ساختار بازار، قدرت خنثی کننده و تبعیض قیمت: مورد فرودگاه

A number of interesting policy questions have arisen regarding airport landing fees. For example, what is the impact of joint ownership of airports? Does airline countervailing power stop airports raising fees? Should airports be prohibited, as an EU directive intends, from charging differential input prices to airlines? We set out a model of upstream airports and downstream airlines with varying countervailing power and pricing structures. Our major findings are: (a) an increase in upstream concentration or in the degree of differentiation between airports always increases the landing fee; (b) the effect of countervailing power, via an increase in downstream concentration, lowers landing fees, but typically does not pass through to consumers; (c) with Cournot competition, uniform landing fees are always higher than discriminatory fees.

The airline market has provided fertile territory for huge numbers of theoretical and empirical papers in economics. Perhaps one reason is that its institutional features span so many interesting phenomena: competition, regulation, networks, auctions, unionized labor markets, the environment, consumer transport choice, etc. But relative to that considerable weight of work there is relatively little on what would seem a rather important complementary input, namely airports.1 Perhaps until twenty years ago, it might be argued that the study of airports was not particularly rewarding either by itself or as something that might inform the study of airline competition. Most airports were public sector owned and regulation or specific agreements held landing fees to non-profit levels. The vast majority of airports held plenty of spare capacity and their location was a historical accident; new entry was almost unheard of. Competition between airports was a fanciful notion regarded as impossible. The position today looks very different. First, market structure. Low cost airlines have brought new, often non-central airports, into effective competition. Even in large cities, competition has emerged between (non-congested) airports: privately-owned rival airports have engaged in documented bidding wars over lower landing fees in cities like Moscow, Belfast, Melbourne, Orlando, Miami and London. Competition has grown even in relatively congested airports where network externalities are important: 33% of London Heathrow passengers only change planes there, leading Heathrow to publicly claim that it competes with hub airports in Paris, Frankfurt and even Dubai. Second, privatization. Fifty-five countries have partially or totally privatized their airports ( IATA, 2007). Third, regulation. There are currently a series of major regulatory changes proposed to airports. In the UK, the Competition Commission in 2009 ruled that BAA, the joint owner of most major UK airports (London Heathrow, London Gatwick, London Stansted, Glasgow, Edinburgh, Southampton and Aberdeen), should be broken up. The EU Airport Charges Directive (2009/12/EC) imposes a host of regulations for large airports, including “non-discrimination”, i.e., an airport must offer the same input charge to all airlines. Other regulatory changes in train include examinations of single and dual till regulation whereby retailing revenue at airports is returned to airlines or airports respectively. Fourth, congestion. Airports have increasingly become more congested, raising additional problems in short-run landing fee pricing but also long-run capacity expansion ( Borenstein and Rose, 2007). New runways have been vetoed in London but the new runway at Frankfurt was completed in 2011. These changes open a number of interesting questions. We cannot study all of them in one paper. This paper sets out a framework that can help in answering answer at least some of the following policy-relevant questions: (a) What is the effect on landing fees of ownership structure up and downstream? For example, should the jointly owned UK airports be split up? (b) Would countervailing power from airlines ever be enough to stop airports charging high landing fees so that even a geographically isolated airport does not need regulation? (c) As airports get more congested, does that alter the nature of the relationship between airports and airlines? (d) Should input price discrimination be allowed by airports? These questions are recognized to be crucial in policy debates regarding the regulatory choices for airports, yet to the best of our knowledge none of them have been answered using a formal model. As we shall see, some of the answers to these questions are exactly those one would expect, but some are remarkably different, providing a strong motivation to our analysis. We also think the paper is of broader interest. First, one ingredient of our model is countervailing power, an issue that is common in cases where more or less concentrated intermediate suppliers and final sellers face each other (e.g., farmers and supermarkets, health insurance companies and hospitals). Second, it turns out that some of the (rather few) existing models in this area have used particular demand functions that do not fully satisfy some requirements such as negative cross-price elasticities of demand. We work with a novel demand system that has not been used before and thus show how we avoid some of the implicit assumptions made in other cases. Thus an innovative contribution of the paper is a new demand system and associated results. We study a vertical industry in which upstream suppliers (airports) provide an essential input (landing rights) to downstream firms (airlines) at a linear price (the per-passenger landing fee). We model various degrees of concentration in the up- and downstream market structure and of substitutability of demand. Upstream, we assume two airports, who have varying derived-demand substitutability between them. The airports may be jointly owned, or be independently owned, to control for upstream concentration. Downstream, we assume up to four products (routes) with varying demand substitutability among them. The flights can be operated by four separate airlines, or by two multiple product airlines that fly from both airports. In this way, we can also investigate the effects of changes in downstream concentration. We also look at different modes of competition downstream. We analyze both cases of airlines competing in quantities and prices, taking these, respectively, as illustrations of more or less congestion of the airport systems; see discussion below. In the last section of the paper, we also consider the effects of a ban on discriminatory landing fees. Our modeling choices attempt to capture the essential features of the industry under analysis.2 Some airport systems are under the same ownership (London before the UK CC decision, but also Paris, New York and Rome) while some other face competitors; in some cases locally (Moscow), in some other cases from remote airports (major international hubs). Competition between airports is also related to the offer of overlapping routes by airlines and to the willingness to travel between alternative airports by consumers, both found to be rather high in recent studies (OFT, 2007). As to the landing fee, it is composed of different elements, most of which vary linearly with the number of passengers. The possibility of discrimination between airlines flying from the same airport is explicitly forbidden in Europe by the recent 2011 EU Airport Charges Directive (Competition Commission, 2009, para 6.15) and Section 41 of the 1986 Airports Act in the UK (CAA, 2006).

In this paper we have set out a model of a vertical industry with up- and downstream competition, and applied it to airports. Our model sheds light on some key policy questions such as: what are the consequences of joint ownership of airports and airlines? Do remote airports need price regulation? Should airports be permitted to price discriminate between airlines? Our results can be summarized as follows. First, should jointly owned airports be broken up? A break up turns out to reduce landing fees ℓ in all cases, as airports compete against each other. But, interestingly, the effects depend in addition on downstream competition, with even lower landing fees if airline routes are more substitutable. Second, what about the countervailing power of airlines? Airports frequently argue that they are stopped from raising ℓ if they are dealing with a dominant airline. Indeed, our model shows that, when a concentrated airline might potentially switch to fly the same route from a nearby airport, the landing fee falls, just as the airports argue. But there are other effects. Airlines succeed in lowering ℓ but keep most of that surplus. So relying on countervailing power is not likely to be welfare-enhancing via reducing prices to consumers. Third, do airlines prefer the new EU rules that appear to make discriminatory pricing harder? Broadly speaking, airlines resist discriminatory pricing more when they are Cournot competitors, so, overall, ℓ is lower in this case compared to uniform landing fees. Thus, if there was Cournot competition at a crowded airport, and the airport moved from a uniform to a discriminatory regime, ℓ would fall. Therefore the model would predict that airports should welcome the proposed EU rule at congested airports, in contrast with airlines.21 The final question we touched on indirectly is how an airport expansion would affect profits of the various parties. There are two broad effects. First, for a given ℓ, an expansion raises airport profits, since quantity rises. But second, an expansion might change the nature of competition and so change ℓ. It is likely that an expansion of a crowded airport changes downstream competition (in our model, from Cournot to Bertrand). The expansion effect always increases airports’ profits under common ownership, though results are more nuanced under separate ownership. As for airlines, they too have more volume, but if competition becomes more intense (in our model, changes from Cournot to Bertrand) their final prices fall as well as their profits. Our approach to the analysis of the vertical structure of oligopolistic up- and downstream industries is amenable to other applications. As mentioned in the Introduction, retailing shows many analogies and similar modeling features. Of course, the model leaves a number of avenues for future work. One is investment incentives in airports. This requires a dynamic investment model and the resulting impact of a change in capacity on competition. Second, there are a number of potentially interesting issues in slot trading. One feature often not realized is that airlines incur substantial sunk costs at airports; thus the grandfather rights to slots may provide appropriate incentives to incur such costs. But this process interacts with competition and so its effects await further analysis. A related issue on slots is that airlines in full airports, with ℓ regulated for example, can switch airports by buying slots from each other: an effect we have not modeled but would require an additional price in the model and bargaining between airlines.