- Email: [email protected]
An economic assessment of airport incentive regulation
Transport Policy 41 (2015) 5–15
Contents lists available at ScienceDirect
Transport Policy journal homepage: www.elsevier.com/locate/tranpol
An economic assessment of airport incentive regulation Nicole Adler a, Peter Forsyth b, Juergen Mueller c,n, Hans-Martin Niemeier d a
Hebrew University, Mount Scopus, Jerusalem 91905, Israel Monash University, Clayton 3800, Australia Berlin School of Economics and Law, Badensche Strasse 50-51, D 10825 Berlin, Germany d University of Applied Sciences, Werderstr. 73, D 28199 Bremen, Germany b c
art ic l e i nf o
a b s t r a c t
Available online 14 May 2015
There has been a gradual trend towards incentive regulation of airports since the privatization of BAA in 1986. Airports are price capped in several countries belonging to the European Union, as well as elsewhere, notably India. However, most of the price caps are not a pure price cap in which the X-factor is set independently of the cost of the regulated airport. Typically, hybrid price caps are used and combined with sometimes complex mechanisms like sliding scales, quality incentives and investment obligations, such that the incentive structures may become distorted. We provide an overview of the changes in the governance structure of airports as a result of privatization, analyze how far the regulatory institutions obey the principles of ‘good’ regulation, such as fairness and transparency, and investigate the performance of the new regulation in terms of allocative efficiency by reviewing benchmarking studies. Finally, we assess the impact of incentive regulation on productive efficiency using data envelopment analysis and second stage regression, which suggests that incentive regulation is superior to cost plus. & 2015 Published by Elsevier Ltd.
Keywords: Airport economic policy Airport performance Regulation Competition and privatization of airports
1. Introduction: why incentive regulation? Incentive regulation (IR) grew out of a dissatisfaction with the way conventional regulation, which had been in place for nearly a hundred years in the US, had been performing. The dominant form of regulation, at least in the US, was Rate of Return (RoR) regulation, which was recognized to have several serious drawbacks, in that it had few incentives for the regulated firm to control costs, and there was evidence that it had led to over-capitalization (the Averch and Johnson effect). In the US, with its long history of regulation, and the UK, which by the 1980s had begun to privatize its public utilities, there was a call for forms of regulation with improved incentive properties. Proposed changes to the form of regulation were debated at both the theoretical and the practical level. At the theoretical level, models of regulation were developed (e.g. Baron and Meyerson, 1982; Laffont and Tirole, 1986) which attempt to provide the firm with greater incentives to pursue productive efficiency. The main focus was on productive efficiency, but there was an expectation that the new forms of regulation would also encourage allocative efficiency, as RoR regulated firms showed little incentive to pursue n
Corresponding author. E-mail addresses: [email protected] (N. Adler), [email protected] (P. Forsyth), [email protected] (J. Mueller), [email protected] (H.-M. Niemeier). http://dx.doi.org/10.1016/j.tranpol.2015.03.008 0967-070X/& 2015 Published by Elsevier Ltd.
profits by adapting their price structures to better reflect costs and demand conditions. At about the same time, especially in the UK, there was a move away from RoR regulation, replaced by simple rules which incentivized the firm to produce more efficiently. The most common of these is the price cap rule, which was devised by Littlechild (1983) (see Beesley and Littlechild, 1989), and under which the firm is given a predetermined price path which sets the maximum price it may charge. If the firm achieves costs lower than the prespecified price path, it enjoys the additional profits, whereas if costs are higher than that defined by the path, the firm will incur losses. While several versions of the price cap rule exist, price caps are the most common form of practical incentive regulation. Incentive regulation of airports began when BAA, which owned the major London airports, along with a group of smaller airports, was privatized. Since then regulation of the London airports has remained in place, although regulation of smaller airports has been discontinued. Australia followed the London model when it privatized its airports, beginning in 1997, though it ceased to regulate via price caps in 2001–02. Since then a number of countries have imposed price regulation, although several of the airports are not regulated by traditional price caps. The primary emphasis of incentive regulation of airports has been on achieving productive efficiency, in addition to avoiding excessive charges. However other aspects of efficiency are expected to be affected by IR. With airports free to maximize profits, there has also been an expectation that price structures will
6
N. Adler et al. / Transport Policy 41 (2015) 5–15
become more influenced by costs, by demand elasticities. Incentive regulation may have negative effects on quality of service, since costs can be reduced by cutting quality. With airports, the distinction between aeronautical and non-aeronautical revenues is an important one whether these are included in the regulated prices (single till) or not (dual till). Finally, IR can be expected to have an impact on profitability and its variability. The objective of this paper is to survey the performance of airports being subjected to different forms of incentive regulation. We first comment on the governance arrangements of airports, and then we try to assess which airports are truly incentive regulated and which are not. In Section 4 we set out what can be expected of incentive regulation. The core of the paper is Section 5, in which we attempt to assess the extent to which incentive regulation has affected performance. In particular we present a benchmarking and econometric study of the impact of incentive regulation on performance. Appendix A details the dataset analyzed and provides summary statistics and Appendix B specifies the type of regulation applied at the 58 airports for which data was available over time. We draw some conclusions in Section 6.
2. Principles of good regulation Attempts to reform regulation in the direction of incentive based price caps also led to a renewed interest in the institutional aspects of regulation, in particular when these reforms were combined with privatization. The key problem of organizing public utilities is to write long term contracts for investments in long lived assets which have value only in a specific exchange relation. This is no problem in a world of perfect foresight, but in reality asymmetric information makes it infeasible to write complete long term contracts covering all contingencies. Opportunistic behavior might lead to hold-up problems so that markets and long term contracts fail and only regulation can create discretionary commitment. However even regulation can fail to provide this long term stability which is captured by changing interest groups and opportunistic behavior (Gomez-Ibanez, 2003). Such failure of regulation led economists1 and international institutions to develop principles of good regulation which are now well defined, precise and, at least in principle, accepted by many states. According to the OECD (2005, p. 35) good regulation should “be carried out by an independent, but democratic authority, because such an institution minimizes opportunistic behavior from the regulated firm, its users, its owners and from policy. An independent authority should try to correct market failure efficiently and avoid regulatory capture.” The concept of good regulation does not prescribe a method of regulation. Cost based regulation can be part of good regulation if it avoids excessive cost inefficiencies. Good regulation, in particular with its demand for an independent regulator, is a necessary though not sufficient condition for incentive regulation. As incentive regulation implies risks for the regulator, it is equally important that firms should be prevented from influencing the regulator in an attempt to mitigate these risks. Traditional cost based regulation leads to problematic incentives which encourage gold plating and cost padding, but may nevertheless be considered a form of good regulation provided any inefficiencies are not overwhelmingly large. 1
Stern (1997) provides a good overview.
3. What is incentive regulation at airports? As the concept of incentive regulation has been developed for all public utilities, we firstly define incentive regulation in general, and then apply it to airports by providing examples of different types of incentives schemes. Thereafter we argue that while ‘yes or no’ categorization is not possible, it is useful to differentiate between different forms of incentive regulation. 3.1. Incentive regulation The regulator must design a contract to set incentives for the regulated firm. If the contract reimburses the firm irrespective of its efforts, the contract is referred to as ‘low’ powered. In this case, regulated prices costs are likely to be high. If the contract only fixes a certain price, so that all cost savings remain with the regulated firm, the contract is referred to as ‘high’ powered (Laffont and Tirole, 1993). Such a regulatory contract which decouples prices from costs and profits are a means to induce efforts to reduce costs. However, cost based regulation would appear to accentuate the central problem for regulators, that is, the regulator has asymmetric information about the demand and cost functions. Given the information asymmetry, the regulated firm will provide the regulator with the information only if in a position to keep some of the informational rents. Regulation will not achieve first best outcomes, but incentive regulation should lead to better outcomes than cost based regulation, depending on how high or low powered the contract is designed. There appear to be three broad forms of incentive regulation that are currently practised. Price cap regulation was developed initially by Beesley and Littlechild (1989), and later evolved into British utility regulation. There are different varieties, but it is useful to distinguish between pure price caps, in which the cap is set independently from the costs of the regulated firm, and hybrid price caps, in which the cap is based every few years on the regulated asset base. Pure price caps are high powered; contracts hybrid caps less so. Revenue cap and revenue sharing agreements limit total revenues rather than prices (Green and Rodriguez-Pardina, 1999). Revenue and profit-sharing concepts define the level of profits the regulated firm is permitted to earn. Once this benchmark is exceeded, the amount by which the firm surpassed the benchmark must be shared between the firm and its consumers (Kunz, 2000). Benchmarking and yardstick competition was developed in particular by Shleifer (1985). The regulator determines the prices for the firm by using the cost levels of comparable firms in order to settle on a regulated price on the basis of costs seemingly unrelated to the regulated firm. It is important to note that these three systems of price level regulation can be combined and even mixed with low powered cost regulation. The former happens, for example, if benchmarking is used to determine the X in the price cap formula. An example of the latter occurs when prices are regulated on a cost based principle and the quality is defined in terms of standards, which are incentivized by a bonus system. 3.2. Forms of incentive regulation at airports Traditionally, for most airports, prices were set on a cost based approach. Even in those countries where governments own, operate and regulate their airports, charges were set in relation to costs, although very often in a non-transparent manner. Price capping of airports started with the UK airports in 1986. The UK Civil Aviation Authority (CAA) followed in the track of the price cap approach applied to other UK utilities, which led to a hybrid price cap. Another important feature of UK incentive regulation is that only a few airports, those that were thought to have market
N. Adler et al. / Transport Policy 41 (2015) 5–15
power, were designated for formal price cap regulation, although all airports could be regulated were it proven that they abused their pricing freedom. There have been several attempts to strengthen such price caps. Setting the X efficiency level through benchmarking was explored, but found not robust enough to be employed in the regulatory process due to data problems (CAA, 2000). The scope of regulation was defined more narrowly in an attempt to strengthen the power of the regulatory system. Attempts to adopt a dual till approach, whereby only aeronautical costs and revenues are regulated, failed in the UK but were adopted at other airports such as Malta and Hamburg (Starkie, 2008). Furthermore, there has been some tendency to move in a step-wise fashion from single till to dual till incentive regulation at airports in Belgium, Hungary and France. Another important feature of price caps is that there is an incentive to reshuffle the price structure. The CAA adopted a cap defined on average revenue, which sets fewer incentives than a tariff basket approach to restructure charges. Hybrid price caps are sometimes adopted along the line of the UK system, but then practised in a way that reduces the incentives further via sliding scales. A sliding scale mechanism generally relates the price to a function of traffic growth, quality and/or inflation rates. This form of revenue sharing agreement allows airports to increase their charges under low growth scenarios but may even reduce charges under high traffic growth, such that overall airport revenue remains relatively stable. Many price capped airports, such as Hamburg and Vienna, have adopted a sliding scale mechanism, stabilizing revenues at a certain level irrespective of airport pricing. The power of incentive regulation was reduced in Paris by adopting the same structure and level of charges for Orly and Charles De Gaulle airports, although both differ in terms of excess demand and supply. Another way to reduce the power of incentive regulation is to officially adopt the UK system, but prevent prices from deviating from costs, as has occurred in India. Revenue sharing elements have been adopted for some time at a number of German airports. The regulator allowed airlines and airports to reach agreements on such contracts in coordination. At first sight this comes close to the policy of price monitoring in Australia and the threat of re-regulation, but the difference is that Germany lacks an independent regulator and there is no review of performance, as there is in Australia. Benchmarking has never been applied as a regulatory tool to develop yard stick competition. The heterogeneity of airports and lack of data has so far prevented airport regulators to follow up the promising examples of other public utilities like energy and water (Reinhold et al., 2009). Benchmarking has however been used to lessen the information asymmetry and to compare charges among groups of airports. Unfortunately the selection of airports reflects fairly low targets such that the incentives are weak. 3.3. Categorization The practice of airport regulation is changing and different forms are emerging, composed of diverse elements of incentive regulation. A ‘yes or no’ classification is not helpful, but somewhere a line must be drawn among the continuum of low to high powered regulation. In the empirical work in Section 5 we distinguish between pure price caps, hybrid price caps, revenue caps and light handed regulation. It may be the case that the power of regulation is lower compared to what could have been achieved and what has indeed been achieved in other industries like energy and water (Reinhold et al., 2009).
7
3.4. Medium to high powered regulation In Australia, prior to monitoring (see below) a price cap with a dual till was set, with an X efficiency level in the range of 1–5% (mostly in the range of 3–4%). Specific provisions passed on the costs of new investment, so that underinvestment could be avoided. Overall, the price cap was set ‘very tightly’ (Forsyth, 2008, p. 84). As this cap was set without reference to the costs of the airport, such a system comes close to a pure price cap. Malta set a dual till price cap with an X of 0.5 for the period 2002–2007. Charges have not been changed since 2007. This cap was not based on costs. Unlike Australia and the UK, the regulator of the Malta airport of which the government holds a share of 30%, is not independent. UK price cap regulation can be divided roughly into two phases. Up until 2002, price cap regulation tried to set strong incentives for efficiency by relatively high X-values over a period of 15 years. Thereafter the price cap was lessened in order to allow for costly investment, in particular at Heathrow. The Irish regulator also needed to handle the issue of investment. However, the publicly owned airport planned investments rather too early and on a substantial scale. Benchmarking the costs of terminals was one of the instruments to control for excessive investments (GAP, 2011). An assessment of Italian airports must consider the ongoing institutional changes. The current Italian regulation is a hybrid price cap system based on a ‘semi’ single till principle, since December 2008 (ENAC, 2007; 2008) on an airport by airport basis.2 A slightly different regulation is applied to Rome (with its two airports), with Milan and Venice-ENAC agreed on a dual till for a five year regulatory period in an individual bargaining process with the airports. In these cases the focus on regulation is very much on investment, for which targets are set and, if not met, fines are imposed.3 While the current regulation is clearly defined from 2008 onwards as a hybrid price cap, the picture is further complicated because there has been a freeze on airport charges (for levels and structure) in effect since 2000. Up to 2000, regulation was cost based, but for the period 2000–2008 the price freeze in nominal terms suggests a pure price cap that has no relation to costs (Sciandra, 2009). 3.5. Low powered regulation Compared to the Anglo-Saxon approach of regulation, the system for Hamburg and Budapest airports set lower incentives due to the additional hybrid elements. Hamburg airport was regulated by a sliding scale price cap based on a dual till. Overall this has decreased the level of charges substantially and to a larger extent than at other German airports with revenue sharing agreements. The hybrid price cap at Budapest lowers the level of charges almost by 20% in real terms for the period of 2006–2011. As the price cap does not regulate the price structure, route development programs with substantial reductions were introduced in a non-discriminatory manner and a low cost terminal with reduced charges was operated for a limited time (GAP, 2011). Incentive regulation for Austrian, Belgian, German and French airports is relatively weak. At Düsseldorf, Frankfurt and Hannover, a sliding scale was used at certain times to stabilize the revenues. As the sliding scale is inversely related to demand, it leads to rising charges in a crisis. Furthermore, it does not set effective incentives 2 At the end of 2011, the following airports signed the ‘contratto di programma’ agreement after being reviewed by ENAC: Bari, Bologna, Brindisi, Naples, Pisa and Palermo. In 2012, Catania, Cagliari and Olbia followed. 3 For example, ADR on the basis of its long-term investment plan for €12 billion was able to negotiate a new regulatory contract valid until the end of the concession in June 2044 (Gemina, 2013).
8
N. Adler et al. / Transport Policy 41 (2015) 5–15
for rebalancing the price structure to manage scarce capacity in periods of high demand. In Austria, the sliding scale formula has no productivity factor, so that productivity gains are not passed onto the users. The level of charges remains relatively high and the incentives for productivity gains are low. At these airports, the structure of charges is left unregulated. The case of Aeroports de Paris highlights this type of inefficiency. Here the same level and structure of prices apply for Charles de Gaulle and Orly airports although the latter has excess demand and the former has excess capacity. A rather heavy handed type of hybrid price cap regulation has been implemented at Delhi, Mumbai and eight Indian airports. The price cap is based on costs and the price structure is set by the regulator. Brussels is regulated by a traditional cost based system with a single till evolving over time into a dual till system. Benchmarking is only used as an additional tool to provide information. As the legally determined set of reference airports are all airports with relatively high charges, the incentives for cost efficiency are rather mild (GAP, 2011). At Copenhagen there are three stages of regulation which have been evolving as a response to the stepwise privatization (25% in 1994, another 24% in 1996 and an additional 17% in 2000). In the phase up to 2003, charges were rather loosely regulated in such a way that efficiency gains were kept by the airport, but it remained unclear for how long (Wolf, 2003). For the period 2003–2005, a hybrid price cap was introduced with an annual increase of 2.75% of charges and a cap for the period 2006–2008 with a reduction of 3% in 2006 followed by an increase of 1% in 2007 and 2008. From 2008, regulation was reformed under the principle that the airport and its users should negotiate, and only if no agreement is reached, regulation of a hybrid price cap with a mixed till sets in. For the period 2009–2010, an increase of 4.2% in charges was allowed and for 2010–2015 a price cap with CPI plus 1% was agreed upon. The future will show whether the dependent regulator will be in a position to act as a mediator. 3.6. Light handed regulation (LHR) In 2002 the Australian government decided not to prolong the price cap approach-rather establish LHR. The Australian Consumer and Competition Commission (ACCC), an independent agency, monitors prices, costs and profits of aeronautical services and aeronautical related services of the airports of Brisbane, Melbourne, Perth, Sydney Kingsford Smith, and Adelaide, (Forsyth, 2004). In 2006 and 2012 the Australian Productivity Commission reviewed the airports' performance positively and recommended continuing LHR. We classify LHR as a form of incentive regulation. It is less high powered than pure price cap regulation, in as much as the regulated airport has the ability to increase its prices, for example, if were costs to rise. However pricing is not unconstrained, because the airport management must convince the monitoring or regulatory body that there has been no abuse of market power. In the normal course of events, if it is possible to reduce costs, profits will rise, hence there is a strong incentive to minimize costs on an ongoing basis. We have also classified nondesignated UK airports as being subject to the credible threat of regulation and therefore as being subject to LHR.4 4
David Starkie suggested to us that the non-designated UK airports should be classified as not-regulated airports because since the airport Act of 1986 no other airports have been designated and it would still have required the government to pass the necessary instrument. While we think that this might reflect well the current situation we prefer our classification for the past because it seems to reflect better the facts that the behavior was more carefully monitored than a normal industry. In response to demands of parliamentarians to regulate the BAA Scottish
In addition to the countries covered in Table 1, a number of other countries have implemented some forms of IR, though detailed information about these is not readily available. These include Portugal, South Africa and Latin American countries – lack of data precludes these countries being included in the empirical study in Section 5.
4. What do we expect from incentive regulation at Airports? It is useful to set out what may be expected from incentive regulation at airports. This is useful when assessing whether it has been successful and whether problems have emerged. Several aspects of performance are relevant including allocative, productive and long run efficiency, as well as quality, and profitability. 4.1. Allocative efficiency Allocative efficiency can always be an issue, even though in the incentive regulation literature it is less of a concern than other aspects. First, there is the relationship of price relative to cost. With regulation, prices are likely to be close to average cost. With non-congested airports, prices may be above marginal cost, but permitted by the regulator to be close to average cost (a first best solution will involve losses). With congested airports subject to excess demand, prices may be kept lower than marginal cost (to the efficient rationing price) but the presence of a slot allocation system may ensure that prices are efficient (most congested regulated airports possess a slot system). Second, there is the issue of the price structure. With noncongested airports given a budget target, Ramsey pricing would be in order. With price caps, there are some incentives for regulated firms to adopt Ramsey pricing, but these incentives are not always present (Giuletti and Waddams Price, 2000). Nevertheless most uncongested airports utilize weight-based charging, which is an approximation to Ramsey pricing. Efficient pricing of airports subject to excess demand should have a uniform pricing structure. A third allocative issue is that of slots. Most congested, regulated airports have slot systems in place. Charges are lower than market clearing levels, and slots take up the allocative function – thus there is a question of how well these slots are allocated. Typically, regulators do not handle slot markets, rather alternative arrangements such as grandfather rights are in place. A final allocative aspect concerns cross subsidies. Airports have the ability to generate non-aeronautical revenues which can be used to cross subsidize operations. Most regulated airports tend to set average prices close to costs thus there is not likely to be a major issue of excessive pricing, although prices may be above marginal cost which cost recovery will require. However some airports, subject to weak incentives, operate with sliding scales. Such scales will have a small negative effect on efficiency, since prices go up when demand falls, and there is more likely to be excess capacity. The price structure is more important than the price level. Most airports operate with a weight-based landing fee structure, possibly combined with a passenger charge. If there is available capacity at the airport, this would be a fairly efficient structure. However, the case for weight(footnote continued) airports BAA voluntarily adopted a price cap of RPI-3 for three years in 1993. A year later the designation of Belfast International Airport and Belfast City played an important role in the attempted merger of the two airports. In 2000 the CAA rejected the proposal of Easyjet to designate Luton Airport. Kunz (2000) and Wolf (2003) see this as sufficient evidence that UK uses the threat of reregulation as part of their regulation and argue that this is credible while the threat for New Zealand airports was not.
N. Adler et al. / Transport Policy 41 (2015) 5–15
9
Table 1 Types of incentive regulation. Airports by country
Australia Adelaide, Brisbane, Melbourne, Perth, Above airports plus Sydney Austria Vienna Belgium Brussels Denmark Copenhagen France Aeroports de Paris Germany Düsseldorf Frankfurt Hamburg Hannover Hungary Budapest
Italy All major airports Aeroporti di Milano Aeroporti di Roma Naples, Venice Other major Italian airports India Delhi, Mumbai plus eight airports Ireland Dublin Malta Malta Airport
UK Heathrow Gatwick Stansted Manchester
Time period
Type of regulation
Independent regulator
Strength/weaknesses
1997 to Price cap with dual till 2001– 02 Since Light handed regulation 2002
ACCC
Low price cap, but instability in crisis with exit of Ansett Airlines
ACCC and Productivity Commission
Strong incentives to reduce costs and differentiate prices subject to assessment of an independent regulator
Since 1998
Revenue cap
Yes, DOT
Stabilizes revenues at high level
Since 2004
Cost based with benchmarking elements
No
Peer group of airports are relatively high cost airports. Cost based thinking prevails
Since 1995
Light handed price cap on a mixed dual till with regulator as arbitrator
No
Long record of incentive agreements. Role of dependent arbitrator so far not tested.
Since 2005
Hybrid revenue cap with bonus malus investment and quality regulation
No
Regulatory capture and inefficient charges
2004– 2009 2002– 2006 Since 2000 2003– 2008
Revenue cap
No
Regulatory capture and instability
Revenue cap
No
Regulatory capture and instability
Revenue cap on dual till
No
Stable and accepted by stakeholders
Revenue cap
No
Regulatory capture and instability
Since 2006
Price cap with quality regulation
Yes, with full privatization in 2011. Na- Tight price cap. Initial conflict of interest retional Transport Authority Civil Aviation solved in 2011. Directorate
2000– 2008 Since 2012
Price freeze
No
Hybrid price cap with dual till, bonus malus investment and quality regulation
No
Since 2009
Hybrid price cap with mixed till
No
Since 2011
Single till hybrid price cap with regulated Yes, in 2008Airports Economic Regprice structure will be developed ulatory Authority of India was established
Price cap is too much cost based. Independent regulator.
Since 2001
Hybrid single till price cap
Yes, Irish CAA
Hybrid price cap. Independent regulator
Since 2002
Price cap on dual till with X ¼ 0.5 for period 2002 to 2007. Since then no changes of charges
No
Strong incentives as cap is not cost based. Role of dependent regulator so far not tested
Since 1986 Since 1986 Since 1986 Till 2008
Price cap
Yes, CAA UK
Hybrid price cap with investment regulation
Pure price cap with strong incentives, but uncertainty about institutional reform Hybrid price cap and regulated investments. Danger of regulatory capture Hybrid price cap Danger of regulatory capture
Price cap
Hybrid price cap
Price cap
Hybrid price cap
Price cap
Hybrid price cap
based charges is not present if the airport it experiencing excess demand, since a uniform charge would be more efficient. There is not much evidence of airports adopting uniform charging. With the London airports, charges are more uniform than most, though weight-based charging remains an element. This may be some evidence of the positive effect of incentive regulation (Forsyth and Niemeier, 2008). In most airports, slot markets are very imperfect. There is reasonably open slot trading for the London airports, but not for most other airports. There is no formal link between slot market
arrangements and regulation. However it may be significant that airports which have the strongest IR are also those which have the most open slot markets. 4.2. Productive efficiency We would expect that productive efficiency would increase as a result of a move to incentive regulation. Greater incentives to produce efficiently are the core rationale of incentive regulation. Given that productive efficiency is relatively easy to measure,
10
N. Adler et al. / Transport Policy 41 (2015) 5–15
using total factor productivity, stochastic frontier analysis or data envelopment analysis measures, it could be hoped that the gains in efficiency would be identifiable, (as discussed in Section 5). There are only a few studies assessing the effects of incentive regulation (Liebert and Niemeier, 2013). This is due to the fact that in the first years after privatization of airports, the emphasis was on testing the effects of different ownerships forms. These analyses suggested that privatization increases efficiency only if airports are fully privatized, such as the UK and Australia. Partially privatized airports were found to be less efficient, even compared to public airports (Oum et al., 2006). This is a rather disappointing result for European airports on the mainland, where partial privatization is very often the typical form of privatization, with a minority share for the private investors. Recent studies analyzing the effects of ownership, regulation and competition, such as Adler and Liebert (2014) based on a comparison of 51 airports, as well as Assaf and Gillen (2012) based on a study of the UK airports, show that competition from nearby airports leads to superior results than regulation. However, at a number of airports, competition is not effective. Adler and Liebert show that cost based regulation leads to lower efficiency than incentive regulation. Assaf et al. argue that regulated airports tend be more efficient than non-price regulated airports and that investment programs at Heathrow, Gatwick and Stansted tend to lower efficiency scores. However, while these studies show that there is some evidence that incentive regulation increases efficiency and that it is superior to cost based regulation, it is equally clear that not all price caps are equal, hence in this research we analyze the forms of incentive regulation in order to understand their relative effects. 4.3. Quality By contrast, incentive regulation is not likely to be helpful in ensuring quality service: in fact it may well reduce quality. It encourages the regulated firm to increase profits, and one way that it can do this is by allowing quality to fall. This problem can be addressed in part or fully through specific regulation of quality, or possibly monitoring of quality (Rovizzi and Thompson, 1992). Quality of service has been an issue in the UK, and after a recent report (Cave, 2009), the government raised the possibility of changes to regulation to give quality a greater weight. 4.4. Investment and long run efficiency Incentive regulation does not have a clear role when it comes to investment and long run efficiency. Indeed, a dynamic theory of incentive regulation has yet to be developed and most models are essentially static. There is some analysis of regulation and investment, in general and for airports, but it is at an early stage of development (Starkie, 2006). Regulators often use a cost based approach to handle additions to capacity, even when they are using formally pure forms of incentive regulation in the short run. The Australian Government ACCC (2001) considered that price caps would lead to underinvestment, and brought in cost passthrough arrangements to encourage necessary new investment. Attempts have been made to develop a form of incentive regulation which includes investment opportunities and an example is that suggested for the London airports. The CAA in the UK attempted to link investment and output into its price cap formula for the London airports (Andrew and Hendriks, 2004). This however did not meet with the approval of competition regulators and the CAA was forced to drop the proposal. Some have suggested that regulation might give rise to excessive investment (rather like a cost plus regulated airport and the Averch and Johnson effect (Starkie, 2006)). Thus there is a complex empirical issue to be resolved.
4.5. Non-aeronautical revenues Another aspect of airport performance concerns the incentive for the airport to develop non-aeronautical revenues, which is connected to whether the airport operates under dual or single till. Whilst incentive regulation is compatible with either a single or a dual till, the two approaches have conflicting implications for different aspects of efficiency. The expectation might be that a dual till approach better encourages productive efficiency, since the airport has a greater incentive to produce unregulated, commercial services. Whilst there has been much theoretical discussion on the merits of the different approaches, there is been little empirical work. A recent study is that of Adler and Liebert (2014) who report that airports which operate under a dual till are more productively efficient. 4.6. Profitability and its variance Finally, an aspect of performance which could be important is that of the airport's profitability and its variability. One would not necessarily expect that an incentive regulated airport be more or less profitable than one subject to other forms of regulation (or no regulation). However there could be a difference in terms of the variability of its profit. Incentive regulation places a greater risk on the regulated airport; under strong demand the airport may expect high profits, whereas under weak demand, the airport may record low profits or losses. By contrast, airports subject to other forms of regulation, particularly those with sharing rules and sliding scales, should record more even profitability, since the regulator tries to set prices close to average costs. Indeed, performance in terms of profitability and its variability may be a good indicator of whether the airport is subject to genuine incentive regulation.
5. Empirical two-stage study of productive efficiency As noted, there is little by way of empirical assessment of the impact of varying levels of incentive regulation on airport management behavior. In this section we provide an assessment based on a DEA estimation of the relative efficiency of airports in the sample, and relate the results to the forms of regulation and other variables of interest, such as whether the regulator is independent or not. Our primary interest is the effects of regulation, not the determinants of individual airport efficiency, thus we use an existing model (Adler et al., 2013) of efficiency, and analyze how regulation affects performance. In order to estimate relative efficiency, we apply a non-oriented, variable returns to scale, bound adjusted measure that minimizes labor and other operating costs and maximizes nonaeronautical revenues given declared runway capacity as a nondiscretionary input and passengers, air traffic movements and cargo as outputs over which airport management have little control. Consequently, this represents a short-term managerial efficiency measurement which defines approximately 8% of the airports in the dataset as relatively efficient. Summary statistics of the 58 airports in the dataset and the number of years for which data was available is presented in Appendix A. The unbalanced data spans the years 1990–2010 and of the 58 airports, 38 are drawn from the United Kingdom, Germany and Italy. The data includes all countries covered in Table 1, except Hungary, India, Ireland, and the two Paris airports (due to data limitations). For the second stage analysis, the explanatory variables defining the regulatory structure include revenue caps (RevCap), hybrid and pure price caps and light handed regulation. We also explore whether an independent regulator influences relative productive
N. Adler et al. / Transport Policy 41 (2015) 5–15
11
Table 2 Second-stage regression analyses explaining relative efficiency estimates. Fixed effects (within) regression
Truncated regression [0,1)
Model 1 Dependent Variable ¼ DEA estimate RevCap Light Hybrid Hybrid/light Pure Independent NA Cap1 Cap2 R2/log likelihood Number of observations
Model 2
Coefficient
t
0.0397 0.0909 0.1218
1.89 0.0386 2.15 2.67 0.1021 3.49 0.1733 1.27 0.0534 6.33 0.0768 4.95 0.0641 8.78 0.2463 Within ¼ 0.2414, between¼ 0.1729, overall¼ 0.2078 707
0.1942 0.0856 0.0786 0.0646 0.2465 Within ¼0.2426, between¼ 0.1613, overall¼ 0.1879 707
efficiency estimates. Furthermore, we include three additional dummy variables, in order to compare the results to those in the literature such as Oum et al. (2006) and Adler and Liebert (2014). Hence we have added a dummy equal to one if the airport earns at least 50% of revenues from non-aeronautical sources (NA), and two dummies to capture capacity utilization. Cap1 receives a score of 1 if the airport is utilized between 50% and 80% of their declared runway capacity on an annual basis and cap2 refers to airports that are serving air traffic movements at a level above 80% of their declared capacity, which includes London Heathrow, Frankfurt, occasionally London Gatwick and Brussels airport at the beginning of the 1990s. In Table 2 we present the results of a fixed effects regression model, which enables an unbiased analysis of the explanatory variables. We also include truncated regressions to test for robustness of the results given that the relative efficiency scores estimated in the first stage are capped at 1, which defines the Pareto frontier. Models (1) and (3) test all four levels of incentive regulation and models (2) and (4) test for weak (RevCap), medium (hybrid price caps and light handed monitoring) and strong (pure price caps) powered incentive regulation. Finally, we include time dummies in order to ensure that these effects do not confound the results. The time dummies, which have been excluded from Table 2 for purposes of parsimony, suggest that airports were relatively more efficient in the 1990s as compared to the 2000s, some of which may be explained by the increased security regulations and their respective costs that developed as a result of 9/11. The variables often analyzed in the literature, namely capacity utilization (cap1 and cap2) and the percentage of non-aeronautical revenues earned out of the total revenues of an airport (NA), behave as expected from the literature. Airports earning more than 50% of their revenues from the landside are approximately 8% more relatively efficient and the higher the capacity utilization, the higher the expected relative efficiency levels at the airport. The results for the incentive regulation also proved to be significant and demonstrate that, in comparison to the cost plus regulatory system in place at half the airports in the sample, incentive regulation leads to higher levels of relative productive efficiency. Revenue caps increase relative efficiency by approximately 4% (although the result is not significantly different to that of the cost based price caps that represent the baseline), light handed regulation by approximately 9%, hybrid price caps by approximately 12% and pure price caps, the highest power form of incentive regulation, by around 19% in this dataset. Interestingly, whether the airport is regulated by an independent regulator or not would seem to have little impact on efficiency because the coefficient,
Coefficient
t
Model 3
Model 4
Coefficient z
Coefficient z
1.84 0.0310 0.1255 0.1375 2.50 3.38 0.1997 0.91 0.0916 6.26 0.0993 4.92 0.0763 8.77 0.4042 925.01 701
1.53 0.0308 2.72 3.05 0.132 3.84 0.1941 1.58 0.0983 8.02 0.0986 5.69 0.0768 10.4 0.4046 924.96
1.52
3.15 3.95 1.8 8.08 5.75 10.42
701
though positive, is strictly insignificant. In summary, there appears to be evidence that incentive regulation has a positive effect on productive efficiency and pure price caps appear to have the strongest impact on short-term productive efficiency.
6. Conclusions The stated objective of incentive regulation of airports (along with other industries) is to increase the efficiency, and in particular, the productive efficiency, with which they operate. Has this objective been met? The objective of this paper is to assess this, taking into account the particular characteristics and features of airports. The evidence is encouraging, though clearly not conclusive. The empirical study reported in Section 5 suggests that a move towards incentive regulation is consistent with increased productive efficiency. These results need to be regarded as a first assessment of the issue. There are several limitations to the study: the dataset is unbalanced over time; there are problems of categorization of what constitutes incentive regulation and what does not; and there are distinct problems in developing a comparable data set for efficiency measurement. Nonetheless, the results are promising. Incentive regulation was designed primarily to increase productive efficiency especially in the short run. There are other aspects to performance – allocative efficiency, quality performance, and long run investment performance. Incentive regulation is not necessarily positive for these aspects of performance. It may improve allocative efficiency, though not necessarily; it may increase the efficiency with which an airport achieves efficiency in nonaeronautical services (depending on the form of regulation); and it may well lead to a reduction in quality. Most of these aspects are more difficult to assess, especially quantitatively, than productive efficiency. This survey indicates that performance in these aspects has been consistent with expectations. So far, at least in terms of the airport industry, the economist's recommendations have been for incentive regulation, rather than for the alternatives such as cost based regulation. This recommendation has been based on theoretical properties, along with empirical results from other industries. The empirical results from our study suggest that this recommendation is well based, and that incentive regulation of airports enhances productive and overall efficiency.
12
N. Adler et al. / Transport Policy 41 (2015) 5–15
Table A1 Descriptive statistics of airport dataset for benchmarking. Airport code
Country
# Years of data
Years
Staff costs
Other operating costs
Non-aeronautical revenue Output
Declared runway capacity Non-discretionary input
PAX
16,372,348
34
2,570,224
88,089
146,279,024 357,230,918
322,480,971
108
41,159,194
406,262 1,376,930
9,285,088
21,038,966
21,895,277
32
4,460,087
46,632
45,928
14,280,879
25,824,549
14,417,160
21
3,872,273
47,777
120,521
27,908,152
41,957,402
47,514,254
40
7,022,637
92,649
17,261
28,591,134
16,225,036
32,834,903
32
2,071,525
50,865
51,049
18,588,783
26,740,211
23,341,322
24
3,753,799
56,071
25,597
14,037,157
12,743,026
12,340,050
30
1,797,959
35,963
1,538
10,611,847
18,527,500
36,730,143
25
5,002,957
57,282
5,356
35,959,993
91,370,754
92,721,069
62
15,093,798 244,536 537,172
14,733,750
26,064,875
39,840,250
45
3,474,271
64,115
34,883
7,054,294
7,026,881
6,605,406
28
1,476,518
25,897
8,120
4,530,473
12,440,878
6,706,287
10
2,455,491
27,570
4,809
93,499,087
102,982,204
67,745,416
52
8,046,062
132,322 550,091
75,918,962
50,152,369
117,014,818
80
17,101,851
251,797 328,843
10,324,185
21,711,589
14,747,676
26
1,724,640
29,216
718
12,672,237
17,784,744
10,754,577
42
1,597,229
30,025
94
101,643,841
118,889,049
97,453,846
39
15,173,211
178,168
59,677
10,149,416
9,712,208
6,521,907
9
1,604,733
27,325
1,728
622,748,900 392,219,289
372,361,727
75
42,363,952 412,403 1,545,454
25,338,762
36,630,483
48,992,034
32
7,781,325
86,758
6,805
9,393,369
11,856,200
10,647,278
18
1,079,835
18,815
5,856
38,876,609
40,743,622
52,209,576
38
7,752,420
121,836 62,317
42,864,248
38,255,528
40,452,757
49
4,519,477
68,790
67,002,211
73,089,924
58,721,338
45
9,494,832
127,790 32,856
8,072,966
7,363,700
8,347,156
20
1,583,944
27,880
7,124
9,873,193
32,003,876
13,911,194
29
2,121,599
63,873
667
13,933,805
26,665,667
12,265,702
36
2,135,663
35,690
7,513
121,817,641
205,797,847
303,911,785
50
31,876,429 247,531 212,574
259,564,477 641,091,745
918,899,347
87
64,950,686 462,718 1,203,856
19,023,039
12,196,601
14,208,494
15
1,281,379
40,194
15,092
7,704,786
12,848,942
11,881,892
30
2,450,136
33,506
26,222
22,650,829
53,320,803
56,668,206
36
7,714,325
69,125
27,862
21,555,574
38,998,616
41,513,709
51
6,533,949
119,809 36,849
74,831,245
159,105,675
194,648,448
57
20,605,230 195,461 125,234
10,159,340
35,979,170
69,066,229
50
18,892,231 170,657 8,215,969
Inputs
ABZ AMS BFS BGY BHX
United Kingdom Netherlands United Kingdom Italy
21 10 9 10
BLL
United Kingdom Denmark
9
BLQ
Italy
10
BRE
Germany
10
BRS
8
BRU
United Kingdom Belgium
21
BSL
France
8
BTS
Slovakia
8
CAG
Italy
10
CGN
Germany
13
CPH
Denmark
21
DRS
Germany
13
DTM
Germany
10
DUS
Germany
21
FLR
Italy
10
FRA
Germany
18
GLA
12
GOA
United Kingdom Italy
10
GVA
Switzerland
21
HAJ
Germany
17
HAM
Germany
21
LBA
United Kingdom United Kingdom Germany
18
LCY LEJ LGW LHR LJU LPL LTN LYS MAN MEL
United Kingdom United Kingdom Slovenia United Kingdom United Kingdom France United Kingdom Australia
21
12 5 13 13 4 21 13 14 13 10
1990– 2010 1998– 2007 2002– 2010 2000– 2009 1990– 2010 2002– 2010 2000– 2009 1998– 2007 2002– 2009 1990– 2010 2002– 2009 2003– 2010 2000– 2009 1998– 2010 1990– 2010 1998– 2010 2001– 2010 1990– 2010 2000– 2009 1990– 2007 1999– 2010 2001– 2009 1990– 2010 1990– 2007 1990– 2010 1990– 2010 1999– 2010 1998– 2002 1998– 2010 1998– 2010 2004– 2007 1990– 2010 1998– 2010 1996– 2009 1998– 2010 1999– 2008
11,509,119
13,560,764
ATM
Cargo
Non-discretionary output
5,910
8,816
N. Adler et al. / Transport Policy 41 (2015) 5–15
13
Table A1 (continued ) Airport code
Country
# Years of data
Years
Staff costs
Other operating costs
Inputs
MLA
Malta
6
MME
1
MRS
United Kingdom France
12
MUC
Germany
13
NAP
Italy
10
NCE
France
12
NCL
21
NUE
United Kingdom Germany
13
OSL
Norway
9
PMO
Italy
10
PRG
3
PSA
Czech Republic Italy
10
RIX
Latvia
9
SOU
12
STR
United Kingdom United Kingdom Germany
13
SZG
Austria
9
TLL
Estonia
9
TRN
Italy
10
VCE
Italy
10
VIE
Austria
13
ZRH
Switzerland
13
STN
11
Non-aeronautical revenue Output
Declared runway capacity Non-discretionary input
PAX
ATM
Cargo
Non-discretionary output
2005– 2010 2002
10,142,025
19,892,769
15,839,484
15
2,965,178
28,748
17,302
6,556,432
5,010,381
4,806,309
14
689,000
16,000
1,000
1998– 2009 1998– 2010 2000– 2009 1998– 2009 1990– 2010 1998– 2010 1999– 2007 2000– 2009 2005– 2007 2000– 2009 2002– 2010 1999– 2010 1998– 2010 1998– 2010 2002– 2010 2002– 2010 2000– 2009 2000– 2009 1998– 2010 1998– 2010
19,459,115
36,153,950
40,161,478
33
6,353,310
90,905
55,276
259,922,880 367,822,664
364,115,866
87
27,447,589
354,135 193,153
12,409,685
17,591,268
15,676,097
28
4,791,686
55,360
21,761,725
64,015,750
64,863,792
49
9,422,466
177,082 16,885
16,290,627
11,898,456
19,452,568
30
3,453,409
44,228
6,390
36,639,374
31,953,076
31,590,056
30
3,549,069
60,323
16,281
36,650,795
80,692,721
137,722,928
65
15,182,364
195,713 81,222
14,432,026
18,560,722
11,189,681
20
3,886,318
43,531
38,862,483
46,063,406
55,535,100
39
11,549,689 167,067 57,628
15,020,757
14,673,942
9,324,162
14
2,535,213
29,488
11,423
11,686,333
6,599,865
8,167,522
27
2,483,934
40,475
10,538
8,755,931
9,187,506
12,349,084
22
1,426,467
37,026
148
58,318,786
88,740,012
137,101,006
47
18,715,728 166,160 365,185
56,152,315
92,029,755
69,654,763
40
8,636,991
129,984 20,978
17,022,727
13,524,445
9,352,647
20
1,608,225
21,182
232
4,616,556
5,715,778
7,337,222
20
1,280,252
25,944
14,722
11,332,354
19,591,109
18,270,833
26
3,095,216
46,012
14,773
19,081,176
29,239,348
23,018,853
29
5,656,134
70,374
23,024
194,628,466 109,584,868
109,097,877
64
14,936,721 215,135
78,690,631
129,397,326
150,973,435
70
20,071,672 257,692 423,019
7,774
5,155
182,812
Table B1 Data on regulation per airport Cost based
RevCap
Hybrid
Light
Pure
Independent regulator
NA
Cap1
Cap2
ABZ AMS
0 1
0 0
0 0
1 0
0 0
1 1
0 1
0 0
BFS BGY
0 0
0 0
0 2009
1 0
1 0
0 0
0 0
BHX BLL BLQ BRE BRS BRU BSL BTS CAG
0 1 0 1 0 1 1 1 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 2009
1 0 0 0 1 0 0 0 0
1 0 0 0 1 0 0 0 0
2009–2010 (2) 1 0 0 2005–2009 (5) 1990–1991 (2) 2002–2003 (2) 0 0
0 0 0 0 0 1993–2008 (13) 0 0 2008
0 0 0 0 0 1990–1992 (3) 0 0 0
CGN CPH
1 1990–1994 (5)
0 0
0 1990/1998 (2)
0 1996–2001 (6)
0 0
1
0
0 1995–2002, 2009– 2010 (10) 0
0 0
DRS
0 2003– 2008 (6) 0
0 2000– 2008 (9) 0 0 1 0 0 0 0 0 2000– 2008 (9) 0 0
0 98, 2000–2001 (3) 2005–2010 (6) 0
0
0
0
0
0
14
N. Adler et al. / Transport Policy 41 (2015) 5–15
Table B1 (continued ) Cost based
RevCap
Hybrid
Light
Pure
Independent regulator
NA
Cap1
Cap2
0 2005– 2008 (4) 0
0 0
0 0
0 0
0 0
0 0
0 1
0 0
FLR
1 1990–2004, 2009– 2010 (17) 0
2009
0
0
0
2000–2002,2005,2007 (5)
0
FRA
1990–2002 (13)
0
0
2000– 2008 (9) 0
0
0
1990–1995, 1997 (7)
0 0 0 0
1 0 0 0
0 0 0 0
1 0 0 0
2008 2001 0 0
0 0 2000–2010 (11) 0
1996, 1998– 2007 (11) 0 0 0 0
0
0
0
0
0
0
0
0 0 0 1
1 1 0 0
0 0 0 0
1 1 0 1
2006–2010 (5) 0 0 1998–2008 (11) 1998–2007 (10) 0 2005–2010 (6) 1 1996–1999 (4) 2005, 2008– 2010 (4) 2008
0 0 0 1998–1999, 2001–2005, 2007–2010 (11) 2009–2010 (2)
0 0 0 2000, 2006 (2)
DTM DUS
LBA LCY LEJ LGW
0 0 1 0
2003– 2007 (5) 0 0 0 2003– 2007 (5) 2000–2010 (11) 0 0 0 0
LHR
0
0
1
0
0
1
LJU LPL LTN LYS MAN
1 0 0 1 0
0 0 0 0 0
0 1 1 0 2007–2010 (4)
0 0 0 0 0
0 1 1 0 1
MEL
0
0
0 0 0 0 1998– 2006 (9) 0
2003–2008 (6)
2003–2008 (6)
MLA 0 MME 0 MRS 1
0 0 0
0 0 0
0 0 0
1999– 2002 (4) 1 0 0
MUC 1 NAP 0
0 0
0 2009
0 0
NCE NCL NUE OSL PMO
1 0 1 1 0
0 0 0 0 0
0 0 0 0 2009
0 1 0 0 0
PRG PSA
1 0
0 0
0 2009
0 0
RIX SOU STN STR SZG TLL TRN
1 0 0 1 1 1 0
0 0 0 0 0 0 0
0 0 1 0 0 0 2009
0 1 0 0 0 0 0
VCE
0
0
2009
0
VIE ZRH
0 1
1 0
0 0
0 0
0 2000– 2008 0 0 0 0 2000– 2008 (9) 0 2000– 2008 0 0 0 0 0 0 2000– 2008 (9) 2000– 2008 (9) 0 0
GLA GOA GVA HAJ
2010 1 1 1990–1996, 1998– 2002 (12) HAM 19901999 (10)
Acknowledgments This paper arose from a research project entitled “Airport Benchmarking by Economic Regulators” produced on behalf of the Netherlands Office of Transport Regulation (NMa). The data were provided by the German Airport Performance research project, which was kindly supported by the German Federal Ministry of Education and Research (see www.gap-projekt.de). We are also grateful to the Australian Research Council and the Recanati Foundation for their support. We thank David Starkie for his ever helpful comments. All errors are our own.
0 1 0
0 0 0 0 1998–2000, 2003–2007 (8)
1998–2008 (11) 0 0 0 0 0
2001, 2005–2008 (5)
0
0 0 2000
0 0 0
0 0
0 0 1998,2000– 2002 (4) 0 0
1999–2010 (12) 0
0 0
0 1 0 0 0
0 2008,2010 (2) 0 0 0
1998–2003, 2006–2008 (9) 0 0 2007 0
0 0 0 0 0
0 0
0 0
1 0
0 0
0 1 1 0 1 0 0
0 0 2003 0 1998–2006 (7) 1999, 2002–2008 (8) 0 2004–2008 (5) 0 0 0 0 0 0
0 0 0 0 0 0 0
0
0
0
0
1 0
0 0
2004–2010 (7) 1998–2004, 2008–2010 (10)
0 0
Appendix A See Appendix A1
Appendix B See Appendix B1 References Adler, N., Liebert, V., 2014. Joint impact of competition, ownership form and
N. Adler et al. / Transport Policy 41 (2015) 5–15
economic regulation on airport performance and pricing. Transp. Res. Part A 64, 92–109. Adler, N., Liebert, V., Yazhemsky, E., 2013. Airport benchmarking from a managerial perspective. Omega – Int. J. Manag. Sci. 41 (2), 442–458. Andrew, D., Hendricks, N., 2004. Airport regulation in the UK In: Forsyth, P., Gillen, D., Knorr, A., Mayer, O., Niemeier, H.M., Starkie, D. (Eds.), The Economic Regulation of Airports. Ashgate, Aldershot. Assaf, A.G., Gillen, D., 2012. Measuring the joint impact of governance form and economic regulation on airport efficiency. Eur. J. Oper. Res. 220, 187–198. Australian Competition and Consumer Commission, 2001. Price Regulation of Airport Services, Submission to the Productivity Commission's draft Report October 2001 〈www.pc.gov.au〉. Baron, D.P., Meyerson, R.B., 1982. Regulating a monopolist with unknown costs. Econometrica 50 (4), 911–930. Beesley, M.E., Littlechild, S.C., 1989. The regulation of privatized monopolies in the United Kingdom. RAND J. Econ. 20 (3), 454–472. Cave, M. Ch., 2009. Report of the Independent Panel on Airport Regulation, DfT January. Civil Aviation Authority (CAA), 2000. The Use of Benchmarking in the Airport Reviews, Consultation Paper, 〈http://www.caa.co.uk/docs/5/ergdocs/benchmark ing(caa122000).pdf〉 (accessed 25.07.13). ENAC, 2007. Capitolato relativo all ‘affidamento dei servizi per la redazione di uno studio sullo Sviluppo futuro della rete aeroportuale nazionale quale componente strategica dell’ organizzazione infrastrutturale del territorio, Rome, Italy. ENAC, 2008. Regolamento per la costruzione e l’esercizio degli aeroporti, Rome, 21 9 2008. Forsyth, P., 2004. Replacing regulation: airport price monitoring in Australia In: Forsyth, P., Gillen, D., Knorr, A., Mayer, O., Niemeier, H.-M., Starkie, D. (Eds.), The Economic Regulation of Airports: Recent Developments in Australasia, North America and Europe. Ashgate, Aldershot. Forsyth, P., 2008. Airport policy in Australia and New Zealand: privatization, lighthanded regulation, and performance In: Winston, C., de Rus, G. (Eds.), Aviation Infrastructure Performance A Study in Comparative Political Economy. Brookings Institution Press, Washington. Forsyth, P., Niemeier, H.-M., 2008. Price regulation and the choice of price structures at busy airports In: Czerny, A., Forsyth, P., Gillen, D., Niemeier, H.-M. (Eds.), Airport Slots. International Experiences and Options for Reform. Ashgate, Aldershot. Gemina, 2013. ADR's New Concession Agreement, ERA to Come into Effect and 2012 Traffic Performance. 〈http://www.atlantia.it/gemina/files/2013/01/04/ENG% 20notice%20presentazioneADR4gennaio2013.pdf〉. Gomez-Ibanez, J.A., 2003. Regulating Infrastructure. Cambridge, Harvard UP.
15
German Aviation Performance (GAP,2011) Airport Benchmarking by Economic Regulators, Study for the Netherlands Office of Transport Regulation, 〈www. gap-project.de〉. Green, R., Rodriguez-Pardina, M., 1999. Resetting Price Controls for Privatized Utilities: A Manual for Regulators. World Bank, Washington, D.C. Giulietti, M., Waddams Price, C., 2000. Incentive Regulation and Efficient Pricing: Empirical Evidence, Research Paper Series No. 00/2, Centre for Management under Regulation. Warwick Business School. Kunz, M., 2000. Regulierungsregime in Theorie und Praxis In: Knieps, G., Brunekreeft, G. (Eds.), Zwischen Regulierung und Wettbewerb – Netzsektoren in Deutschland. Physica, Heidelberg. Laffont, J.-J., Tirole, J., 1986. Using cost observation to regulate firms. J. Polit. Econ. 94 (3), 614–641. Laffont, J.-J., Tirole, J., 1993. A Theory of Incentives in Procurement and Regulation. MIT Press, Cambridge. Liebert, V., Niemeier, H.-M., 2013. A survey of empirical research on the productivity and efficiency measurement of airports. J. Transp. Econ. Policy 47, 157–189. Littlechild, S.C., 1983. Regulation of British Telecommunications' Profitability. HMSO, London. OECD, 2005. OECD Guiding Principles for Regulatory Quality and Performance, Paris, Organization for Economic Cooperation and Development. Oum, T.H., Adler, N., Yu, C., 2006. Privatization, corporatization, ownership forms and their effects on the performance of the world's major airports. J. Air Transp. Manag. 12 (3), 109–121. Reinhold, A., Niemeier, H.-M., Kamp, V., Müller, J., 2009. An evaluation of yardstick regulation for European airports'. J. Air Transp. Manag. 16 (2), 74–80. Rovizzi, L., Thompson, D., 1992. The regulation of product quality in the public utilities and the Citizen's Charter. Fiscal Stud. 13 (3), 74–95. Sciandra, L., 2009. Regulatory assessment of Italian airports: some critical issues. Mercato Concorrenza Regole 1, 51–78. Shleifer, A., 1985. A theory of yardstick competition. RAND J. Econ. 16 (3), 319–327. Starkie, D., 2006. Investment incentives and airport regulation. Util. Policy 14, 262–265. Starkie, D., 2008. A critique of the single-till In: Starkie, D. (Ed.), Aviation Markets. Ashgate, Aldershot. Stern, J., 1997. What makes an independent regulator independent? Bus. Strateg. Rev. 8 (2), 67–74. Wolf, H., 2003. Privatisierung im Flughafensektor. Eine ordnungspolitische Analyse. Springer, Berlin.
Contents lists available at ScienceDirect
Transport Policy journal homepage: www.elsevier.com/locate/tranpol
An economic assessment of airport incentive regulation Nicole Adler a, Peter Forsyth b, Juergen Mueller c,n, Hans-Martin Niemeier d a
Hebrew University, Mount Scopus, Jerusalem 91905, Israel Monash University, Clayton 3800, Australia Berlin School of Economics and Law, Badensche Strasse 50-51, D 10825 Berlin, Germany d University of Applied Sciences, Werderstr. 73, D 28199 Bremen, Germany b c
art ic l e i nf o
a b s t r a c t
Available online 14 May 2015
There has been a gradual trend towards incentive regulation of airports since the privatization of BAA in 1986. Airports are price capped in several countries belonging to the European Union, as well as elsewhere, notably India. However, most of the price caps are not a pure price cap in which the X-factor is set independently of the cost of the regulated airport. Typically, hybrid price caps are used and combined with sometimes complex mechanisms like sliding scales, quality incentives and investment obligations, such that the incentive structures may become distorted. We provide an overview of the changes in the governance structure of airports as a result of privatization, analyze how far the regulatory institutions obey the principles of ‘good’ regulation, such as fairness and transparency, and investigate the performance of the new regulation in terms of allocative efficiency by reviewing benchmarking studies. Finally, we assess the impact of incentive regulation on productive efficiency using data envelopment analysis and second stage regression, which suggests that incentive regulation is superior to cost plus. & 2015 Published by Elsevier Ltd.
Keywords: Airport economic policy Airport performance Regulation Competition and privatization of airports
1. Introduction: why incentive regulation? Incentive regulation (IR) grew out of a dissatisfaction with the way conventional regulation, which had been in place for nearly a hundred years in the US, had been performing. The dominant form of regulation, at least in the US, was Rate of Return (RoR) regulation, which was recognized to have several serious drawbacks, in that it had few incentives for the regulated firm to control costs, and there was evidence that it had led to over-capitalization (the Averch and Johnson effect). In the US, with its long history of regulation, and the UK, which by the 1980s had begun to privatize its public utilities, there was a call for forms of regulation with improved incentive properties. Proposed changes to the form of regulation were debated at both the theoretical and the practical level. At the theoretical level, models of regulation were developed (e.g. Baron and Meyerson, 1982; Laffont and Tirole, 1986) which attempt to provide the firm with greater incentives to pursue productive efficiency. The main focus was on productive efficiency, but there was an expectation that the new forms of regulation would also encourage allocative efficiency, as RoR regulated firms showed little incentive to pursue n
Corresponding author. E-mail addresses: [email protected] (N. Adler), [email protected] (P. Forsyth), [email protected] (J. Mueller), [email protected] (H.-M. Niemeier). http://dx.doi.org/10.1016/j.tranpol.2015.03.008 0967-070X/& 2015 Published by Elsevier Ltd.
profits by adapting their price structures to better reflect costs and demand conditions. At about the same time, especially in the UK, there was a move away from RoR regulation, replaced by simple rules which incentivized the firm to produce more efficiently. The most common of these is the price cap rule, which was devised by Littlechild (1983) (see Beesley and Littlechild, 1989), and under which the firm is given a predetermined price path which sets the maximum price it may charge. If the firm achieves costs lower than the prespecified price path, it enjoys the additional profits, whereas if costs are higher than that defined by the path, the firm will incur losses. While several versions of the price cap rule exist, price caps are the most common form of practical incentive regulation. Incentive regulation of airports began when BAA, which owned the major London airports, along with a group of smaller airports, was privatized. Since then regulation of the London airports has remained in place, although regulation of smaller airports has been discontinued. Australia followed the London model when it privatized its airports, beginning in 1997, though it ceased to regulate via price caps in 2001–02. Since then a number of countries have imposed price regulation, although several of the airports are not regulated by traditional price caps. The primary emphasis of incentive regulation of airports has been on achieving productive efficiency, in addition to avoiding excessive charges. However other aspects of efficiency are expected to be affected by IR. With airports free to maximize profits, there has also been an expectation that price structures will
6
N. Adler et al. / Transport Policy 41 (2015) 5–15
become more influenced by costs, by demand elasticities. Incentive regulation may have negative effects on quality of service, since costs can be reduced by cutting quality. With airports, the distinction between aeronautical and non-aeronautical revenues is an important one whether these are included in the regulated prices (single till) or not (dual till). Finally, IR can be expected to have an impact on profitability and its variability. The objective of this paper is to survey the performance of airports being subjected to different forms of incentive regulation. We first comment on the governance arrangements of airports, and then we try to assess which airports are truly incentive regulated and which are not. In Section 4 we set out what can be expected of incentive regulation. The core of the paper is Section 5, in which we attempt to assess the extent to which incentive regulation has affected performance. In particular we present a benchmarking and econometric study of the impact of incentive regulation on performance. Appendix A details the dataset analyzed and provides summary statistics and Appendix B specifies the type of regulation applied at the 58 airports for which data was available over time. We draw some conclusions in Section 6.
2. Principles of good regulation Attempts to reform regulation in the direction of incentive based price caps also led to a renewed interest in the institutional aspects of regulation, in particular when these reforms were combined with privatization. The key problem of organizing public utilities is to write long term contracts for investments in long lived assets which have value only in a specific exchange relation. This is no problem in a world of perfect foresight, but in reality asymmetric information makes it infeasible to write complete long term contracts covering all contingencies. Opportunistic behavior might lead to hold-up problems so that markets and long term contracts fail and only regulation can create discretionary commitment. However even regulation can fail to provide this long term stability which is captured by changing interest groups and opportunistic behavior (Gomez-Ibanez, 2003). Such failure of regulation led economists1 and international institutions to develop principles of good regulation which are now well defined, precise and, at least in principle, accepted by many states. According to the OECD (2005, p. 35) good regulation should “be carried out by an independent, but democratic authority, because such an institution minimizes opportunistic behavior from the regulated firm, its users, its owners and from policy. An independent authority should try to correct market failure efficiently and avoid regulatory capture.” The concept of good regulation does not prescribe a method of regulation. Cost based regulation can be part of good regulation if it avoids excessive cost inefficiencies. Good regulation, in particular with its demand for an independent regulator, is a necessary though not sufficient condition for incentive regulation. As incentive regulation implies risks for the regulator, it is equally important that firms should be prevented from influencing the regulator in an attempt to mitigate these risks. Traditional cost based regulation leads to problematic incentives which encourage gold plating and cost padding, but may nevertheless be considered a form of good regulation provided any inefficiencies are not overwhelmingly large. 1
Stern (1997) provides a good overview.
3. What is incentive regulation at airports? As the concept of incentive regulation has been developed for all public utilities, we firstly define incentive regulation in general, and then apply it to airports by providing examples of different types of incentives schemes. Thereafter we argue that while ‘yes or no’ categorization is not possible, it is useful to differentiate between different forms of incentive regulation. 3.1. Incentive regulation The regulator must design a contract to set incentives for the regulated firm. If the contract reimburses the firm irrespective of its efforts, the contract is referred to as ‘low’ powered. In this case, regulated prices costs are likely to be high. If the contract only fixes a certain price, so that all cost savings remain with the regulated firm, the contract is referred to as ‘high’ powered (Laffont and Tirole, 1993). Such a regulatory contract which decouples prices from costs and profits are a means to induce efforts to reduce costs. However, cost based regulation would appear to accentuate the central problem for regulators, that is, the regulator has asymmetric information about the demand and cost functions. Given the information asymmetry, the regulated firm will provide the regulator with the information only if in a position to keep some of the informational rents. Regulation will not achieve first best outcomes, but incentive regulation should lead to better outcomes than cost based regulation, depending on how high or low powered the contract is designed. There appear to be three broad forms of incentive regulation that are currently practised. Price cap regulation was developed initially by Beesley and Littlechild (1989), and later evolved into British utility regulation. There are different varieties, but it is useful to distinguish between pure price caps, in which the cap is set independently from the costs of the regulated firm, and hybrid price caps, in which the cap is based every few years on the regulated asset base. Pure price caps are high powered; contracts hybrid caps less so. Revenue cap and revenue sharing agreements limit total revenues rather than prices (Green and Rodriguez-Pardina, 1999). Revenue and profit-sharing concepts define the level of profits the regulated firm is permitted to earn. Once this benchmark is exceeded, the amount by which the firm surpassed the benchmark must be shared between the firm and its consumers (Kunz, 2000). Benchmarking and yardstick competition was developed in particular by Shleifer (1985). The regulator determines the prices for the firm by using the cost levels of comparable firms in order to settle on a regulated price on the basis of costs seemingly unrelated to the regulated firm. It is important to note that these three systems of price level regulation can be combined and even mixed with low powered cost regulation. The former happens, for example, if benchmarking is used to determine the X in the price cap formula. An example of the latter occurs when prices are regulated on a cost based principle and the quality is defined in terms of standards, which are incentivized by a bonus system. 3.2. Forms of incentive regulation at airports Traditionally, for most airports, prices were set on a cost based approach. Even in those countries where governments own, operate and regulate their airports, charges were set in relation to costs, although very often in a non-transparent manner. Price capping of airports started with the UK airports in 1986. The UK Civil Aviation Authority (CAA) followed in the track of the price cap approach applied to other UK utilities, which led to a hybrid price cap. Another important feature of UK incentive regulation is that only a few airports, those that were thought to have market
N. Adler et al. / Transport Policy 41 (2015) 5–15
power, were designated for formal price cap regulation, although all airports could be regulated were it proven that they abused their pricing freedom. There have been several attempts to strengthen such price caps. Setting the X efficiency level through benchmarking was explored, but found not robust enough to be employed in the regulatory process due to data problems (CAA, 2000). The scope of regulation was defined more narrowly in an attempt to strengthen the power of the regulatory system. Attempts to adopt a dual till approach, whereby only aeronautical costs and revenues are regulated, failed in the UK but were adopted at other airports such as Malta and Hamburg (Starkie, 2008). Furthermore, there has been some tendency to move in a step-wise fashion from single till to dual till incentive regulation at airports in Belgium, Hungary and France. Another important feature of price caps is that there is an incentive to reshuffle the price structure. The CAA adopted a cap defined on average revenue, which sets fewer incentives than a tariff basket approach to restructure charges. Hybrid price caps are sometimes adopted along the line of the UK system, but then practised in a way that reduces the incentives further via sliding scales. A sliding scale mechanism generally relates the price to a function of traffic growth, quality and/or inflation rates. This form of revenue sharing agreement allows airports to increase their charges under low growth scenarios but may even reduce charges under high traffic growth, such that overall airport revenue remains relatively stable. Many price capped airports, such as Hamburg and Vienna, have adopted a sliding scale mechanism, stabilizing revenues at a certain level irrespective of airport pricing. The power of incentive regulation was reduced in Paris by adopting the same structure and level of charges for Orly and Charles De Gaulle airports, although both differ in terms of excess demand and supply. Another way to reduce the power of incentive regulation is to officially adopt the UK system, but prevent prices from deviating from costs, as has occurred in India. Revenue sharing elements have been adopted for some time at a number of German airports. The regulator allowed airlines and airports to reach agreements on such contracts in coordination. At first sight this comes close to the policy of price monitoring in Australia and the threat of re-regulation, but the difference is that Germany lacks an independent regulator and there is no review of performance, as there is in Australia. Benchmarking has never been applied as a regulatory tool to develop yard stick competition. The heterogeneity of airports and lack of data has so far prevented airport regulators to follow up the promising examples of other public utilities like energy and water (Reinhold et al., 2009). Benchmarking has however been used to lessen the information asymmetry and to compare charges among groups of airports. Unfortunately the selection of airports reflects fairly low targets such that the incentives are weak. 3.3. Categorization The practice of airport regulation is changing and different forms are emerging, composed of diverse elements of incentive regulation. A ‘yes or no’ classification is not helpful, but somewhere a line must be drawn among the continuum of low to high powered regulation. In the empirical work in Section 5 we distinguish between pure price caps, hybrid price caps, revenue caps and light handed regulation. It may be the case that the power of regulation is lower compared to what could have been achieved and what has indeed been achieved in other industries like energy and water (Reinhold et al., 2009).
7
3.4. Medium to high powered regulation In Australia, prior to monitoring (see below) a price cap with a dual till was set, with an X efficiency level in the range of 1–5% (mostly in the range of 3–4%). Specific provisions passed on the costs of new investment, so that underinvestment could be avoided. Overall, the price cap was set ‘very tightly’ (Forsyth, 2008, p. 84). As this cap was set without reference to the costs of the airport, such a system comes close to a pure price cap. Malta set a dual till price cap with an X of 0.5 for the period 2002–2007. Charges have not been changed since 2007. This cap was not based on costs. Unlike Australia and the UK, the regulator of the Malta airport of which the government holds a share of 30%, is not independent. UK price cap regulation can be divided roughly into two phases. Up until 2002, price cap regulation tried to set strong incentives for efficiency by relatively high X-values over a period of 15 years. Thereafter the price cap was lessened in order to allow for costly investment, in particular at Heathrow. The Irish regulator also needed to handle the issue of investment. However, the publicly owned airport planned investments rather too early and on a substantial scale. Benchmarking the costs of terminals was one of the instruments to control for excessive investments (GAP, 2011). An assessment of Italian airports must consider the ongoing institutional changes. The current Italian regulation is a hybrid price cap system based on a ‘semi’ single till principle, since December 2008 (ENAC, 2007; 2008) on an airport by airport basis.2 A slightly different regulation is applied to Rome (with its two airports), with Milan and Venice-ENAC agreed on a dual till for a five year regulatory period in an individual bargaining process with the airports. In these cases the focus on regulation is very much on investment, for which targets are set and, if not met, fines are imposed.3 While the current regulation is clearly defined from 2008 onwards as a hybrid price cap, the picture is further complicated because there has been a freeze on airport charges (for levels and structure) in effect since 2000. Up to 2000, regulation was cost based, but for the period 2000–2008 the price freeze in nominal terms suggests a pure price cap that has no relation to costs (Sciandra, 2009). 3.5. Low powered regulation Compared to the Anglo-Saxon approach of regulation, the system for Hamburg and Budapest airports set lower incentives due to the additional hybrid elements. Hamburg airport was regulated by a sliding scale price cap based on a dual till. Overall this has decreased the level of charges substantially and to a larger extent than at other German airports with revenue sharing agreements. The hybrid price cap at Budapest lowers the level of charges almost by 20% in real terms for the period of 2006–2011. As the price cap does not regulate the price structure, route development programs with substantial reductions were introduced in a non-discriminatory manner and a low cost terminal with reduced charges was operated for a limited time (GAP, 2011). Incentive regulation for Austrian, Belgian, German and French airports is relatively weak. At Düsseldorf, Frankfurt and Hannover, a sliding scale was used at certain times to stabilize the revenues. As the sliding scale is inversely related to demand, it leads to rising charges in a crisis. Furthermore, it does not set effective incentives 2 At the end of 2011, the following airports signed the ‘contratto di programma’ agreement after being reviewed by ENAC: Bari, Bologna, Brindisi, Naples, Pisa and Palermo. In 2012, Catania, Cagliari and Olbia followed. 3 For example, ADR on the basis of its long-term investment plan for €12 billion was able to negotiate a new regulatory contract valid until the end of the concession in June 2044 (Gemina, 2013).
8
N. Adler et al. / Transport Policy 41 (2015) 5–15
for rebalancing the price structure to manage scarce capacity in periods of high demand. In Austria, the sliding scale formula has no productivity factor, so that productivity gains are not passed onto the users. The level of charges remains relatively high and the incentives for productivity gains are low. At these airports, the structure of charges is left unregulated. The case of Aeroports de Paris highlights this type of inefficiency. Here the same level and structure of prices apply for Charles de Gaulle and Orly airports although the latter has excess demand and the former has excess capacity. A rather heavy handed type of hybrid price cap regulation has been implemented at Delhi, Mumbai and eight Indian airports. The price cap is based on costs and the price structure is set by the regulator. Brussels is regulated by a traditional cost based system with a single till evolving over time into a dual till system. Benchmarking is only used as an additional tool to provide information. As the legally determined set of reference airports are all airports with relatively high charges, the incentives for cost efficiency are rather mild (GAP, 2011). At Copenhagen there are three stages of regulation which have been evolving as a response to the stepwise privatization (25% in 1994, another 24% in 1996 and an additional 17% in 2000). In the phase up to 2003, charges were rather loosely regulated in such a way that efficiency gains were kept by the airport, but it remained unclear for how long (Wolf, 2003). For the period 2003–2005, a hybrid price cap was introduced with an annual increase of 2.75% of charges and a cap for the period 2006–2008 with a reduction of 3% in 2006 followed by an increase of 1% in 2007 and 2008. From 2008, regulation was reformed under the principle that the airport and its users should negotiate, and only if no agreement is reached, regulation of a hybrid price cap with a mixed till sets in. For the period 2009–2010, an increase of 4.2% in charges was allowed and for 2010–2015 a price cap with CPI plus 1% was agreed upon. The future will show whether the dependent regulator will be in a position to act as a mediator. 3.6. Light handed regulation (LHR) In 2002 the Australian government decided not to prolong the price cap approach-rather establish LHR. The Australian Consumer and Competition Commission (ACCC), an independent agency, monitors prices, costs and profits of aeronautical services and aeronautical related services of the airports of Brisbane, Melbourne, Perth, Sydney Kingsford Smith, and Adelaide, (Forsyth, 2004). In 2006 and 2012 the Australian Productivity Commission reviewed the airports' performance positively and recommended continuing LHR. We classify LHR as a form of incentive regulation. It is less high powered than pure price cap regulation, in as much as the regulated airport has the ability to increase its prices, for example, if were costs to rise. However pricing is not unconstrained, because the airport management must convince the monitoring or regulatory body that there has been no abuse of market power. In the normal course of events, if it is possible to reduce costs, profits will rise, hence there is a strong incentive to minimize costs on an ongoing basis. We have also classified nondesignated UK airports as being subject to the credible threat of regulation and therefore as being subject to LHR.4 4
David Starkie suggested to us that the non-designated UK airports should be classified as not-regulated airports because since the airport Act of 1986 no other airports have been designated and it would still have required the government to pass the necessary instrument. While we think that this might reflect well the current situation we prefer our classification for the past because it seems to reflect better the facts that the behavior was more carefully monitored than a normal industry. In response to demands of parliamentarians to regulate the BAA Scottish
In addition to the countries covered in Table 1, a number of other countries have implemented some forms of IR, though detailed information about these is not readily available. These include Portugal, South Africa and Latin American countries – lack of data precludes these countries being included in the empirical study in Section 5.
4. What do we expect from incentive regulation at Airports? It is useful to set out what may be expected from incentive regulation at airports. This is useful when assessing whether it has been successful and whether problems have emerged. Several aspects of performance are relevant including allocative, productive and long run efficiency, as well as quality, and profitability. 4.1. Allocative efficiency Allocative efficiency can always be an issue, even though in the incentive regulation literature it is less of a concern than other aspects. First, there is the relationship of price relative to cost. With regulation, prices are likely to be close to average cost. With non-congested airports, prices may be above marginal cost, but permitted by the regulator to be close to average cost (a first best solution will involve losses). With congested airports subject to excess demand, prices may be kept lower than marginal cost (to the efficient rationing price) but the presence of a slot allocation system may ensure that prices are efficient (most congested regulated airports possess a slot system). Second, there is the issue of the price structure. With noncongested airports given a budget target, Ramsey pricing would be in order. With price caps, there are some incentives for regulated firms to adopt Ramsey pricing, but these incentives are not always present (Giuletti and Waddams Price, 2000). Nevertheless most uncongested airports utilize weight-based charging, which is an approximation to Ramsey pricing. Efficient pricing of airports subject to excess demand should have a uniform pricing structure. A third allocative issue is that of slots. Most congested, regulated airports have slot systems in place. Charges are lower than market clearing levels, and slots take up the allocative function – thus there is a question of how well these slots are allocated. Typically, regulators do not handle slot markets, rather alternative arrangements such as grandfather rights are in place. A final allocative aspect concerns cross subsidies. Airports have the ability to generate non-aeronautical revenues which can be used to cross subsidize operations. Most regulated airports tend to set average prices close to costs thus there is not likely to be a major issue of excessive pricing, although prices may be above marginal cost which cost recovery will require. However some airports, subject to weak incentives, operate with sliding scales. Such scales will have a small negative effect on efficiency, since prices go up when demand falls, and there is more likely to be excess capacity. The price structure is more important than the price level. Most airports operate with a weight-based landing fee structure, possibly combined with a passenger charge. If there is available capacity at the airport, this would be a fairly efficient structure. However, the case for weight(footnote continued) airports BAA voluntarily adopted a price cap of RPI-3 for three years in 1993. A year later the designation of Belfast International Airport and Belfast City played an important role in the attempted merger of the two airports. In 2000 the CAA rejected the proposal of Easyjet to designate Luton Airport. Kunz (2000) and Wolf (2003) see this as sufficient evidence that UK uses the threat of reregulation as part of their regulation and argue that this is credible while the threat for New Zealand airports was not.
N. Adler et al. / Transport Policy 41 (2015) 5–15
9
Table 1 Types of incentive regulation. Airports by country
Australia Adelaide, Brisbane, Melbourne, Perth, Above airports plus Sydney Austria Vienna Belgium Brussels Denmark Copenhagen France Aeroports de Paris Germany Düsseldorf Frankfurt Hamburg Hannover Hungary Budapest
Italy All major airports Aeroporti di Milano Aeroporti di Roma Naples, Venice Other major Italian airports India Delhi, Mumbai plus eight airports Ireland Dublin Malta Malta Airport
UK Heathrow Gatwick Stansted Manchester
Time period
Type of regulation
Independent regulator
Strength/weaknesses
1997 to Price cap with dual till 2001– 02 Since Light handed regulation 2002
ACCC
Low price cap, but instability in crisis with exit of Ansett Airlines
ACCC and Productivity Commission
Strong incentives to reduce costs and differentiate prices subject to assessment of an independent regulator
Since 1998
Revenue cap
Yes, DOT
Stabilizes revenues at high level
Since 2004
Cost based with benchmarking elements
No
Peer group of airports are relatively high cost airports. Cost based thinking prevails
Since 1995
Light handed price cap on a mixed dual till with regulator as arbitrator
No
Long record of incentive agreements. Role of dependent arbitrator so far not tested.
Since 2005
Hybrid revenue cap with bonus malus investment and quality regulation
No
Regulatory capture and inefficient charges
2004– 2009 2002– 2006 Since 2000 2003– 2008
Revenue cap
No
Regulatory capture and instability
Revenue cap
No
Regulatory capture and instability
Revenue cap on dual till
No
Stable and accepted by stakeholders
Revenue cap
No
Regulatory capture and instability
Since 2006
Price cap with quality regulation
Yes, with full privatization in 2011. Na- Tight price cap. Initial conflict of interest retional Transport Authority Civil Aviation solved in 2011. Directorate
2000– 2008 Since 2012
Price freeze
No
Hybrid price cap with dual till, bonus malus investment and quality regulation
No
Since 2009
Hybrid price cap with mixed till
No
Since 2011
Single till hybrid price cap with regulated Yes, in 2008Airports Economic Regprice structure will be developed ulatory Authority of India was established
Price cap is too much cost based. Independent regulator.
Since 2001
Hybrid single till price cap
Yes, Irish CAA
Hybrid price cap. Independent regulator
Since 2002
Price cap on dual till with X ¼ 0.5 for period 2002 to 2007. Since then no changes of charges
No
Strong incentives as cap is not cost based. Role of dependent regulator so far not tested
Since 1986 Since 1986 Since 1986 Till 2008
Price cap
Yes, CAA UK
Hybrid price cap with investment regulation
Pure price cap with strong incentives, but uncertainty about institutional reform Hybrid price cap and regulated investments. Danger of regulatory capture Hybrid price cap Danger of regulatory capture
Price cap
Hybrid price cap
Price cap
Hybrid price cap
Price cap
Hybrid price cap
based charges is not present if the airport it experiencing excess demand, since a uniform charge would be more efficient. There is not much evidence of airports adopting uniform charging. With the London airports, charges are more uniform than most, though weight-based charging remains an element. This may be some evidence of the positive effect of incentive regulation (Forsyth and Niemeier, 2008). In most airports, slot markets are very imperfect. There is reasonably open slot trading for the London airports, but not for most other airports. There is no formal link between slot market
arrangements and regulation. However it may be significant that airports which have the strongest IR are also those which have the most open slot markets. 4.2. Productive efficiency We would expect that productive efficiency would increase as a result of a move to incentive regulation. Greater incentives to produce efficiently are the core rationale of incentive regulation. Given that productive efficiency is relatively easy to measure,
10
N. Adler et al. / Transport Policy 41 (2015) 5–15
using total factor productivity, stochastic frontier analysis or data envelopment analysis measures, it could be hoped that the gains in efficiency would be identifiable, (as discussed in Section 5). There are only a few studies assessing the effects of incentive regulation (Liebert and Niemeier, 2013). This is due to the fact that in the first years after privatization of airports, the emphasis was on testing the effects of different ownerships forms. These analyses suggested that privatization increases efficiency only if airports are fully privatized, such as the UK and Australia. Partially privatized airports were found to be less efficient, even compared to public airports (Oum et al., 2006). This is a rather disappointing result for European airports on the mainland, where partial privatization is very often the typical form of privatization, with a minority share for the private investors. Recent studies analyzing the effects of ownership, regulation and competition, such as Adler and Liebert (2014) based on a comparison of 51 airports, as well as Assaf and Gillen (2012) based on a study of the UK airports, show that competition from nearby airports leads to superior results than regulation. However, at a number of airports, competition is not effective. Adler and Liebert show that cost based regulation leads to lower efficiency than incentive regulation. Assaf et al. argue that regulated airports tend be more efficient than non-price regulated airports and that investment programs at Heathrow, Gatwick and Stansted tend to lower efficiency scores. However, while these studies show that there is some evidence that incentive regulation increases efficiency and that it is superior to cost based regulation, it is equally clear that not all price caps are equal, hence in this research we analyze the forms of incentive regulation in order to understand their relative effects. 4.3. Quality By contrast, incentive regulation is not likely to be helpful in ensuring quality service: in fact it may well reduce quality. It encourages the regulated firm to increase profits, and one way that it can do this is by allowing quality to fall. This problem can be addressed in part or fully through specific regulation of quality, or possibly monitoring of quality (Rovizzi and Thompson, 1992). Quality of service has been an issue in the UK, and after a recent report (Cave, 2009), the government raised the possibility of changes to regulation to give quality a greater weight. 4.4. Investment and long run efficiency Incentive regulation does not have a clear role when it comes to investment and long run efficiency. Indeed, a dynamic theory of incentive regulation has yet to be developed and most models are essentially static. There is some analysis of regulation and investment, in general and for airports, but it is at an early stage of development (Starkie, 2006). Regulators often use a cost based approach to handle additions to capacity, even when they are using formally pure forms of incentive regulation in the short run. The Australian Government ACCC (2001) considered that price caps would lead to underinvestment, and brought in cost passthrough arrangements to encourage necessary new investment. Attempts have been made to develop a form of incentive regulation which includes investment opportunities and an example is that suggested for the London airports. The CAA in the UK attempted to link investment and output into its price cap formula for the London airports (Andrew and Hendriks, 2004). This however did not meet with the approval of competition regulators and the CAA was forced to drop the proposal. Some have suggested that regulation might give rise to excessive investment (rather like a cost plus regulated airport and the Averch and Johnson effect (Starkie, 2006)). Thus there is a complex empirical issue to be resolved.
4.5. Non-aeronautical revenues Another aspect of airport performance concerns the incentive for the airport to develop non-aeronautical revenues, which is connected to whether the airport operates under dual or single till. Whilst incentive regulation is compatible with either a single or a dual till, the two approaches have conflicting implications for different aspects of efficiency. The expectation might be that a dual till approach better encourages productive efficiency, since the airport has a greater incentive to produce unregulated, commercial services. Whilst there has been much theoretical discussion on the merits of the different approaches, there is been little empirical work. A recent study is that of Adler and Liebert (2014) who report that airports which operate under a dual till are more productively efficient. 4.6. Profitability and its variance Finally, an aspect of performance which could be important is that of the airport's profitability and its variability. One would not necessarily expect that an incentive regulated airport be more or less profitable than one subject to other forms of regulation (or no regulation). However there could be a difference in terms of the variability of its profit. Incentive regulation places a greater risk on the regulated airport; under strong demand the airport may expect high profits, whereas under weak demand, the airport may record low profits or losses. By contrast, airports subject to other forms of regulation, particularly those with sharing rules and sliding scales, should record more even profitability, since the regulator tries to set prices close to average costs. Indeed, performance in terms of profitability and its variability may be a good indicator of whether the airport is subject to genuine incentive regulation.
5. Empirical two-stage study of productive efficiency As noted, there is little by way of empirical assessment of the impact of varying levels of incentive regulation on airport management behavior. In this section we provide an assessment based on a DEA estimation of the relative efficiency of airports in the sample, and relate the results to the forms of regulation and other variables of interest, such as whether the regulator is independent or not. Our primary interest is the effects of regulation, not the determinants of individual airport efficiency, thus we use an existing model (Adler et al., 2013) of efficiency, and analyze how regulation affects performance. In order to estimate relative efficiency, we apply a non-oriented, variable returns to scale, bound adjusted measure that minimizes labor and other operating costs and maximizes nonaeronautical revenues given declared runway capacity as a nondiscretionary input and passengers, air traffic movements and cargo as outputs over which airport management have little control. Consequently, this represents a short-term managerial efficiency measurement which defines approximately 8% of the airports in the dataset as relatively efficient. Summary statistics of the 58 airports in the dataset and the number of years for which data was available is presented in Appendix A. The unbalanced data spans the years 1990–2010 and of the 58 airports, 38 are drawn from the United Kingdom, Germany and Italy. The data includes all countries covered in Table 1, except Hungary, India, Ireland, and the two Paris airports (due to data limitations). For the second stage analysis, the explanatory variables defining the regulatory structure include revenue caps (RevCap), hybrid and pure price caps and light handed regulation. We also explore whether an independent regulator influences relative productive
N. Adler et al. / Transport Policy 41 (2015) 5–15
11
Table 2 Second-stage regression analyses explaining relative efficiency estimates. Fixed effects (within) regression
Truncated regression [0,1)
Model 1 Dependent Variable ¼ DEA estimate RevCap Light Hybrid Hybrid/light Pure Independent NA Cap1 Cap2 R2/log likelihood Number of observations
Model 2
Coefficient
t
0.0397 0.0909 0.1218
1.89 0.0386 2.15 2.67 0.1021 3.49 0.1733 1.27 0.0534 6.33 0.0768 4.95 0.0641 8.78 0.2463 Within ¼ 0.2414, between¼ 0.1729, overall¼ 0.2078 707
0.1942 0.0856 0.0786 0.0646 0.2465 Within ¼0.2426, between¼ 0.1613, overall¼ 0.1879 707
efficiency estimates. Furthermore, we include three additional dummy variables, in order to compare the results to those in the literature such as Oum et al. (2006) and Adler and Liebert (2014). Hence we have added a dummy equal to one if the airport earns at least 50% of revenues from non-aeronautical sources (NA), and two dummies to capture capacity utilization. Cap1 receives a score of 1 if the airport is utilized between 50% and 80% of their declared runway capacity on an annual basis and cap2 refers to airports that are serving air traffic movements at a level above 80% of their declared capacity, which includes London Heathrow, Frankfurt, occasionally London Gatwick and Brussels airport at the beginning of the 1990s. In Table 2 we present the results of a fixed effects regression model, which enables an unbiased analysis of the explanatory variables. We also include truncated regressions to test for robustness of the results given that the relative efficiency scores estimated in the first stage are capped at 1, which defines the Pareto frontier. Models (1) and (3) test all four levels of incentive regulation and models (2) and (4) test for weak (RevCap), medium (hybrid price caps and light handed monitoring) and strong (pure price caps) powered incentive regulation. Finally, we include time dummies in order to ensure that these effects do not confound the results. The time dummies, which have been excluded from Table 2 for purposes of parsimony, suggest that airports were relatively more efficient in the 1990s as compared to the 2000s, some of which may be explained by the increased security regulations and their respective costs that developed as a result of 9/11. The variables often analyzed in the literature, namely capacity utilization (cap1 and cap2) and the percentage of non-aeronautical revenues earned out of the total revenues of an airport (NA), behave as expected from the literature. Airports earning more than 50% of their revenues from the landside are approximately 8% more relatively efficient and the higher the capacity utilization, the higher the expected relative efficiency levels at the airport. The results for the incentive regulation also proved to be significant and demonstrate that, in comparison to the cost plus regulatory system in place at half the airports in the sample, incentive regulation leads to higher levels of relative productive efficiency. Revenue caps increase relative efficiency by approximately 4% (although the result is not significantly different to that of the cost based price caps that represent the baseline), light handed regulation by approximately 9%, hybrid price caps by approximately 12% and pure price caps, the highest power form of incentive regulation, by around 19% in this dataset. Interestingly, whether the airport is regulated by an independent regulator or not would seem to have little impact on efficiency because the coefficient,
Coefficient
t
Model 3
Model 4
Coefficient z
Coefficient z
1.84 0.0310 0.1255 0.1375 2.50 3.38 0.1997 0.91 0.0916 6.26 0.0993 4.92 0.0763 8.77 0.4042 925.01 701
1.53 0.0308 2.72 3.05 0.132 3.84 0.1941 1.58 0.0983 8.02 0.0986 5.69 0.0768 10.4 0.4046 924.96
1.52
3.15 3.95 1.8 8.08 5.75 10.42
701
though positive, is strictly insignificant. In summary, there appears to be evidence that incentive regulation has a positive effect on productive efficiency and pure price caps appear to have the strongest impact on short-term productive efficiency.
6. Conclusions The stated objective of incentive regulation of airports (along with other industries) is to increase the efficiency, and in particular, the productive efficiency, with which they operate. Has this objective been met? The objective of this paper is to assess this, taking into account the particular characteristics and features of airports. The evidence is encouraging, though clearly not conclusive. The empirical study reported in Section 5 suggests that a move towards incentive regulation is consistent with increased productive efficiency. These results need to be regarded as a first assessment of the issue. There are several limitations to the study: the dataset is unbalanced over time; there are problems of categorization of what constitutes incentive regulation and what does not; and there are distinct problems in developing a comparable data set for efficiency measurement. Nonetheless, the results are promising. Incentive regulation was designed primarily to increase productive efficiency especially in the short run. There are other aspects to performance – allocative efficiency, quality performance, and long run investment performance. Incentive regulation is not necessarily positive for these aspects of performance. It may improve allocative efficiency, though not necessarily; it may increase the efficiency with which an airport achieves efficiency in nonaeronautical services (depending on the form of regulation); and it may well lead to a reduction in quality. Most of these aspects are more difficult to assess, especially quantitatively, than productive efficiency. This survey indicates that performance in these aspects has been consistent with expectations. So far, at least in terms of the airport industry, the economist's recommendations have been for incentive regulation, rather than for the alternatives such as cost based regulation. This recommendation has been based on theoretical properties, along with empirical results from other industries. The empirical results from our study suggest that this recommendation is well based, and that incentive regulation of airports enhances productive and overall efficiency.
12
N. Adler et al. / Transport Policy 41 (2015) 5–15
Table A1 Descriptive statistics of airport dataset for benchmarking. Airport code
Country
# Years of data
Years
Staff costs
Other operating costs
Non-aeronautical revenue Output
Declared runway capacity Non-discretionary input
PAX
16,372,348
34
2,570,224
88,089
146,279,024 357,230,918
322,480,971
108
41,159,194
406,262 1,376,930
9,285,088
21,038,966
21,895,277
32
4,460,087
46,632
45,928
14,280,879
25,824,549
14,417,160
21
3,872,273
47,777
120,521
27,908,152
41,957,402
47,514,254
40
7,022,637
92,649
17,261
28,591,134
16,225,036
32,834,903
32
2,071,525
50,865
51,049
18,588,783
26,740,211
23,341,322
24
3,753,799
56,071
25,597
14,037,157
12,743,026
12,340,050
30
1,797,959
35,963
1,538
10,611,847
18,527,500
36,730,143
25
5,002,957
57,282
5,356
35,959,993
91,370,754
92,721,069
62
15,093,798 244,536 537,172
14,733,750
26,064,875
39,840,250
45
3,474,271
64,115
34,883
7,054,294
7,026,881
6,605,406
28
1,476,518
25,897
8,120
4,530,473
12,440,878
6,706,287
10
2,455,491
27,570
4,809
93,499,087
102,982,204
67,745,416
52
8,046,062
132,322 550,091
75,918,962
50,152,369
117,014,818
80
17,101,851
251,797 328,843
10,324,185
21,711,589
14,747,676
26
1,724,640
29,216
718
12,672,237
17,784,744
10,754,577
42
1,597,229
30,025
94
101,643,841
118,889,049
97,453,846
39
15,173,211
178,168
59,677
10,149,416
9,712,208
6,521,907
9
1,604,733
27,325
1,728
622,748,900 392,219,289
372,361,727
75
42,363,952 412,403 1,545,454
25,338,762
36,630,483
48,992,034
32
7,781,325
86,758
6,805
9,393,369
11,856,200
10,647,278
18
1,079,835
18,815
5,856
38,876,609
40,743,622
52,209,576
38
7,752,420
121,836 62,317
42,864,248
38,255,528
40,452,757
49
4,519,477
68,790
67,002,211
73,089,924
58,721,338
45
9,494,832
127,790 32,856
8,072,966
7,363,700
8,347,156
20
1,583,944
27,880
7,124
9,873,193
32,003,876
13,911,194
29
2,121,599
63,873
667
13,933,805
26,665,667
12,265,702
36
2,135,663
35,690
7,513
121,817,641
205,797,847
303,911,785
50
31,876,429 247,531 212,574
259,564,477 641,091,745
918,899,347
87
64,950,686 462,718 1,203,856
19,023,039
12,196,601
14,208,494
15
1,281,379
40,194
15,092
7,704,786
12,848,942
11,881,892
30
2,450,136
33,506
26,222
22,650,829
53,320,803
56,668,206
36
7,714,325
69,125
27,862
21,555,574
38,998,616
41,513,709
51
6,533,949
119,809 36,849
74,831,245
159,105,675
194,648,448
57
20,605,230 195,461 125,234
10,159,340
35,979,170
69,066,229
50
18,892,231 170,657 8,215,969
Inputs
ABZ AMS BFS BGY BHX
United Kingdom Netherlands United Kingdom Italy
21 10 9 10
BLL
United Kingdom Denmark
9
BLQ
Italy
10
BRE
Germany
10
BRS
8
BRU
United Kingdom Belgium
21
BSL
France
8
BTS
Slovakia
8
CAG
Italy
10
CGN
Germany
13
CPH
Denmark
21
DRS
Germany
13
DTM
Germany
10
DUS
Germany
21
FLR
Italy
10
FRA
Germany
18
GLA
12
GOA
United Kingdom Italy
10
GVA
Switzerland
21
HAJ
Germany
17
HAM
Germany
21
LBA
United Kingdom United Kingdom Germany
18
LCY LEJ LGW LHR LJU LPL LTN LYS MAN MEL
United Kingdom United Kingdom Slovenia United Kingdom United Kingdom France United Kingdom Australia
21
12 5 13 13 4 21 13 14 13 10
1990– 2010 1998– 2007 2002– 2010 2000– 2009 1990– 2010 2002– 2010 2000– 2009 1998– 2007 2002– 2009 1990– 2010 2002– 2009 2003– 2010 2000– 2009 1998– 2010 1990– 2010 1998– 2010 2001– 2010 1990– 2010 2000– 2009 1990– 2007 1999– 2010 2001– 2009 1990– 2010 1990– 2007 1990– 2010 1990– 2010 1999– 2010 1998– 2002 1998– 2010 1998– 2010 2004– 2007 1990– 2010 1998– 2010 1996– 2009 1998– 2010 1999– 2008
11,509,119
13,560,764
ATM
Cargo
Non-discretionary output
5,910
8,816
N. Adler et al. / Transport Policy 41 (2015) 5–15
13
Table A1 (continued ) Airport code
Country
# Years of data
Years
Staff costs
Other operating costs
Inputs
MLA
Malta
6
MME
1
MRS
United Kingdom France
12
MUC
Germany
13
NAP
Italy
10
NCE
France
12
NCL
21
NUE
United Kingdom Germany
13
OSL
Norway
9
PMO
Italy
10
PRG
3
PSA
Czech Republic Italy
10
RIX
Latvia
9
SOU
12
STR
United Kingdom United Kingdom Germany
13
SZG
Austria
9
TLL
Estonia
9
TRN
Italy
10
VCE
Italy
10
VIE
Austria
13
ZRH
Switzerland
13
STN
11
Non-aeronautical revenue Output
Declared runway capacity Non-discretionary input
PAX
ATM
Cargo
Non-discretionary output
2005– 2010 2002
10,142,025
19,892,769
15,839,484
15
2,965,178
28,748
17,302
6,556,432
5,010,381
4,806,309
14
689,000
16,000
1,000
1998– 2009 1998– 2010 2000– 2009 1998– 2009 1990– 2010 1998– 2010 1999– 2007 2000– 2009 2005– 2007 2000– 2009 2002– 2010 1999– 2010 1998– 2010 1998– 2010 2002– 2010 2002– 2010 2000– 2009 2000– 2009 1998– 2010 1998– 2010
19,459,115
36,153,950
40,161,478
33
6,353,310
90,905
55,276
259,922,880 367,822,664
364,115,866
87
27,447,589
354,135 193,153
12,409,685
17,591,268
15,676,097
28
4,791,686
55,360
21,761,725
64,015,750
64,863,792
49
9,422,466
177,082 16,885
16,290,627
11,898,456
19,452,568
30
3,453,409
44,228
6,390
36,639,374
31,953,076
31,590,056
30
3,549,069
60,323
16,281
36,650,795
80,692,721
137,722,928
65
15,182,364
195,713 81,222
14,432,026
18,560,722
11,189,681
20
3,886,318
43,531
38,862,483
46,063,406
55,535,100
39
11,549,689 167,067 57,628
15,020,757
14,673,942
9,324,162
14
2,535,213
29,488
11,423
11,686,333
6,599,865
8,167,522
27
2,483,934
40,475
10,538
8,755,931
9,187,506
12,349,084
22
1,426,467
37,026
148
58,318,786
88,740,012
137,101,006
47
18,715,728 166,160 365,185
56,152,315
92,029,755
69,654,763
40
8,636,991
129,984 20,978
17,022,727
13,524,445
9,352,647
20
1,608,225
21,182
232
4,616,556
5,715,778
7,337,222
20
1,280,252
25,944
14,722
11,332,354
19,591,109
18,270,833
26
3,095,216
46,012
14,773
19,081,176
29,239,348
23,018,853
29
5,656,134
70,374
23,024
194,628,466 109,584,868
109,097,877
64
14,936,721 215,135
78,690,631
129,397,326
150,973,435
70
20,071,672 257,692 423,019
7,774
5,155
182,812
Table B1 Data on regulation per airport Cost based
RevCap
Hybrid
Light
Pure
Independent regulator
NA
Cap1
Cap2
ABZ AMS
0 1
0 0
0 0
1 0
0 0
1 1
0 1
0 0
BFS BGY
0 0
0 0
0 2009
1 0
1 0
0 0
0 0
BHX BLL BLQ BRE BRS BRU BSL BTS CAG
0 1 0 1 0 1 1 1 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 2009
1 0 0 0 1 0 0 0 0
1 0 0 0 1 0 0 0 0
2009–2010 (2) 1 0 0 2005–2009 (5) 1990–1991 (2) 2002–2003 (2) 0 0
0 0 0 0 0 1993–2008 (13) 0 0 2008
0 0 0 0 0 1990–1992 (3) 0 0 0
CGN CPH
1 1990–1994 (5)
0 0
0 1990/1998 (2)
0 1996–2001 (6)
0 0
1
0
0 1995–2002, 2009– 2010 (10) 0
0 0
DRS
0 2003– 2008 (6) 0
0 2000– 2008 (9) 0 0 1 0 0 0 0 0 2000– 2008 (9) 0 0
0 98, 2000–2001 (3) 2005–2010 (6) 0
0
0
0
0
0
14
N. Adler et al. / Transport Policy 41 (2015) 5–15
Table B1 (continued ) Cost based
RevCap
Hybrid
Light
Pure
Independent regulator
NA
Cap1
Cap2
0 2005– 2008 (4) 0
0 0
0 0
0 0
0 0
0 0
0 1
0 0
FLR
1 1990–2004, 2009– 2010 (17) 0
2009
0
0
0
2000–2002,2005,2007 (5)
0
FRA
1990–2002 (13)
0
0
2000– 2008 (9) 0
0
0
1990–1995, 1997 (7)
0 0 0 0
1 0 0 0
0 0 0 0
1 0 0 0
2008 2001 0 0
0 0 2000–2010 (11) 0
1996, 1998– 2007 (11) 0 0 0 0
0
0
0
0
0
0
0
0 0 0 1
1 1 0 0
0 0 0 0
1 1 0 1
2006–2010 (5) 0 0 1998–2008 (11) 1998–2007 (10) 0 2005–2010 (6) 1 1996–1999 (4) 2005, 2008– 2010 (4) 2008
0 0 0 1998–1999, 2001–2005, 2007–2010 (11) 2009–2010 (2)
0 0 0 2000, 2006 (2)
DTM DUS
LBA LCY LEJ LGW
0 0 1 0
2003– 2007 (5) 0 0 0 2003– 2007 (5) 2000–2010 (11) 0 0 0 0
LHR
0
0
1
0
0
1
LJU LPL LTN LYS MAN
1 0 0 1 0
0 0 0 0 0
0 1 1 0 2007–2010 (4)
0 0 0 0 0
0 1 1 0 1
MEL
0
0
0 0 0 0 1998– 2006 (9) 0
2003–2008 (6)
2003–2008 (6)
MLA 0 MME 0 MRS 1
0 0 0
0 0 0
0 0 0
1999– 2002 (4) 1 0 0
MUC 1 NAP 0
0 0
0 2009
0 0
NCE NCL NUE OSL PMO
1 0 1 1 0
0 0 0 0 0
0 0 0 0 2009
0 1 0 0 0
PRG PSA
1 0
0 0
0 2009
0 0
RIX SOU STN STR SZG TLL TRN
1 0 0 1 1 1 0
0 0 0 0 0 0 0
0 0 1 0 0 0 2009
0 1 0 0 0 0 0
VCE
0
0
2009
0
VIE ZRH
0 1
1 0
0 0
0 0
0 2000– 2008 0 0 0 0 2000– 2008 (9) 0 2000– 2008 0 0 0 0 0 0 2000– 2008 (9) 2000– 2008 (9) 0 0
GLA GOA GVA HAJ
2010 1 1 1990–1996, 1998– 2002 (12) HAM 19901999 (10)
Acknowledgments This paper arose from a research project entitled “Airport Benchmarking by Economic Regulators” produced on behalf of the Netherlands Office of Transport Regulation (NMa). The data were provided by the German Airport Performance research project, which was kindly supported by the German Federal Ministry of Education and Research (see www.gap-projekt.de). We are also grateful to the Australian Research Council and the Recanati Foundation for their support. We thank David Starkie for his ever helpful comments. All errors are our own.
0 1 0
0 0 0 0 1998–2000, 2003–2007 (8)
1998–2008 (11) 0 0 0 0 0
2001, 2005–2008 (5)
0
0 0 2000
0 0 0
0 0
0 0 1998,2000– 2002 (4) 0 0
1999–2010 (12) 0
0 0
0 1 0 0 0
0 2008,2010 (2) 0 0 0
1998–2003, 2006–2008 (9) 0 0 2007 0
0 0 0 0 0
0 0
0 0
1 0
0 0
0 1 1 0 1 0 0
0 0 2003 0 1998–2006 (7) 1999, 2002–2008 (8) 0 2004–2008 (5) 0 0 0 0 0 0
0 0 0 0 0 0 0
0
0
0
0
1 0
0 0
2004–2010 (7) 1998–2004, 2008–2010 (10)
0 0
Appendix A See Appendix A1
Appendix B See Appendix B1 References Adler, N., Liebert, V., 2014. Joint impact of competition, ownership form and
N. Adler et al. / Transport Policy 41 (2015) 5–15
economic regulation on airport performance and pricing. Transp. Res. Part A 64, 92–109. Adler, N., Liebert, V., Yazhemsky, E., 2013. Airport benchmarking from a managerial perspective. Omega – Int. J. Manag. Sci. 41 (2), 442–458. Andrew, D., Hendricks, N., 2004. Airport regulation in the UK In: Forsyth, P., Gillen, D., Knorr, A., Mayer, O., Niemeier, H.M., Starkie, D. (Eds.), The Economic Regulation of Airports. Ashgate, Aldershot. Assaf, A.G., Gillen, D., 2012. Measuring the joint impact of governance form and economic regulation on airport efficiency. Eur. J. Oper. Res. 220, 187–198. Australian Competition and Consumer Commission, 2001. Price Regulation of Airport Services, Submission to the Productivity Commission's draft Report October 2001 〈www.pc.gov.au〉. Baron, D.P., Meyerson, R.B., 1982. Regulating a monopolist with unknown costs. Econometrica 50 (4), 911–930. Beesley, M.E., Littlechild, S.C., 1989. The regulation of privatized monopolies in the United Kingdom. RAND J. Econ. 20 (3), 454–472. Cave, M. Ch., 2009. Report of the Independent Panel on Airport Regulation, DfT January. Civil Aviation Authority (CAA), 2000. The Use of Benchmarking in the Airport Reviews, Consultation Paper, 〈http://www.caa.co.uk/docs/5/ergdocs/benchmark ing(caa122000).pdf〉 (accessed 25.07.13). ENAC, 2007. Capitolato relativo all ‘affidamento dei servizi per la redazione di uno studio sullo Sviluppo futuro della rete aeroportuale nazionale quale componente strategica dell’ organizzazione infrastrutturale del territorio, Rome, Italy. ENAC, 2008. Regolamento per la costruzione e l’esercizio degli aeroporti, Rome, 21 9 2008. Forsyth, P., 2004. Replacing regulation: airport price monitoring in Australia In: Forsyth, P., Gillen, D., Knorr, A., Mayer, O., Niemeier, H.-M., Starkie, D. (Eds.), The Economic Regulation of Airports: Recent Developments in Australasia, North America and Europe. Ashgate, Aldershot. Forsyth, P., 2008. Airport policy in Australia and New Zealand: privatization, lighthanded regulation, and performance In: Winston, C., de Rus, G. (Eds.), Aviation Infrastructure Performance A Study in Comparative Political Economy. Brookings Institution Press, Washington. Forsyth, P., Niemeier, H.-M., 2008. Price regulation and the choice of price structures at busy airports In: Czerny, A., Forsyth, P., Gillen, D., Niemeier, H.-M. (Eds.), Airport Slots. International Experiences and Options for Reform. Ashgate, Aldershot. Gemina, 2013. ADR's New Concession Agreement, ERA to Come into Effect and 2012 Traffic Performance. 〈http://www.atlantia.it/gemina/files/2013/01/04/ENG% 20notice%20presentazioneADR4gennaio2013.pdf〉. Gomez-Ibanez, J.A., 2003. Regulating Infrastructure. Cambridge, Harvard UP.
15
German Aviation Performance (GAP,2011) Airport Benchmarking by Economic Regulators, Study for the Netherlands Office of Transport Regulation, 〈www. gap-project.de〉. Green, R., Rodriguez-Pardina, M., 1999. Resetting Price Controls for Privatized Utilities: A Manual for Regulators. World Bank, Washington, D.C. Giulietti, M., Waddams Price, C., 2000. Incentive Regulation and Efficient Pricing: Empirical Evidence, Research Paper Series No. 00/2, Centre for Management under Regulation. Warwick Business School. Kunz, M., 2000. Regulierungsregime in Theorie und Praxis In: Knieps, G., Brunekreeft, G. (Eds.), Zwischen Regulierung und Wettbewerb – Netzsektoren in Deutschland. Physica, Heidelberg. Laffont, J.-J., Tirole, J., 1986. Using cost observation to regulate firms. J. Polit. Econ. 94 (3), 614–641. Laffont, J.-J., Tirole, J., 1993. A Theory of Incentives in Procurement and Regulation. MIT Press, Cambridge. Liebert, V., Niemeier, H.-M., 2013. A survey of empirical research on the productivity and efficiency measurement of airports. J. Transp. Econ. Policy 47, 157–189. Littlechild, S.C., 1983. Regulation of British Telecommunications' Profitability. HMSO, London. OECD, 2005. OECD Guiding Principles for Regulatory Quality and Performance, Paris, Organization for Economic Cooperation and Development. Oum, T.H., Adler, N., Yu, C., 2006. Privatization, corporatization, ownership forms and their effects on the performance of the world's major airports. J. Air Transp. Manag. 12 (3), 109–121. Reinhold, A., Niemeier, H.-M., Kamp, V., Müller, J., 2009. An evaluation of yardstick regulation for European airports'. J. Air Transp. Manag. 16 (2), 74–80. Rovizzi, L., Thompson, D., 1992. The regulation of product quality in the public utilities and the Citizen's Charter. Fiscal Stud. 13 (3), 74–95. Sciandra, L., 2009. Regulatory assessment of Italian airports: some critical issues. Mercato Concorrenza Regole 1, 51–78. Shleifer, A., 1985. A theory of yardstick competition. RAND J. Econ. 16 (3), 319–327. Starkie, D., 2006. Investment incentives and airport regulation. Util. Policy 14, 262–265. Starkie, D., 2008. A critique of the single-till In: Starkie, D. (Ed.), Aviation Markets. Ashgate, Aldershot. Stern, J., 1997. What makes an independent regulator independent? Bus. Strateg. Rev. 8 (2), 67–74. Wolf, H., 2003. Privatisierung im Flughafensektor. Eine ordnungspolitische Analyse. Springer, Berlin.