We regroup two presentations dealing with the possible merger of Air and Canadian International Ltd. The first was the kick-off presentation at a student lunch-debate at the University of on November 8, 1999, and the second was the basis of the two-hour testimony given in front of the Permanent Transport Committee of the House of Commons of Canada on September 28, 1999. As the Annex is part of the testimony, it is included here despite the fact that it appeared as “Airport Subsidies and Congestion in North America : the Need for Accounts and a Regulator within Virtual World Trade Organisation Rules”, Journal of Air Transport Management, 7, 35-41, 2001.

Should the Two Largest Canadian Companies be Merged?

by

Marc Gaudry

• Bureau d’économie théorique et appliquée (BETA) Université Louis Pasteur, Strasbourg et • Département des sciences économiques Université de Montréal, Montréal [email protected]

Agora Jules DupuitPublication AJD-60 Centre de recherche sur les transportsPublication CRT-99-54

December 1999, September 2002 RÉSUMÉ

Nous regroupons deux textes traitant de la fusion possible d’ et des Lignes aériennes Canadien International Ltée: le premier donnait le coup d’envoi au débat-midi étudiant du 8 novembre 1999 à l’Université de Montréal et le second a été présenté au Comité permanent des transports de la Chambre des communes du Canada le 21 octobre 1999 dans le cadre d’un témoignage de deux heures.

Mots-clés: Air Canada, Lignes Aériennes Canadien International Ltée, Onex Corporation, Chambre des Communes, Université de Montréal.

ABSTRACT

We regroup two related texts pertaining to the potential merger of Air Canada with Canadian International Airlines Ltd. The first one was the kick-off for a student lunch-debate at the University of Montreal on November 8, 1999, and the second one was deposited with the Permanent Transport Committee of the House of Commons of Canada on October 21, 1999, within the context of a two-hour long interrogation.

Keywords : Air Canada, International Ltd, Onex Corporation, House of Commons, University of Montreal.

SOMMAIRE Résumé...... ….. ii Abstract...... …... ii Four questions about the two airline issue…………………………………………………………… 1 Why is the idea of an AC-CAIL merger a public policy issue of national interest?...... 1 What is meant by a middle-size player? ……………………………………...... 1 Are there other economic issues of general interest? ……...... 2 What process leads to a successful result?……...... 2 On the Onex proposal……...... 4 Context……...... 4 Key elements of the air industry environment...... 4 A handle to influence outcomes...... 6 Framework elements...... ……...... 8 Other issues……...... 9 References...... 10 Annexe 1. Subventions aéroportuaires et congestion aérienne en Amérique du Nord : plaidoyer pour un régulateur et pour l’établissemement d’une comptabilité collective...... 11 About the author...... ………………… 23

ii Four Questions about the Two-Airline Issue1 by Marc Gaudry

We need a framework to evaluate from a general interest perspective the potential net gains arising from the various proposals to integrate Air Canada (AC) and Canadian Airlines (CA), or from the current and forthcoming variants of these or new proposals. We focus on net gains to Canada, irrespective of the location of underlying gains or losses. Players and proposals are mentioned in stylised fashion, without due precision or formality.

1. Why is the idea of an AC-CA merger a public policy issue of national interest?

We may still have the opportunity to foster the establishment of a healthy middle-size carrier, with its head-office in Canada and chances of surviving and growing within international alliances. This opportunity arises from the fact that these two airlines currently hold almost all rights to serve international routes within bilateral air treaties signed by Canada. Without the “ national designation ” and these treaties, we risk being overtaken by our geography and becoming the mere spokes on American hubs that we might well have been had Canada been sold to the U.S., as Louisiana and were in their time.

But the value of these rights is threatened by the increasing liberalisation and “ open skies ” policies pursued principally by the . These policies increase the relative attractiveness of itineraries passing through American hubs on the way from Canada to other continents and could well lead to a collapse of the entire structure of restrictive bilateral treaties generating the profitable franchises. In addition, even foreign ownership rules imposed on airlines might be profoundly affected by a parallel movement, recently initiated by the American Transportation Research Board, to scrap the 25% limit generally applied.

2. What is meant by a middle-size player?

As Professor Oum of the University of pointed out in 1992-1993, airline head office functions are foot-loose and can be gained and lost easily; moreover, in this still relevant view, a major head office would have its best chances of survival if a Canadian player of reasonable size could be formed and became an anchor-airline within the growing world of notoriously unstable international air alliances.

Although AC does not have its close Continental Airlines partner anymore, a healthy and growing anchor based on truly merged firms is still worth a try: note how Canadian National Railways (CN) moved from the status of maintained consumer of public subsidies, responsible for perhaps 11-15 % of the national debt between its inception and its privatisation, to that of contributor to the public purse and healthy Class I railway competitor. The form of the CN network is rapidly being transformed from a market fringe line-shape to a market core T-shape reaching ever further through alliances across all of the North American continent. CN has become in 3 years an anchor-railway for Canada. In a few years, it might have been too late to start.

1 Kick-off notes prepared for the November 8, 1999, students’ lunch-debate (“ Débat-midi ”) at the University of Montreal. Fuller notes are found in “ On the Onex Proposal ” , a summary of comments made to the House of Commons ad hoc and Permanent committees on September 28 and October 21, respectively. This summary is available from the author ([email protected]) in both French or English versions. The author is professor of Economics, co-founder and occasional director of the multi-university Centre for Research on Transportation (CRT) in Montreal and a former commissioner of the Royal Commission on National Passenger Transportation (1989-1992). 1

3. Are there other economic issues of a general interest?

First, there is the subsidy issue. The Royal Commission on National Passenger Transportation recommended in 1992 against subsidies to mature networks. But in our problem, AC officials are reported to have said a number of times that, in their offer to maintain two firms under separate brand names, they might simply pull the plug under CA after paying shareholders their agreed amount, should the creditors of that airline not go along with the “ restructuring ” of the value of their claims, betting that the Canadian Government would step in with a subsidy to avoid the bankruptcy. If correct, this information shows that the expectation of a subsidy can still be part of a take-over bid.

Second, there is the economic efficiency argument. Clearly, only a true merger will yield both the revenue and the cost savings necessary to maximise combined gains. Also, all other formulas restrict the ability of the merged firm to create new and better services with less resources, to the benefit of Canada. One should not underestimate the ability of a larger firm to create new “ goods ” : all feasible paths and services among all origins and destinations are termed the “ scope ” of the firm and generate synergies (economies of scope) not available to firms serving few points. The Onex proposal obtains efficiencies of all kinds without the threat of subsidies.

But cost synergies can be negative. How could AC create a low-fare airline with high-cost CA employees in Hamilton—a place where this idea has been tried without success? Of course, this type of question has been asked and answered before—in the case of short rail lines spun off the trunk railways. Labour rule flexibility has been central to the success of the short line railways, of which there are now tens in Canada. How would the “ short line ” airline achieve this result ? The experience of Westjet in suggests ways this could be done, but these ways did hurt CA yields.

Finally, the credibility of the AC low-price airline proposal is also as dubious in other ways: if one is worried about competition, how could one expect that the Competition Bureau would accept it, given that it might well require some divestiture of some feeder airlines implicated in a real merger touching over 80 % of flights? It is obvious that this attempt to occupy the “ low end ” of the market could have the effect of keeping at bay firms like Westjet, that has already registered the name Eastjet for such purposes, to say nothing and other airlines.

One might then expect the Bureau of Competition to examine the merger and suggest that, even in the absence of subsidies from the main to the low-cost airline, competition would not be well served by such a move. It is true that competition can be dosed almost at will by the regulators, for instance by allowing within-Canada service by foreign carriers, perhaps even without foreign reciprocation of this “ cabotage ” allowance, or by requiring divestiture. But it is not true that competition would be well served by this cabotin proposal. No wonder many think that the only clear part of the proposal is the buyback by AC of 35 % of its shares in order to thwart the Onex bid!

4. What process leads to a successful result?

Basically, in view of the relative urgency of the matter, we need a clear and fast process because our window of opportunity might fade away.

Firstly, Canadian control has to be clearly defined in law. For instance, it is obviously incompatible with any veto power held by foreign stakeholders. Second, the maximum share of voting capital

2 (currently 10 %) allowed any shareholder cannot apply only to one company (AC) but not to the other (CA), as is the case today.

More importantly, full disclosure of long-term alliance conditions with foreign airlines should be required because, independently from any issues of responsibility to shareholders, such agreements may involve backroom deals dependent on government-owned international route permits: e.g. an unfavourable route profit sharing agreement with a foreign airline traded against fewer voting shares than warranted by the foreigner’s investment, or against some number of shares at an apparently excessive price.

Secondly, due process has to be fast process in the case of large firms whose bankruptcy would disrupt the economy and whose merger may (with conditions) be in the public interest. It is therefore understandable that some work is required on the part of government to define a fast track, or critical path, through the regulatory triangle defined by the Transportation Act, the National Transportation Agency and the Bureau of Competition. This important homework is of some urgency before the window of opportunity evaporates and our geography finally catches up with us.

3 On the Onex Proposal by Marc Gaudry, F.R.S.C.

Professor of Economics, Université de Montréal, Montreal, QC ([email protected])

Context2

These notes are intended to provide some background to the current proposal by Onex to merge Air Canada (AC) and Canadian Airlines International (CAIL).

Key elements of the air industry environment

International structure. The key to the air industry structure world-wide is the legal limit on foreign ownership and operation of airlines (currently 25 % in Canada). In every country, the ensuing « national » designation makes it possible to obtain rights to exploit international air routes assigned within the framework of bilateral treaties under the Chicago Convention (1944) and complementary Bermuda I (1946) and Bermuda II (1977) Agreements. There are over 1600 of these bilateral treaties among countries, with each country typically assigning (for free, i.e. without auctioning them) the rights-to-operate to one or more of her national carriers, without the country losing property of these rights (they can in principle be recalled, according to recognised legal expert John Hamilton, Q.C.).

This multi-national structure explains that, because of the impossibility of merging firms belonging to different countries, companies try to form alliances to get around the rules and indirectly form consortia that can to some extent function as if the firms were merged. But as this is almost impossible without ownership, the alliances are unstable, as shown in Table 1 from Berechman and de Wit (1999) :

Table 1 Airlines Alliances 1997 1996 1995 1994 Number of alliances 363 390 324 280 ■ with equity stakes 54 62 58 58 ■ without equity stakes 309 327 266 222 New alliances 72 71 50 - Number of airlines 177 159 153 136 Source: Airline Business, June 1997. The difficulty of forming stable alliances partly explains why the international airline industry has been notoriously unprofitable during the last 8 years (and before) : in contrast with, say the oil firms,

2 These few pages principally present in extended form the comments I made at the end of the liberal caucus meeting on Monday 27th September, 1999, at the Airport Hilton Hotel, after hearing presentations by Hon. D. Collenette, Mr. K. Benson (Canadian Airlines), Mr. R. Milton (Air Canada) and Mr. G. Schwartz (Onex). The comments were repeated and expanded on during a one hour question period at the request of the House of Commons members of the Ad Hoc Committee on Air Transportation on Tuesday 28th September 1999 in . Some additional information based on documentation I distributed to the Committee members, or of a complementary nature, is also marginally used. 4 air carriers do not generally earn an economic return on their capital and often survive due to large direct or indirect (tax expenditure) subsidies.

These carrier subsidies exist over and above those provided by national, regional and local governments for air infrastructure and air traffic control, as shown for instance for Canada by the Royal Commission on National Passenger Transportation (1989-1992) whose numbers are quoted in Appendix 1 (Gaudry, 1999).

National structure. In Canada, the 10 % maximum legal limit on individual ownership of AC shares (but not of CAIL shares) partly explains the difficulty of merging Air Canada and Canadian : it is difficult to form a stable coalition to that end, despite the apparent advantages of such a merger on the revenue side, notably joint use of the presently assigned rights to international routes, as well as on the cost side, where the perceived deficit of both of these firms in terms of international competitiveness might be partially overcomed. The most sophisticated work on the relative competitiveness of airlines is due to Oum and Yu (1995) who estimated the relative productivity and input price standings of 22 large airlines, taking due account of differences in average stage length and mix of outputs (scheduled passengers, non-scheduled services, scheduled freight services and non-core airline outputs). Results for 1993, reproduced from a large study carried out for , are found in Table 2, drawn from Hickling Lewis Brod Inc. (1997). Results shown for Air Canada and Canadian Airlines are for the trunk carriers and exclude affiliates.

Table 2. Indexes for Cost Calculations, 1993 American = 1.00 # Airline Total Factor Productivity, Input Price Index, Input price per or input per unit of output or price per unit of input unit of output (A) (B) (B)/(A) 1 American 1.000 1.000 1.00 2 United 1.045 1.035 ... 3 Delta 1.069 1.075 ... 4 Northwest 1.124 1.036 0.92 5 US Air 0.832 1.034 ... 6 Continental 1.005 0.891 ... 7 Air Canada 0.807 0.881 1.10 8 Canadian 0.860 0.911 1.05 9 air Lines 0.851 1.421 1.67 10 All Nippon 0.777 1.432 ... 11 Singapour Airlines 0.958 0.813 ... 12 Korean Air 0.988 0.781 ... 13 Cathay 0.969 0.926 ... 14 0.875 0.897 ... 15 Thai 0.647 0.520 ... 16 0.956 1.190 ... 17 British Air 0.893 0.974 1.09 18 Air 0.875 0.974 1.25 19 Alitalia 0.840 1.250 ... 20 SAS 0.838 1.289 ... 21 KLM 0.946 1.098* ... 22 Swissair 0.950 1.360* ... * Based on 1992 data.

In many countries, the deregulation that started in the United States in 1978, and to which we now turn briefly, partly explains the differing performance of national airlines, but the work by Oum and 5 Yu does not try to explain such differences : it simply attempts to measure performance in a uniform way. Despite the fact that they change continuously, and that in Canada the evolution of productivity of the affiliates also matters, the indices are extremely helpful in understanding each country’s current position on « Open Skies », for instance the strong Japanese opposition to any significant liberalisation of bilateral agreements between Japan and abroad.

A handle to influence outcomes

International route permits : a tool of vanishing value ? Many airlines would be bankrupt if national governments suddenly recalled their permits to serve international routes in order to put them out to tender. It is always a possibility : the Government giveth, the Government taketh away and blessed be the markets that discount it all ! But even if the markets discount it to some extent, a sudden change in policy to that effect would be unexpected and quite traumatic for many, and not only for Canadian, as the international routes are the most profitable routes.

This is natural : monopolies are both profitable and a source of a quiet life. But this happy state of affairs might not last forever, or even very long, primarily due to the initiative of the United States, who, as shown in Table 3, have signed 36 « Open Skies » agreements since the first with The Netherlands in 1992.

Table 3. Open Skies Agreements signed by the US* Year Eastern or Western Other country 1992 1 1995 9 1 1996 1 1 1997 1 14 1998 1 3 1999 (8 months) 0 4 *These data include only the 36 « Open Skies » agreements allowing airlines to fly anywhere in each country, as often as they like and at any fare they wish to charge (but all without cabotage rights—the right to serve points beyond the first in the foreign country3). The data exclude major liberalisations, such as those signed with France (in April 1998), as well as major deals providing simply for combinations of increased routes, frequencies, designated airlines, overflight rights, code sharing or 5th Freedom4 improvements. Source : US Ambassy, Ottawa.

It is important to realise that the value of Canadian bilateral rights « assigned » to designated Canadian carriers depends crucially on the availability of better competing paths, for instance to Europe or to the Far East. As Frequency of service is a crucial element in the choice of paths by passengers and freight, the relative importance of high frequency hubs in places like New York, Chicago and the US West Coast, will increase with liberalisation. These accelerating developments put at risk the monopoly rents associated with current international route permits in Canada. This is true whether slots at foreign destinations are made tradable everywhere or not, although the value of excess profits to be earned by those that have the Canadian permits will of course be affected by such changes and how they are brought about. As access between the US and abroad improves, the market pressures to feed US hubs from Canada will increase, reducing the value of our permits.

3 For 250 years, cabotage in shipping has traditionally designated local, as opposed to open-seas (« long cours » in French) services. Here however cabotage means the right of an airline to carry passengers from a foreign airport to another airport in the same foreign country and is formally designated as « the Eighth Freedom », when the flight originates in the home country and as « the Ninth Freedom » when the flight originates anywhere. 4 Fifth Freedom rights are the rights of an airline, for a flight originating in the home country, to pick up passengers in a foreign country and carry them to another country. 6

Our recent cross-border treaty with the United States, which excludes cabotage rights, implemented gradually during the last 3 years 1996-1999, will serve these purposes well. For many years before that treaty, traffic within the US and traffic within Canada had been growing much faster than traffic between these two countries : as the catch-up now rapidly occurs, the competitive position of US hubs will further improve, relative to ours, for many competing paths to Europe and Asia as US Open Skies policies make further access gains across both oceans. Are we to become mere spokes on US hubs ? Are we about to catch up in this respect also?

Geography is destiny ? If Canada were merely « the upper 10 » , without national boundaries with « the lower 48 », how would our air networks look like ? Hubs are generally in central locations and next to large cities : , , an apparent exception to the centrality rule, currently has compensating features, but it will be interesting to see what happens as the 22 international hubs imposed by national governments in Europe restructure5 in a liberalising market (as this is now the case at least within the EU 15) where all big countries progressively sign Open Skies treaties (even without 5th Freedom rights), as did in February 1996 and in November 1998. It is a safe bet to state that our national borders have so far prevented us from being mere spokes on US hubs, with at least one big regional hub, in .

A handle for what ? Let us grant that our international permits are still worth something for some time. Then how can they help us make net gains, if rationalisation is to occur in any case, as is clear to the naked eye ? It has been pointed out by many, in particular by Prof. Tae Oum some years ago, that headquarters are very footloose and that one can gain them or lose them, often for minor reasons that do not fit easily into the economic theory of static comparative advantage.

In this case, our chance is to build a credible middle-size player that might well survive and prosper in the evolving world of competition and alliances by facilitating a merger of Air Canada and Canadian. As the sum of these two is more that their current sizes, seizing this chance is a proper public policy objective : success could give us net value-added, i.e. headquarters jobs6 to offset the necessary job losses at hand. International liberalisation is changing the game without us.

Air markets become progressively more demanding, if only because US dynamism and sheer size are rapidly prying open the covers of protection enjoyed by locations and firms functioning under restrictive bilateral agreements. It is not just that US delegates, whining about liberalisation, have « got religion » : they are succeeding. What are our options ?

5 Does anyone believe that direct flights from Vienna and Switzerland to New York will both survive? 6 And a better handle on conformity with The Official Languages Act. 7 Framework elements

The costs of inaction. Insufficient action might lead to the bankruptcy of Canadian. Some say :

« They took their chances in merging Eastern Provincial Airways, , CP Air, and in the way they did it. Overmortgaged, they cannot make it despite government loan garantees (1992), favouritism through the government purchase of their old Airbuses 310, forced on our armed forces (19937), cash injections by (1994), fiscal adjustments for the spreading of airline losses/profits introduced at their request to help them (1996-1997) and the granting of new routes to Asia that they would not have won in a competitive tender environment (another hidden subsidy !)—let them fail according to market rules, as they clearly deserve !»

One should resist this position, however much of a free marketeer one might be, because there are other costs than those of just deserts.

Firstly, one should remember that there are disruption costs to the public that arise from the destruction of a large firm, particularly in places, such as Toronto (to say nothing of elsewhere), where there are many employees and air services that in this case have to be partly replaced rapidly. Many of these costs can be mitigated by a fusion, despite the obvious complexity of the operation, as the firm simultaneously meets other constraints, such as making all aircraft8 conform to Year 2000 noise regulations or phasing them out.

Secondly, doing nothing puts Air Canada at risk because, if there is no merger, it is progressively more difficult to seize our chance to construct our middle-size player : smaller size players are more likely to end up as mere spokes and then lose their real headquarter functions. Our geography will have caught up with us : all spokes, no headquarters.

Requirements for action : a facilitating policy. Section 47 of the Canada Transportation Act may be useful, in that it brings government approval upfront in any merger talk, but there are reasons to think that it is not sufficient to ensure a neutral framework in which Onex or other comparable proposals can be assessed. Why ? Because there are many requirements to the achievement of a successful merger, when there is some urgency to the situation. What properties should the mechanism update display ?

DEFINE. Canadian ownership might have to be redefined. For instance, in the world of Canadian telecommunications, 33% of foreign ownership is allowed if a financial holding is involved. Why not also in transportation ? New arguments have to be considered : in the US, the recently updated version of the 1991 Transportation Research Board report, Winds of Change : Domestic Air Transport Since Deregulation, urges an end to federal limits on foreigners owning and operating US-based airlines. Can the 10 % AC maximum (authorised by the Word Trade Organisation) be changed and effective Canadian control maintained ? For instance, would not the veto power that the (minority) representatives that American Airlines currently have on the board of Canadian Airlines violate the spirit of the existing law, even if the 10 % threshold were met for that firm (as it must be for AC, by law)? It is hard to see how effective Canadian control can be appropriately9 defined to facilitate fusion proposals without some modifications to the current legislative framework.

7 The deal was probably done in October 1992 but it was kept hidden by the then minister of transport, Mr. Corbeil, who did not announce it with the loan garantees on November 24, 1992. 8 In 1996, none of the 46 such aircraft could have met the Year 2000 noise limits. 9 For instance by excluding such veto powers. The Onex proposal is consistent with this exclusion. 8

DISCUSS. The legal mechanism should admit of merger discussions without violation of the Competition Act. Is it better to invoke Section 47 every time a fusion proposal is talked about ? How can it get on the table without violation of this act?

DISPOSE. Assume that a merger proposal, sent to the Competition Bureau, is opposed. Even if their decisions, that can currently take up to 24 months, are made within a 5-6 month range, this is not fast when bankruptcy looms. Although, the Commissioner of the Competition Bureau can in theory make any recommendation to the Government, such recommendations are not obvious if they are inconsistent with existing Canadian treaties. In practice, could the Bureau say :

« This merger before us is anticompetitive, but it would be fine if company X divests itself of affiliate Y and the Canadian Government unilaterally allows cabotage between points Montreal and Toronto by Air France and gives Fifth Freedom right to SAS Airlines to stop at Halifax on the way to New York, as we understand they have asked by speaking to journalists» ?

At first, why is the Canadian Transportation Agency, clearly already competent to adjudicate on Canadian ownership, not also empowered to manage in addition both fusion and competition conditions within a combined fast-track (quick disposition) cum due process mechanism? It could play a role akin to that of our telecommunications regulator and admit all merger requests within a clear legal framework. This proposal simply formalises the recent idea expressed by Minister Collenette :

« ...The Canadian Transportation Agency may well look into the ramifications of a possible merger » (Press Conference, 28th September 1999).

For instance, one can further imagine the following additional role, dealing with the monitoring of competition, subsequently to having allowed mergers to occur:

« The Agency is empowered to unilaterally allow cabotage by foreign airlines or other ways of sustaining competition in specific markets, such as allowing foreign airline 5th Freedom rights, in markets where competition is insufficient. »

An aggressive policy. But the matter if of some urgency, so it is appropriate to ask whether a yet more aggressive policy is possible, to focus minds. Clearly, the minister could say :

« International route permits belong to all Canadians. At the end of 2002, I will auction them off to the highest bidders, as done recently in the US for pollution permits. All airlines, including Charter airlines and coalitions of firms that meet Canadian ownership and control conditions may bid. The proceeds will go to the Consolidated Fund. »

Other issues

Generally speaking, it is not possible to solve all problems simultaneously. The above suggestions, no doubt modified after a deeper examination with legal issues in mind, would not solve the problem of remote communities. These issues have to be dealt with separately, through normal subsidy programs. They are already with us and should not detract us from rapidly setting up complements to existing legislation that make it possible to deal rapidly with urgent matters, lest we risk losing all of the chance offered by Onex or similar fusion offers perhaps in the works.

9

References

Berechman, J. and J. de Wit, « On the Future Role of Alliance », Ch. 17 in Gaudry, M. and R. R. Mayes, eds, Taking Stock of Air Liberalization, Kluwer Academic Press, 282 p., 1999. French version: « Rôle futur des alliances », forthcoming in Gaudry, M. et R. R. Mayes, ed., Le Point sur la Libéralisation des Échanges Aériens, Les Presses de l’Institut du Transport Aérien, , 1999.

Gaudry, M., « Airport Subsidies and Congestion in North America : the Need for Accounts and a Regulator », Publication CRT-99-26, Centre de recherche sur les transports, Université de Montréal, 9 p., July 1999. French version : Gaudry, M. « Subventions aéroportuaires et congestion aérienne en Amérique du Nord: plaidoyer pour un régulateur et pour l'établissement d'une comptabilité collective », Bureau d’économie théorique et appliquée, Working Paper No 99XX, Université Louis Pasteur, Strasbourg, et Publication CRT-99-34, Centre de recherche sur les transports, Université de Montréal, 13 p , Août 1999.

Hickling Lewis Brod Inc., Assessing the Benefits and Costs of International Air Transport Liberalization, Final Report, 96 p., Transport Canada, 26th February 1997.

Oum, T. and C. Yu, « A Productivity Comparison of the World’s Major Airlines », Journal of Air Transport Management, 4, 181-195,1995.

10

A first version of this paper, presented on December 3, 1998, as a Conférence Jules Dupuit at École Nationale des Ponts et Chaussées in Paris, was also supported by the author’s tenure as a 1998 Centre National de la Recherche Scientifique (CNRS) researcher at BETA, Université Louis Pasteur and UMR CNRS 7522. The author thanks Benedikt Mandel of MKmetric GmbH, Karlsruhe and Centre de recherche sur les transports (CRT), Université de Montréal, who allowed generous quotation of his recent papers on airport competition, as well as David Gillen and John Panzar for helpful comments.

Appendix 1

Airport Subsidies and Congestion in North America : the Need for Accounts and a Regulator

by

Marc Gaudry

• Bureau d’économie théorique et appliquée (BETA) Université Louis Pasteur, Strasbourg, and • Département des sciences économiques et Centre de recherche sur les transports (CRT) Université de Montréal

Bureau d’économie théorique et appliquéeWorking Paper No 99XX Centre de recherche sur les transportsPublication CRT-99-26

July 1999

11 ABSTRACT We examine the potential role of airport slot trading as a congestion internalization mechanism in North America. We argue that the air transportation system is massively subsidized both in Canada and in the United States and that the imposition of prices that recover full direct costs, including environmental costs, would severely reduce the potential need for any slot trading scheme. We also argue that in many airports a severe problem of monopoly regulation would arise if current pricing control institutions were modified, as Canada has started to do. Slot values are a buffer that would also residually reflect true demand driven scarcity not captured by carrier hub dominance, landing fees and airport passenger charges.

Keywords : airport competition, Canada, United States, air transport subsidies, slot trading.

RÉSUMÉ Nous analysons le rôle éventuel d’un marché de créneaux aériens comme remède à la congestion aérienne en Amérique du Nord. Nous prétendons que la suppression des subventions cachées très importantes dont bénéficie le transport aérien et l’imposition de prix qui recouvrent tous les coûts directs de l’aviation, y compris les coûts environnementaux, réduirait l’utilité d’un tel marché, tant au Canada qu’aux États-Unis. Nous prétendons aussi que des réformes institutionnelles affaiblissant la régulation des prix des services aéroportuaires, comme on peut en déceler la trace au Canada, pose de sérieux problèmes de régulation de nombreux aéroports aux caractéristiques monopolistiques. La valeur de créneaux transigés sur un marché est une valeur tampon ou résiduelle qui incorporerait la partie de la rente de rareté négligée par les transporteurs dominants des hubs ou les tarifs aéroportuaires imposés aux aéronefs ou aux voyageurs.

Mots-clés: concurrence inter-aéroports, Canada, États-Unis, subventions aériennes, marché des crénaux aériens.

TABLE OF CONTENTS Abstract...... ii Résumé...... iii 1. Introduction : trade, transport and airports...... 1 Bringing transportation into trade ...... 1 All transport system layers matter...... 1 From carrier to airport competition...... 1 Slot prices : the straw, the beam, or the shock absorber ?...... 1 Cost, demand and the slot price buffer as residual...... 2 Peak charges and 2 efficiency...... Necessary cost 3 recovery...... 2. Virtual air system accounts...... 3 Public sector accountability without accounts ...... 3 Large air system subsidies in Canada ...... 3 New local airport authorities or stronger local monopolies in 4 Canada...... Probably comparable air system subsidies in the United States...... 5 Other considerations that would affect slot prices...... 5 3. The value of the franchise ...... 6 Trends in demand and supply ...... 6 Competition among airports ...... 6 4. Conclusion...... 8 5. References...... 9 About the author...... 10

12 1. Introduction : trade, transport and airports

Bringing transportation into trade. Transport costs have impacts on resource allocation that can be formulated and thought about in the same way as taxes and subsidies: they affect the (c.i.f.) relative prices that determine apparent spatial comparative advantage. As transport costs can represent a large proportion of the delivered price of goods, and incorporate large subsidies, they could be expected to be a proper object of World Trade Organization (WTO) discussions and at the core of discussions on the liberalization of particular modes, such as those pertaining to air liberalization.

But this is not the case yet. Countries, such as the United States and Canada, will have serious quarrels about stumpage fees on lumber exports (perhaps 2-3 % of the total cost of delivered lumber) but will ignore the potentially much larger implicit transport subsidy received by the 62- ton 8-axle trucks carrying this wood on roads for which they might not be paying their due share of fully allocated costs, for instance in Canada (Lawson, 1993). Similarly, service industries are entering the WTO arena, but no mention is made of the fact that many such industries, such as head-office or consulting services, depend much on the availability of underpriced airports and air traffic control services.

All transport system layers matter. If one thinks of the three-layer construct of the air transport system, consisting in carriers, air traffic control and airport infrastructure, there is a clear need to deconstruct it and to ask the right questions about each layer, as each poses distinct problems of regulation. In terms of competition, deregulating carriers is one thing, because a strong case can be made for the presence of competition subject to the normal anti-trust regulation. But deregulating natural monopolies such as air traffic control systems and airports is another.

It is not sufficient to desire efficiency for a particular layer : obtaining « economic » prices for a layer, while neglecting the others, might move us further away from optimal spatial comparative advantage. For instance, trucking deregulation in North America led to sharply lower charges but accelerated the destruction of roads—yielding very ambiguous results : in the United States, the recent decision to spend 230 billion $ coming mostly from the Highway Trust Fund, to rebuild the interstate network almost certainly involves a disproportionate contribution by the lighter vehicles that caused very little destruction of the road pavements. Similarly, some will whine about carrier subsidies at the OECD discussions on air liberalization (Michalski, 1999) but simultaneously remain totally silent about huge hidden airport infrastructure and air traffic control subsidies (including tax expenditures).

From carrier to airport competition. The prima facie presence of large (capital) subsidies or tax expenditures, combined with large sunk costs, relatively long planning times and significant spatial monopoly power at airports makes it difficult to disagree with Panzar’s wrap up comment (Panzar, 1999) at a conference we organized to take stock of 20 years of air liberalization that « airport competition is going to replace flag carrier competition as the national rivalry. The European Union is going to have to engage in some serious thinking about airport anti trust policy for want of a better term. »10.

10 A French version of the papers presented at that conference is in press (Gaudry et Mayes, 1999). 13 Slot prices : the straw, the beam, or the shock absorber ? All layers matter for the full price. The same applies to congestion or peak usage charges in any form, for instance in the form of slot prices of particular interest to the organizers of this Lecture. Slot and gate charges can be thought of as « value buffers » reflecting in part both relatively high consumer demand and relatively low landing and terminal charges and their structure. Their value will depend on infrastructure capital subsidies, landing fee price structure for various aircraft types and services, such as general and commercial aviation, and capture any uncaptured pure location rent associated with the zoning authorizations (the value of the permit to operate an airport) minus any market power rent associated with any local asset dominance of a hub-dominant carrier.

Cost, demand and the slot price buffer as residual. If one excludes time-varying landing charges—they would function as peak-load charges and be substitutes for slot costs— in any form, the value of slots is a sort of buffer or residual varying within two implicit boundaries defined primarily by the answer to two questions: (a) are the airports recovering their full direct costs ? (b) if airports were privatized, would the permit to operate (the zoning designation) without a regulatory cap be worth much ? Positive slot prices may be less a straw, compared to the beam of infrastructure subsidies, than a global measure of governement failure to balance supply and demand at prices that both recover costs and reflect full spatial scarcity rents.

We treat the « slot price buffer » as a scarcity fee, analogous to high season, day-of-the-week or other existing time-of-day airline surcharges. As airlines already extensively practice peak charging on many forms, they could be accused of hypocrisy if they refused time-varying landing charges in order to protect the local component of hub-dominance rents.

Peak charges and efficiency. Such peak charges are easier to understand, specify and implement than optimal short-run marginal cost charges but are clearly positively correlated with them. Indeed, we will not formally discuss optimality and the links between peak charges and classical externality taxes aimed at the difference between short run average and marginal costs. This classical point of view consists in representing, for each link of a network, the congestion characteristics of this link : in Figure 1, the average and marginal times increase beyond point H and a system optimal assignment required to obtain a marginal cost assignment at q2 implies an «externality tax» equivalent t3t2. This theory, stated ad nauseam by economists, for instance again by Mohring (1999), has had no implementation success except in the guise of a peak load surcharge. Marginal time D

t3 Mean time

Time t1 per vehicle t2 D H

Number of vehicles per unit of time

Figure 1. Volume-delay curve q q1 14

But peak charges differ from externality charges conceptually based on short run marginal cost calculations. And not only in the case of congested facilties. Many believe that airports benefit from economies of scale that, in the presence of congestion, require subsidization if prices are to be equalized with marginal costs. Necessary cost recovery. We believe that this view of economies of scale is neither true, nor warranted if true, and that full cost recovery is, as a rule, always desirable for airports : we believe that eficient prices provide for full financing as well.

In this context, we want to ask two questions about slot charges, defined as a buffer or residual : (i) to what extent are they an implicit function of public subsidies ? (ii) would they frequently exist if all airports were privatized under a «licence to steal » regime ? Our suspicion on the first issue is that full direct cost recovery by North American airports would considerably reduce room available for such charges because current prices would have to rise significantly for existing subsidies to be discarded. Our suspicion concerning the second is that, were the complicated institutional restraints on airport pricing to be lifted, many locations with considerable « theft » potential could extract spatial monopoly rents in one form or another, including the sale of slots.

We shall try to provide an intutitive answer to each question in turn. The first question requires a system of accounts. The second some quantitative measure of spatial maket power. Both are missing, in an analytical sense, but some guesses can be made.

2. Virtual air system accounts

Public sector accountability without accounts. Congestion is associated both with the capacity of the Air navigation system and with the capacity of commercial airports, but we focus here on airport congestion and on its relative worth. So, do commercial airports pay their way ? In the United States and Canada, there is no complete set of accounts allowing a fair answer to this question by a proper accounting of aviation-side revenues and resource costs in all forms. It is more difficult still to compute cost recovery for commercial and general aviation or to obtain such measures by aircraft type. And it even more difficult to estimate the costs of noise, pollution and, if appropriate, of excess congestion—of the amount of congestion beyond the optimal amount. In Canada, this situation might soon change : The Canada Transportation Act, 1996 (Section 52 (a, b and c)) obliges the Minister to report to Parliament on the financial viability of the modes and on all resources provided directly or indirectly at public expense to all carriers and modes of transportation. This revolutionary obligation of accountability requires the development of accounts. And a broad view of subsidies should include externalities. We are not aware of similar obligations in the European Union. This means that the 1995 European Commission Green Paper on Fair and Efficient Transport Pricing (Commission, 1995), intended to be an essential reference for the development of the Common Transport Policy, has no hands. To have hands—even if a little dirty, accounts are necessary. For this reason, no doubt, the European Union’s 5th Framework Program for transport (DG VII), available since March 1999, includes Task 2.1.1/2 « Transport network accounts and marginal costs in relation to fair payment for infrastructure use ».

Large air system subsidies in Canada. Fortunately, the Canadian Royal Commission on National Passenger Transportation (Hyndman, L.D. et al., 1992) has elaborated a first set of such

15 «virtual » accounts for the year 1991, in order to provide a rough measure of the cost recovery of commercial aviation making use of the 98 airports in which Transport Canada has a significant interest. This set includes almost all airports with scheduled services and excludes the hundreds of small community airports in Canada. The results, for the system as a whole, and for the Montreal-Toronto link, are shown in Table 1 where « others » designates the population in general in the case of environmental costs and designates government in the case of other types of costs and fees. The infrastructure line includes all approriate costs, including the oportunity cost of capital, depreciation, administration, etc. for both the infrastructure and the traffic control layers of the air system. It appears from these figures that users pay about 82 % of all costs of the system as a whole, including environmental costs but excluding any excess congestion costs, and a lower proportion (70 %) of the same costs on the largest market route. A major element of this subsidy is due to the tax expenditures implied by the relatively small rents paid by users to taxpayers for land and terminals. Although these numbers can be disagreed with in their specifics, as is true of all estimates, and clearly vary across airports, their order of magnitude is reasonable.

Currently, both and Toronto airports are said to be significantly congested. But they are also the airports with the highest tax expenditures on the value of land and terminals : in the calculations of Table 1, the value of land for these two airports is conservatively estimated at 1 250 million $ (1991). Only a small part of the congestion is due to general aviation (excluded from Table 1 calculations) currently generating costs estimated at 200 million $ per year at these two airports. So, if airports were obliged to recover all costs, perhaps by raising the landing fees by 20-30 %, and without special provisions for general aviation, how much congestion would remain in equilibrium? In Canada, not very much. The « residual buffer » size will clearly depend on the specifics of the cost-recovery formula, such as the relative roles of per movement, per passenger and gross weight charges. Of course, a 30 % increase in fee income is compatible with the simultaneous introduction of time of day peak charges, or slot prices in their absence, but the buffer might be very thin.

Table 1. Annual Costs of Intercity Domestic Air Travel, Paid by Users and Others, 1991, in 1991 Dollars, Canada and Toronto to Montreal Totals : $ millions System-wide Toronto-Montreal Route (25 billion pass km) Type of cost Users Others Total Users Others Total Infrastructure 556 845 1 401 22 20 42 Environmental 0 247 247 0 6 6 Accidents 2502500 0 Special trans. tax/fee 149 -149 0 3 -3 0 Vehicle/carrier 3 595 0 3 595 111 35 146 Total 4 325 943 5 268 136 58 194

New local airport authorities or stronger local monopolies in Canada? In Canada, the answer is even more difficult to give than implied above because the devolution of federal airports to non-profit local airport authorities (LAA), announced in 1987 and effected principally since 1992, and the creation in 1998 of a separate self-funded organization for air traffic control, called NAV Canada, has removed the clear « essential services » classification of many of these airport and air navigation employees that effectively prevented them from striking since 1982. The consequences of this have not yet unravelled. In the absence of a similar new mechanism

16 mimicking to the extent possible a solution that reflects the opportunity costs of these employees, they may well use this sudden gift of monopoly power to extract remunerations and working conditions well in excess of their social opportunity cost, as happened generally in urban transit systems. This extraction of a new pound of flesh, over and above the new pension plan and employee benefit packages offerred shortly after devolution (for instance in Montreal) would further increase the average cost of air travel and reduce the available pure peak load buffer space of interest here.

Probably comparable air system subsidies in the United States. How different would comparable figures be for the U.S. system ? In 1989, there were about 400 primary airports in the U.S., accounting for over 95 % of air passenger emplanements on scheduled domestic air carriers and over 99 % on certified air carriers. But the system of « public use » airports had 5 680 members, including about 1 700 privately owned and the rest generally the property of local and state administrations (« Authorities »).

It is possible that the implicit capital subsidy in American airports is of the same order of magnitude as in Canadian airports, because these airport authorities: (i) generally finance investments with State-garanteed tax-free revenue bonds—an advantage that can be equivalent to about a one third capital subsidy, as Bombardier calculated for the Texas High Speed Rail project; (b) receive direct capital subsidies from state, local and federal administrations.

Consider for instance federal grants. For the Airport Improvement Program (AIP), in 1989, 3 272 airports (as well as 412 proposed new airports) were listed as eligible for these federal capital improvement grants and included in the National Plan of Integrated Airport System (NPIAS); and about half of the federal capital subsidies available under that AIP program went to the primary commercial airports. As for the National Airspace System (NAS), including air traffic control, federal funding covers almost all capital and operating costs. In toto, between 1982 and 1989, only 57 % of all federal funding for aviation came from the Airport and Airways Trust Fund, a revenue account supported by various aviation user fees and charges : the rest (43 %) was covered by the General Fund of the U.S. Treasury (U.S. D.O.T., 1990).

This basically means that both subsidies and tax expenditures are pervasive and very large. So the other costs of airport operations, however complicated (for a summary, see Langer, 1996) and varied their financing, are to be understood on a beam of capital and operating subsidies, from all levels of government.

Other considerations that would affect slot prices. As also well explained by Langer (op. cit.), the two forms of attainment of budget equilibrium in U.S. airports (called the compensatory and the residual structures), involve varied arrangements on landing fees (often the object of long- term 30-year contracts), gates (often rented by firms for 40-year contracts, and sub-let afterwards to other firms) and ticket counters (92 % of which are on long-term exclusive rental basis). Her description of the management of U.S. airports gives a proper summary of each element of this multi-layer problem, with many interesting tables showing the frequency distributions of the various long-term arrangements.

Concerning slot trading, Langer’s summary of the history, monitoring and prices of slot trades and their values at the four « slot coordinated » high density airports in the U. S. (O’Hare in Chicago, La Guardia and J.F. Kennedy in New York, and Washington National in Washington, D.C.), based on the available information, suggests that air fare might increase 5-10 % after the 17 introduction of generalized slot trading. These estimates, based on CD Scicon (1991) study for the , do not precisely take into account the fact that air fares at high density airports already reflected the scarcity of slots even before the implementation of slot trading at those airports in 1986.

Full cost direct recovery at all primary U.S. airports might therefore easily imply 20-40 % increases in air ticket prices, sharply reducing the possible share of peak load charges.

3. The value of the franchise

Trends in demand and supply. But surely, the evolution of demand, and the spatial monopoly power of various locations, are also important in making our guesses. Over the last 20 years, air travel has grown considerably, driven in particular by tourism that has grown faster than international trade, itsef growing about 60 % faster that Gross World Product. All forecasts, for instance those of the Airports Council International (ACI) representing 523 authorities running 1300 airports in 160 countries (25 % in developing countries), for the next 15 years suggest a doubling of demand from current levels, with passenger growth around 4,9 % worldwide and freight about 8,4 % per year (or, as usual, about 50 % faster than passengers). Aircraft movements are expected to grow by 3,2 % per year, due to the shift to larger aircraft. On the freight side, for instance small packages, representing almost 0 % of freight 20 years ago in the U.S., now account for more than 60 % of total freight : similar developments at the world scale can be expected. On the supply side, about 350 billion $ of large airport capital investments are planned during the same period. This will barely suffice to meet the expected demand. Generally speaking then, the market power of individual airports is not expected to fall. But how large is it now ?

Competition among airports. Although many researchers have studied air demand, the most sophisticated work on airport competition is perhaps that of Mandel (1999a, 1999b) for Germany. His models take into account time-of-service and airline choice, choice of access/egress mode to/from airports, choice of airport and choice among air services at airports, mode choice and destination choice. As all networks in Germany are well developed, market power estimates for Germany should underestimate market power in countries with less dense networks and longer distances.

In consequence, we draw from one of his many scenarios, « Local pricing strategy », dealing with a potential increase in passenger fares at an international airport—the « passenger facility charge », in U.S. jargon, increasing to 50 DM at Hamburg airport. And we draw from his figures to illustrate the results. Figure 2 depicts the simulated market share losses for that airport and Figure 3 summarises the passenger shifts among the competing alternatives, with 197 000 passengers shifting to other airports and 638 000 to other modes.

18 0% to 1% 1% to 1% 1% to 2% 2% to 3% 3% to 5% 5% to 7%

Hamburg

Berlin ic

Dusseldorf r

Frankfurt © MKmet

Figure 2. Market share losses Hamburg 1991: all destinations

scsceennaarriioo:: +D+DMM 50,-/50,-/ppaassengssengerer onon HamburgHamburg AAiirprportort 19911991

HHAAM:M: -- 835835 TsTsdd ttoo ototherher modmodees:s: ++ 638638 TTssdd

losers winners

FRA

DUS shifts at the airport MUC

HAJ

BRE

CGN passengers changing airport STR

KEL

Berlin

CPH shifts on flights from/to Hamburg AMS ic r

FMO

BLL © MKmet

-200 -150 -100 -50 0 50 100 shifts of passengers in Tsd.

Figure 3. Passengers shifts

Mandel further analyzes the changes by trip purpose, destination, etc. as well as the derived demand for the different aircraft types (Figure 4) and, finally, the changes in airport revenues

19 (Figure 5) as the passenger « tax » is increased, without the other airports or modes changing their prices. Although Mandel does not explicitly introduce the formal measures of market power found in the industrial organization literature, for instance the ability of an airport to raise charges by 10 % and not suffer more than a 5 % loss in market share, this and other measures are all derivable from his refined models.

In Europe, airport rates and charges account for about 15-18 % of total airline operating costs. In this context, privatization with a « licence to steal », i.e. without anti-trust rules, could clearly lead to unusually high location rents associated with the relative scarcity of airport franchises. In North America11, such charges represent about 4-6 % : there is therefore even more room for increases before elasticities of demand increase much.

In the U.S., both the Federal Aviation Administration (FAA) and the Airport and Airways Trust Fund play an important role in controlling prices, over and above the role of the local administrations : for instance, FAA approval of peak charges is necessary if an airport has used money from the Fund, and courts have forced airports like Boston and Los Angeles to cancel peak charges. These institutional constraints currently limit the exercise of spatial monopoly power. But in many locations the power is there.

in total

Airbus A340 t f a

r c r ai

f

Boeing 737 ic o tr s e as m cl K © M ATR72

0% -5% -10% -15% -20% aircraft movements

Figure 4. Aircraft movements Hamburg Airport 1991 (percentages)

11 This was pointed out by David Gillen. 20 400 Mill.

e stim a te d co rrid o r

300 Mill.

200 Mill. actual aviation revenues ic Ham burg Airport 1991 metr

(DM 27.61/passenger) K

100 Mill. © M + 0 + 50 + 100 + 150 + 200 [D M ] additional charge per passenger

Figure 5. Aviation revenues depending on passenger charges

4. Conclusion

We must conclude that the imposition of full direct airport cost recovery would significantly increase air prices in North America and reduce opportunities for residual slot trading. However, it is also clear that the considerable location rents could be extracted from many locations if the control over airport pricing was reduced, and that slot prices can capture such rents.

Conversely, we can conclude that air network accounts and airport monopoly regulation are both relevant to any transport policy and that airport and air traffic control devolution in Canada raise monopoly control issues that have yet to be fully addressed. Overall, generalized slot trading without such reforms would deal more with a symptom than it would solve a problem.

21 5. References

Commission des Communautés Européennes, « Vers une tarification équitable et efficace dans les transports », COM (95) 691 final, 87 p., 12th December 1995.

Gaudry, M. and R. R. Mayes, ed. La Libéralisation des Échanges Aériens : Bilan et Perspectives, Les Presses de l’Institut du Transport Aérien, Paris, 312 p., 2002..

Hyndman, L.D. et al. "Directions: The Final Report of the Royal Commission on National Passenger Transportation" Minister of Supply and Services, Ottawa, Canada, 1718 p., 1992.

Langer, S.J., The Allocation of Slots in the Airline Industry, Nomos Verlagsgesellschaft, Baden- Baden, 212 p., 1996.

Lawson, J., “Highway Infrastructure Costing and Revenue Comparisons”, Economic Analysis, Transport Canada, 1993.

Mandel, B., « The Interdependency of Airport Choice and Travel Demand », in Gaudry, M. and R.R. Mayes, eds, Taking Stock of Air Liberalization, Kluwer Academic Publishers, 189-222, 1999a.

Mandel, B., « Measuring Competition in Air Transport », Publication 99-18, Centre de recherche sur les transports, Université de Montréal, 25 p., 1999b.

Michalski, W., « The OECD Project on International Air Transport », in Gaudry, M. and R.R. Mayes, eds, Taking Stock of Air Liberalization, Kluwer Academic Publishers, 133-141, 1999, et Gaudry, M. et R. R. Mayes, La Libéralisation des Échanges Aériens : Bilan et Perspectives, Les Presses de l’Institut du Transport Aérien, Paris, 219-232, 2002.

Mohring, H., « Congestion », in Gómez-Ibáñez, J., Tye, W.B. and C. Winston, eds, Essays in Transportation Economics and Policy, 181-221, The Brookings Institution, 1999.

Panzar, J., « Wrap-up : the Way I see it », in Gaudry, M. and R.R. Mayes, eds, Taking Stock of Air Liberalization, Kluwer Academic Publishers, 281-282, 1999, et R. R. Mayes, La Libéralisation des Échanges Aériens : Bilan et Perspectives, Les Presses de l’Institut du Transport Aérien, Paris, 309-312, 2002.

SD Scicon, « Study on Airport Slot Allocation », Final Report, A study commissioned by the U.K. Department of Transport, Camberley, 1991.

U. S. Department of Transportation, Moving America, February 1990.

22 About the Author 01/01/98

SUMMARY OF MARC GAUDRY'S CURRICULUM VITAE

Occupation. Marc Gaudry is Full professor at Université de Montréal (Economics department and Centre de recherche sur les transports, CRT) and Associate researcher at Université Louis Pasteur (Bureau d'économie théorique et appliquée, BETA), Strasbourg. He was recently a member of the Royal Commission National Passenger Transportation (1989-1992) and invited professor at École Nationale des Ponts et Chaussées (Paris), Université Robert Schuman (Strasbourg), Universität Dresden and Universität Karlsruhe (1993-1995).

Education. Born in , Canada, in 1942, he obtained a B.A. in Montreal (1961), an M.A. from Oxford University as a Rhodes Scholar (1961-1964), a B.Th. (Theology) from the Institut Catholique de Paris (1967) and a Ph.D. (Economics) from Princeton University (1973).

Teaching activities. At Université de Montréal, he has taught simultaneously in Economics for 25 years and in Political Science (2 years) or Civil Engineering (18 years). He has directed 46 Master's theses and 16 Ph.D. theses (in Economics, Civil Engineering, Operations Research, Geography and Psychometrics), including 7 M.A. and 12 Ph.D. theses in European universities (Karlsruhe, Bamberg, Dresden, Erlangen- Nürnberg, Oslo, Strasbourg and Reims).

Research activities. He has published since 1974, mainly in applied econometrics or transportation, 49 articles in scientific journals, 20 chapters of books and 60 technical papers (with 42 co-authors). His team has developed documented statistical estimation programs (over 100 users in 22 countries) and data banks. Since 1974, he has made 336 presentations in 17 countries: 104 in scientific conferences and 232 as invited seminar lecturer.

He has over 1000 citations in scientific journals and is associate editor of Transportation Research B, of Les Cahiers Scientifiques du Transport and of Recherche Transports Securité. Elected Fellow of the Royal Society of Canada in 1987, he won in 1990 the Quebec Transportation Research and Development Prize and an Alexander von Humboldt Research Award (Forschungspreis), and was listed in Canadian Who's Who. He is a member of the jury for the INSPIRIT '95 International Research Prize given by the Initiativkreis uhrgebiet in Essen and won the 1995 Ernst Blickle Research Prize (SEW-Eurodrive Foundation, Germany).

Administration activities. Associated with Centre de recherche sur les transports since its beginning in 1971, he contributed as founding member, assistant or acting director, to the CRT status of world leadership in transportation planning. Recently, he had a decisive role in obtaining the largest of the 43 "Structuring Grants" (1986-1991) awarded to scientific teams by the Quebec Government, including 5 new permanent positions for the Université de Montréal.

Professional activities. He has implemented large scale demands models in transportation agencies Montreal, Toronto, Province of Quebec) and developed such models for national ministries of transport (France, 1987; Germany, 1989-1994; Canada 1992-1994; European Commission 1996-1998). He has been used as expert by two provincial royal commissions (Quebec, ) and in over 80 media interviews (radio or TV) across Canada. He is a member of the Advisory Committee on Nuclear Safety of the Atomic Energy Control Board of Canada.

1460 Dr. Penfield Ave. Fax : +1 514 842 0693 203A rue du Verger, Apt. 205 a/s Dr. Stoll Montreal , QC Tel. : +1 514 284 0156 Westhoffen Tel. : +33-3-88-50-58-92 Canada H3G 1B8 [email protected] F-67 310 France

23