Nienke Hendriks1 and Doug Andrew
Aviation is often regarded as the first of the major regulatory reforms in Western economies. Unlike the US where airlines were already privately owned, in the UK de-regulation and privatisation were closely linked. The main reasons for privatisation were a reassessment of the role of government, budget constraints and greater awareness of the relationship between ownership structure and governance.
In the UK aviation market, privatisation started with the privatisation and deregulation of British Airways, The benefits of airline deregulation are seen to be substantial in terms of efficiency gains, innovation and consumer welfare. In the US, Steven Morrison and Cliff Winston's recent work estimated a 27% reduction in airfares due to de-regulation.2 They also stress the benefits of competition or the threat thereof. In the UK, competition has resulted in the introduction of innovative business models by the low frills carriers, which has led to major benefits for European travellers.
Going beyond the US practice, the UK extended the reform model to infrastructure with the privatisation of BAA in 1987, following the British Telecom privatisation in 1984. Given the large market share of the BAA London airports, they were made subject to price cap regulation. The airport services market was liberalised and gradually many airports were put onto a more commercial footing. At the time, price cap regulation was seen by experts as a temporary measure: sector-specific economic regulation was introduced to protect consumers "until competition arrived".
One of the main problems faced by the downstream market is the lack of responsiveness in the upstream infrastructure market. The downstream market in the UK is highly competitive in the sense of few or no government-imposed economic entry barriers. The upstream market is still characterised by monopolies in the South East. This raises issues about the impact of market structure, ownership structure and regulation.
In developing a public policy framework, it is desirable to adopt a comparative institutional approach where market failures and government/regulatory failures are taken into account in designing optimal policy responses in a world where perfect competition (e.g. no fixed, let alone, sunk costs) is merely a pedagogical tool and information asymmetries are inherent.
The two main types of models available are the traditional de-centralised market approach on the one hand and public sector management models on the other hand. There are a range of variants in-between; Buchanan's user clubs, franchise bidding and not-for-profit models, for example.
The market model is based on the fact that commercial firms, aiming as they generally are at maximising shareholder value, have strong incentives to define and price outputs. However, theoretically when a commercial firm has market power it might restrict output as a corollary to increasing its prices. Its incentive to invest optimally (from a social perspective) is therefore weakened and it is likely to invest late, even when accounting for real option values.3 This becomes even more of an issue when dealing with lumpy investments.
The traditional critique of the market approach to infrastructure was highlighted by Dupuit's case of a proposed bridge: high fixed and sunk costs, zero marginal costs. Private provision funded by user charges would lead to under-utilisation as prices would exceed marginal costs, once the bridge was built. This results in a loss of economic efficiency. Private provision may be later than ideal as the private developer may not be able to price discriminate sufficiently to obtain all the consumer surplus that would feature in the social cost benefit. However, a tax-funded subsidy could address this problem, although taxes themselves also impose costs to the economy which must be accounted for.
The public sector model is built around the premise that governments act in the public interest and are therefore more likely to achieve economic efficiency, i.e. ensuring best use of scarce capacity, least cost production and optimal investment levels, the aim being for projects to proceed as long as their marginal social costs equate to marginal social benefits. Under public provision the publicly interested government would undertake a cost benefit analysis, trying to judge the expected demand "schedule" for the bridge and assess the costs through to implementation, all against the next best alternative. If the bridge was built, no charge would be levied, achieving best use.
However, this is based on three key assumptions. Firstly, it assumes that the government is able to obtain the best information on demand and costs. However, the problem of asymmetric information is a considerable obstacle in doing so. Secondly, it assumes that the government would be able to remain "public interested". Public choice theory suggests that interest group capture of public policy diverts governments from this goal: local construction interests for example. Niskanen's model of bureaucratic behaviour4 suggests that empire building is an important motive explaining government behaviour: the bridge building agency for example. And thirdly there is the issue of the deadweight costs of taxation to finance the bridge.
From a public policy view, the objective of economic regulation should be economic efficiency: the aim being to maximise the size of the national pie in standards of living terms. Practically, this suggests that the policy problem is how to ensure:
The main policy issue then becomes having sufficient instruments to address each of the targets and, where competition is seriously limited, the trade-off between market failure versus regulatory failure. This raises the question of how far is regulation more likely to achieve the public policy objectives compared with relying on competition law.
Economic regulatory models can be distinguished by the degree to which they rely on the use of incentives as the regulatory equation shows:
P = k + βc
P is the maximum price allowed by the regulator, k are costs independent from the regulated firm's own costs and C represents the firm's own costs. β represents the cost pass-through coefficient.
On one side of the spectrum there are the so-called Cost-of-Service (CoS) regulation models (β= 1 and k=0). On the other side of the spectrum, there are the high-powered pure incentive regulation models (β=0). Clearly, there is a continuum of options in between (0<β<l).
Under the pure Cost-of-Service model (i.e. rate of return regulation) all the firm's costs are remunerated. The firm is ensured a fair rate of return ex-post: the firm will ultimately be able to pass all its costs to its customers. This has the advantage of protecting the firm's investment from the risk of expropriation, respecting private property rights. However this also raises issues with respect to efficiency. The firm knows that all its investment, no matter whether profitable, or desirable from a customer's point of view, will ultimately end up in its cost-base on which it will be allowed to earn a fair rate of return. This might result in 'gold-plating': the firm might act as if its cost of capital is lower than it really is.
In the US, regulators have spent considerable resources examining investment programmes to address the resulting adverse selection problem ex ante. At the extreme is the state regulation of investment decisions by hospitals in the US. Certificate of Need programmes require a hospital to obtain approval before undertaking certain investment programmes.5
Thus, although in theory the firm will recoup all its costs as long as the regulatory cost of capital is equal or greater than the firm's actual cost of capital, in practice CoS regulation seldom allows an automatic cost pass-through of all operating expenditure and capital expenditure. In practice, this tends to be subject to prudency tests, which might result in certain investments being disallowed ex post. Given the regulatory lag, depending on the period for which the regulator sets the rate of return, regulatory risk such as the exclusion of some investment from the allowed regulatory cost base ex post may force regulated firms to behave more prudently or excessively risk-adversely. This might result in sub-optimal investment decisions.
The dominant economic regulation model in the UK is RPI-X regulation. This form of regulation, pioneered by Stephen Littlechild in the 1980s, is based on the premise that prices rather than profits should be regulated.
RPI-X regulation allows the firm to adjust its prices by the retail price index (which measures consumer price inflation) plus/minus an ex ante determined X-factor. The maximum prices are determined ex ante for a set period (e.g. five years in the case of UK airports) and not meant to be adjusted until the next control period.6 This implies that if the firm is able to outperform the regulator's expectations, the firm keeps the efficiency gains until the next price review.
Given the information asymmetry between regulator and regulatee, one of the key objectives is to incentivise the firm to reveal its true costs by allowing the firm to keep efficiency gains within the price control period. The next period's prices are adjusted for inflation and the X-factor. 'X' depends on how the regulator assesses the trade-offs under her objectives given the cost and demand climate in which the firm operates. A high positive X-factor (i.e. RPI-X, resulting in lower real prices) might indicate that the firm revealed substantial cost savings in the past or it may indicate that the regulator sees considerable scope for (further) efficiency improvements during the next control period. A high negative X-factor (i.e. RPI+X, enabling a real price increase) is an indicator that the regulator might be placing more emphasis on the firm's planned investments and the firm facing rising incremental costs.
The degree to which an RPI-X regime provides strong incentives depends on the objectives of the regulated firm and the credibility and the duration of the control period, and the risk sharing specified in the price cap. A high level of cost pass-through results in a regulatory regime with weaker incentives, close to CoS regulation, whereas a credible regime that allows significantly less automatic cost pass-through could be more high powered. If a regulator is particularly concerned about rents then pass-throughs allow tighter caps without risking the viability of the firm. Generally speaking, RPI-X price cap regulation in the UK tends to a large extent to be cost-based, i.e. like CoS regulation it is mainly based on inputs rather than outputs. Low quality output specification, including service quality dimensions, is often a constraint.
The more the focus shifts to outputs, the more high powered the regulatory approach could become. One of the main reasons why regulators might be reluctant to use high powered incentives is the limited insight they have in the firm's true cost structure together with the concern about rents being earned.
By setting a price cap solely based on the 'most efficient' firm in the same or similar industry without taking the firm's own costs into account, the firm faces maximum incentives to operate as cost-efficiently as possible during the control period. However, if the firm cannot achieve the predicted efficiencies it might get into financial difficulties and approach the regulator for an interim review. This raises the issue of regulatory credibility.
One of the main issues when setting a price cap is the trade-off between the risk of bankruptcy if the cap is set too low and the risk of high rents if it is set too high. If the firm were to earn abnormal returns there might be (political) pressure on the regulator to intervene, and similarly if the firm were to get into financial difficulty the regulator might be pressured to intervene, especially when security of supply is perceived to be an issue.
The main problem is therefore asymmetric information between regulator and regulatee, which gives rise to considerable scope for gaming. Where the regulator positions herself will clearly depend on her main objectives. Where the regulator sets the cost pass-through coefficient will depend on how much she focuses on economic rents rather than incentives. This will at least in part be determined by the regulator's main objective given the statutory framework.
If cost reduction is the prime concern then it is often argued that the price cap might be tight, whereas if investment is the main concern the price cap is likely to be looser. If the level of cost-pass through is relatively low then a tight price cap might increase the risk faced by the firm as part of the profit curve is chopped off compared with the profit curve of an unregulated firm. This would potentially result in a higher cost of capital (and thus potentially looser cap). However, even a looser price cap might not necessarily provide the right incentives for investment. The main risk if the firm was to make excess profits is regulatory intervention during a control period. This gives rise to regulatory uncertainty for the regulatee and its financiers (potentially increasing the cost of capital if this risk is regarded as being non-diversifiable), it might also deter potential new entrants, given the potential regulatory threat.
Regulatory credibility and commitment are key determining factors of the firm's behaviour. The length of the period for which the price cap is set is therefore of importance. The shorter the interval between the setting of price caps, the closer RPI-X becomes to CoS regulation.
Long-term regulatory commitment increases incentives. The longer the control period, the stronger the incentives, because the firm will be able to benefit for a longer time from its cost savings. This especially applies when dealing with long-term assets.
The risk with relatively frequent resetting of the price cap is ex-post opportunism of the regulator. Future regulators might ex-post decide to disallow certain capital investments and thus exclude them from the Regulatory Asset Base (RAB). In a situation with high sunk costs, this is likely to result in the firm investing below the socially optimal level. In situations where infrastructure investment is seen as one of the prime objectives, a long-term price cap commitment is therefore desirable.
However, it is difficult for the present regulator to indicate her commitment to a long-term cap, especially considering that she might not be able to commit her successors. Regulatory commitments can never be as strong as contractual obligations. The time inconsistency problem is an issue, but one that is addressed by the establishment of regulators independent of day-to-day political control.
The forgoing stresses the importance of the institutional framework in which the regulator operates. The institutional framework will depend on how the government defines its role, the implementation of this definition and its perceived sustainability.
The relation between the regulator and the government should be transparent and set out in statute, ensuring the regulator's independence whilst also ensuring that the regulator can be held accountable. The government appoints the regulators and determines how much discretion the regulators have (by statute and/or other means).
Critically, the UK regulators, generally as single persons, were statutorily independent of the Government, although appointed and with budgets controlled by the Government. On the surface this was seen as important to reduce the risk of dominant market positions being abused but there was also the objective of reducing the risk of the Government intervening post-privatisation to keep prices too low and so at least partially expropriate the investment by private investors.
The idea behind the UK model is that as long as the regulator is given a clear statute she can be held accountable against her statutory objectives without compromising her independence. This enables the Government (which is the regulator's 'owner') to hold the regulator accountable against its statute. However, the other side of the coin is that the regulator is given considerable operating autonomy in carrying out her duty under the Act.
Regulatory governance can be regarded as the degree of the regulator's discretion and therefore the extent to which the regulator can be held accountable. Best practice in the UK is evolving with the operation of monetary policy via the Monetary Policy Committee perhaps leading the way in implementation while Ofcom represents current policy thinking, e.g. separation of the Chair from the Chief Executive Officer.
In how far the regulator is able to impose an effective incentive structure on the regulatee on behalf of society overall, will depend to a great extent on the institutional framework, including market structure, in which both operate. Incentive compatibility is the objective. Management of private firms, which aim at maximising shareholder value, face stronger incentives to improve performance than managers of state-owned firms. However, a new type of firm has been (re)-emerging at the interface between the public and the private sector: the so-called not-for-profit/not-for-dividend firms.
These types of firms are likely to face considerably fewer incentives even when operating in the private sector.7 These types of firms are tending to be highly geared which, when subject to economic regulation and other on-going "government involvement", gives rise to a whole range of issues. For the remainder of this chapter the focus is on "for profit" firms when referring to privately owned firms.
In order to design and implement a regulatory framework based on high-powered incentives, regulatory credibility is essential. However, regulatory credibility will, to a large extent, depend on how far the regulator can pre-commit and this will depend on the institutional framework and the approach to regulation from the government and regulator alike (i.e. how far is the ultimate aim the phasing out of economic regulation, duration of the control period, etc).
Like other UK regulators, the CAA (Civil Aviation Authority) is an autonomous body, set up by the UK government with all its board members appointed by the Secretary of State. The CAA consists of four groups: the safety regulatory group, economic regulatory group, directorate of airspace policy and the consumer protection group. The CAA only sets a price cap on airport charges at airports that have been designated by the Secretary of State. These are the three BAA-owned London airports and local government owned Manchester airport. The CAA can investigate complaints about the conduct of other UK airports (if their annual turnover exceeds £1 million) but cannot set caps on airport charges. It also manages the licence with NATS including the price control arrangements.
Under the Airports Act (1986), the CAA is required to set price caps for a five-year period, extendible by one year. The CAA formulates proposals for price caps but then has to refer the airports to the Competition Commission. The Competition Commission then reports back to the CAA with its recommendations. The final decision is made by the CAA, after consultation with interested parties.
The main weakness in this process is that there is no appeal against the CAA decision apart from resorting to judicial review. It potentially introduces greater regulatory uncertainty as two regulatory bodies are dealing with the one problem, which they might approach in a rather different way. This is not surprising given their different statutory objectives. The Government policy is to change this arrangement into the conventional one whereby the Commission is the appeal body that the regulatee can appeal to over a regulator's decision.
In determining the price cap the CAA is subject to its objectives under the Transport and Airports Acts, The CAA's statutory objectives in the case of airport regulation are as follows:
The CAA has also to take account of international obligations advised to the CAA by the Government.
A key difference with the Competition Commission is that whereas the Competition Commission can come with public interest findings, this is not in the remit of the CAA. If in the view of the Competition Commission the designated airports have acted against the public interest, it can come with a public interest finding under which the CAA has to take action to remedy the situation.
This process is rather different compared with that followed by the UK utility regulators (OFWAT, OFGEM, OFTEL8). These regulators determine the price cap without referring to the Competition Commission, as the latter is the appellate body. This process is more transparent. The expectation is that in the near future the Airports Act will be amended in line with the Utilities Act, thus bringing airport regulation more in line with utility regulation.
The provision of airport services in the South East of the UK is dominated by the BAA London airports (Heathrow, Gatwick and Stansted). In 1987, BAA was sold with a dominant position in two regions (Scotland also) and a golden share, making hostile take-overs more difficult. The BAA London airports serve approximately 90% of the South East market.9 Local-government owned Manchester Airport (MA) also has a considerable degree of local market power.
In both cases, market power seems to be the main rationale for subjecting these airports to economic regulation. Privately owned BAA, operating with a profit motive, and having considerable market power, might be expected to engage in monopolistic behaviour, i.e. increasing prices as a corollary to restricting output. How far this applies to local-government owned MA is debatable, as MA shareholders probably operate with public sector objectives. If, for example, regulation allowed MA to monopoly price, would MA's shareholders not intervene to address such pricing following airline complaints given its adverse effects on local/regional business and travellers?
In both cases, market structure is clearly an issue and ownership structure is potentially an issue as well. For a considerable time now, Heathrow, and to a lesser extent, Gatwick, have been characterised by excess demand. It might be argued that given the lumpiness of aviation infrastructure, the lack of seemingly adequate levels of capacity is not necessarily always inefficient. When real option costs are allowed for, it might be efficient for the market to undertake the investments "late" compared with the timing that a benevolent social planner would prefer. So some excess demand is probably optimal with efficient prices.
However, especially in the case of Heathrow, the persistence of excess demand is likely to come at considerable welfare costs. In a competitive market, the price mechanism not only ensures best use of scarce capacity but also functions as a signal to where and when new investment is needed. On the other hand, firms with market power, which operate under a profit motive, might want to restrict output and increase prices as a corollary. The firm's objectives are of importance in this context and hence its governance structure should be taken into account when defining the problem. The situation with respect to local-government owned MA is rather different compared with privately owned BAA, as MA operates under public sector objectives. It might therefore not be surprising that whereas BAA seems to invest too late, MA might have been investing too early, which is too costly given that it now has considerable excess capacity.
The key problems in relation to airport infrastructure in the UK are:
Given the variety of problems, it should be noted that the regulator has only one instrument: the price cap.
This raises the question of how far price cap regulation is the best available mechanism to address these problems. Price cap regulation limits price adjustment, preventing best use of scarce capacity and potentially distorting investment signals. The inefficient functioning slot market further exacerbates the problem. The lack of clear property rights with respect to slots is the key obstacle to more efficient slot allocation systems, such as slot trading. The current system prevents best use of scarce slot capacity.
Airport regulation in the UK is based on the single till RAB-based approach. The starting point of this approach is the regulated firm's projected cost base. The price cap is set to ensure that the expected return of the firm covers all its projected (and, by the regulator, allowed) costs (including the regulatory cost of capital) over the duration of the control period, A key issue for the regulator is therefore to determine the firm's RAB.
The valuation of the RAB is a key factor in determining:
Initial valuations of the asset base have generally been made by reference to market values at privatisation, as a proxy for value in use. Another way of determining initial valuations has been to use the present values of future cash flows assuming a pre-existing price regime, based on incremental or benchmarked costs. Optimised depreciated replacement costs would be another option.
However, valuation of the asset base is not just a one-off issue, in all successive price reviews the regulator will again have to establish the starting value of the RAB. In addition, the regulator will have to decide how to treat depreciation and how to deal with capital expenditure. The regulator will also have to determine the regulatory cost of capital.
In the case of previously regulated firms, the issue for the regulator is how to roll the regulatory asset base forward. In the UK, regulators use the real value of financial net investment as the basis for rolling forward asset base valuations. The starting RAB consists of assets rolled forward from the previous control period, i.e. the opening book value plus allowed capital expenditure minus depreciation for the period in question maintained in real terms.
The key advantage of using the real10 value of financial net investment is that prospective returns on investment will be driven by the cost of capital assessment and the incentive mechanisms put in place by the regulator. This implies that the users (or more generally the consumers) carry the inflation risk as the RAB will be maintained in real terms. It also implies that the cost of capital applied is in real terms. This approach ensures greater certainty over future valuations of the asset base and hence should aid investor confidence. Therefore, whenever firms are privately or part-privately owned, this approach is likely to be preferable over a current cost approach. The latter approach is based on the replacement cost of assets. From an investor point of view this implies that investors are penalised for technological advancements where incremental costs are lower than average costs. From a regulator's point of view, this could result in perverse incentives with respect to innovation, and dynamic efficiency, as financing new investment might become more costly under such an approach.
The CAA has adopted an approach based on projected rather than actual depreciation. If actual depreciation is adopted the airport might have an incentive to adjust its capital expenditure plans, or engage in accounting 'gaming', in order to maximise the RAB at the next review. Since depreciation is an accounting construct rather than related to any actual economic costs, the CAA considers this undesirable.
Using actual depreciation also provides an added incentive to reduce capital expenditure between reviews, which is inconsistent with the CAA's general view that investment incentives are the critical challenge faced by the South East airports.
However, in practice, price controls are often based on a combination of economic data (i.e. future cash flows) and accounting data. Initial RABs are adjusted in part by accounting depreciation. Future cash flows include projected operating expenditure and capital expenditure.
Airport regulation in the UK is currently based on a single till RAB based approach. In this approach not only the aeronautical but also the commercial revenues and costs of the airport are taken into account when determining the price cap for airport charges. The RAB therefore incorporates all airport assets regardless of their function and characteristics. The objective under the single till is to determine the level of airport charges so that the expected return across all activities reflects the regulatory cost of capital.
In practice, this involves costs and volume projections being translated into a multi-year price cap that is re-set periodically (five yearly) on the basis of newer information. The single till operates to offset the costs of monopoly airport services by the projected net profits on retail activities (less losses on certain public transport schemes). Depending on the level of commercial profits, this approach may result in the commercial activities bearing a substantial share of an airport's common costs. It potentially results in cross-subsidisation of the aeronautical activities if these fail to cover their incremental costs. However, it does not address excessive retail pricing.
The key drivers in the single till RAB-based approach are projected to be:
The charges are set to allow the airport to earn its regulatory cost of capital and thus depend on the difference between the total revenue required to do so and the projected revenue from sources other than airport charges. Under the single till RAB-based approach, the price cap does not only apply to revenues of monopoly activities but also to the revenue of commercial activities. This mixing of aeronautical activities with commercial activities could potentially result in distorted decisions not only in relation to the monopoly activity but also in relation to the commercial activities: investments in the latter have the same policies applied to them over the long-term as aeronautical investments.
The regulated airport in its investment analysis will allow for the single till RAB-based regulated price plus the expected net profits from the unregulated activities over the first five years multiplied by projected volumes. The investment and operating costs are a cost to the firm as in the commercial model in the first period. In second and subsequent periods, the single till RAB approach "takes away" any excess earnings including profits from non-regulated activities and sets regulated prices relevant to the project to allow on-going operating expenditure and sunk capital expenditure to be recovered.
One critical point is that this approach distorts investment incentives because it turns costs into benefits, in taking away any super profits. The price cap increases incentives for regulatory gaming with regulated firms having incentives to under-spend projected capital programmes to gain the cash-flow and hence business value benefit within the control period, thus managing the regulator's response at the next review.
It is therefore in the interest of the firm to ensure that its projected cost base is as high as possible when the regulator sets the price cap. As a result, the firm is likely to enter in gaming to present its projected costs and its projected cost of capital as high as possible as the firm will be able to keep the difference between its regulatory cost of capital and its true cost of capital during the control period.
This regime has generated real price reductions but has provided weak incentives for allocative efficiency and weak investment incentives compared to a situation where prices reflect much higher short or long-run social marginal costs and values that are likely actually to prevail.
Given the fact that airport infrastructure projects tend to be lumpy, one of the key issues is how to treat assets in the course of construction (AICC) under a RAB approach. The two main options are either to account for these costs when these assets become operational as indicated by an output focus or to remunerate them while still in construction. In the former case, prices would be lower while the assets are still in construction: the incentives for the airport to ensure that the assets come on stream as soon as possible are clearly strong. However, from an economic perspective it would be counterintuitive as prices should be relatively higher during the periods of congestion and vice versa.
The other approach is to remunerate the assets while still in construction. This implies that airlines would at least in part pre-fund the assets. Remunerating AICC might reduce the need for price hikes when new capacity comes on stream, although it might not eliminate the need for price increases altogether. From an economic perspective, it is more sensible that prices are higher when resources are scarce (i.e. situation of excess demand).
The second issue for the regulator to address when dealing with large and costly infrastructure projects is whether profiling should be used in determining the price caps. Profiling might entail revenue advancement from a later price control period to an earlier control period. The advantage of profiling is that it results in more stable prices and gives some assurance about regulatory credibility and risk. Without profiling, the regulator would have to commit herself to large price increases when the new assets become operational. This is potentially an obstacle to raising funds for these types of investments, given the fact that nothing is known about what the world looks like at that point in time, and the investment is substantially sunk.
Airlines are likely to exert maximum pressure on the regulator to limit price increases, thus preventing the regulatee and its financiers from recouping their costs. Given the fact that control periods are set for a relatively short-term (5 years in the UK), in relation to the economic lives of the assets, this is a real issue when dealing with long-lived assets. The risk of such an approach is that investment becomes either very costly (financiers might view the project as very risky and might want to be remunerated if they expect that this type of regulatory risk is not diversifiable) or it might not proceed at all (which is likely to result in considerable welfare losses), because under the UK regime the regulator cannot commit her successor.
In order to raise finance for major and potentially risky infrastructure projects with long-lead times, regulatory credibility and regulatory commitment are of utmost importance. Profiling therefore aids the sustainability of a regulatory regime.
In its price cap review, the CAA recognised that the application of single till cost-based price caps at the congested Heathrow and Gatwick airports has become increasingly problematic, forcing prices down and seriously out of line with the value of that capacity (as indicated by slot values) and the costs of new capacity.
However, setting caps to allow market-clearing prices, while permitting better use of scarce capacity, would provide weak incentives for development by the dominant BAA. The resulting rent transfers would be substantial, raising questions about the sustainability of such a policy.
In February 2002, the CAA proposed a compromise. The CAA proposed to maintain (modified) cost-based pricing for existing outputs at Heathrow and Gatwick while introducing more powerful incremental cost incentives for the delivery of new capacity. This would allow real prices to rise at the congested airports, reflecting the economic fundamentals and the long-term user interest. Nevertheless, although real prices would be allowed to rise, even under the proposed new regulatory regime, prices would have remained below (marginal) costs/values. Price profiling was to be used to prevent very large price hikes. Instead, over time, prices were to rise until in line with (marginal) costs.
The idea was to minimise regulatory risk through a simple and transparent foundation for the major investment programme planned, by bridging the gap between investment projects with long lead times and pay back periods and price caps that are fixed for only five years.
The price path commitment would have entailed a 20 year commitment on the price caps linked to current capacity and to Terminal 5, with arrangements for incentives for new investments, such as runways, to be considered separately and added on to the proposed commitment in a similar manner. The price path commitment was to be based on a dual till approach in order to provide headroom against negative shocks.
The objective was to give strong incentives to BAA to improve service quality and address the capacity-demand imbalance; however, the CAA was well aware of the regulatory risk involved. As the CAA set out in its original proposals, a move to a more high powered regulatory regime would hinge on regulatory commitment and hence on regulatory credibility. The move away from the single till was part of this commitment.
The fact that the Competition Commission has not recommended the dual till and has not recommended the price path commitment to be adopted and the fact that the current regulator cannot bind her successors, would have made the price path commitment less credible, which would have severely weakened its incentive properties. This has resulted in the CAA continuing with the RAB-based single till approach in line with the Competition Commission's recommendation. However, the CAA considers that the case for moving to a higher powered incentive regime in future would be substantially stronger if the airports were better able to demonstrate the potential benefits that it could bring, and considers that a range of possible mechanisms for providing enhanced investment incentives, and superior alternatives to both the single till or dual till, could be considered.
Regulation is a means to an end, not an end in itself. At the over-arching public policy level it is therefore important periodically to examine the trade-off between market failure and regulatory/government failure. Given the variety of problems regulation might seek to address, it should be noted that the regulator has only one instrument: the price cap.
Regulation is always second best to competition. Price cap regulation tends to limit price adjustment. The regulated price might not be consistent with long-run incremental costs, over the long-run. This might result in inefficient capacity levels resulting in excess demand and deteriorating service quality (e.g. delays). This would not only result in a loss of allocative efficiency but also of dynamic efficiency. Therefore, market structure options, while respecting property rights, should be used to the maximum extent if economic regulation can be avoided as a consequence.
Commercial firms, operating under a 'for profit' motive, face strong incentives to define and price outputs. However, if they have market power, they might restrict outputs with increased prices as a corollary. However, first best solutions may be unobtainable, but the potential downsides of economic regulation are significant. Regulated firms might spend a considerable amount of time and effort gaming the regulator rather than responding to the preferences of their customers.
In order to design and implement a regulatory framework based on high-powered incentives, regulatory credibility is essential. However, regulatory credibility will, to a large extent depend, on how far the regulator can pre-commit and this will depend on the institutional framework and the approach to regulation from the government and regulator alike.
Thus, when looking at the social loss function, regulation might not necessarily be the preferred option. It might be worthwhile to consider the potential for contracting between the airport and other parties. This would allow for risk (and return) sharing. Airlines would be an obvious contract counter-party for a major airport terminal development. It is well known that in US airports airlines often own terminals. Retailers would be other potential contractors with whom to lay off the project and volume risk.
1 The views presented in this paper are these of the authors and do not necessarily represent the views of their respective organisations.
2Steven Morrison and Clifford Winston (2000), "The Remaining role for Government policy in the de-regulated airline industry" in Deregulation of Network Industries, AEI-Brookings.
3A real option is the right but not the obligation to take an action that might consist of deferring, expanding, downsizing or abandoning the project at a pre-determined price for a pre-determined period of time. This flexibility is valuable and should therefore be taken into account in the cost of the project last sent. For example, a cost of undertaking a mutually exclusive project at time t is not being able to undertake it at time t+1.
4 Niskanen, W. (1971) Bureaucracy and Representative Government, Chicago; Aldine.
5 W Kip Viscusi, John M. Vernon and Joseph E. Harrington (2000) Economics of Regulation arid Antitrust, 3rd edition, p. 310
6 In practice only the regulated firms can seek an interim review.
7 E.L. Glaeser (2002) "The Governance of Not-for-Profit Firms", Harvard Institute of Economic Research, Discussion Paper Number 1954, April
8 Office of Water Services (OFWAT), Office of Gas and Electricity Markets (OFGEM), Office of Telecommunications (OFTEL).
9 The South East market being defined as Heathrow, Gatwick, Stansted, London City and Luton.
10 Hence the RPI component in the pricing formula.