeffectiveness of alliance relationships varies directly with the extentof exchange of equity between the participating airlines.49’50 Thereexists a conflict between the efficiency gains (and thereby consumerbenefits) that would be realised if Canada-United States airlinemergers would be permitted, versus the general consensus that one ofthe objectives of any new bilateral should be to preserve a viable,Canadian airline industry. As the aim of a specified rights regime is(partly) to promote the survival of the Canadian industry, thelimitation of equity exchanges will be a component of this type ofagreement. Nonetheless, there is considerable support for therelaxation of foreign ownership restrictions. The proposed foreignownership guidelines would permit up to 49 per cent of Canadian carriervoting equity to be held by non-Canadians.51’2 Such an amendmentwould facilitate closer alliances between Canadian and foreign193carriers. Though the National Transportation Act (1987) would need tobe amended to permit increased foreign ownership, the scope of theapplication of the new regulations would need to be moderated by thereciprocity of opportunity for Canadian interests to make similaracquisitions of foreign countries’ airlines. To accomplish this, theadoption of conditions within bilateral agreements specifying theserights would be appropriate. As the United States is considering theadoption of a similar policy53 (as discussed in Chapter III, Section3.3), this condition could well be adopted as part of a specifiedrights agreement.Related to the issue of alliances between Canadian and US aircarriers is the practice of code sharing. Code sharing allows acarrier to provide nominal services without actually flying the routesinvolved.There is concern, where code sharing is practiced for internationalservices and no underlying route rights exist for the non-operatingcarrier, that it may undermine the intended equity of route specificbilateral agreements.54’5The ability to effectively offer services ona transborder route through code sharing is viewed as being parallel tomultiple designation status for the route.56 This may have the effectof diverting traffic away from the designated domestic carrier wherecode sharing exists between a non-designated domestic carrier and a194carrier of the other country.57Supporting the allowance of code sharing are the beneficial effectsthat the practice is deemed to have for both carriers and consumers.Code sharing benefits include greater choice of flights, reducedconnection time, improved CRS positioning, and the opportunity forcustomers to take advantage of more extensive frequent flierprograms.58’9 While the practice may divert traffic away from thedesignated domestic carrier for the transborder portion of the service,it encourages competition between the home country’s carriers for thedomestic portions of the international markets.On balance, code sharing should be permitted as part of a newspecific rights bilateral. Canada and the United States should includea condition that code sharing be permitted, with the proviso that theregulatory authorities in each country have the power to review, and ifdeemed necessary, disallow particular code sharing situations. This isin accord with the position taken by the Ministerial Task Force, whichrecommended that code sharing be subject to the approval of theNational Transportation Agency where the code sharing carrier has nounderlying route authority.60 Implicitly, code sharing authorisationwould thereby be on a double approval basis, with approval beingassumed unless a regulatory body took issue concerning a particularcase. This would provide the consumer and carrier benefits195anticipated, whilst safeguarding designated carriers from commercialarrangements that undermine the equity objectives of the bilateral.In summary, the use of any of the alternative specified rightssolutions will result in a transborder market structure with severalcommon characteristics. Individual transborder markets will have alimited number of potential or actual competitors; the maximum numberof competitors will be double the number of Canadian carriers that areable to provide transborder services. Most routes will be highlyconcentrated--this will have implications for fare structures, whichwill be explored in the following section. The nature of the demandfor transborder travel is such that uS carriers will continue to haveadvantages over their Canadian counterparts. US carriers will retainthe ability to connect a greater number of cross-border city pairsbecause of their extensive hub-and-spoke networks. Overall,transborder markets will be more competitive than Canadian domesticmarkets because of the larger number of actual and/or potentialcompetitors involved, yet will remain less competitive than US domesticmarkets for similar reasons.1963.3 Expected Conduct and Performance of Carriers3.3.1 PricingIt has been shown, in both Canada and the United States underregulation, in current United States markets where a dominated hub isan end-point, in the current transborder environment, and in otherinternational markets governed by traditional bilaterals, that wheremarket entry is restricted fare structures tend to result in somedegree of rents to the airlines, with an accompanying decrease inconsumer welfare.61’234 The entry restrictions inherent withspecified rights bilaterals can be expected to maintain price premia tosome extent.Oum (1990) determined that, under the existing agreement, prices fortransborder flights of 540 miles or less are lower than those fordomestic flights of equivalent distance within either Canada or theUnited States.65 These results are probably a reflection of themarket-driven fares associated with the adoption of the Regional, Localand Commuter Air Services Agreement in 1984. Average transborder fareswere seen to be lower than average Canadian domestic fares forequivalent stage lengths up to 1580 miles; average transborder fareswere determined to be greater than average fares for US domesticservices for equivalent stage lengths of greater than 540 miles.66 The197origins of the differences in average fares between the intra-Canadaand intra-United States markets, and those of the transborder marketsmust be understood in order to determine the likely outcome of futurebilateral structures.There are two possible explanations for the transborder farestructures that have been observed: that costs associated with theproduction of transborder services are higher than for most intraUnited States services, and for some intra-Canada services; that thenature of competition on transborder routes has produced the observedfares. Of course, there could be interaction effects between these twoinfluences.There is no obvious reason that the costs associated with theproduction of transborder air services should be any greater than thosefor the production of domestic services in either Canada or the UnitedStates. As discussed in Chapter VI, Section 2.2, standard productiontechnologies are employed in air transport, regardless of the marketsbeing served.67 Differential costs and factor productivities have beenshown to exist between countries, and between individual carriers.68Canadian carriers produce available-seat-miles at a higher coststructures than their U.S. counterparts.69 Furthermore, it has beenshown that real average domestic fares in Canada are higher than thosein the United States.70’ This presents the possibility that198transborder fare levels are higher than US domestic fares because ofthe proportion of transborder services provided by Canadian carriers,who charge generally higher fares for their services; by implication,United States carriers serving transborder routes would then set faresaccording to the “umbrella effect” afforded by the high fares ofCanadian carriers.That the observed differential between transborder and intra-UnitedStates fare levels are the result of Canadian carrier participation inthe production of these services is unlikely. Although several studieshave shown that Canadian fares are higher than those in the UnitedStates, Morrison (1992) pointed out that there are complicationsinvolved in comparing Canadian and US average domestic fares due to thenature of the data supplied for fares in the respective countries.72Jordan (1991) suggested that the use of exchange rate adjustments formaking international comparisons would lead to inaccurate conclusions;purchasing power parity data were indicated as the appropriate means ofinternational comparison.73The extent to which fare discounts are available and utilised wouldhave an effect on the actual average fares experienced in any market.Oum, Stanbury, and Tretheway (1991) observed that United Statesdomestic markets featured greater discounts, and that the rate ofutilisation of these discounts was greater than in Canada.74 Unless199the actual average yields for individual markets are known, it isdifficult to compare average fares as between the two countries.Dresner and Tretheway (1990) concluded that liberalised bilateralshad no significant effect on the average fares for non-discretionarytravellers.75 By implication, traditional, specific rights bilateralssupport premium fares for this market segment. If there is a largeproportion of business travel between Canada and the United States, therelatively high fares observed could be the result of a greaterproportion of patrons of these services being non-discretionarytravellers, having low elasticities of demand.76Finally, Canadian carriers would have to charge higher fares at allroute distances in order for the claim that they are responsible forhigh transborder fares to be valid. Oum (1990) has shown that this isnot evident: average intra-Canada fares are lower than averagetransborder fares for flights of over 1580 miles.77 Therefore, itfollows that the source of the differential between intra-linited Statesand transborder average fares is accounted for by the difference instructure between these markets.The structure of specified rights bilaterals prevents the entry ofadditional firms that would otherwise compete away economic rents insuch markets. Dresner and Tretheway (1990) have shown that competitive200bilaterals have led to reduced prices for travellers utilising discountfare classes.78 Their definition of a competitive bilateral included aprovision that additional airlines were free to enter routes covered bythe agreement.79 The implication of their research is that theadoption of a specified rights bilateral for Canada-United Statesservices should not be expected to lead to lower (more competitive)fares for transborder services.80There is extensive evidence to show that, wherever there are barriersto entry, airline markets will feature price premia. We have reviewedthe effects of barriers to entry and the resultant market concentrationin previous Chapters of this paper. Assuming that the nature of theentry barriers is non-significant--that only the existence of barriersis meaningful--the implications of the results of studies conductedconcerning non-competitive markets in general can be applied to thetransborder situation.Applying the conclusions of Adrandi, Chow, and Gritta (1989),Borenstein (1989), Graham, Kaplan, and Sibley (1983), Huston and Butler(1988), and Morrison and Winston (1987) to transborder markets suggeststhat it is the inherent concentration in these markets that hasresulted in relatively high fares, and therefore the use of specifiedrights regimes will inevitably lead to fare premia.201Borenstein (1989) studied the advantages of airport and routedominance by United States carriers. He calculated that, for every 1per cent increase in a carrier’s route traffic share, it was able toincrease its prices by between 0.03 and 0.22 per cent.81 He observedthat such price increases tended to by applicable to low-end fares82(Note that Dresner and Tretheway (1990) found that liberal bilateralstended primarily to influence low-end fares.)83Adrandi, Chow, and Gritta (1989) concluded that there was a positiverelationship between a carrier’s market share, and its profitability.They asserted that this finding supported the implications of thestructure-conduct-performance hypothesis: that market power, ratherthan firm efficiency, determined price levels.84 It should be notedthat their conclusions may be less applicable to transborder marketsbecause their study considered airline market shares, costs, andprofitability on a system-wide basis.Graham, Kaplan, and Sibley (1983) tested the extent to which airlineswere able to set fares in concentrated markets above fares in lessconcentrated but otherwise similar markets. They pointed out that theability of airlines to mark up fares above long run marginal cost was afunction of the market elasticity of demand.85 Entry barriers reduceddemand elasticity, allowing for higher margins; transborder services,under route specificity, feature explicit entry barriers, hence permit202price premia. The authors found that concentration affected averagefares, but the effects of increases in concentration were greatest formarkets having Herfindahi indexes of 0.5 or less.86 Their findingssuggest that, once route duopoly has been achieved, the affects ofchanging route shares on raising fare levels have been substantiallyrealised.Morrison and Winston (1987) investigated the implications of thecontestability hypothesis in the context of the United States domesticairline industry. They analysed the extent to which actual andpotential competitors influenced the difference between the optimal andactual welfare distributions between producers and consumers. Morrisonand Winston determined that each additional actual competitor on aroute reduced the gap between actual and optimal welfare by $0.0044 perpassenger mile.87 Their calculations of the effects of potentialcompetitors on welfare optimisation showed that each potentialcompetitor lessened the difference between actual and optimal welfareby $0.0015 per passenger mile.88 Their findings would suggest that,where transborder markets are governed by specified rights, and thenumber of carriers is restricted to below that which traffic densitieswould otherwise profitably support, there will be fare premia.In summary, there is extensive evidence that, wherever entry barriersare present, airline markets will feature fare premia. Moreover, the203connection between the number of actual and/or potential competitorsand the extent to which airlines are able to mark up fares abovemarginal cost has been established. The average higher fares forintra-Canada, and transborder markets versus those for intra-UnitedStates markets can be explained by the greater number of averagecompetitors per market in the United States. Morrison (1992)calculated that the average number of effective competitors at theroute level in Canada was 1.7 in the second quarter of 1987, and 1.6 inthe fourth quarter of iggo.89 In contrast, we have calculated that theaverage number of such competitors in US domestic markets was 2.8.90When applied to transborder services, these findings suggest that thestructure of these markets, with the explicit limitations on entry, hasresulted in observed fare levels greater than those for otherwisesimilar, intra-United States markets. Therefore, the continuation of aspecified rights regime can be expected to maintain premium fares intransborder markets.While double tracking and multiple designation will provide forgreater competition in transborder markets, the achievement of theequity objectives of a new transborder regime will limit thecompetitive benefits of these agreements.2043.3.2 Behaviour in Non-Price AreasWindle (1990) found that Canadian firms were slightly less efficient,in terms of total factor productivity, than US carriers.91 SinceCanadian firms have the opportunity to exact price premia in singletracked-single designated transborder markets, they can be expected tocapitalise on this ability to enhance overall revenues. Therefore, anincrease in the proportion of Canadian-provided services, under thespecified rights option, could well mean an increase in average faresfor transborder services.An increase in the number of double-tracked and double-designatedroutes should serve to discipline prices. Under the existing regime, acarrier having the sole authority to serve a transborder city-pair haslimited incentive to minimise fare levels. The incentive that existsis in the form of the option that the traveller has to selectalternative routings. However, there is a relative scarcity ofacceptable alternatives because of the limited number of transborderroutes available. Moreover, there is a good chance that anyalternative routing will mean significantly greater total travel time.Recognising this, the airlines can price their transborder services tocapitalise on the value of the passenger’s travel time alternatives,keeping fares just below the level at which diversion to alternateroutings would occur. If there was an increase in double-tracking or205double-designation, there would be an increase in intra-routecompetition rather than the current inter-route competition. Intraroute competition exerts downward pressure on fares in a market, thusincreased double-tracking and double-designation would be a desirablefeature of a specified rights regime.With the increase in the number of cross-border city pairs receivingdirect services, there should be an increase in inter-routecompetition. Having a greater number of alternative routings to choosefrom, the consumer should be able to maximise the value of his airtravel expenditure.The increase in alternative routings should serve to exert downwardpressure on transborder fare levels in general.A confound to the benefits expected from an increase in the number oftransborder city-pairs should be recognised in that there is likely tobe a time cost involved in the selection of alternative routings.There is no guarantee that the new route alternatives will bedistributed amongst the carriers in such a way that the passenger willbe able to avoid inter-lining on any new transborder origin-destinationcombination. If carriers are successful in securing the service rightsto new transborder routes such that they could take advantage of theirmarketing and service alliances, the cost to the consumer with respectto the costs of inter-lining would be reduced. There is, however, no206guarantee that the routes would be so-awarded. Furthermore, since thecarriers currently have somewhat “weak alliances,” the route allocationthat best avoided inter-lining at the inception of a new specifiedrights regime might not continue to do so with changes in alliances inthe future. Yet again, the practice of multiple-designation wouldeliminate this problem: carriers could operate on routes that bestsuited their overall network and alliance structures.As discussed in section V.3.3, there will be an obstacle to theefficient provision of transborder service under a specified rightsschema, even if that regime should incorporate multiple-designations.This obstacle is the existence of fortress hubs in the United States.Airport dominance will, yet again, allow dominating carriers to chargeprice premia: in this case, for transborder services involving a hub,regardless of the number of alternative routings available. Thiseventuality should be of particular concern to policy makers as many ofthe new routes being contemplated under the specified rights optioninvolve establishment of direct services between Canada and dominatedUS hub airports such as Atlanta.Transborder prices can therefore be expected to continue to be higher(per mile) than those of all United States domestic flights ofcomparable stage lengths, and of many domestic services in bothcountries.207The allocation of the new routes created under the specified rightsoption is an important consideration in evaluating the expectedperformance of the transborder sector. Both of Canada’s major carrierswould like to have authority to operate on as many of the new routes aspossible. On superficial consideration, it should be expected thatboth Air Canada and Canadian Airlines International would attempt tosecure authority for operating on all new transborder routes. Thecomprehensive geographic coverage that both carriers’ networks have inCanada suggest that neither would attempt to emphasize the acquisitionof route authority on new transborder routes which originate/terminateat any particular Canadian airport. More careful assessment of routebenefits suggests that, because of alliances with US carriers, therecould be some differential in the attractiveness of particulartransborder routes to the two Canadian firms.Air Canada, due to its marketing alliances with United Airlines, andits interest in Continental Airlines, would probably urge that it begranted the rights to serve the Canada-Chicago, Canada-Denver andCanada-Houston markets. Canadian Airlines would seek authority toserve the Canada-Dallas/Fort Worth and Canada-Chicago markets due toits evolving relationship with American Airlines. Conflict would ariseover the Canadian designation of which Canadian carrier would receivenew routings between Canada and Chicago as both Canadian companies have208an interest in securing these routes because of their alliances with USairlines. A solution to this problem would be for Canada to doubledesignate the Canada-Chicago routes so that both Canadian carrierscould operate them.The question of route awards as between US carriers is much simplerthan the Canadian situation. The United States has made multipledesignation a part of its transborder policy to date. Furthermore, theUS carrier will seek those routes that connect Canadian cities with UScenters at which the carrier operates a hub. The small number ofCanadian urban centres makes the value of routes more related to thecharacteristics at the US end rather than the Canadian terminus. Aswith domestic routes, the US carrier is primarily concerned withintensifying the returns from its hubbing operations. US firms havelittle incentive to seek particular routes, other than to further addto the scope economies that they presently enjoy from their domestichubbing operations.An achievement of absolute equity between Americans and theirCanadian competitors requires a negotiated increase in Canadiancarriers’ market shares. As this form of regulatory revision is notlikely to be traffic generating,92 market share gains for Canadiancompanies must come at the loss of US market share. US carriers areexperiencing difficult times, so it is doubtful that they would be209agreeable to such legislation.Transborder passengers may also not approve of this solution. Sincethe specification of routes retains protection of the airlines fromopen competition, fares can be expected to be higher than economicallyjustified.The net effects of a revised specified rights regime should be asfollows:1. The greater number of route choices will promote decreases inreal transborder fares by increasing inter-route competition.2. The impact of increased routing options on fares will belimited by the time cost of alternative routings where suchroutings increase total distance flown. Where these routingsnecessitate inter-lining, their attractiveness will bereduced.3. Double-tracking and multiple-designation of routes willpromote intra-route competition, exerting downward pressure onfares.2104. The award of routes to Canadian carriers on an exclusiveservice basis may result in relatively higher fares on theseroutes due to the cost structure of Canadian firms. Inter-route competition will, however, mitigate these fare premia toa significant extent.5. The actual allocation of new routes will affect efficiency andperformance in transborder markets.a). If airlines can secure routes that integrate well intotheir existing networks (including services provided byalliance partners), efficiency will be enhanced, andfares should reflect these efficiency gains. However,where this results in the strengthening of market power,the net effect on fares will be negative (from theconsumer’s perspective).b). If routes are distributed somewhat randomly, with noprovision for the firms to exchange rights on a marketbasis, there will be a negative effect on efficiency.The value of service for the consumer will be reduced bythe complications that will arise because of the need forcircuitous routings and/or inter-lining.211The selection of the specified rights option will produce moderatebenefits to interested parties in Canada and the United States.Canadian air carriers should benefit from the acquisition of rights toserve more cross-border city pairs, in many cases on an exclusiveservice basis. To make such an agreement acceptable, it is essentialthat US carriers also receive new route authorities as well; they willbenefit to some extent with these new route awards. Transbordertravellers will be able to fly to more cross-border destinationsdirectly as the result of these new services, thereby increasing thevalue of travel expenditures through time savings.There are obvious disadvantages associated with the specified rightsoption. First and foremost, it does not allow market forces todetermine the structures of both route networks and transborder fares.The benefits that have been realised from domestic deregulations inboth Canada and the United States will not be fully available in thetransborder markets. Fares should experience real decreases, but theseprice reductions will be limited by the restrictive nature of theproposed agreement. The regime would confer additional rights toCanadian carriers, but will do so at the expense of transbordertravellers. Whether Canadian carriers actually capture the benefitsassociated with a greater share of transborder rights is uncertain astravellers may elect to patronise US services. Gillen, Hansen, andRamos (1990) anticipate that the pursuit of the specified rights option212will not meaningfully increase the Canadian carriers’ traffic shares.93The main advantage of the specified rights option is that itaccommodates the pursuit of the equity principle as per the ChicagoConvention. This equity concept was established in the 1940s, andshould be scrutinised as to its applicability to the North Americanmarket of today. With the emergence of the Canada-United States FreeTrade Agreement and the North American Free Trade Agreement, and thegeneral policy goal of creating a tariff-free economy on the continent,the national notion of equity should perhaps be weighed against anequity that is considered at an organisational or individual level ofscrutiny. The historic concept of equity (in air law) and the implicitprotectionism of managed competition in air transportation is vestigialin the emerging world of free trade areas. The European Community hasabandoned the equity consideration from its internal aviation policy.Canada and the United States should consider following the sameideology. If they do, the sole advantage of the specified rightssolution is lost. Therefore, with a view to the future of trade inNorth America, the specified rights scheme should not be pursued.The eventual goal of Canada and the United States (and, for thatmatter, Mexico) is to enjoy a continental economy free of regulatoryimpediments to efficiency. The conclusion of a new specific rightsregime today will serve only to delay the time at which free market213discipline will allow and encourage companies to maximise efficiency.The choice of this option does not address the fundamental reasons forthe inability of Canadian carriers to effectively compete with their UScounterparts. The greatest long term difficulty facing the Canadianairline companies in continental competition stem from demographicinevitabilities, the Canadian government’s policies regarding taxationand capital accumulation, and from the paucity of dense air travelmarkets in Canada. Alternatives that address these issues will promoteincreased efficiency for airlines in Canada, and more competitiveservices for consumers in both Canada and the United States.To encourage airline competition and airline efficiency, regimesinvolving greater integration of the air transport industries of bothcountries warrant consideration. As indicated above, these methodsinvolve opening transborder markets to free competition, and/orallowing cabotage rights. Recognising the disadvantages facing theCanadian industry at the present time, the method of implementation ofmore liberal approaches would have to be done in such a way as to averta crisis for Canadian carriers. To accomplish this, it has beensuggested that the approach taken should be to confer differentialrights of access. This would be accomplished by granting one or acombination of transborder, cabotage, and/or fifth and sixth freedomrights to carriers in some prescribed manner. Of course the suggestionis to give these expanded rights to Canadian airlines first in order to214promote a “leveling of the playing field.11We now turn to a discussion of these more liberal transborderstructures.2154. The Open Border Option4.1 The Structural Implications of an Open Border Regime4.1.1 Basic PhilosophyThe open border concept allows any carrier from either Canada or theUnited States to establish direct, transborder services without theneed for governmental approval. This strategy would allow marketforces to dictate which cross-border city-pairs would enjoy directservices, what prices would prevail, and which carriers would provideservices. The United States promoted the open border policy in itsposition paper of 1985.The underlying premise of the Open Border approach is that marketforces can allocate air transportation resources more efficiently thancan the efforts of any regulatory authority. This belief was the basisfor the deregulation of domestic air transport industries in Canada,the United States, and other countries. In the case of the Open Borderstrategy, the deregulation would only apply to cross-border services;domestic operations in each country would remain contestable only bycarriers that were nationals of the particular country.The aim of an Open Border regime would be to increase the efficiency216of transborder air services through the application of marketdiscipline to this sector. The benefits of such a policy can beexpected to accrue to transborder travellers, who should receive higherquality transportation at reduced prices, and to those airlines thatcan most efficiently serve transborder markets.4.1.2 Regulatory StructureThe introduction of an open border regime would mean the end of routeallocation through regulation. An open border policy would permit anycarrier, whether of Canadian or United States nationality, to offerservices on any transborder route. This principle would apply to bothscheduled and charter operations. Attempts to provide for theequitable distribution of routes between the airlines of Canada and theUnited States, and between specific firms in each country, would beimplicitly abandoned.To fully realise the benefits of a deregulated market structure, itwould be necessary to deregulate pricing in transborder markets.Therefore, the open border policy would include the end of the existingscheme of double approval pricing; fares would no longer need to beapproved by the relevant regulatory agencies in each country. Instead,a double disapproval provision could be employed, so that if both217countries’ agencies objected to the fare(s) being offered, the farewould be disallowed.As further protection to consumers and firms, competition legislationin Canada and the United States would continue to provide a structurethrough which claims of anti-competitive pricing could be investigatedand, if need be, remedied. The US Department of Transportation andTransport Canada would probably be given the right to review fareofferings, when they have received a complaint from an interestedparty. If one of these agencies found the fare(s) in question to beeither predatory or excessive, they would have the ability to challengethe fare through means provided for under competition legislation.Complications may arise under such a scheme however, as the issue ofjurisdiction over transborder fare offerings would have to beaddressed.The open border regime could include provisions for “blind sectors,”co-terminalisation, and change-of-gage operations. The intention ofthese provisions is to enhance the number of transborder markets thatcan be served by carriers. With these rights airlines can establishquasi-hubs in the other country’s territory allowing for trafficconsolidation that will increase traffic density through scopeeconomies. The inclusions of such provisions into an Open Borderagreement would be of particular significance to Canadian carriers.218“Blind sectors” are flights that are consecutive to transbordersegments, departing from the first airport of entry into the foreigncountry, and terminating at a second airport in the same foreigncountry. The carrier is not permitted to offer seats to thosetravellers journeying strictly between the points in the foreigncountry.94 For example, Air Canada could operate a flight from Torontoto Kansas City, thence on to New Orleans. It could sell seats forToronto-Kansas City and Toronto-New Orleans, but not for Kansas City-New Orleans.Related to blind sector operations are co-terminalisation hubs. Coterminalisation hubs are hubbing operations involving interchanges ofpassengers between several transborder and blind sector flights. Thisallows the transborder carrier to capture economies of scope anddensity by consolidating traffic from diverse origins at the coterminal hub, and distributing it between various flights to numerousdestinations.95 For example, Canadian Airlines could establish a coterminal hub at Salt Lake City. Transborder flights from Vancouver andCalgary would arrive at Salt Lake City roughly simultaneously. Blindsector flights would subsequently depart from Salt Lake City to Phoenixand Dallas. This would give Canadian effective services for VancouverPhoenix, Vancouver-Dallas, Calgary-Phoenix, and Calgary-Dallas. Theadvantage of this structure is that it allows for higher load factors219(thus economies of density) over the operation of four separate, nonstop services.While blind sector operations would be attractive to both Canadianand US carriers, co-terminalisation would appeal almost exclusively toCanadian carriers due to the nature of the demographics of, and demandfor, transborder services.It must be noted that co-terminalisation is not a form of cabotage asit does not entail the rights to offer seats to domestic passengers atthe co-terminal airport.Finally, the practice of United States Customs pre-clearance atCanadian airports can be expected to continue under an open borderregime. While the disadvantages of pre-clearance, as it has beeninstituted, have been incident on Canadian airline companies, benefitsof the program have been realised by transborder travellers.96 Sincethe abolition the present manifestations of pre-clearance has not beenadvocated, even by Canadian carriers,97 it can be anticipated toendure. However, recommendations have been made that pre-clearance notbe extended to other Canadian airports, so as to avoid giving furtheradvantages to US carriers.98 It should be recognised that the failureto permit the establishment of additional pre-clearance facilities willundermine the advantages of an open border policy from the perspective220of the consumer.4.1.3 The Structure of Transborder Air Service DemandThe demographic and geographic structures of transborder markets willnot be affected by the adoption of an open border regime. While theliberalisation of transborder markets should lead to reduced fares dueto efficiency gains, and thereby induce added demand, the essentialnature of these markets will be unaltered. Forces related to economicand cultural interactions between Canada and the United States haveshaped the flow of air travel between these countries. The characterof such travel is that most passengers are Canadian nationals. Thedemographics and geography of travel see many more destinations in theUnited States than in Canada. An open border policy should be expectedprimarily to influence the distribution of traffic between carriers,with some degree of traffic inducement occurring due to improvedefficiency.An important demand-related consideration iseconomies of scale. Economies of scale have beenlargely absent from air transportation from a costAs discussed in Chapter IV, economies of scale havebeing significant regarding their effects on carrierthe influence ofdemonstrated to beperspective.99’100been recognised aschoice on the part221of consumers.’0’ Because the more comprehensive networks of U.S.carriers will, for the foreseeable future, allow them to serve agreater number of transborder markets, they can be expected to be moreappealing to transborder travellers under an open border regime thantheir Canadian rivals.4.1.4 The Structure of Transborder Air Service SupplyThe ultimate market structure of transborder airline services willdepend on the conduct of the individual carriers, in conjunction withthe influence of the exact regulatory regime that is adopted.Nonetheless, there are factors that will influence the supply of suchservices that can be readily anticipated.The advantages of hubbing operations have been discussed previously.The hub operations of US carriers will be extended to further includeCanadian points in these networks. The retention of pre-clearancefacilities in Canada can be expected to be a part of an open borderregime.In conjunction, these factors will confer competitive advantages toUnited States carriers; they can be expected to dominate thetransborder sector (to an extent even greater than under the currentregime) with the adoption of an open border agreement. US firms have222cost advantages: the average cost per available-seat-kilometre in theUnited States in 1990 was $0.0665; the average cost per available-seat-kilometre in Canada was $0.1053 in iggo.b02 These differences in costper available-seat-kilometre necessitate the achievement of higher loadfactors for Canadian carriers in order to make a given serviceeconomic. It must be concluded that, in open competition, particularlyon relatively low volume routes, US carriers have a competitiveadvantage that should allow them to capture the majority of traffic.Gillen, Hansen, and Ramos (1990) have estimated that an open borderagreement would reduce the share of transborder traffic carried byCanadian companies by 0.04 per cent, given no traffic-inducing effectswith the introduction of the policy.103The resultant general market structure of transborder airlineservices under an open border regime will almost certainly be an “n°firm oligopoly--as in the present situation. The absolute number ofparticipating firms should be little affected by the change in policy.What can be expected to change is the number and identity of airlinesthat are operating any particular transborder route.The nature and composition of alliances between Canadian and U.S.carriers may be affected by the emergence of an open border regime.The existing alliances between Air Canada, Continental, and United, andbetween Canadian Airlines International and American Airlines could be223either strengthened or strained by the emergence of an open borderenvironment. We believe that they are more likely to be strengthenedas the result of an open border policy; the basis for this belief willbe discussed in the following section on airline conduct.The number of individual transborder routes can be anticipated toincrease with the opening of the Canada-United States border tounfettered airline competition.To gage the nature of the eventual structure of transborder airlineservices that are likely to emerge under an open border policy, we turnnow to a consideration of the expected conduct that airlines willfollow. In turn, the conduct of the carriers will determine theperformance of the transborder sector. We will also discuss theprobable performance of the industry under the contemplated regime.2244.2 Expected Conduct of Carriers and Performance of theTransborder Industry4.2.1 General Network StrategiesAn open border agreement should result in a considerablerestructuring of routes between Canada and the United States. Airlinesof both countries can be expected to re-configure their transborderroutes. In general, the route networks should conform to the structurethat is suggested by gravity models of spatial interaction. The routeconfiguration strategies of carriers should differ according to thecarrier’s nationality.Gravity models of spatial interaction suggest that the number oftrips between centres is a function of the attractiveness of thosecentres, in terms of their population and economic activities, and ofthe distance between the centres.’°4 Direct air services will beestablished where the size (in terms of population) and economicimportance of the two centres in question generate sufficient trafficdensity to warrant it. (Note that the existence of a major hub at acity would be one such important economic factor.) Where densities areinsufficient to justify direct services, markets will be served throughthe use of hub-and-spoke systems, co-terminalisation or throughbilateral partnerships.225Canadian carriers may be expected to add more transborder flights totheir networks. This should be especially true of Canadian AirlinesInternational, which has few transborder services under the presentagreement. In general, Canadian carriers can be expected to link majorCanadian airports to co-terminalisation hubs in the United States.(Note that we have presumed that the Canadian carriers would be giventhe right to establish co-terminalisation hubs, and change-of-gageoperations, under an open border agreement.) These co-terminalisationhubs would be located at US cities having the greatest measure ofattractiveness to Canadian travellers. (A superficial examination ofthe traffic between Canada and the United States would suggest that NewYork, Tampa-St. Petersburg, and San, Francisco would be likelylocations).’05 Blind sector flights would then distribute passengersto their ultimate U.S. destinations. For services between Canadianpoints and northern tier cities in the United States, direct flightscan be expected, often being operated by Canadian level II carriers.Toronto’s significance as a hub for Canadian carriers should furtherincrease. 106U.S. airlines would reorganise their transborder flights to betterfit with their domestic hub operations. They will introduce serviceslinking Canadian cities with these airports to offer more comprehensiveconnections to transborder travellers through the hub networks. They226may be expected to introduce direct transborder services where trafficvolumes justify such flights and where the time cost associated withconnecting through hubs is anticipated to lose passengers tocompetitors who offer direct services. In general, US carriers willlink Canadian cities to their nearest major hub airport. Oum (1990)has shown that this form of hubbing, termed “Entry Point Hubbing,” issuperior to “Central Hubbing,” wherein the airline connects the cross-border city to a centrally located hub, because it minimises totalpassenger travel time.107Allied carriers can be expected to follow the network organisationstrategy of US carriers. A distinctive characteristic of allied-carrier operations would be the connection of smaller Canadian citiesto northern-US hubs through the services of Canadian level II carriers.4.2.2 Significance of Hubs in the United StatesU.S. airlines can be expected to offer transborder services throughextensions of their hub networks. This will allow them to connect alarge number of cross-border origin/destination pairs most efficiently.In addition, the US carriers will be able to provide services betweenCanadian centres, using mirror image routings associated with this typeof hubbing operation.’08227In order for Canadian carriers to be competitive, they would have toestablish co-terminalisation hubs in the United States. These hubswould allow the Canadian carriers to replicate the number oftransborder city-pairs that can be served by their US counterparts.The creation of these hubs will be complicated by the problems ofaccess at US airports, as outlined in Chapter VII. Although there aresomewhat underutilised airports at large cities in the United Statesthat have been suggested as likely candidates for Canadian hubs,’09”°the effectiveness of hub systems depends on the total number of pointsthat can be connected via the hub. Without an extensive number of USdestinations associated with the co-terminalisation hub, the Canadiancarriers’ returns to their operations from potentially attractive hublocations will be much less than those enjoyed by US firms.Allied-carrier networks would also be affected by the location of UShubs. As a US partner can usually more effectively serve customersthrough hubbing, its Canadian partner will likely feed traffic to, orcollect traffic from, the US hub rather than choosing to operate astand-alone service. By implication, this calls into question thesustainability of Canada-US alliances.2284.2.3 Canada-U.S. Carrier Alliances and the Open BorderEnvironmentCanadian carriers could choose to compete with their present U.S.affiliates in providing direct services on such origin-destinationpairs. The economics suggest that they should not elect to do so, asdirect services will, in many markets, not realise the traffic volumesnecessary to make such operations profitable. By cooperating with a USpartner, and taking advantage of its hubbing operations, the Canadiancarrier may share in the economies of scope associated with the hub-and-spoke system. However, the trade-off would be limited growthopportunities for Canadian carriers.Conversely, a U.S. carrier could choose to simply operate transborderservices on its own. However, the continuation of an alliance with aCanadian firm gives it better geographic coverage. As discussed inChapter VI, demand-based returns to scale exist in terms of the numberof points served. To have the same number of Canadian points served,US carriers would have to have inter-line agreements with alternativeCanadian carriers in the event that they should defect from existingalliances. There are few unaffiliated Canadian regionals for UScarriers to establish feeder arrangements with. US carriers would haveto establish new feeders in Canada, or at least underwrite existingsmall Canadian charter carriers to acquire the necessary assets to be229able to inaugurate such services. This would be problematic given theguidelines of the National Transportation Act (1987) that limit theamount of foreign equity holdings in a Canadian transport company, asdiscussed in Chapter II, Section 2.As noted in Chapter IV, Section 3, the evidence surrounding thedemand for airline services indicates that FFPs have a significanteffect on carrier choice. Airlines enjoy economies of scale from ademand perspective in that the firm having the ability to carry thepassenger to the greatest number of destinations is likely to be thefirst consulted by the consumer when he wishes to travel. In an openborder environment, the comprehensive geographic coverage offered bythe Type 3 Carrier should make it the first choice for the transbordertraveller seeking passage. By flying with the Type 3 Carrier, thecustomer will be able, not only to access most North Americandestinations more readily, but will also be able to maximise hisfrequent flier benefits. These benefits will be maximised because hecan patronise one such carrier to almost all destinations in Canada andthe United States, collecting frequent flier points as he goes; when hewishes to utilise free travel, he can do so to the extensive number ofdestinations served by the Type 3 Carrier, not only in North America,but anywhere these allied carriers fly to.A related advantage favouring the Type 3 Carrier is the CRS. The230allied carriers happen to include the two airlines having the mostwidely used computer reservations systems in the United States:American Airlines and United Airlines. Though the Gemini CRS currentlyprevails in Canada, it is legitimate to anticipate that this situationwill not persist given the wishes of Canadian Airlines International toparticipate in the Sabre CRS of its U.S. partner, American Airlines.This will leave Gemini with Air Canada and Covia as its remainingpartners. Since Covia is the CRS of United Airlines, and is one of themost widely used systems in the United States, it is expected thatGemini will be converted to Covia in Canada as well. Presuming thatthe benefits associated with CRS ownership endure, it will be difficultfor any non-Type 3 carrier to compete effectively in the provision ofmost transborder services.4.2.4 Anticipated Characteristics of Transborder MarketsThe distribution of traffic between carriers will be determined bythe characteristics of the travellers’ origin-destination centers, andby the competitive advantages of the carriers.231These origin-destination pairs can be divided into six generalcategories, giving the following O/D matrix:TRANSBORDER MARKET CATEGORIESCanadian CentreI Large I Small IU.S. CenterDominated Hub Category 1 Category 4....Large Category 2 Category 5....Small Category 3 ....Category •••IWithin each of the foregoing categories, it is anticipated that aparticular type of carrier will have a service advantage that willallow it to capture a majority share of the traffic on that type oforigin-destination pair. These four types of carriers are defined asfollows:Type Description ExamplesType 1: US Carriers American, Delta, United,USAir, Northwest, TWAType 2: Canadian Carriers Air Canada, CanadianAirlines InternationalType 3: Allied Carriers Air Canada-ContinentalUnited, American-CanadianType 4: Multiple Carriers Any Type 1, 2, or 3Carrier232For Category I services, it should be expected that Type 1 carrierswill predominate. The advantages that have been identified with hub-and-spoke operations in the domestic markets of the United States willbe extended to transborder services in an open border environment.Flights connecting large Canadian centres with United States hubairports will have relatively high frequencies. There will be somecompetition on such routes, primarily from Type 2 Carriers. Where theservices in question link large Canadian centres with the dominated-hubs of American, Continental, and United Airlines, Type 3 carriers canbe expected to be the primary operators, assuming that the existinginter-airline alliances can be sustained.It should be noted that, to the extent that the alliances betweenCanadian and US carriers change, so too will the dominant Type ofcarrier on these transborder services. Moreover, if traffic isdistributed between allied carriers in such manner that one (or more)partners is displeased, the resulting strain on the alliance may leadto its termination. Given the relatively “weak” nature of the existingalliances, this eventuality has a reasonable probability.Category 2 services should be the most competitive of all transbordermarkets. These markets should have sufficient volumes of traffic thatmany airlines could profitably participate; Type 4 Carriers are233expected to offer services in these markets.Due to economies of scope, Type 1 Carriers should enjoy an advantagein Category 2 markets. These scope economies are related to theconsolidation of traffic through major centres, as noted by Morrisonand Winston (1990). They found that the competitiveness of a carrierin a market was directly related to the number of network connectionsassociated with the end-point airports of a particular city-pair.’1’US carriers will be able to offer high quality services in terms ofreduced schedule delays due to increased flight frequencies by therouting of Category 2 passengers through U.S. hub airports. Thebusiness traveller will find such increased flight frequenciesparticularly appealing; scope economies should allow for discretionarytravellers to take advantage of deep discount fares in these markets.Category 3 markets can be expected to be dominated by Type 1carriers. The most efficient means of serving these markets will be bytraffic consolidation, utilising the domestic hub networks of U.S.airlines. These carriers need only introduce spoke services to largeCanadian centres in order to link most small U.S. communities to thesecities. Those U.S. firms having the most extensive domestic networkswill enjoy the greatest success in these markets.Category 4 markets are expected to be dominated by Type 3 Carriers.234Canadian carriers, with their extensive domestic networks, will servethe smaller Canadian communities. Connections with US hub airportswill be provided by the allied US carrier that is dominant at therelevant hub airport. This arrangement will provide the traveller witheffectively on-line service between these points. Most of thesemarkets will be served through connecting flights involving airports atlarger Canadian centres. Where the 0/0 pair is in close proximity tothe Canada-United States border, these markets may be connected withdirect flights by the hub-dominating US carrier, using commuter-typeaircraft.In cases where, there is close proximity between the transborderpoints, the anticipated turboprop services commuter services willusually be provided by the feeder airlines of one or more of the Type 3Carriers. Where the uS hub in question is operated by a US airlinehaving no alliance with a Canadian firm, the US carrier will likelyoffer the only service in the market. Such services may be conductedby the hub operator itself, by one of its existing feeder carriers, orby contract with a small, local operator in Canada. For example,markets in the vicinity of, and involving, Minneapolis/St. Paul can beanticipated to be served in such manner by Northwest Airlines.Category 5 markets will have Type 3 Carriers as favoured producers.Once again, allied carriers can consolidate traffic involving small235Canadian centres, and thereby provide the most efficient means ofconnecting these communities with large, United States cities. In mostcases, these markets will be served through connecting flights, oftenbeing routed through a hub airport- -either in Canada or the UnitedStates. There should be reasonably good competition in these marketsdue to the alternatives that the consumer will have in selectingbetween Type 3 Carriers who will participate in most such markets.Advantage to a particular Type 3 Carrier will be to the extent of thetotal time differences that arise due to routing differences based onhub locations. For services linking several large US cities withsmaller Canadian points in the Saskatchewan, Manitoba, western Ontarioregions, Northwest Airlines can be expected to be a significant, non—Type 3 competitor.Category 6 markets are likely to be dominated by Type 3 Carriers.Yet again, the ability of these allied carriers to consolidate trafficoutbound from, and destined to, smaller communities will give them acompetitive edge in these markets. Northwest Airlines can also beexpected to be a significant player in the provision of some suchservices, in the geographic areas alluded to in the foregoingCategories.There should be relatively high levels of competition in Category 6markets due to their contestability.112 Competition will be limited236where traffic volumes make it efficient for only one carrier to existin the market.In summary, for most transborder travel, passengers will be bestserved by alliance (Type 3) carriers. The Air Canada-Continental-United and American-Canadian groups will therefore be foremost in theprovision of transborder services in Categories 4, 5, 6, and, to alarge extent, in Category 1. Type 3 Carriers should be expected tocapitalise on advantages related to FFPs and CRSs, as well as theiroperational advantages.Though it is notsignificant playerMinneapolis-St. Pauland western Ontario,connections throughalliance with DutchEurope, Africa, and allied carrier, Northwest Airlines may be ain transborder markets due to the proximity of itshub to Canadian centers in Saskatchewan, Manitoba,and because of its ability to supply internationalits own services to the Far East, and through itscarrier KLM, which has extensive services toCanadian carriers may realise market dominance in certain situations,not indicated in the foregoing discussion. Though it has beensuggested that Category 2 markets will be widely contested, Canadianfirms have some degree of advantage that may make them pre-eminent inthese markets. The fact that the majority of transborder passengers237are Canadian residents, coupled with the inclination of Canadians topatronise Canadian carriers where possible, should yield these firms acompetitive advantage with respect to attracting customers in thesemarkets. Canadian carriers should be able to continue their advantagein serving Canadian holiday travellers, who fly from various points inCanada to vacation destinations in the southwestern and southeasternUnited States.4.2.4 PricingWe have explored the bases for airline price structures in ChapterVI, Section 3. Fares were seen to be dependent on three generalfactors: input prices, productivity, and the extent of competition inthe individual market. The adoption of an open border regime wouldhave implications for all three of these factors, and therefore shouldinfluence transborder fares.The least impact will be on input prices. These effects will beconfined to reduced fuel prices experienced by Canadian carriersoperating transborder services via co-terminalisation hubs. Costsavings may result from the utilisation of US-supplied fuel purchasedat the hub site. The realisation of these savings will be dependentupon the trade-off between the lower price of aviation fuel in the238United States, and the extra fuel consumption associated with huboperations.With the introduction of an open border regime, the extent ofhubbing operations incorporating transborder services can be expectedto increase. In Chapter VI, Section 3, we explored the effects of theextent of hubbing practices on air fares. We found that there was asignificant negative relationship between the two. Therefore, weanticipate that the additional use of hubbing to serve transbordermarkets would serve to reduce average fares for these services.Liberalisation would not result in lower fares in all transbordermarkets. Certain markets would feature premium fares, usually as theresult of dominated hubs in the United States. Markets where theorigin or destination is a US carrier’s fortress hub would be difficultto contest, thereby entrenching market concentration. These marketscan be expected to be associated with fare premia. Since manytransborder markets are presently served by a lone carrier, theinstances where this phenomenon would represent a potential welfareloss to consumers, due to fare premia, would be relatively few.239Expected Pricing Behaviour by Market CategoryFare structures for Category 1 markets should be similar to those forUS domestic services involving dominated hubs, with price premia beingcharged.Fares in Category 2 markets should be the most competitive of allcategories of transborder services, reflecting the wide number ofalternatives facing the consumer. These should be the leastconcentrated of all transborder markets.Fares in Category 3 markets are anticipated to be moderatelycompetitive. The least competitive fares will prevail where theconsumer has limited choices regarding airlines serving the smallcommunity in the United States. The possibility of entry by otherfirms, and the viability of alternative modes of transport to/fromthese small US centers will help to discipline prices in these markets.Category 4 markets can be expected to have high fares relative toother transborder services. These price premia are expected becausethese markets involve a US hub airport. The fare premia that have beenobserved in United States domestic markets involving dominated hubsshould also prevail in this type of transborder market.240Fares in Category 5 markets should be highly competitive. The degreeof competitiveness of fares will be a reflection of the number ofairline and routing choices facing the consumer. The largestdeterminant of market competitiveness will be the number of airlinesthat are serving the small, Canadian community: where there is amonopoly, fares can be expected to be higher. However, thecontestability of these markets should exert downward pressure onaverage fares. Also, the option to travel via an alternative mode to alarger Canadian center at which Category 2 services are available, willdiscipline prices.Fares in Category 6 markets can be expected to be very competitive.The variety of routing, and carrier, options for the consumer to choosefrom will keep fares relatively low. As with Category 5 markets, thereshould be significant downward pressure on fares due to the possibilityof substituting other modes for portions of the trip so as to takeadvantage of alternative services aimed at other market Categories.241Table 6.SUMMARY OF EFFECTS OF OPEN BORDER REGIME ON MARKET CATEGORIESCategory Dominant Service FareCarrier Type Characteristics Characteristics1 Type 1 High Frequency Fare Premia2 Type 4 High Frequency, HighlyVarious Routings Competitive3 Type 1 Various Routings Competitive4 Type 3 Captive Markets Fare Premia5 Type 3 Various Routings HighlyCompetitive6 Type 3 Various Routings, CompetitiveAlternative ModesSignificant4.3 An Open Border and Traffic Flows4.3.1 Implications of an Open Border Regime for Intra-CanadaTrafficIn addition to the effects that an open border agreement would haveon conduct and performance in transborder markets, the policy wouldalso influence intra-Canada airline markets. While theoretically anopen border situation could affect travel patterns for domestic marketsin both Canada and the United States, there is reason to believe that242only intra-Canada movements would be significantly affected by thispolicy.U.S. carriers have both cost and network advantages that should allowthem to successfully contest de facto intra-Canada markets. Since theopen border policy would permit U.S. firms to connect many Canadiancenters to US hub airports, it would be possible for these carriers tooffer intra-Canada services through hub connections. U.S. carriersoperate hubs at sites close enough to the Canada-United States borderthat the addition to total travel time facing the consumer would beacceptable, particularly when weighed against the probable advantagesin terms of ticket prices and flight times.The use of hub connections allows a U.S. carrier to reduce per-seat-mile costs, permitting the sale of services at lower yields. Thisshould attract discretionary, intra-Canada passengers through lowerprices. The diversion of this traffic away from Canadian carriers maybe sufficient to reduce flight frequencies for Canadian-provided,intra-Canada services.113 Ironically, the reduction in flightfrequencies for these services may prompt non-discretionary travellersto opt for passage on US carriers, as the added total travel timeassociated with hubbing may be more than compensated for by the reducedschedule delay time US carriers can offer due to greater flightfrequency. This effect would be particularly significant in long-haul,243intra-Canada markets in which the passenger is already faced withhaving to make a connection using Canadian carriers’ services.114In contrast, the attractiveness of Canadian-provided services forintra-United States passengers is relatively minimal. While there issignificant east-west travel in the United States, the prospect ofrouting through Canada presents substantial additional travel time.Moreover, the existing services of US carriers are of such quality, interms of both price and frequency, that Canadian carriers would havelittle hope of contesting these markets. Assuming that the higheryields in Canada would be associated with any Canadian-providedservices introduced, the fares for these services would not becompetitive with those of US carriers. The frequencies of US carriers’services, made possible by density and scope economies associated withtheir hubbing operations, would be unattainable by Canadian carriers.Even if Canadian carriers successfully establish co-terminal hubs inthe United States, they would be prevented from utilising thesefacilities to service intra-US traffic as these would be cabotageoperations: this would not be legal under the terms of the contemplatedopen border agreement. Canadian carriers can therefore not be expectedto capture significant shares of intra-United States traffic.2444.4 Summary and ConclusionsThe anticipated effects of the introduction of an open border regimeare as follows:1. An increase in the number of cross-border city-pairs receivingdirect services.2. The introduction of flights linking major Canadian centerswith U.S. hub airports where such routes are not currentlyreceiving services.3. Increased competition on transborder routes, particularly onthose where the origin or destination is not a major U.S. hubairport.4. Lower real fares for transborder services due to increasedintra- and inter-route competition.5. Diversion of significant volumes of intra-Canada traffic awayfrom Canadian carriers to U.S. carriers through mirror imageroutings.245In summary, the open border concept has merits. It would lead togreater efficiency in the provision of transborder air services, andhence to lower average real fares. Both Canadian and United Statescarriers have strengths that should allow them to benefit from an openborder regime. There are problems that cast a shadow of doubt over theacceptability of this policy.First and foremost, the nature of the geography of transborderairline markets gives persistent advantages to US air carriers.Morrison and Winston (1990) determined that the ability to contest aparticular air transport market was directly related to the number ofconnections that were possible at each end-point of that market.15The vastly greater number of possible destinations in the United Statesversus Canada, coupled with the extensive route networks and hubbingoperations of US carriers, give them an enormous advantage overCanadian companies in this regard. Even if Canadian carriers canestablish co-terminalisation hub operations in the United States,without duplicating the comprehensive route networks of US carriers,their ability to contest all transborder markets will be inferior.Second, a capacity problem exists at many of the key airports in theUnited States. This scarcity of capacity may prevent the increase incompetition that the advantages of the open border strategy are relianton. Without assured access to important US airports, Canadian firms246will not be able to expand their transborder operations to reap thebenefits of an open border regime. Furthermore, the ability ofCanadian firms to establish co-terminalisation hubs in the UnitedStates is crucial to their success in an open border environment. Theenormous cost of creating these facilities may be beyond the capabilityof the financially-strapped Canadian carriers.Third, although alliances between Canadian and US air carriers havebeen argued to enhance the competitiveness of Canadian-providedservices, the nature of the extent of alliances that are currentlypermitted can be expected to relegate Canadian carriers to theeffective role of regional carriers for their US partners, should aplain, open border structure be adopted. Having noted the advantagesof allied carriers, it should be noted that the scheme of revenuedistribution between the Canadian and US firms would have a bearing onthe desirability of sustaining such partnerships. If revenuedistribution was based on flight segments provided by each carrier, theCanadian firms would, if the most operationally advantageous routestructures were adopted, receive a relatively small proportion of thetotal transborder fare. Difficulties in arranging a suitabledistribution of revenues should be anticipated, and would undermine theintroduction and/or persistence of agreements between Canadian and USairlines.247In conclusion, although an open border policy would be beneficial totransborder travellers of both Canada and United States, it hassignificantly adverse implications for the overall viability ofCanadian carriers. As one of the stated aims of Canadian air transportpolicy is to ensure the viability of national airlines, another form oftransborder agreement must be explored.2485. The Cabotage Option5.1 Basic PhilosophyUnder cabotage, all markets within, and between, Canada and theUnited States could be served by any Canadian or U.S. carrier. Theinstitution of cabotage would necessarily include the structuralchanges associated with the open border strategy, as discussed in thepreceeding section of this chapter. Pricing is already deregulated onall services within the two countries; price deregulation would occuron all services between Canada and the United States with cabotage.The pursuit of a regime of cabotage in North American airtransportation would provide both Canadian and U.S. air carriers withthe regulatory freedom to enter markets in either country. It isintended to increase the amount of potential, if not actual,competition on both transborder routes between, and on domestic routeswithin, the two countries. This added competition, whether potentialor actual, should benefit consumers in both countries by exertingdownward pressure on prices.6’782495.2 Structural Implications of Cabotage Regimes5.2.1 Regulatory StructureCabotage could be applied either unilaterally, allowing only one ofthe parties to operate services on purely domestic routes in the othercountry, or bilaterally, wherein airlines of both countries couldoperate on routes in either country. The adoption of a cabotageagreement would not change the regulatory authority of the respectivegovernments over air carriers that are identified as national to therespective countries. The existing regulations regarding the grantingand retention of carrier operating certificates, personnel standards,and safety requirements, would continue in force. Canada and theUnited States have largely similar regulations in the aforementionedareas. Therefore, the desirability of a cabotage regime should not bedetracted from by the possibility of conflict between Canadian andUnited States interests in these areas.Nationality of CarriersCarriers would continue to have distinct nationality. A carrier thatwas owned at least 75 per cent by citizens of Canada (the UnitedStates) would be considered a Canadian (United States) carrier. Though250there would be no restriction on the entry by either country’s firmsinto markets anywhere in the two countries, the regulations of theairline’s country of nationality, pertaining to safety, employment, andtaxation, would continue in force. Cabotage is not an agreement tofacilitate the corporate integration of the airline industries ofCanada and the United States, but rather an agreement strictly onaccess rights to transborder and domestic markets. In the absence ofreforms to Canadian and United States regulations regarding taxation,and airport access schemes, Canadian carriers should be expected to beat a competitive disadvantage under a cabotage regime.Other aspects of regulatory policies in Canada and the United Statescan be expected to inhibit the acceptability of a cabotage regime. Theintroduction of cabotage without some substantial modifications toseveral air transport-related policies in both Canada and the UnitedStates will pose problems to the achievement of the potential socialwelfare benefits of cabotage. Cabotage is intended to increase theefficiency of air transportation within and between both countries.The means of increasing efficiency is to introduce or enhancecompetition in air services markets. There are existing policies inboth Canada and the United States which, if unaltered at the inceptionof cabotage, will undermine the goal of competition enhancement onterms fair to both Canadian and U.S. interests.251In Canada, the tax regime confronting Canadian air carriers adverselyaffects their cost competitiveness relative to United States airlines.As discussed in Chapter IV, Section 2.3, there are higher taxes facingCanadian firms with respect to fuel and capital. While taxes on fuelwould equally apply to U.S. carriers operating in Canada, thedifferential in capital taxes would adversely affect only Canadiancarriers. In order for cabotage to allow fair competition, it would benecessary for Canadian policy to change regarding capital taxes for aircarriers: they would have to be reconciled with U.S. tax rates.The significant United States policy that would inhibit therealisation of potential benefits from cabotage may be more difficultto change. This is the method of capacity allocation at U.S. airportsthrough the use of private property rights. As discussed in ChapterIV, Section 3.3, airline control over airport infrastructure, and theexisting method of slot allocation by the U.S. Federal AviationAdministration at the four slot controlled airports in the UnitedStates, pose serious barriers to entry to many U.S. airlinemarkets.’19’201 The fact that U.S. airport facilities have beenfinanced by the controlling carriers, that, where not explicitlyfinanced, airlines have long-term contractual rights to their use, andthat airports in the United States are locally, rather than federally,financed and controlled, pose serious obstacles to the solution of theaccess question. It is therefore essential that a cabotage agreement252between the United States and Canada contain provisions to ensureCanadian access to U.S. airports.In summary, there are several areas of air transport regulation thatare essentially common to both Canada and the United States. Ingeneral, these are in the realms of carrier licensing, and personneland operational standards. These areas of regulation should not beexpected to complicate the introduction of a cabotage regime whereinboth Canadian and U.S. firms can equitably participate. However,Canadian tax laws and United States policies regarding air transportinfrastructure allocation, if maintained in their present forms, wouldnot allow for equality of opportunities for Canadian carriers under acabotage regime.5.2.2 Barriers to EntryCanadaThe nature of the Canadian air transport environment is such thatU.S. carriers wishing to operate cabotage services in Canada shouldencounter few entry barriers. The greatest problem facing U.S. firmsshould be the feeder carrier networks of the established Canadiancarriers. Both Air Canada and Canadian Airlines International have253extensive domestic networks in Canada. These are serviced by regionalairlines who are either wholly or substantially owned by the two majorairlines. However, the bulk of Canadian air traffic is between largeCanadian centres, therefore it would certainly be feasible for U.S.airlines to establish operations serving most Canadian travellerswithout creating comprehensive networks in Canada.The United StatesAs noted in the foregoing discussion of regulatory structure, asizeable obstacle facing Canadian carriers wishing to compete in U.S.domestic markets is the control that U.S. carriers have over theallocation of capacity at airports in the United States. This controlover airport facilities has prevented U.S. firms from contesting manyairline markets in the United States. Canadian entrants would be in aneven worse position since they have higher costs per seat-mile thantheir U.S. counterparts.Morrison and Winston (1990) examined the factors that determined thatthe ability of an airline to enter a particular airline market in theUnited States. Using a probit regression, they found that theprobability of entry into a market was positively influenced by theextent of that airline’s activity at the end-point airports of the254market, and negatively influenced by yields for the market. They foundthat the elasticity of the probability of entry with respect to anairline’s share of total enplanements at the end-point airports was0.1; the carrier’s own network strength was found to be the greatestdeterminant of entry.122 The somewhat counter-intuitive finding thatmarket yield was negatively related to the probability of entry wasexplained as an indication of the effects of market dominance on fares:where yields were higher, the market was less contestable.123 Fromthese findings, we should conclude that Canadian carriers would havedifficulty contesting domestic U.S. markets because of the entrybarrier presented by the necessity of having a well-developed network.Due to their limited existing presence at U.S. airports, Canadiancarrier entry would be discouraged as they would have to createextensive networks immediately and simultaneously to be competitivewith U.S. carriers.The effects of frequent flier programs, and of CRS dominance in thevicinity of many major U.S. metropolitan also reduces thecontestability of many U.S. airline markets.’24’1256 The conduct ofU.S. carriers has largely been to avoid contesting markets whereanother carrier is dominant. Performance has therefore suffered inthat premium prices have been charged for these services. Althoughcabotage rights would afford Canadian carriers theoretical access toall U.S. markets, those markets that have proved incontestable by U.S.255carriers should prove to be similarly incontestable by Canadiancompanies.5.2.3 Implications of Cabotage for Air Transport DemandA cabotage agreement should be expected to influence the total levelof demand for airline services in Canada and the United States. Thepolicy can be expected to increase the overall volume of air travel inCanada and the United States due to induced demand. This induceddemand would be the result of lower average air fares. Lower averageair fares are anticipated because of the increase in competition incity-pair markets in both countries. Canadian domestic markets can beexpected to be most affected by cabotage, as these markets currentlyfeature the highest average yields in North America due to therelatively low number of average competitors in Canada.2565.2.4 Implications of Cabotage for the Supply of AirTransport ServicesEffects on the Number of Potential CompetitorsCabotage permits carriers access to domestic markets in the othercontracting party’s territory. The total number of airlines operatingin Canada and the United States can be expected to remain the sameunder cabotage. However, we may expect that there would be an increasein the average number of competitors per market in both Canada and theUnited States. The potential for an increase in the number ofcompetitors should be greatest for transborder markets. Canadiandomestic markets should also experience an increase in the number ofairlines providing services. U. .S markets can be expected to realisethe least effects regarding the number of competing carriers because ofthe relative weakness of Canadian carriers in being able to enter thesemarkets. The average number of competitors on all North Americanroutes can be expected to be in the range of 1.6, the current levelin Canada, and 2.8, the current level in the United States. Theaverage number of competitors in Canada should be most greatlyaffected.257Horizontal and Vertical Integration of Airline FirmsThe adoption of a cabotage agreement should have an impact on theintegration of airline firms. The policy should not affect the statusof vertical integration between firms of the same nationality: forexample, AirBC can be expected to remain a part of Air Canada. Wherecabotage should influence integration is in the area of the horizontalarrangements between Canadian and United States carriers.It has been argued, in Chapter II, that the principle motivation forthe formation of alliances between Canadian and American air carriersis that such alliances capitalize on the demand-related effects ofscale economies. These agreements have been mutually beneficial to theCanadian and American carriers alike for, without them, thecomprehensive geographic coverage these arrangements permit would notbe possible. With the introduction of cabotage rights, the likelihoodof the continuation of these alliances is questionable.Discouragingalliances ismarkets havingairline marketsincentive forcompete againstthe continuation of the existing Canada-United Statesthe desire of individual carriers to capture shares ofhigh traffic densities. There are many more densein the United States than in Canada. This gives anCanadian carriers to defect from alliances in order toU.S. alliance partners in high-density, U.S. markets.258Conversely, U.S. firms will have an interest in entering those intraCanada markets having attractive traffic densities: those connectingVancouver, Toronto, and Montreal. Thus, there would be a divergence ofinterests between the Canadian majors and their current United Statesallies.As outlined in section II, Canadian companies will find that, in theabsence of meaningful changes in U.S. policies regarding the allocationof airport and airways capacities, their prospects for entering most ofthe most lucrative U.S. markets will be slim. The control over accessto these facilities that is enjoyed by major U.S. carriers will makeeffective participation by Canadian companies very difficult. TheCanadian firms will therefore have an incentive to sustain theirpartnerships with their US. allies.The major U.S. airlines are in a better position to benefit fromcabotage in the absence of partnerships with Canadian allies. Thebarriers to entering Canada’s major markets are readily surmountable byAmerican carriers. Capacity rationing in Canada is done through usercharges rather than property rights. U.S. firms should continue to beable to access Canadian facilities with little difficulty: they needonly be able to pay the appropriate user charges. Canadian carriershave no direct control over airport facilities in Canada, thereforethey would be unable to block U.S. entry, save for through political259means. U.S. participation in attractive intra-Canada markets ispromoted by this ease of access.At face value, a cabotage policy appears to give U.S. carriers boththe incentive and ability to cream-skim intra-Canada routes. Becauseof a lower cost structure, it will be possible for U.S. carriers tooffer lower prices to consumers; they should be expected to capture asizeable traffic share. This participation will divert traffic fromCanadian carriers, causing inevitable strain on cross-border alliancesas a consequence.There are benefits to carriers that militate in favour of thecontinuation of these agreements. These benefits accrue to bothCanadian and U.S. firms.By maintaining their alliances with Canadians, U.S. companies couldcontinue to offer the traveller service to a greater number of Canadiandestinations of all sizes, without the need to create their own feeder-carrier networks. While the introduction of a cabotage regime wouldallow Americans to create their own intra-Canada feeder-carriers, thecost of so-doing would be substantial. They are of a magnitude thatwould represent a considerable financial obstacle, even for large U.S.airlines. Demand-related economies of scale, the cost of creating newfeeder-carriers, and the fact that this market niche favours existing260Canadian carriers, will promote the continuation of Canada-U.S.partnerships.The essence of the motivation for the continuation of existingalliances between Canadian and U. .S carriers is in the requirements foradditional aircraft that increased competition on all North Americanroutes will necessitate. There are three ways that the additionalaircraft necessary to provide greater capacity in this setting could beobtained: through better capacity utilisation (increased flying timeper aircraft per diem), through acquisition (whether by purchase orlease), or through operating contracts with other carriers. The lattermethod being one of the very bases of airline alliances.An ironic outcome of cabotage could be that Canadian carriers couldbe forced out of some Canadian domestic markets due to the costadvantages enjoyed by American firms. This cessation of services wouldrender some portion of the Canadian companies’ fleets superfluous.They might find that a sale or lease of these aircraft to U.S. carrierswould be in order; the U.S. carriers would need this flight equipmentin order to make up for aircraft devoted to the conduct of theirCanadian cabotage operations. Unfortunately for Canadian investors andlabour, their factors of production would not be as readilytransferable as is flight equipment.261It is possible that new Canada-U.S. alliances would be formed withthe inception of cabotage. As indicated previously, the establishmentof Canadian co-terminalisation hubs in the United States would presentan opportunity for alliance with local U.S. commuter carriers. Afurther alliance possibility is that new alliances would emerge betweenCanadian carriers, and U.S. regional carriers.The characteristics of an ideal U.S. partner for the Canadiancompanies would be one which has a strong presence in markets in thesouthern United States, yet which has a relatively weak presence inmany northern U.S. markets. Such a partnership would benefit bothcompanies by allowing them to complete a continent-wide network. TheCanadian firm would serve Canada and the northern U.S. from hubs atCanadian cities; the US airline would serve the southern U.S. usingits hubs in those regions of the United States. Transborder flightswould connect the respective hubs.Both Canadian majors would be suited to entering into such apartnership. An obvious U.S. carrier to form such an alliance withwould be Southwest Airlines. Southwest operates hubs at San Diego,Houston, Phoenix, El Paso, Austin, Albuquerque, Oklahoma City, SanAntonio, Tulsa, and New Orleans. Alliances involving Canadian carriersand Southwest would provide feasible network layouts to servecontinental markets. However, the thus allied carriers’ presence would262be relatively weak in the important northeast region of the UnitedStates.The desirable alliances that would be formed with the inception ofcabotage can be predicted (albeit crudely) from the merger activitythat occurred as the result of deregulation in the United States.Several significant U.S. carriers were absorbed into larger firms overthe period 1979 to 1991. It appears that the purpose of theseacquisitions/mergers was both to form more comprehensive networks inthe United States, and to gain greater control over important hubairports in that country.Under cabotage, a similar pattern of behaviour may occur. Therewould, however, be two significant differences in conduct between thepost-deregulation and post-cabotage situations.First, as has been extensively discussed throughout this paper,airlines operating in the United States face market entry barriers inthe form of accessing airport facilities. Where a U.S. carrieracquires or merges with another, it thereby gains the feeder-carrierrelations and, more importantly, access to those airports that theacquired or mergee carrier held the property rights at. In thecabotage situation, the motivation related to airport access issomewhat unidirectional: only the Canadian carrier will need to enter263into the alliance to secure U.S. airport access. U.S. carriers mayfreely access almost all Canadian airports because of the means ofcapacity allocation in Canada. The U.S. carrier will nonetheless bemotivated to alliance by the comprehensive national coverage of theCanadian carriers.The second difference that will be affect post-cabotage conduct isthat the outright merger of Canadian and U.S. companies is prohibitedby legislation--as discussed in Chapter II. Any form of such analliance must therefore remain largely contractual in nature.Canadian carriers would have an advantage in this setting as there isa greater number of alternative U.S. partners with whom to formalliances with. They would be able to seek the most profitable suchrelationship, offer the associated U.S. firm the advantage of completeCanada-U.S. coverage. As has been previously asserted, the ability toprovide a comprehensive network has significant demand-side effects inairline economics. In contrast, the Canadian companies will likelyfind themselves in the position that Canadian regional carriers were inafter Canadian deregulation: they will have no markets into which togrow. This eventuality suggests that, from the point of view ofCanadian carriers, the cabotage option will be no more beneficial thanthe adoption of an open border regime. Furthermore, with the prospectof non-allied, U.S. majors electing to enter Canadian mainline markets,264Air Canada and Canadian Airlines International may find their domestictraffic base being gradually eroded under cabotage.On balance, it will probably be better for the Canadian firms to allythemselves with the strongest U.S. carriers. They will find that themarketing advantages (in terms of frequent flier programs and CRSs)urge this conduct. This will also reduce (though certainly noteliminate) the threat to their intra-Canada operations from U.S.competition.5.3 Expected Conduct Under a Cabotage Regime5.3.1 PricingWe have previously discussed the influences that market structureshave on airline pricing behaviour. Having seen evidence regarding theextent to which airline fares are determined by input prices, factorproductivity, and the extent of concentration in individual airtransport markets, we can make the following contentions regarding thenature of air fares that would emerge with the adoption of a cabotageagreement.First, the allowance of cabotage should not be expected to influence265the level of input prices facing the airlines. Therefore, the cost peravailable-seat-mile produced by the carriers would presumably not beaffected by the introduction of a cabotage regime in so far as thepolicy’s effects on input prices are concerned. For example, althoughfuel prices are higher in Canada than in the United States, bothCanadian and U.S. carriers operating point-to-point services withinCanada would face the same fuel prices. Therefore, the bearing ofinput prices on air fares should be expected to be neutral with respectto the adoption of cabotage.Second, the permission of cabotage can be expected to increase factorproductivity. The expected productivity growth is actually aderivative of the open border component of the policy regime. In astrict sense, cabotage itself should have little impact on factorproductivity, save for the effect of induced traffic on densityeconomies.The greatest impact of cabotage should come from the increase in thenumber of potential and/or actual competitors for individual airtransport markets. We have discussed the significance that the numberof competitors has had on price levels in airline markets. Theexpectation is that cabotage rights should increase the number ofpotential and/or actual competitors in those airline markets to whichcabotage rights would be applicable. Since we have argued that the266greatest impact that cabotage rights would have would be for the numberof competitors in Canadian markets, the greatest influence of thepolicy on air fare levels should be expected for intra-Canada services.The difficulties facing Canadian carriers wishing to enter should be expected to preclude their entry into most suchmarkets. Therefore, the effects of cabotage rights on prices in U.S.domestic markets can be expected to be minimal.Where Canadian carriers prove able to enter intra-United Statesmarkets, they can be expected to assume the role of price-takers. Thehigher cost per available-seat-mile of the Canadian carriers shouldprevent them from offering lower prices to entice customers. Moreover,the ready prospect of price retaliation by U.S. firms would appear topreclude this strategy. Presumably, the Canadian carriers would haveno ability to charge premium prices in U.S. markets; they would beinterested in establishing a market presence, and relatively higherprices would compromise this objective. The pricing behaviour of U.S.firms operating in intra-Canada routes presents a less obvious picture.U.S. carriers enjoy an overall cost advantage over Canadian carriers.However, Windle (1991) has attributed this cost advantage toproductivity differences between carriers in the two countries.’27 Forcabotage operations in Canada, the U.S. carriers would lose somemeasure of this productivity differential, as it is based on thetraffic density advantages of operating in the United States. Although267the cost per available-seat-kilometre produced by United Statescarriers in 1991 was $O.0665 versus $O.1053 for Canadian firms,128 thetransferability of these figures to U.S. operations in Canada islimited by the differences in traffic densities between the twocountries’ markets, and the input price differences (particularly forfuel) between Canada and the United States. When these costdifferences are taken into consideration, the apparent cost advantagesof US carriers may not be as great for cabotage operations in Canada.The significance of a U.S. firms’ cost advantage is that it affordsthe U.S. carrier the option to undercut Canadian fares, without thepenalty of operational losses, and reduces the probability ofsuccessful prosecution of potential predatory pricing suits by Canadiancompetitors.The eventual conduct of U.S. firms operating intra-Canada servicescould presumably be that of a price leader, in which case they wouldtake advantage of lower costs to offer fares below those of Canadiancarriers, or that of a price taker, in which case they would providecapacity at the same prices as their Canadian rivals. For thefollowing analyses, we consider an average Canadian airline market,assume that there is sufficient traffic volume to permit entry to thatmarket by a lone U.S. carrier, and that the characteristics ofthe nature of the competitive conduct of Canadian air carriers wouldremain the same in the face of entry by a U.S. carrier.268A crude estimation of the influence that U.S. carrier entry, on thebasis of assuming a price-taking position, would have on individualintra-Canada markets can be made by considering the relationship thatwas found between average yield and market concentration for UnitedStates markets by Graham, Kaplan, and Sibley (:1983). The relationshipthey found between the Herfindahi index for a market and its fare permile suggested that the decrease of the H2 value from 0.50 to 0.33(which we use to approximate the effect of U.S. entry into intraCanada markets based on the assumption that the number of carrierswould be increased from two to three, and that all participatingcarriers would realise equivalent market shares), would result in aroughly 6 per cent reduction in the market fare.129We base a more sophisticated analysis of the potential impact of U.S.carrier entry to Canadian domestic markets on the conjecturalvariations model of oligopoly conduct)30 The conjectural variationsmodel provides a means by which a firm may estimate the reaction of itscompetitors to changes in the firm’s output and price offerings. Themodel formalises the relationships between the variables price, marketprice elasticity of demand, the firms’ market shares and coststructures, and a reaction term indicating the expected conduct of thecompetitors.The model is developed from the profit, price, quantity, and cost269relationship,= x. p(X) - C’(xi)The first order condition24P+X. L-4=Odx 1 dx1 dx dx1gives rise to the definition of the conjectural variation term. Since= + Z = 1 + V Pthe term vi is the variation in output that firm i conjectures wouldoccur for the other j firms’ combined output as the result from analteration in the output of i. Thus, v1 is the conjectural variationof firm i.131Brander and Zhang (1990,1993) used a conjectural variations model toanalyse the conduct of airlines in a duopoly market. Specifically,they examined the conduct of American Airlines and United Airlines inChicago-based airline markets for which the two carriers’ combinedtraffic share was at least 75 per cent.270Brander and Zhang used the model in its formv. = (p - C.)(e)(X)1 (p)(s13 — 1alternatively expressed asp -= s (1 + v)p ewhere, s. = the market share for firm i, and e is the absolute value ofthe market price elasticity of demand.Brander and Zhang (1990) estimated the conjectural variations termsfor American Airlines and United Airlines. The quantum of theconjectural variations term would provide an indication of the natureof the conduct of these firms in their operations in the Chicago-basedmarkets. A conduct parameter value of -1 would imply Bertrandoligopoly behaviour: the firms’ cost structures would determine themarket price, with that price being equal to marginal cost, and themarket being split between the firms in zero-profit equilibrium.’32 Aconduct parameter of 0 would imply Cournot oligopoly behaviour: normalprofits being secured by the firms through a process of adjustments tothe quantity supplied to the marketplace, with each firm making itsquantity decisions based on its rivals supplied quantities remaining at271existing levels. A conduct parameter of 1 would imply that there wascollusion between firms. This value would indicate a cartel situation,with monopoly rents being divided between the cartel members.133”34Brander and Zhang (1990) found that the behaviour of AmericanAirlines and United Airlines for the routes under study tended towardCournot duopoly. 135Following the approach of Brander and Zhang, we investigated theconduct of Canadian air carriers to determine the nature of competitionin the Canadian airline industry. The purpose of our investigation wasto provide an ad hoc assessment of the form of Canadian competition onwhich to base an estimation of the likely response of Canadiancarriers, and thus Canadian airline markets, to entry by U.S.competitors. We used the conjectural variations model, as in Branderand Zhang (1990), using parametric data from a number of sources.272The National Transportation Act Review Commission reported thefollowing information for Canadian and United States air carriers foriggi:136Table 7.REVENUES AND COSTS OF CANADIAN AND U.S. AIR CARRIERSCanada USCost per available seat km. $O.1053 $0.0665Revenue per revenue pass. km. 0.1529 0.1039Interest expense per ASK 0.0040 0.0015Wage expense per ASK 0.0324 0.0229Fuel expense per ASK 0.0147 0.0096Assuming that U.S. carriers would face equivalent market-relatedcosts to Canadian firms where cabotage operations are conducted, weimpute a cost per ASK for U.S. carriers operating such services. Thecalculations are based on the premise that U.S. firms would retain costadvantages with respect to interest and wage expenses, while all othercosts would be identical to those experienced by Canadian operators.273Based on these assumptions, the cost per ASK for a US carrier’scabotage operations would be:Canadian cost per ASK $0.1053.less: (in expCdfl - mt. exp ) - 0.0025/ASK(wage expCdfl - wage exp) - 0.0095/ASKImplied US cost per ASK in Canada $0.0933....Our analysis proceeds on the assumption that the average prices andcosts for Canadian airline markets can be represented by the foregoingfigures. We therefore are analysing Canadian carrier conduct for ahypothetical Canadian airline market. We assumed that thishypothetical market would be equally divided between the two majorCanadian carriers, Air Canada and Canadian Airlines International, andthat these carriers would have identical cost structures.We rearrangep-C = s. (1 + i’m)p eto yield the conjectural variations model such that the conductparameter, v., is the objective variable:v. = (p - C1) - 1.p274Utilising the cost and price figures from the NTA study, andmaking the assumptions as previously noted, and taking the estimates ofCanadian airline markets’ average price elasticity of demand from Oumand Gillen (1983),137 the above model yields the following symmetricalconduct parameters for the Canadian carriers:if:e=1.1 v=-0.32e = 1.2 = -0.25e = 1.3 = -0.19These conduct parameters suggest that the Air Canada and CanadianAirlines International have tended toward Cournot duopoly behaviour:the conduct parameters are closest to 0.0 as opposed to 1 or -1, whichwould imply Bertrand and collusive oligopoly conduct, respectively.While this analysis has been crude at best, the reality of the Canadianindustry is that it has been characterised by over-capacity in terms ofavailable-seat-kilometres)38 This over-capacity suggests a form ofquantity-based competition between the Canadian air carriers.Returning to the matter of the effect on Canadian airline prices ofentry by a U.S. competitor, we make the assumption that quantity-basedcompetition would continue to prevail in Canadian airline markets, thusthe conduct parameter of the participating carriers would be assumed tobe 0. We further assume an equal distribution of traffic amongst thetwo Canadian carriers and the lone U.S. entrant carrier in this275hypothetical Canadian airline market. By rearranging the conjecturalvariations model (recalling that we have assumed a conduct parametervalue of 0), we arrive at the following equation for the expected price*(p ) for this hypothetical Canadian market:*p = - (e)(C)Si - éThe figures for elasticity are from Oum and Gillen (1983). The cost C1was the hypothesized cost for U.S. carriers arrived at through our owncalculations. The expected elasticity-dependent prices for thehypothetical market would be:*For: e = 1.1, p = 13.33e = 1.2, p = 12.87e = 1.3, p = 12.50Note that these “prices” are in fact revenues, in terms of cents, perrevenue-passenger-kilometres, or yields.In comparison, the existing average Canadian market yield is 15.29cents. The implication of our calculations is that the entry of cost-advantaged, U.S. firms into the hypothetical Canadian market wouldreduce prices by between 12.8 and 18.2 per cent, depending on themarket’s price elasticity of demand. These reductions are double totriple those predicted by our previous consideration, which, toreiterate, was based on Graham, Kaplan, and Sibley (1983) for the276experience in US airline markets.We considered that an explanation for the difference in magnitudebetween the expected prices based on the two approaches considered maybe that U.S. firms would have had relatively similar cost structures.In contrast, Canadian firms, and a U.S. firm operating cabotageservices in Canada, had been hypothesized to have significantlydifferent cost structures. As a check on the price results obtainedfrom the estimation based on Graham, Kaplan, and Sibley (1983), weestimated the expected price for the hypothetical Canadian market underthe assumption that all firms in that market had identical coststructures. We therefore estimated=- (e)(C)s. - eusing as C. the cost per available-seat-kilometre realised by Canadianfirms, and retained the hypothesized division of traffic between thetwo Canadian firms, and the U.S. entrant. The elasticity-dependentexpected yields, and percentage reductions were:*For: e = 1.1, p = 15.04 %p = -1.6e = 1.2, P = 14.52 %p = -5.0e = 1.3, p = 14.11 %p = -7.7Note that these results are closer in magnitude to those suggested byGraham, Kaplan, and Sibley (1983).277It may be inferred from this analysis that the extent of costadvantages enjoyed by U.S. entrants operating in Canadian domesticmarkets could be directly translated into the magnitude of farereductions for Canadian domestic markets in which U.S. carriers electto offer services. Since U.S. carriers would presumably retain atleast some portion of cost advantage over Canadian carriers, it shouldbe expected that the magnitude of the price reductions would be in theregions between 1.6 and 18.2 per cent, with the reductions in thevicinity of 5.0 to 6.0 per cent having good probability according toour ad hoc analysis. Variations can be expected from these figures tothe extent that individual market characteristics in Canada differ,particularly in terms of the price elasticity of demand, to the extentof cost differentials between U.S. and Canadian carriers, and to theextent that the conduct of Canadian carriers would be affected by theprospect of entry by U.S. firms.Regardless of whether U.S. entrants into Canadian domestic marketselect to behave as price takers or price cutters, the net effect oftheir participation would be to reduce average fares for the marketsthat they decide to enter. We have assumed that most U.S. domesticmarkets would be incontestable by Canadian air carriers. Those marketsthat are entered by Canadian carriers should realise minimal priceeffects, as the market share of the Canadian carriers can be expected278to be small.It must be noted that, aside from the impact of direct U.S. carrierparticipation in Canadian domestic markets, that their ability to servethese markets through the use of mirror image routings would alsoinfluence intra-Canada fares. Due to the qualitative characteristicsthese services, it should be anticipated that mirror image optionswould be most attractive to discretionary travellers.5.3.2 Expected Network EffectsCanadian Carrier NetworksWe have argued that cabotage rights would present limitedopportunities for Canadian air carriers to enter U.S. markets. Thedifficulties facing Canadian firms with regard to contesting most U.S.airline markets should confine them to providing cabotage serviceswhere the nature of such services is that they are consecutive cabotageextensions of transborder flights. In essence, Canadian carriers wouldtake advantage of cabotage rights to fill empty seats on U.S. domesticflight stages for what are essentially transborder services. Thesewould be associated with co-terminalisation hubs in the United States.This practice would allow Canadian participation in related U.S.279domestic markets by building on the traffic densities associated withCanadian transborder services.’39 The Canadian carriers would therebyreap density economies for the transborder and cabotage services.Canadian carriers can be expected to establish co-terminalisation hubsat U.S. locations having proven attractiveness for Canadian travellers.US, cities that would be appealing co-terminal hub locations on thispremise would include New York, Chicago, Tampa-St. Petersburg, SanFrancisco, and Houston. (These cities currently account for a largeproportion of transborder travel) However, U.S. airlines’ controlover facilities at airports in the vicinity of these centers maycompromise the ability of Canadian firms to establish coterminalisation hubs at these locations; New York and Chicago would beparticularly difficult to access.The Canadian carriers would benefit from the use of change of gageoperations associated with the co-terminalisation hubs. They couldemploy smaller aircraft to operate between the co-terminal hubs andsubsequent U.S. destinations. This could be accomplished through theuse of the Canadian carriers’ own aircraft, or through contractualarrangements with local U.S. carriers. The latter schema would beparticularly advantageous in that the Canadian carriers would benefitfrom the local carriers’ marketing activities.280United States CarriersU.S. carriers can be expected to enter major markets in Canada. Thelone entry barrier to U.S. participation in Canadian domestic marketsis the traffic consolidating advantages of Canadian air carriers viatheir feeder-carrier networks. However, there are several intra-Canadamarkets that have sufficient traffic volumes to allow U.S. entrywithout derogation due to lack of feeder traffic. In particular, theToronto-Montreal and Toronto-Vancouver markets should be attractive toU.S. firms.Other intra-Canada markets can be expected to receive service by U.S.carriers, but through the use of routings involving hub connections inthe United States. This method would allow U.S. carriers to serve manyCanadian city-pairs with both relatively low fares and relatively highflight frequencies. The attractiveness of one-stop services by U.S.carriers is enhanced by the general provision of intra-Canada servicesby Canadian carriers through the use of connecting flights--oftenthrough Toronto. The drawback is the potential extra flying timeassociated with routing through the United States, and the need forpassengers to clear Canadian customs on arrival at the destinationairport. These complications should limit the appeal of such routings,with their primary attractiveness being to discretionary travellers;business travellers could be enticed by frequent flier programs and281schedule advantages. Only U.S. carriers having hub operations atairports in close proximity to the Canada-United States border shouldbe expected to be able to contest such markets. American Airlines,Northwest Airlines, and United Airlines should be able to engage insuch services due to the locations of their northern hubs.5.4 Anticipated Performance Under a Cabotage Regime5.4.1 Division of Traffic Between CarriersGeneral Division of TrafficGillen, Hansen, and Ramos (1990) estimated the division of trafficbetween Canadian and U.S. carriers that would occur in the event of theadoption of bilateral, and both forms of unilateral, cabotage rights.They calculated that, if bilateral cabotage was permitted, the share oftotal North American traffic carried by Canadian companies would bereduced from 4.68 per cent to 3.34 per cent.’4° If cabotage waspermitted for Canadian carriers only, the Canadian carriers’ share oftotal North American traffic would increase to 5.08 per cent.’4’282Division of Traffic For a Hypothetical Canadian MarketUsing the conjectural variations model, we can estimate the trafficshares that may be realised by carriers in a hypothetical Canadianmarket. Rearranging the model, as described above, so that theexpected carrier share is the objective variable, and assuming that theconduct parameter is common across carriers, having the value 0, wehave the following equation:s1 = e (p- C)We estimated the U.S. entrant’s market share for two cases:1. Where the U.S. entrant would have cost advantages (asdescribed in the foregoing subsection, based on capital andwage cost advantages over Canadian carriers);2. Where the U.S. entrant would have the same cost structure asits Canadian counterparts.The expected price used in these calculations is that suggested byour calculation based on Graham, Kaplan, and Sibley; it is 94 per centof the current average yield in Canada.283Case 1.: (wherein = $0.0933 per ASK)For: e1.1 38.6%1.2 42.1%1.3 45.6%Case 2.: (wherein C = $O.1053 per ASK)For: e1.1 29.4%1.2 32.1%1.3 34.7%Our calculations (in Case 2) suggest that Canadian carriers wouldrealise total traffic shares of between 3.30 and 3.06 per cent. Themagnitude of the traffic share varying with the market elasticity ofdemand. In comparison, Gillen, Hansen, and Ramos (1990) suggested thatbilateral cabotage rights would reduce the Canadian carriers’ trafficshare (in terms of enplanements) by 28.8 per cent: from 4.68 per cent,to 3.34 per cent of total, intra-North America air traffic.’42We have omitted a consideration of the traffic share changes withrespect to intra-United States markets. As stated previously, webelieve that Canadian carriers’ cabotage operations in the UnitedStates would divert only a marginal amount of traffic away from U.S.carriers. On this premise, we made no calculation of the effect ofsuch services on Canadian carriers’ share of overall, North Americantraffic.2845.4.2 Implications of Cabotage for Market EfficiencyThe allowance of U.S. entry into domestic airline markets in Canadacould improve the long term efficiency of the provision of airtransport services in those markets which U.S. firms participate.However, the efficiency gains would be realised largely on the basis oflower cost structures of U.S. airlines that have been enabled byoverall network effects (ie. returns to hub-and-spoke networks) and dueto differential input costs between the two countries created in largepart by regulatory (particularly taxation) policies)43 Canadiancarriers could supply lower cost services within Canada if Canadianregulatory policy was similar to that in the United States.Windle (1991) has shown that the area of advantage for U.S. carriersis in total factor productivity, whereas overall unit costs actuallyfavour Canadian carriers.’44 Caves, Christensen, Tretheway and Windle(1987) determined that U.S. carriers have been able to realisesuperior productive efficiency, relative to carriers of othercountries, due to operating characteristics and technical efficiencies.In particular, operating efficiencies in the form of traffic densities,firm size, and capital utilisation were determined to have beenimportant sources of advantage for U.S. carriers in comparison toairlines of other countries.’45 Their conclusions suggest that U.S.285carriers should be able to offer more efficient intra-Canada servicesto the extent that such services can be integrated into their overallnetworks.From the findings of Caves, Christensen, Tretheway, and Windle (1987)it should also be recognised that entry by U.S. firms may have anadverse effect on efficiency. Caves et. al. have identified theimpact of the interactive effects of points-served and output, ameasure of scale economies, as making a contribution to reductions invariable costs.’46”7 With the entry of U.S. firms into Canadiandomestic markets, and the expectation that Canadian carriers would losesignificant market shares to these airlines, the dilution of traffic inintra-Canada markets would translate into dis-economies for Canadiancarriers’ network operations. In other words, U.S. entry into plumintra-Canada routes may detract from Canadian carriers’ overall trafficdensities so as to make the maintenance of their overall networks, attheir present scale, unviable. Therefore, while U.S. entry may lead tobeneficial effects in the specific markets in which they participate,the overall effect for Canadian markets may well be negative.The expected nature and magnitude of Canadian carrier participationin United States domestic markets should produce limited efficiencybenefits in these markets.286The allocative efficiency in Canadian airline markets may be improvedby the introduction of cabotage rights. With U.S. entry, it should beexpected that U.S. carriers would seek to expand their computerreservations systems presence in Canada. By implication, theconditions imposed on the Gemini CRS by the Canadian CompetitionTribunal regarding the operation of that system reflected the belief ofthe Tribunal that a virtual CRS monopoly in Canada would have adverseeffects on allocative efficiency.’48 With an increased presence ofU.S. airlines’ CRSs, it should be expected that greater competition inthe marketing of airline services would result. An increased presenceof alternative CRS systems is already be on the way, due to the rulingof the Competition Tribunal with respect to the exit of CanadianAirlines International from the Gemini CRS. This ruling allowsCanadian Airlines International to join the CR5 system of AmericanAirlines (AMR Corporation).The eventual structure of CRS services in Canada can be expected tobe a duopoly: Sabre (American Airlines and Canadian AirlineInternational) and Covia (Air Canada and United Airlines) should be theprincipal providers of CRS services. The participation of Covia inGemini should give the Covia system an initial advantage in marketpresence within Canada. However, Covia’s advantage may well be shortlived: with the exit of Canadian Airlines International from Gemini,several Canadian travel agency executives surveyed indicated that they287viewed the Sabre system as having consumer-related advantages, implyingthat agents would benefit from adopting the Sabre system.149 Whileadoption of a cabotage regime is not a necessary condition to increasecompetition between these two CRS systems, it can be expected toincrease the potential extent of CRS competition in Canada as other UScarriers may elect to participate in intra-Canada airline markets,concomitantly introducing their CRS systems to Canada.5.5 Summary and ConclusionsIn summary, the anticipated effects of the adoption of cabotagerights in Canada-United States air transport services are as follows:1. Where airlines of either Canada or the United States are ableto introduce viable cabotage services, consumers would benefitfrom the increase in market competition through expecteddecreases in fares. We have calculated that, for ahypothetical Canadian market, fare decreases would be in theneighbourhood of 6 per cent; decreases in United Statesmarkets can be expected to be of lesser magnitude due to theexisting presence of a greater average number of competitors.2. Cabotage rights may increase the extent of potential, if not288actual, competition between CRS suppliers in Canada. Thepresent CRS dominance of Gemini can be expected to be brokenregardless of whether U.S. carrier cabotage in Canada ispermitted.3. The form of cabotage operations expected of Canadian carriersserving United States markets is that associated with coterminalisation hub operations in the United States. Weanticipate that only through associating cabotage operationswith transborder services will Canadian carriers be able torealise the traffic densities necessary to compete in U.S. carriers can be expected to introduce cabotage servicesin Canada on routes having relatively high traffic densities.We anticipate that routes linking Toronto with Montreal andVancouver would be particularly appealing to U.S. entrants.5. Access problems with respect to airports at most large citiesin the United States will prevent Canadian carriers fromcreating meaningful (if not comprehensive) networks in theUnited States. Therefore, Canadian cabotage operations can beexpected to concentrate on regions of the United States thatare presently destinations of a high number of transborder289travellers, to take advantage of scope economies related totransborder services.6. U.S. carriers should have little difficulty in accessingfacilities at airports in Canada.7. Bilateral cabotage rights are expected to reduce the Canadiancarriers’ share of total traffic by roughly 29 per cent oftheir current level. This loss of traffic share, particularlyin intra-Canada markets, could compromise the overallviability of the Canadian carriers; their operations may haveto be scaled-back from their present form.8. Unilateral cabotage rights for Canadian carriers are expectedto increase their share of overall North American traffic byroughly 8.5 per cent.In conclusion, there are potential benefits from the adoption of acabotage rights regime for Canada-United States air transportation.Consumers, particularly in Canada, would receive benefits in the formof reduced air fares due to increased competition. Unilateral cabotagerights for Canadian carriers would allow these carriers to attract agreater (and more equitable) share of transborder travellers.Bilateral cabotage rights would further enhance the competitive290position of the already advantaged U.S. carriers. Therefore, anyadoption of cabotage rights as a part of a new Canada-United States airtransportation agreement should be restricted to unilateral cabotagerights for Canadian air carriers, assuming that the preservation ofviable, Canadian airline companies remains a policy objective.Our analysis suggests that, while changes in regulatory policieswithin Canada with respect to the taxation of factor inputs used by aircarriers should move the cost of producing such services closer to thelower cost structures of U.S. carriers. The real problem in providingefficient air transportation in Canada is rooted in the nature ofdemand in Canada--particularly in relation to traffic densities and thelinear direction of air travel in Canada. Permitting U.S. carriercabotage in Canada would provide these lower cost producers theopportunity to offer some Canadians lower fares, yet this would come atthe detriment to Canadian air carriers, who, on balance, would receiveless benefits in a bilateral cabotage environment. Canadian consumerswho were not patrons of airline services in markets entered by U.S.carriers could well be disadvantaged as Canadian carriers can beexpected to reduce the overall scale of their operations following theerosion of their important traffic bases by U.S. entrants.Therefore, only unilateral cabotage, permitting Canadian carriers tooperate in the United States, should be considered as a legitimate291policy option. Even if such rights are secured, Canadian carriers mayfind the establishment of such operations difficult if not impossiblein the United States due to the air transport infrastructure accessproblems in the U.S. These barriers are substantial, and therefore itmay be the case that any form of cabotage rights will prove unworkable.2926. Strategies Utilising Phase-Ins of Liberal Regimes6.1 Motivation for Alternative Forms of Introducing NewCanada-United States Air Transport AgreementsCanada and the United States have experienced difficulty in reachingagreement on a new bilateral air services regime. This difficulty hasbeen rooted in the desire of both nations to ensure that they do notenter into a relationship that will have unacceptable negativeconsequences for their respective airline industries. At the sametime, they have recognised that the current regime is inherentlyinefficient, with the burden of inefficiency being borne by consumersin both countries.It is evident that consumer welfare would be enhanced by theintroduction of a more liberal regime of transborder air services.However, it is also evident that the current options being proposedcould have negative ramifications for both countries’ air transportindustries. These adversities should prove particularly severe forCanadian carriers; they are confronted by significant structuraladvantages possessed by U.S. airline firms.It may be possible to introduce a liberalised regime in such a way asto provide the expected benefits to consumers, while at the same time293minimising the adverse effects on the Canadian carriers. Properlyconceived, such a scheme would achieve the objectives of increasingefficiency while satisfying the equity criterion. The phasedintroduction of an open border regime has been suggested as a suitablesuch method.’50’’16.2 A Phased Introduction of an Open Border Regime BetweenCanada and the United States6.2.1 Basis of the Phase-In ApproachThe purpose of phasing-in a liberalised regime is to give protectionto Canadian air carriers during the short term. This protection isconsidered essential as the U.S. airlines enjoy distinct advantagesover their Canadian counterparts. If a liberal regime was introducedon a single event basis, the advantages enjoyed by the U.S. carrierscould be expected to allow them to dominate most, if not all, newtransborder routes. Therefore, the advantages of U.S. carriers wouldbe counter-balanced by advance access for Canadian carriers under thephase-in strategy. This would meet the objective of equity ofopportunity that is one of the requirements of a new Canada-UnitedStates air services agreement.2946.2.2 Application of the Phase-In Strategy to theIntroduction of an Open Border RegimeThe phase-in approach would be applied to the creation of an openborder regime; any form of cabotage regime poses too many accessproblems to be seriously considered at this time. These accessproblems are the result of the nature of the means of capacityallocation at airports in the United States. As discussed in ChapterIII, Section 3.3, the issues related to the accessibility of U.S.airports appear to be far from solution at this time. Furthermore, theextensive, existing hub networks of the U.S. carriers should continueto pose a significant entry barrier to Canadian firms wishing tointroduce cabotage operations in United States markets. Incombination, these factors make the adoption of a cabotage regime, evenwith the safeguard of a phase-in approach, untenable for theforeseeable future.In contrast, an open border regime appears to be feasible. It hasbeen argued, in Chapter IV, Section 4.3, that an open border regime maypresent opportunities for traffic growth for the Canadian carriers. Inparticular, Canadian carriers can be expected to benefit from beingable to offer more services to destinations in the southern United295States, to which Canadian travellers are attracted for reasons of bothbusiness (for example, San Diego) and pleasure (for example, NewOrleans). The alliances of the Canadian carriers with U.S. airlinesshould also prove to be advantageous for the Canadian firms with anopen border regime.6.2.3 The Phased Approach and Canada-United States CarrierAll lancesThe contention in Chapter V, Section 4.3 was that allied carriersshould be expected to be the dominant service providers in many marketsunder an open border regime. It was noted therein that the immediateintroduction of an open border regime could lessen the value of Canada-United States carrier alliances. The phased approach should actuallyenhance their continuance, as the allied carriers could be first tointroduce new, transborder services, giving them advance marketpresence.It should be recognised that the phased introduction of an openborder regime cannot guarantee that the ultimate distribution ofbenefits will favour Canadian carriers allied with U.S. airlines.While the phase-in period can be expected to be associated with theintroduction of new transborder services by Canadian allied-carriers,296in the long term these operations may be supplanted by U.S. allied-carriers. While there is probably no liberal policy that can guaranteethat Canadian carriers will capture and sustain a greater marketpresence, the allowance of greater equity exchange between Canadian andU.S. airlines might help to cement alliances. With stronger alliances,the position of the Canadian firms may be more secure with regard tooverall North American operations. In particular, this would giveCanadian firms secure access to U.S. airports; this would give U.S.carriers secure access to Canadian feeder networks and a moreadvantageous position with Canadian policy-makers.6.2.4 The Duration of the Phase-In PeriodA most important dimension of the phase-in concept is thedetermination of the duration of the phase-in period. This time framecould vary from zero time (immediate introduction), to indefinite(sustained advantage for Canadian carriers). The duration of thephase-in period should effectively be that over which Canadian carrierscould be realistically expected to establish services in newly-access ible, transborder markets. There are several determining factorsaffecting this time horizon.At the very least, the phase-in period must be sufficient for the297Canadian carriers to acquire the necessary flight equipment to conductthe new transborder operations. As the Canadian carriers haverelatively small fleets, it will probably be necessary for them toacquire additional aircraft in order to operate new transborderservices. As noted in Chapter III, Section 1.1, additional aircraftcan be obtained via purchase or leasing. As it may not always bepossible to acquire lease-aircraft due to fluctuations in theiravailability, the phase-in period should probably be based on theassumption that the requisite aircraft would have to be purchased. Ithas been noted that as much as two years may elapse between the date anairliner is ordered, and the date of its delivery.In order to operate these aircraft, the carriers will have to haveflight crews trained on the aircraft types. The greatest concernregarding crew training is with the pilots. They may require extensivetraining. As a guide, it should be noted that some airlines take pilotrecruits, with absolutely no flying experience or credentials, and putthem through an intensive training program from which they graduatewith first officer qualification. Such programs are of 18 to 24 monthsduration.’52 Using this as a reference figure, the Canadian carriers,with access to a pool of relatively experienced pilots, should be ableto fulfill all of their crew requirements in 18 months or less.Other considerations are relevant to the determination of the298duration of the phase-in period. The accessibility of airportfacilities should influence the time it takes carriers to establishthese new services. It was noted in Chapter III, Section 3.3, thataccess to gates and slot times is problematic at some United Statesairports. The acceptability of the phase-in agreement may have tohinge on provisions ensuring that such access is possible. (Thealliances between Canadian and U.S. carriers should help to minimisethe difficulties posed by capacity problems at U.S. airports.) Accessprovisions must be included in the new bilateral agreement to make thephased approach viable. The access provisions could be engineered tofit within the time frame dictated by the flight equipment acquisitionconsiderations.Finally, the disadvantages faced by non-allied, U.S. carriers mustbe factored into the timing of the introduction of the generally openborder. With the size advantages of these carriers (eg. Delta,Northwest), provision should be made to allow the Canadian carriers tohave the first opportunity at entering new markets. With two yearsbeing anticipated for aircraft acquisition considerations, this timeframe should be maintained as the exclusive period for the Canadiancarriers. Therefore, at the end of two years, the non-allied U.S.carriers should be permitted free entry into transborder markets.The ultimate duration of the phase-in period should be based on the299time frame necessary to acquire aircraft by purchase. As such, itshould be based on the maximum of the two years that can be necessaryfor aircraft acquisition, and include some time for the inception ofmarketing activities, and the establishment of market presence. Givena one year period for the establishment of market presence, a threeyear period should suffice to accommodate all of these requirements.Therefore, we advocate a three year period for the phase-in, afterwhich time a general open border environment would be instituted.6.2.5 Structural Implications of the Phased-In, Open BorderRegimeThe phased introduction being proposed above should affect thestructure of transborder airline markets in two primary ways. First,it should result in the creation of direct services between currentlyindirectly-linked city-pairs. Instead of having to travel bycircuitous routings, transborder travellers in many Canada-UnitedStates airline markets would be able to enjoy direct services. Second,it will result in an increase in the number of potential, if notactual, competitors in existing transborder airline markets. Theseeffects are qualitatively identical to the plain open border option.The differential will be with regard to the quantitative level ofcompetition.300The limitation of entry during the phase-in period will establish, asan upper bound, the number of Canadian carriers that are able to offertransborder services. As a crude estimate of this number, the averagenumber of carriers per route in Canada could suffice, at approximately1.7. Alternatively, it could be assumed that both Canadian majorcarriers would contest most transborder markets, bringing the averagenumber of competitors in new city-pairs to 2.0. Finally, theavailability of new opportunities, restricted to Canadian carriers,could encourage airlines such as Air Transat or Canada 3000 to opt toprovide scheduled services in selected transborder markets. This mightbe particularly appealing for those markets where the traffic base isCanadian vacationers. This scenario would bring the average number ofpotential competitors to greater than 20.Once the phase-in period has expired, U.S. firms would be able towidely contest transborder markets. As noted in the discussion of thestructural implications of extending the specified rights regime, thecurrent average number of competitors in U.S. markets is 2.8, It couldbe expected then that the post-phase-in number of competitors would bein the vicinity of this number for most transborder routes.The foregoing comments on market structure are founded on system-wideaverages. It must be recognised that the most attractive new301transborder markets will be those for which there is a good deal oforigin-destination traffic (for example, Ottawa-Washington DC), orwhere the service connects a large Canadian centre to an important U.S.hub airport (for example, Dallas-Vancouver). For city-pairs in thelatter category, it should be expected that, even during the phase-inperiod, the number of competitors would be limited by expectations ofthe establishment of services by the hubbing airline. This should beexpected to keep the number of competitors to the vicinity of Expected Conduct With a Phased-In, Open Border RegimeDuring the phase-in period, it should be anticipated that pricingbehaviour will reflect the type of ultimate market structure expected.For those markets wherein multiple firms are anticipated toparticipate, both Canadian carriers can be expected to introduceservices, and competitive prices should prevail. During the phase-in,these fares should be similar to those for intra-Canada services,reflecting the restricted number of potential competitors available tothese markets. At the conclusion of the phase-in, fares can beexpected to decrease to reflect the increase in the number ofpotential/actual competitors in the marketplace. They should be loweron average than both intra-Canada and existing transborder fares,reflecting the higher level of competition in these markets.302For those markets that are anticipated to ultimately become dominatedby a U.S. carrier hubbing at the U.S. end-point airport, or by analliance of carriers whose U.S. partner is the hub operator, fares canbe expected to be relatively high during the phase-in period. With theonset of open competition, fare premia should be retained, as has beenthe case for U.S. domestic services involving dominated hubs. The farepremia can be expected to be applied only to travel for which the U.S.hub airport is an origin or destination. For connecting services,average fares can be expected to be somewhat lower, reflecting inter-route competition. This has been the observed pricing behaviour inU.S. domestic markets involving flights through hubs.In terms of non-price conduct, it can be anticipated that during thephase-in period, alterations to airline networks will be the mostprevalent activity. Many new cross-border city-pairs will beconnected. The major Canadian carriers will introduce new serviceslinking Canada’s major cities to the hub airports of their U.S.alliance partners. They can be expected to establish new directservices between U.S. and Canadian centres having sufficient origindestination traffic, yet which do not enjoy direct air links under thecurrent regime. It is also possible that Canadian carriers couldprovide services between Canadian centres and the hub airports of nonallied, U.S. carriers, where direct such services are not currently303permitted. Over the short-term, this presents the possibility of creamskimming markets that ultimately would be expected to be dominated bynon-allied, U.S. carriers. The prospect of this type of behaviourshould militate against the unanimous acceptance by the United Statesindustry of the phase-in strategy.The expectation that the hub-dominating U.S. firms would inauguratecompetitive services at the conclusion of the phase-in period maydiscourage Canadian carriers from starting such services. At thattime, the hub-dominating firm would be able to commence its own, newtransborder services. The advantages of connecting the Canadian centrein question to the airline’s overall network should serve to yield thehubbing carrier eventual dominance in the relevant transborder market.In many cases, the density and scope economies associated with thehubbing operation may wrest enough transborder traffic away from theCanadian firm that it would have to exit from the transborder service.It should also be remembered that the introduction of any new serviceswill be costly for the Canadian carriers. The need to maximise the netpresent value of returns for all new routes could well, of its ownaccord, dissuade Canadian carriers from taking advantage of phase-inperiod entry barriers in this manner.6.2.7 Expected Performance Under a Phased-In, Open BorderRegime304The forms of performance benefits, in terms of increasedefficiencies, that are anticipated for a plain open border regime,should be realised through the phase-in strategy. Allocativeefficiency would be enhanced. The increase in the average number ofcompetitors in transborder markets should reduce average fares in thesemarkets; economic rents due to route monopoly power would be reduced.The value of travel time savings that would occur because of moredirect routings between transborder city-pairs would increase consumerwelfare. Travellers would waste less time in making flight connectionsthat are now required because of circuitous transborder routings.During the phase-in period, there may be some accumulation ofeconomic rents for Canadian carriers introducing new direct services.These rents would arise due to the absence of direct competition. Theinter-route competition should be characterised by having higheraverage fares due to longer flight distances. Direct services offeredby Canadian firms may be provided at lower average fares made possibleby direct, lesser distance routings. However, the determining factorwith regard to market performance in this regard will be the density oftraffic permitted by the market. Without sufficient densities, theCanadian-supplied, direct services may not be operationally feasible.Moreover, their appeal to consumers may be limited, as connectingservices offered by U.S. competitors may remain more attractive due to305reduced schedule delay time associated with greater flight frequenciesthrough hubbing.6.2.8 The Phased-In Open Border Regime and EquityOne of the major objectives that must be met by a new bilateralregime is the provision of equity of opportunity for carriers of bothCanada and the United States. The phased-in open border strategy hasbeen advocated largely because it addresses the imbalance ofopportunities that are posed by other forms of bilateralliberalisation. It would allow the Canadian carriers advance rights soas to be competitive with U.S. carriers having superior abilities toreap economies of density and scope. However, the pursuit of thephase-in approach would pose problems for non-allied, United Statescarriers.Those U.S. carriers allied with Canadian carriers would share in theshort term benefits captured by the Canadian firms. For non-allied,U.S. airlines, the phase-in period represents then a period ofcompetitive disadvantage versus their allied compatriots. In order toaddress this disparity, it may be desirable to allow non-allied, U.S.airlines to be able to contest transborder markets at a date prior tothat on which free entry is permitted.3066.2.9 Summary and ConclusionsThe phase-in strategy of introducing an open border regime has beenproposed as a means of capturing the efficiency gains associated withan open border strategy, while at the same time meeting the equityobjectives of a new bilateral air services agreement. The policy wouldallow Canadian carriers to freely access transborder markets in advanceof the same rights being applicable to U.S. carriers. After aspecified phase-in period, airlines of both countries would be free tointroduce transborder services.The duration of the phase-in period could be determined on severaldifferent bases. The most appropriate such basis would be the expectedtime frame over which Canadian carriers could realistically acquire thenecessary aircraft, flight crews, and airport facilities to conductsuch services. Some time should also be allotted for the developmentof market presence. We believe that a 3 year period would beappropriate.The phase-in approach is intended to provide an equitable means oftransborder service liberalisation. Recognising that it may conferbenefits to those U.S. carriers having alliances with Canadian carriers307at the expense of non-allied U.S. airlines, it may be necessary topermit these non-allied carriers general entry rights in advance ofallied U.S. airlines.7.1 Summary of Analyses of the Proposed Models7.1.1 The Specified Rights OptionThe specified rights option would extend the current bilateral formatto additional transborder routes. This is the most conservative optionof the three under consideration. Its strength is that it maintainsthe equity of the division of benefits between Canada and the UnitedStates. The drawback to the specified rights option is that it doesnothing to enhance the efficiency of the air transport systems of bothcountries, or of the services offered that connect Canadian and UScommunities. Adoption of this solution will prolong the premium pricesthat consumers currently endure for transborder travel.3087.1.2 The Open Border OptionThe open border option would allow any Canadian or U.S. air carrierto offer service on any cross-border city-pair. While theoreticallysuch a regime should provide equitable opportunity for airlines ofeither nationality to establish themselves in transborder markets,there are other factors that will undermine the bi-nationalcontestability of such markets.Canadian air carriers will find themselves at a competitivedisadvantage with the introduction of a plain, open border policybecause of the unfavourable tax regime in Canada which raises theircost per ASM relative to U.S. firms. There is the difficulty posed bythe control over U.S. airport capacity that is held by U.S. carriers.Finally, there is the issue of beyond the gateway traffic, and theeconomies of scope that are available to American carriers. Thesescope economies can be capitalised on by U.S. lines, through their hubnetworks. Canadian carriers will find themselves at a competitivedisadvantage as the result of only being able to provideorigin/destination traffic for transborder services. In contrast, U.S.firms will have effective access to Canada’s most important internalairline markets. This will be accomplished through mirror imageservices. Such services will be provided through the extension ofspoke routes to Canadian centers from existing U.S. hubs. U.S.309carriers should be able to capture a greater share of transbordertraffic, and furthermore a significant share of intra-Canada trafficunder an open border regime.Canadian consumers would benefit from the introduction of an openborder agreement. They would be able to fly to the United States lessexpensively, and would have more choices for intra-Canada travelbecause of the alternative of travelling through U.S. hubs. Americanconsumers would be largely unaffected by the open border regime.Canadian carriers would experience an erosion of their traffic base,without being able to necessarily benefit from the open border due toU.S. airline control over airport facilities. U.S. carriers can beexpected to benefit from the policy since it will allow them to capturea greater share of overall North American air traffic.7.1.3 CabotageAs with the open border option, cabotage presents the opportunity forenhanced consumer benefit in air transportation. Cabotage would permitcarriers of either country to operate on any domestic or transborderroute. The aim of the policy is to maximise the potential number ofcompetitors in all North American markets. Average real fares can beexpected to be reduced with the introduction of cabotage on many310transborder and intra-Canada routes, with fares in intra-United Statesservices being largely unaffected.Cabotage would have adverse consequences for the Canadian airlineindustry. Although the Canadian carriers can be expected to maintaintheir alliances with U.S. carriers, thus strengthening their long-termviability, they are likely to be forced out of many transborder andintra-Canada markets by non-aligned U.S. competition. This will havenegative effects on Canadians who are interested in participating inthe airline industry.7.1.4 Phased-In Liberalisation StrategiesIn order to capture the benefits of transborder liberalisation, thepossibility of phasing-in of such regimes was considered. Thedesirability of phasing-in a cabotage regime was quickly recognised tobe minimal; phasing-in of cabotage was therefore discarded as a policyoption. In contrast, the phasing-in of an open border regime should bean advantageous solution.The phased-in, open border would provide the efficiency gainsexpected for the general open border strategy. Unlike the generalstrategy, however, it would safeguard the well-being of the Canadian311carriers until they have a chance to establish themselves in newtransborder markets. To fully meet the equity requirements of a newpolicy, the phase-in approach would have to accommodate United Statesairlines that are non-allied with Canadian carriers. Allowing thesefirms unfettered access in advance of the remainder of the U.S.industry would probably be an acceptable equity provision.312Footnotes to Chapter Five1. Dresner and Tretheway, p.82. Dresner and Tretheway, p.8,93. Dresner and Tretheway, p.164. Dresner and Tretheway, p.165. Dresner and Tretheway, p.166. Dresner and Tretheway, p..14n7. Windle, p.44n8. Gillen et.aJ.[1985], p.979. See Tretheway [19891 for a complete discussion of the anti-competitive nature of frequent flier programs.10. G.A.O. [1990], p.2211. G.A.O. [1990], p.3112. G.A.O. [1990], pp.33-36 13. G.A.O. [1990], p.3714. G.A.O. [1990], p.3315. G.A.O. [1990], p.4316. G.A.O. [1990], p.4317. G.A.O. [1990], pp.59,6018. G.A.O. [1990], p.6219. G.A.O. [1990], p.6220. G.A.O. [1990], p.6321. G.A.O. [1990], p.6322. G.A.O. [1990], p.6423. G.A.O. [1990], p.6524. The non-linear nature of the reward structure made it advantageousfor an agent to book as many passengers with a single carrier aswas possible.25. G.A.O. [1990], p.6726. G.A.O. [1990], p.6727. Transport [1991c], p.3528. Oum [1990], pp.6,729. Transport [1991c], pp.38,11730. Statistics Canada, Catalogue 51-205 [1992], p.1331. Transport [1991c], p.5332. House of Commons, p.1033. Transport [1991c], p.5834. Transport [1991c], p.6435. Transport [1991c], p.6336. Transport [1991c], p.6337. Transport [1991c], p.6338. N.T.A. (II), p.18439. Directions, p.25240. Transport [1991c], p.9641. House of Commons, p.1442. Directions, p.74643. N.T.A. (I), p.124313Footnotes to Chapter Five (Cont.)44. Graham, Kaplan, and Sibley, p.12945. Transport [1991b], p.10446. Canton, p.7347. For example, the proximity of the respective carriers’ gates atthe airport of transfer will affect the time savings expected overmultiple carrier services.48. Transport [1991b1, p.7449. Gillen, Stanbury, and Tretheway, p.1950. Patterson and Tretheway, p.551. N.T.A. (I), p.12152. Transport [1991b], p.7353. This is the intent of Bill 926, House of Representatives, 1993.This proposal is commonly known as “The Clinger Bill,” after itsauthor, Representative William F. Clinger, Jr., of Pennsylvania.54. Hadrovic, p.20955. Transport [1991b], p.3356. Transport [1991b], p.3557. Hadrovic, p.21158. G.A.O. [1990], pp.66,6759. Transport [1991b1, p.3360. Transport [1991bj, p.3561. Borenstein, pp.357,35862. Onesnen and Tretheway, p.1863. Morrison and Winston [1987], p.6164. Oum, p.2465. Oum, p.2366. Oum, p.2367. Once again, this is due to the similarity of flight equipment, andterminal-related technologies worldwide.68. Windle, pp.36,3869. Transport [1991a], pp.73,7470. Directions, p.116171. Oum, pp.19,2272. Directions, pp.1157,1160,116173. Jordan, p.32274. Oum, Stanbury, and Tretheway, p.1375. Dresner and Tretheway, p.1676. None of the studies we reviewed reported a breakdown of CanadaUnited States transborder traffic by consumption group- - le. byutilisation of fare discounts.77. Oum, p.2378. Dresner and Tretheway, p.1679. Dresner and Tretheway, p.19314Footnotes to Chapter Five (Cont.)80. Dresner and Tretheway (1990) used a random selection ofinternational air routes in their analysis. The characteristicsof their data set may make their conclusions less applicable tothe Canada-United States situation. They used routes of 4,000 km(2,500 ml) or more: most Canada-US routes are shorter than this.Most international markets are such that there are few, if any,attractive alternative routings: most Canada-US markets can beserved by a variety of different routings, which may include theuse of alternative modes to cross the border, and flyingexclusively on domestic services in the United States.81. Borenstein, p.35782. Borenstein, p.35883. Dresner and Tretheway, p.1684. Adrandi [1989], p.22585. Graham, p.12986. Graham, p.13587. Morrison and Winston [19871, p.6188. Morrison and Winston [1987], p.6189. Directions, p.115190. Based on a random sample of the top 1000 markets in the UnitedStates in 1988. See Appendix I. for details.91. Windle, p.4592. Dresner and Windle, p.28293. Gillen [19901, p.2294. Transport [1991c], p.7695. Transport [1991c], p.7696. Transport [1991c], p.2897. House of Commons, p.1498. Transport [1991c], p.3199. Gillen, Oum, and Tretheway, pp.131, 133100. Gillen, Stanbury, and Tretheway, p.20101. Gillen, Stanbury, and Tretheway, p.20102. N.T.A. (I), p.77103. Gillen[1990j, p.23104. Wilbanks, pp.218,219105. Statistics Canada, Catalogue 51-205 [19921, p.13106. Korenic, pp.64,71,84,85107. Oum, pp.46,48,49108. Oum, pp.46,52,53109. Gillen et.aI. [1990], pp.15,16110. Oum, p.46n111. Morrison and Winston [1990], p.391112. Oum, p.45113. Oum, pp.52,53114. Oum, p.53315Footnotes to Chapter Five (Cont.) and Williams,G.A.O. [1990], p.69Morrison and WinstonWindle, p.45N.T.A. (I), p.77Graham, p.135Waterson, pp.18,19Waterson, p.18Waterson, p.25p.392p.61p.393p.137[1987], p.59[1990], p.392[1990], p.392pp.182,189[1987], p.59Morrison and Winston [1990],Morrison and Winston [1987],Morrison and Winston [1990],Graham, Kaplan, and Sibley,Borenstein [1989], p.362G.A.O. [1990] pp.31,43Morrison andMorrison andMorrison andBrander and Zhang [1990], p.569Waterson, p.23Brander and Zhang [1990], p.577,578N.T.A. (I), p.77Tretheway [1991], p.7Erickson, p.35Transport [1991c], p.83Gillen, Hansen, and Ramos, p.23Gillen, Hansen, and Ramos, p.23These calculations are based on transborder market share data fromThe Ministerial Task Force on International Air Policy Report,Volume III (1991), p.37, and on total market share data fromGillen, Hansen, and Ramos (1990), p.23. Case 2 is compared withGillen because in both approaches it was presumed that U.S.carriers would face identical costs to Canadian firms for intraCanada operations.Directions, pp.1660-1664Windle, pp.37,39Caves, pp.304,305Caves, p.310Gillen [1985], p.118Directions, p.736Erickson, p.36House of Commons, p.12Transport [1991cj, p.98Petrie, p.26316CHAPTER SIX.SUMMARY AND CONCLUSIONS1. The Need for a New Structure for Canada-United States Air TransportServicesThere is general recognition in Canada and the United States that thecurrent bilateral agreement regarding air transport services betweenthe two countries is inadequate. There are artificial limitations onthe way in which airlines of both countries may offer travel betweenthe countries to the consumer. Current transborder services feature alack of direct routings between important destinations, and some degreeof monopoly pricing. A more liberal regime, based on marketdiscipline, would result in an increase in consumer welfare.Canada and the United States have attempted to negotiate a newbilateral agreement. They have been unsuccessful in so-doing becauseof differences in each nation’s evaluation of the distribution of thecosts and benefits that would result from the adoption of the threegeneral solutions that have been proposed. While it has beenacknowledged that a more liberal structure for these services would bebeneficial to consumers in Canada and the United States alike, theconcerns of industry participants in both countries have complicated317the negotiations. The concerns of the industry are legitimate and mustbe satisfactorily addressed if a good solution is to be achieved.2. The Three Solutions Under Consideration2.1 The Specified Rights OptionThe first format for a new bilateral that is being considered is thatof an extension of the status quo. This can be termed the specifiedrights option. Under this plan, a new bilateral would be negotiatedthat would be similar in nature to the existing agreement, but whichwould specify new transborder routes that could be flown. Each countrywould then allocate these new routes to their carriers. Multipledesignation of carriers would be a possibility; this would serve toincrease the level of competition in transborder markets.The drawbacks to the specified rights option are that it does notallow the airlines to construct route networks that take advantage ofthe economies of scope that have been demonstrated to exist in airtransportation. It would also not address the disparity of benefitsthat exists between US and Canadian carriers. Lack of improvements inefficiency should relegate this strategy to being that of one of lastresort.3182.3 The Cabotage OptionAnother general form that has been suggested for a new bilateralregime is that of cabotage. A cabotage agreement would allow carriersfrom either country to provide both transborder flights, and domesticservices in the other country. Proponents argue that cabotage wouldincrease the level of competition in all North American markets,thereby benefitting Canadian and US consumers. The airlines would beable to truly optimise their route layouts, and would be able to takeadvantage of currently under-utilised airport facilities in theprocess.Going a against the cabotage option are the same arguments lodgedagainst the open border policy. The cabotage regime also has specialproblems all its own. There is the concern in both Canada and theUnited States over the precedent that the granting of cabotage rights,one to the other, would establish for both countries’ relationshipswith third parties. The United States is particularly concerned thatthey could be forced to grant similar privileges to third countrieswhere US interests would stand little to gain in the way of reciprocalrights. There have been legal opinions expressed, however, that Canadaand the United States would be bound only to permit other parties toengage in negotiations toward creation of a cabotage agreement.319The greatest problem regarding cabotage is that the equity objectivesof a new agreement would probably not be achieved by its adoption. Thedifficulty is that the superior network-related economies of scope anddensity associated with the operations of U.S. carriers could beexternalized to transborder and Canadian domestic markets to thedecided detriment of Canadian carriers.Concerns over extension of cabotage rights to third parties coupledwith Canadian anxieties over the survival of a Canadian airlineindustry make cabotage an unacceptable option.2.3 The Open Border OptionCanada and the United States could institute an open border regimefor airline services. Under this policy, the carriers of eithercountry could offer services on any cross-border city-pair that theywished. The aim of the open border solution is to enhance competitionin transborder markets. With a more competitive structure, thereshould be a rationalisation of routes, and an increase in efficiencythat will lead to lower prices for consumers.The open border strategy has been criticised as posing a threat tothe continued viability of the Canadian carriers. Critics claim that320these airlines cannot compete on such terms with their US counterpartsbecause of the advantages the latter have with respect to costs,geography, and demographics.2.4 Recommended Policy: The Phased-In, Open Border RegimeWhile the foregoing options meet either the efficiency or the equityobjective severally, the phased-in, open border option wouldaccommodate both policy objectives jointly. The efficiency gainsexpected from the open border strategy would be realised. The phased-in nature of the regime would allow the Canadian carriers advanceability to establish market presence in new transborder operations.This should serve to level the terms of competition between them andtheir U.S. counterparts. Provisions would be made to address thedisparity between Canadian-allied, U.S. airlines, who would participatein the first-mover advantages of the Canadian carriers, and non-allied,U.S. carriers.As part of the agreement, it would be necessary to ensure thatCanadian carriers have access to significant U.S. airports. Accessprovisions regarding gates and slots will be vital prerequisites to thesuccessful introduction of new services by Canadian firms. Theseconditions must be attached to the agreement.321The duration of the phase-in period should be three years. This timeframe should allow the Canadian carriers to assemble the necessaryflight and ground equipment, and personnel to be able to conduct suchservices.3. ConclusionsThis study has explored the desirability of adopting the alternativeschemes for a new Canada-United States air services bilateralagreement. It has analysed the three general alternative strategiesusing the structure-conduct-performance paradigm as the method ofanalysis. On the basis of two important performance criteria,efficiency and equity, we have concluded that a modified version of theopen border strategy should be pursued in the new regime.Evaluation of the Structure-Conduct-Performance Paradigm as Applied toAirline MarketsThe use of the structure-conduct-performance model in this analysiswas invaluable. The model allows for hypothetical situations to beevaluated from a well-founded set of assumptions. The assumptions usedin this analysis were that airline markets of similar structure would322produce the same conduct and therefore similar performance. Inparticular, it was assumed that the existing bilateral agreement hasproduced some degree of economic rents for carriers. Proposedderegulation of the transborder sector should increase decrease theconcentration of airline markets. This should give rise to performanceresults similar to those realised in domestic markets in both Canadaand the United States after those markets were deregulated.The model also suggests that entry barriers may be significantdeterminant of market performance. In application to airlineeconomics, it has been shown that, where a carrier dominates a marketbecause of control over access to airport facilities and or computerreservations systems, market contestability may be blockaded. Thus,the plain institution of deregulation may not provide a path to theenhancement of consumer welfare.The application of the structure-conduct-performance method ofanalysis to airline markets demonstrates that the model is bestutilised in its more complex form: that causality should be expected toflow in both directions. In particular, the conduct of air carriershas been shown to have had a definite impact on the structure ofairline markets. This has been most evident where dominant-carriercontrol over operations at a hub airport has deterred entry bycompetitors that, ceteris paribus, should be competitive in such323markets.The foregoing observation illustrates a potential short-coming of thestructure-conduct-performance paradigm: it may be susceptible toproblems regarding the identifiability of causality. If causalitycannot be identified, the applicability of the paradigm as a means offorecasting market performance would be undermined.Another potential problem arises with regard to the scope of thestructure that should be considered. If too broad a number ofstructural factors is included in the model, the applicability of theparadigm becomes questionable; there should be a manageable number ofexplanatory factors to make the model useful. If too great a number ofelements of structure are included, the likelihood of observingpatterns of causality can be expected to be reduced. The net result isthat the qualitative and quantitative variety of situations changes thedeterministic, structure-performance relationship to one of novelty.This is the essence of the case study approach, which is an alternativemethod of analysis.Policy RecommendationThis study has considered the prominent, available alternative324strategies for a new Canada-United States air transport agreement. Theassessment of these alternative has shown that the desirable option isthe phased-in, open border approach. It entails giving Canadiancarriers immediate access to any transborder route they desire tooperate. After this period, which we estimated should be three years,all transborder city-pairs could be freely entered by airlines ofeither country.This study has concluded that adoption of this policy would benefitCanadian and American consumers alike. These benefits would be theresult of several factors:1. Efficiency would be enhanced in that there would be traveltime savings due to more direct routings for transbordertravel;2. Allocative efficiency would be improved as there would begreater potential, if not actual, competition on transborderair routes. This can be expected to lead to reductions inaverage fare levels;3. Productive efficiency would be promoted as market disciplinewould extend into the transborder sector;3254. Equity objectives would be realised as the policy addressesthe disparity that exists between Canadian and U.S. airlineswith respect to their differential abilities to capitalise oneconomies of scope and density.Further research would be valuable. What is needed is a rigouroustreatment of the solution to the matter of the length of theappropriate periods for the phase-in of the regime. Also, the point intime at which the non-allied, U.S. airlines should be given completefreedom to enter transborder markets should be so determined.326BIBLIOGRAPHYAdrandi, Bahram; Chow, Garland; and Gritta, Richard D. (1989). “TheMarket Share, Cost and Profit Relationship in the AirlineIndustry,” Journal of the Transportation Research Forum, VolumeXXIX, Number 2, 1989, pp.223-227.Adrangi, Bahram; Gritta, Richard D.; and Chow, Garland (1986).“Deregulation and Concentration in Air Transportation,” Journal ofthe Transportation Research Forum, Volume XXVII, Number 1, 1986,pp.6-il.Balfour, John (1991). “Factortame: The Beginning of the End forNationalism in Air Transport?”, Air Law, Volume 16, Number 6,1991, pp.251-266.Bonsor, Norman C. (1984). Transportation Economics: Theory andCanadian Policy. Toronto: Butterworths, 1984.Borenstein, Severin (1989). “Hubs and High Fares: Dominanace andMarket Power in the US Airline Industry.” Rand Journal ofEconomics, Volume 20, Number 3, pp.344-365.Brander, James A. (1992) Government Policy Toward Business, SecondEdition, Vancouver: Butterworths, 1992.Brander, James A., and Zhang, Anming. (1993) “Dynamic OligopolyBehaviour in the Airline Industry.” International Journal ofIndustrial Organization, Volume 11, 1993, pp.407-433.Brander, James A., and Zhang, Anming. (1990) “Market Conduct in theAirline Industry: An Empirical Investigation.” Rand Journal ofEconomics, Volume 21, Number 4, 1990, pp.567-583.Canada. House of Commons Special Committee on Canada - United StatesAir Transport Services (1991). Report of the Special Committee onCanada-United States Air Transport Services. Ottawa: Queen’sPrinter, 1991.Canada. National Transportation Act Review Coninission (1993).Competition in Transportation: Policy and Legislation inReview, Volumes I and II. Ottawa: Minister of Supply and ServicesCanada, 1993.327Canada. Royal Commission on National Passenger Transportation.Directions: the final report of the Royal Commission on NationalPassenger Transportation. Ottawa: Minister of Supply and ServicesCanada, 1992.Canada. Statistics Canada (1992). Air Passenger Origin andDestination, Catalogues 51-002, 51-206, 1992. Ottawa: StatisticsCanada, 1992.Canada. Transport Canada. (1994). Aeronautical InformationPublication (A.I.P.). Ottawa: Supply and Services Canada, 1994.Canada. Transport Canada (1991a). “Report of the Ministerial TaskForce on International Air Policy, Volume I.” Ottawa: TransportCanada, 1991.Canada. Transport Canada (1991b). “Report of the Ministerial TaskForce on International Air Policy, Volume II.” Ottawa: TransportCanada, 1991.Canada. Transport Canada (1991c). “Report of the Ministerial TaskForce on International Air Policy, Volume III.” Ottawa: TransportCanada, 1991.Canada. Transport Canada (1981). “Economic Regulation and Competitionin the Domestic Air Carrier Industry,” Discussion Paper TP 2743E,1981, pp.9-28.Canton, Dennis W., Landes, William M., and Posner, Richard A. (1980).“Benefits and Costs of Airline Mergers: A Case Study.” BellJournal of Economics, Volume 11, Number 1. pp.65-83.Caves, Douglas W.; Christensen, Laurits R.; Tretheway, Michael W.; andWindle, Robert J. (1987). “An Assessment of the EfficiencyEffects of U.S. Airline Deregulation via an InternationalComparison.” In Public Regulation: New Perspectives onInstitutions and Policies. Ed. E.E. Bailey. Cambridge, Ma.: TheM.I.T. Press, 1987.Chesen, Joseph R. Canadian - American American Air ServiceNegotiations: Ending the Gridlock. Ed. Charles D. La Fond: J.R.Chesen Associates. Washington: 1989.328Close, George. (1990). “External Relations in the Air TransportSector: Air Transport Policy or the Common Commercial Policy?”Common Market Law Review, Volume 27, 1990, pp.107-127.Corbeil, Jean. “Turbulence in the Air.” Wings, June/July 1993, p.70.Coyle, J.J. (1965). “Cost-of-service Pricing in Transportation.”Quarterly Review of Economics and Business, Volume 5, Number 1,Spring, pp.63-74.Crans, Berend J.H. (1992). “EC Aviation Scene.” Air and Space Law,Volume XVII, Number 4/5, 1992, pp.217-223.Dempsey, Paul Stephen. (1987), “Turbulence in the ‘Open Skies’: TheDeregulation of International Air Transport.” Transportation LawJournal, Volume XV, Number 2, pp.305-388.Diebold, William, Jr., ed. Bilateralism, Multilateralism. and Canadain U.S. Trade Policy. Cambridge, Mass.: Ballinger, 1988.Diederiks-Verschoor, I.H.Ph., Dr., (1983). An Introduction to Air Law.Boston: Kiuwer Law and Taxation Publishers, 1983.Dierikx, Marc L.J., Dr. (1991). “Bermuda Bias: Substantional Ownershipand Effective Control 45 Years On,” Air Law, Volume 16, Number 3,1991, pp.118-124.Dresner, Martin, and Tretheway, Michael W., (1990), “Modelling andTesting the Effect of Market Structure on Price: The Case ofInternational Air Transport,” Journal of Transport Economics andPolicy, Volume 26, 1992, pp.171-184.Dresner, Martin, and Windle, Robert J.., (1992), “The Liberalization ofU.S. International Air Policy: Impact on U.S. Markets andCarriers,” Journal of the Transportation Research Forum, Volume32, Number 2, 1992.Ebke, Werner, and Wenglorz, Georg W., (1991), “Liberalizing ScheduledAir Transport Within the European Community: From the First Phaseto the Second and Beyond,” Transportation Law Journal, Volume 19,1991, pp.417-452.Erickson, Rick. (1993). “Canada’s Air Carrier Industry 1993: TheFuture Beckons.” Wings, November, 1993, pp.34-37.Feldman, Joan M. (1989). “The global alliance game: More style thansubstance?” Air Transport World, November, 1989, pp.24-31.329Feldman, Joan M. (1991). “Scrambling Well,” Air Transport World,July, 1991, pp.57-60.Gaibraith, John Kenneth (1973). Economics and the Public Purpose. NewYork: Houghton Mifflin, 1973.Gidwitz, Betsy (1980). The Politics of International Air Transport.Toronto: Lexington, 1980.Gillen, David W.; Hansen, Mark; and Ramos, Robson (1990). “Free Tradein Airline Services: Assessing the Proposals to Liberalize theCanada-U.S. Air Transport Bilateral.” Berkeley: University ofCalifornia Department of Economics, 1990.Gillen, David W.; Oum, Tae Hoon; and Tretheway, Michael W. (1985).Airline Cost and Performance: Implications for Public andIndustry Policies. Vancouver: University of British Columbia,Centre for Transportation Studies, 1985.Gillen, David W.; Stanbury, W.T.; and Tretheway, Michael W. (1988).“Duopoly in Canada’s Airline Industry: Consequences and PolicyIssues,” Canadian Public Policy, Volume XIV:1, pp. 15-31, 1988.Graham, David R.; Kaplan, Daniel P.; and Sibley, David S. (1983),“Efficiency and Competition in the Airline Industry,” Bell Journalof Economics, Spring 1983, pp.118-138.Haanappel, Peter P.C. (1991). “Recent Regulatory Devleopments inEurope,” Annals of Air and Space Law,” Volume 16, 1991, pp.107-121.Haanappel, Peter P.C. (1992a). “Europe 1992 and Airline(De)regulation,” Annals of Air and Space Law,” Volume 17,Number 1, 1992, pp.271-280.Haanappel, Peter P.C. (1992b). “Recent European Air TransportDevelopments,” Annals of Air and Space Law,” Volume 17, Number 2,1992, pp.217-235.Hadrovic, Carolyn, (1990), “Airline Globalization: A CanadianPerspective,” Transportation Law Journal, Volume 19, 1990, pp.193-217.Huston, John H. and Butler, Richard V. (1988), “The Effects of FortressHubs on Airline Fares and Service: The Early Returns.” Logisticsand Transportation Review, Volume 24, Number 3, 1988, pp. 203-215.330Jordan, William A. (1991). ‘U.S. Airline Performance From 1978 to1988: The Bridge From Regulation to Deregulation.” Journal of theTransportation Research Forum, Volume XXXI, Number 2, 1991,pp.317-325.Kasper, Daniel M. Deregulation and Globalization: LiberalizingInternational Trade in Air Services. American EnterpriseInstitute/Ballinger Publications. Cambridge, Ma.: 1988.Korenic, John A. (1991). “Canada-US Air Traffic: Analysis of thePotential Impact of a Liberalized Air Services BilateralAgreement.” Vancouver: UBC Centre for Transportation Studies,1991.Kraft, Dennis J.H.; Oum, Tae H.; and Tretheway, Michael W. (1986),“Airline Seat Management,” The Logistics and TransportationReview, Volume 22, Number 2, June 1986, pp.115-130.McShan, Scott and Windle, Robert. (1989), “The Implications of Hub-and--Spoke Routing for Airine Costs and Competitiveness.” ]jLogistics and Transportation Review, Volume 25, Number 3, pp.209-230, September 1989.Morrison, Steven, and Winston, Clifford (1987). “EmpiricalImplications and Tests of the Contestability Hypothesis,” Journalof Law and Economics, Volume XXX, April, 1987, pp.53-66.Morrison, Steven, and Winston, Clifford (1986). The Economic Effectsof Airline Deregulation. Washington: Brookings Institution, 1986.Needham, Douglas (1978). The Economics of Industrial Structure,Conduct, and Performance. New York: St. Martin’s Press, 1978.O’Connor, William E. (1985). An Introduction to Airline Economics,Third Edition. New York: Praeger, 1985.Oum, Tae H. (1990). “North America Traffic Pattern,” Vancouver:University of British Columbia, 1990.Oum, Tae H.; Stanbury, W.T.; and Tretheway, Michael W. (1991).“Airline Deregulation in Canada and Its Economic Effects,”Transportation Journal, Volume 30, Number 4, 1991, pp.4-22.Oum, Tae H., and Tretheway, Michael W. (1989). “Airline Hub and SpokeSystems,” Journal of the Transportation Research Forum, VolumeXXX(2), 1989, pp.380-393.331Oum, Tae H., and Tretheway, Michael W. (1984). “Reforming CanadianAirline Regulation,” The Logistics and Transportation Review,Volume 20, Number 3, 1984, pp.261-284.Patterson, Brett, and Tretheway, Michael W. (1992). “Globalization ofthe Airline Industry: Solutions to the Holding Company Problem.”University of British Columbia, 1992.Petrie, Fred, (1993), “The Pilot Shortage,” Wings, November, 1993,p.26.Petsikas, George, (1992), “‘Open Skies’- North America,” Annals of Airand Space Law, Volume 17, Number 1, 1992, pp.281-291.Petsikas, George, (1991), “A Canadian Perspective on ‘Open Skies’Between Canada and the United States: Unprecedented Opportunitiesor the End of a Viable Canadian Airline Industry?,” Annals of Airand Space Law, Volume 16, 1991, pp.161-176.Platt, Jeffrey R. (1992). “The Creation of a Community Cabotage Areain the European Community and its Implications for the U.S.Bilateral Aviation System.” Air and Space Law, Volume VXII,Number 4/5, 1992, pp.183-198.Reid, Gavin C. (1987). Theories of Industrial Organization. Oxford:Basil Blackwell Ltd., 1987.Sassone, Peter G., and Schaffer, William A. (1978). Cost-BenefitAnalysis: A Handbook. Toronto: Academic Press, Inc., 1978.Scherer, F.M. (1980). Industrial Market Structure and EconomicPerformance, 2nd ed. Chicago: Rand McNally, 1980.Schmid, Ronald. (1992). “Air Transport Within the European SingleMarket: How Will It Look After 1992?” Air and Space Law, VolumeXVII, Number 4/5, 1992, pp.199-205.Taneja, Nawal K. (1989). Introduction to Civil Aviation. Toronto:Lexington, 1989.Thurow, Lester C. (1992). Head to Head: The Coming Economic BattleAmong Japan Europe. and America. New York: Warner, 1992.Tirole, Jean. (1988). The Theory of Industrial Organization.Cambridge, Mass.: The M.I.T. Press, 1988.332Tretheway, Michael W. (1991a). “The Characteristics of Modern Post-Deregulation Air Transport.” Vancouver: University of BritishColumbia, 1991.Tretheway, Michael W. (1991b). “European Air Transport in the 1990s:Deregulating the Internal Market and Changing Relationships withthe Rest of the World.” Vancouver: University of British Columbia,1991.Tretheway, Michael W. (1989). “Frequent Flier Programs: MarketingBonanza or Anti-Competitive Tool?” Journal of the TransportationResearch Forum, Volume XXX (I), pp.195-201.United States. Department of Transportation. (1990). Secretary’s TaskForce Report on Competition in the U.S. Domestic Airline Industry,Volume II. Washington: Department of Transportation, 1990.United States. General Accounting Office, (1990). Airline Competition:Industry and Marketing Practices Limit Market Entry. Washington:General Accounting Office, 1990.United States. General Accounting Office, (1989a). AirlineCompetition: DOT’s Implementation of Airline Regulatory Authority.Washington: General Accounting Office, 1989.United States. General Accounting Office, (1989b). “Barriers toCompetition in the Airline Industry: Statement of Kenneth M. Mead,Director, Transportation Issues, Resources, Community, andEconomic Development Division.” Washington: General AccountingOffice, 1989.United States. House of Representatives. Committee on Public Works andTransportation. Hearings Before the Subcommittee on Aviation ofthe Committee on Public Works and Transportation. February 5 and6., 1991. Washington: U.S. Government Printing Office, 1991.United States. United States Code Service (USCS). USCS, Volume 10.Vincent, Daniel, and Stasinopoulos, Dinos, (1989). “The AviationPolicy of the European Community,” Journal of Transport Economicsand Policy, Volume 24, Number 1, 1990, pp.95-100.Waterson, Michael (1984). Economic Theory of the Industry. New York:Cambridge University Press, 1984.Weber, Ludwig, (1992), “EEC Air Transport Liberalization and theChicago Convention,” Annals of Air and Space Law, Volume 17,Number 1, 1992, pp.245-262.333Weisman, [than. Trade in Services and Imperfect Competition:ApDlication to International Aviation. Kiuwer Academic Press.Boston: 1990.Wilbanks, Thomas J. (1980). Location and Well-Being: An Introductionto Economic Geography. New York: Harper & Row, 1980.Windle, Robert J. (1991). “The World’s Airlines: A Cost andProductivity Comparison,” Journal of Transport Economics andPolicy, Volume 25, Number 1, 1991, pp. 31-50.334APPENDIX I.“Calculation of Average Number of Competitors Per Route inUnited States Airline Markets, 1988”PurposeThe purpose of these calculations is to determine the average number ofairlines operating per route in United States domestic airline markets.DataThe data used in this calcluation were obtained from the United StatesDepartment of Transportation, “Secretary’s Task Force Report onCompetition in the U.S. Domestic Airline Industry, Volume II,” Appendix“S.” To be included in the “Report’s” data set, an airline had to havehad at least a ten per cent market share. The data were for the top1000 markets (in terms of traffic volumes) for 1988.MethodA random sample of 30 markets was drawn from the available data. Itshould be noted that, because of the influence of density economies, wemay expect that there would be a greater average number of competitorsin larger airline markets. Therefore, the methodology employed mayupwardly bias the calculated average for all United States markets.ResultsHistogram of Average Competitors per Route N = 30Midpoint Count1 2 **2 103 10 ********4 8N MEAN STDEV SE MEAN 95.0 PERCENT C.I.AverageCompetitorsPer Route 30 2.800 0.925 0.169 (2.455, 3.145)335


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items