VI Pricing by public bodies and economic regulation

Section Contents

• Introduction
• Basic economic principles
• Competitive market
• Dissemination cost pricing
• Classic monopoly
• Natural monopoly
• Price differentiating monopolist
• Practical considerations
• Externalities in practice
• Regulation
• Geospatial information providers
• Conclusions

Introduction

In Section II, we reviewed the economic and other arguments that can be advanced for government provision of certain types of information, including geospatial information, given the public good, externalities, natural monopoly, and national interest considerations. These arguments raise complex questions about charging for information and about the funding of government information providers.

As we have seen, the principal UK information providers are currently government agencies and so subject to public sector constraints and guidelines on funding and pricing. Privatisation is, however, a possibility under current government policy for some providers and is already being actively considered for HMSO.

There are two bodies of economic argument relevant to pricing and funding. One is the large volume of literature on public sector pricing which has been developed over many years; the other relates to the regulation of privatised enterprises which has been developed over the last fifteen years.

Both strands of analysis are based, however, on certain fundamental concepts of economic welfare and efficiency. In this section, therefore, we:

• explain the relevant economic concepts and then bring together the main elements of both strands of analysis; and then
• consider the implications for the pricing and funding of government information.

We have, in this report, restricted ourselves to the core concepts and key results of economic theory in order to develop practical arguments. A bibliography to this report highlights key sources for further reading.

Basic economic principles

A fundamental concept in economics is that of economic welfare. The underlying principle is that best use should be made of the scarce resources in an economy. Welfare is "maximised", and an outcome is defined as economically efficient, where no one person can be made better off without someone else being worse off. Defining welfare "maximisation" in this weak form avoids the need to make any interpersonal comparisons of welfare or utility.

Maximising welfare relies on achieving:

• "allocative" efficiency - resources being allocated to the activities where they are most valued; and
• "productive" efficiency - resources being used in the most effective way possible.

A key conclusion of welfare economics is that the "free market" will result in an optimal outcome only if certain conditions are fulfilled including, for example, the presence of a large number of producers and consumers, all with full information, in each market. These and the other necessary conditions are very demanding ones. Furthermore, the concept of economic efficiency does not help to address any issues relating to the distribution of income, ie equity considerations. Nonetheless, the "perfect competition" model provides a useful starting point from which to explain the role of governments in markets where the necessary assumptions do not hold, and in particular in markets where there are fewer buyers and sellers and these buyers and sellers do not have full information.

We accordingly consider below three basic economic models: a competitive market, classic monopoly and natural monopoly. In each case, we examine the implications for output, prices and economic welfare. Each model is illustrated by a diagram. These diagrams show combinations of demand curves (D), supply curves (S), average cost curves (AC), short run or long run marginal cost curves (MC) and a marginal revenue curve (MR). Each of these concepts is explained further below. These models only consider prices and quantities; they do not extend to other important dimensions such as quality. They are, nevertheless, a key source of important economic insights.

The demand curve (D) shows the relationship between output demanded by customers and price. This reflects the willingness and ability of customers (or users) to pay for the product; this will depend on their preferences, budget constraints, and relative prices. The gap between actual price and the price that a consumer would be willing to pay gives a measure of consumer welfare and is called "consumer surplus".

The supply curve (S) shows the volume that producers are prepared to supply at a given price. The difference between the actual price and the price that the producer would be willing to supply at gives us a measure of " producer surplus".

Total economic welfare is calculated by adding total producer and consumer surplus.

The marginal cost curve (MC) shows the additional cost of producing an extra unit of output. The marginal revenue curve (MR) shows how the total revenue earned by producers changes following a change in the price and, hence, the output demanded.

Competitive market

In the perfectly competitive model, it is assumed that there is a large number of well-informed buyers and sellers. Under these conditions, producers are "price-takers". By this we mean that producers must accept the prevailing market price, which will be determined by the overall balance in the market between supply and demand; with a large number of sellers any attempt by one to raise prices would result in customers switching to other suppliers.

Under these conditions, a profit-maximising firm will increase its output to the point where marginal cost equals the market price; so long as the market price exceeds marginal cost a producer can increase profits by expanding output.

Figure 6.1 shows that, with price equal to marginal cost, producers will cover total costs including a "normal" return on capital. For simplicity, we have assumed a constant average cost curve as, in this case, there are no fixed costs and so the marginal cost will be equal to the average cost. It should, however, be noted that the results generated by this model do not rely on this assumption. In Figure 6.1, therefore, the average cost curve (AC) is the same as the marginal cost curve (MC) and the industry supply curve (S).

In Figure 6.1, consumer surplus is the shaded area, revealing all consumers who would have been prepared to pay more than the marginal cost price. There is no "producer surplus" at all.

This outcome is welfare maximising; we can see from the diagram that:

• if price was increased above marginal cost, (say to P1), the level of output would fall to (O1 ) and some consumers would no longer buy the good even though they are prepared to pay the cost of supplying them (P);
• if price was below marginal cost (say at P11), the level of output would rise (to O11) and some consumers would buy the good when the value they attach to it is less than the cost of production (P).

The key results from this model are, therefore, that under perfect competition:

• price is equal to marginal cost;
• producers earn a normal profit (zero producer surplus); and
• economic welfare is maximised, so the outcome is economically efficient.

These results can be generalised for the whole economy where similar conditions prevail in every market. Complex mathematical proofs supporting this conclusion are amongst the most important conclusions of microeconomic theory. In more popular terms, this is the conclusion that the "invisible hand" of the market will ensure an efficient allocation of the resources in an economy.

Dissemination cost pricing

Figure 6.2 compares the competitive market outcome with the case where price is set at the marginal cost of dissemination and the line through ED effectively becomes the supply curve (S). A constrained pricing rule like this may, for example, be imposed on a public body.

The lower price means that demand and output will increase but, with price below the total marginal cost of supply, the producer will make a loss. The figure shows that this will result in an overall welfare loss because the additional benefit to users will be more than offset by a reduction in "producer" surplus. On the diagram:

• the area ABDE is the additional benefit to consumers (ie, the increase in consumer surplus);
• the rectangle ACDE is the loss to producers (ie the negative producer surplus); and
• the triangle BCD is the net welfare loss.

Dissemination cost pricing, supported by government subsidy, would therefore appear to result in a sub-optimal allocation of resources, since the government subsidy required to support this would be greater than the benefit to users of the lower price.

In some cases, however, dissemination cost pricing may be justified on economic grounds if there are significant "external" benefits resulting from the wider provision. Figure 6.3 illustrates that, where there are "external" benefits, dissemination cost pricing can be welfare maximising.

In Figure 6.3, the "external" benefits are represented by additional demand and a "social" demand curve is illustrated. This "social" demand curve crosses the marginal cost curve at C. If the provider is constrained to price at dissemination cost, then the outcome at D generates greater welfare than the "free market" outcome at B. The increase in consumer surplus is the area ACDE plus the area BFC. This more than exceeds the decrease in producer surplus which is given by the area ACDE alone. The net increase in economic welfare is the triangle BCF.

In the example, the level of cost recovery generated has fallen from 100% to around, say, 50% but overall economic welfare has increased.

Classic monopoly

We turn next to the case of classic monopoly. By definition, there is only one producer who can influence the price. The classic monopoly model is shown in Figure 6.4. For simplicity, we have again assumed a constant average cost curve which will be equal to the marginal cost.

A monopolist will maximise profits by setting production at the level where marginal revenue equals marginal cost (O) and then charging the price which the market will bear (P). The diagram shows that, at lower levels of output, the marginal cost is below marginal revenue, which means that profit can be increased by increasing output. Beyond this point, however, marginal revenue is below marginal cost. The price is given by the demand curve.

Compared with the competitive outcome, the equilibrium here involves:

• a lower level of output (O compared with O1);
• a higher price (P compared with P1);
• profits in excess of those required to earn a reasonable return (ABDE is "monopoly" profit or producer surplus); and
• a reduction in economic welfare; the loss of consumer surplus - resulting from higher prices is ABCE, which is more than the benefit to the producers in terms of higher profits (ABDE); this net loss, represented by the triangle BCD, is called dead-weight loss.

The classic monopoly outcome is not, therefore, efficient in overall economic welfare terms. This is one of the starting points for economic regulation.

Natural monopoly

In the case of the natural monopolist, we relax the assumption that there are no fixed costs. Instead, we assume very high fixed costs and low marginal costs such that it would be inefficient for more than one firm to supply the market because of the duplication in fixed costs involved. Typically, there are also types of public goods involved.

Natural monopoly is illustrated in Figure 6.5. It shows:

a falling average cost curve (AC), because as output increases the fixed costs can be spread over a greater number of units; and

a marginal cost curve (MC) that is always lower than the average cost curve (AC) because, mathematically, if average costs are falling they must exceed marginal costs.

The implication of this is that marginal cost pricing (price set equal to P), the simple pricing rule that results from the analysis of the competitive model, would result in significant operating losses for the monopoly producer (area ABCD). These losses could be eliminated by setting prices equal to average cost (shown as the Full Cost Recovery price, P1). In practice, the profit-maximising monopolist, if unregulated, would reduce output to the point where marginal revenue equals marginal cost thus increasing the price further (to P11). At this output, the producer would earn monopoly profits (EFGH).

One solution, in principle, is to enforce pricing at marginal cost with the losses met by the taxpayer. The justification for this is that average cost pricing would result in a welfare loss because people prepared to pay the marginal (or additional) cost of supplying them are excluded from the service (shown as the difference of output between O and O1). There are, however, other problems with marginal cost pricing in this situation because:

• there are economic costs associated with taxation which may create distortions in the economy;
• there is a risk of inefficient allocation of resources if losses exceed the true value to users of providing the product or service;
• there will be over-consumption by individuals not prepared to pay the "true" costs of supplying the product (including investment in maintaining supply);
• there are risks of under-provision because of the lack of incentives to invest in capacity.

The last two points are related and provide strong arguments for what is called long run marginal cost pricing. Long run marginal cost pricing provides appropriate signals to users; it also provides an incentive for providers to respond to the needs of the market.

This approach still leaves the issue, however, of how to structure a tariff, when short-term marginal costs are very low and there are benefits from high output levels. A solution to this problem can be a "two- part" tariff where a customer is charged a fixed fee that reflects the long-run costs associated with maintaining services and a unit charge based on short-run marginal costs. Well-known examples of this practice include telephone, gas and cable TV charges.

Two-part tariffs can also be helpful in achieving a balance between cost recovery and marginal cost pricing and are, therefore, also used:

• where long run marginal costs (LRMC) are falling and so pricing at LRMC will still result in losses to the organisation; and
• to fund activities with a social rather than commercial imperative, such as supplying customers in rural areas.

Using two-part tariffs to tackle the second issue can, however, raise problems of "cream skimming". Here, competitors enter some market segments and can undercut the "monopoly" supplier, even if they are less efficient; they do not have to cross-subsidise non-commercial activities. The result can be that a "monopoly" supplier loses more and more customers and is eventually left with only loss-making activities.

Price differentiating monopolist

An alternative solution is price differentiation, where the fixed costs of supply are recovered from higher value users, but users with a lower willingness to pay are charged a lower price. In Figure 6.6, we consider the case of "perfect price differentiation" by which economists mean that the producer charges each customer the full value he or she attaches to the product or service, irrespective of the costs of supply.

The ability to differentiate prices in this way relies on certain key assumptions being fulfilled, namely:

• the producer knowing how much value each user attaches to a product or service; and
• the producer being able to prevent users from reselling (or leaking) to other users (ie no arbitrage).

In this model, the producer will supply all consumers prepared to pay the short run marginal costs or more. As a result:

• output will increase to the perfectly competitive level (O);
• there is no consumer surplus as this will all be captured by the producer as monopoly profit or producer surplus; and
• the total of producer and consumer surplus will be the same as that in the competitive model, but there is a large direct transfer of surplus from consumers to producers.

An implication of this is that, on strict "economic welfare" grounds, one should be indifferent as between the perfectly competitive and the perfect price differentiation outcomes. In practice, however, greater weight tends to be attached to consumer welfare than to producer welfare, and such a transfer of surplus from consumers to producers may be regarded by policy-makers as a deterioration in social welfare. Whilst this is in part an issue of equity and fairness rather than a purely economic issue there is also some economic justification for this view.

In economic parlance, although the perfect differentiation model assures allocative efficiency, it may not result in productive efficiency. The strength of the monopoly may mean that attention to costs is lax and resources are not utilised efficiently by the monopolist.

Price differentiation combined with economic regulation can provide a solution to the natural monopoly problem where marginal cost pricing results in a loss; it should be noted, however, that the economic costs of productive inefficiency may, in some circumstances, be lower than the costs of the regulation necessary to eliminate them.

Some additional arguments can, however, also be advanced against organisations which practice price differentiation not based on differences in the costs of supply. These include the following:

• there may be a distortion in competition in downstream markets or implications for the competitiveness of user sectors; this is particularly relevant for geospatial information which is often used as an input to another product or service; and
• an incumbent may be using selective price cutting to deter or prevent potential competitors entering a market.

This last practice is called "predation" and is one of the key reasons economic regulation is used to control the behaviour of natural monopolies.

Practical considerations

Economic theory provides an important framework for policy-makers concerned with pricing by publicly-owned bodies including geospatial information providers. It provides useful guidelines on the types of pricing policy that are likely to be broadly consistent with maximising economic welfare.

There are, however, major practical considerations when translating these key principles from economic theory into actual pricing decisions. For example:

organisations do not operate in the static world of many economic models; nor do they have perfect knowledge of the demand curves they face or the "external" benefits of their products or services;

• cost structures are often much more complex than those of the simple theoretical models, with large elements of cost being capable of being allocated or shared between different outputs on many different defensible bases;
• the simple theoretical models do not take into account the quality of the product or service that is provided; they consider only the quantity;
• the models do not take into account the scope of provision of a "public good" which supports many users with similar but not identical needs (see below); and
• there are anyway likely to be major distortions elsewhere in the economy constraining the maximisation of economic welfare which tend to undermine the importance of achieving precise optimal pricing in any single market.

As we saw in Sections III and IV, the HM Treasury Fees and Charges Guide attempts to bridge this gap between economic principles and practical policy-making on pricing issues.

A key practical issue here for many public bodies is, when there are "external" benefits from the provision of information and both public sector and private sector users with similar but not identical needs, how should the "fixed" costs of provision be shared between them (this is an example of the public good scope issue highlighted above)?

Policy-makers and information providers usually have to tackle this issue:

• without full knowledge of how individual users value their product or service;
• with limited scope to set differential prices because of arbitrage, statutory limitation or the avoidance of hidden taxation; and
• without an agreed estimate of the "external" benefits of the provision of the product or service.

Externalities in practice

As a consequence of all of the above, the "external" benefits of many geospatial databases, especially OS' National Topographic Database which forms the framework on which much other data is collected, certainly can not all easily be valued. This is because they are highly disparate and often inextricably linked to the provision of other public goods. It is possible to approach the problem however, through the development of a matrix or impact statement which represents a first stance in linking options to outputs. This analysis could be developed by reviewing the implications for user groups in more depth through, for example, questionnaires and then possibly using a weighting system to develop a single output measure. It may also be possible to value some of the specific "external" benefits drawing on

Table 6.1: "External" benefits of consistent national geospatial dataset: an example using the British National Topographic

Types of Economic Benefit	Who benefits	Area of Benefit	Key Beneficiaries
Support to emergency services	All GB citizens	Support to defence of the realm	All GB citizens
Support to better business and policy-making	All GB citizens	Higher business profits	Benefit estimates from sample business and policy decisions
	GB businesses	Higher economic benefits or cost savings from policy decisions	Questionnaire to identify nature of benefits recognised by major Ordnance Survey users
Facilitate planning and co-ordination between key users	Local authorities	Savings on major infrastructure projects	Assessment of types and scale of projects where Ordnance Survey data is vital
	Utilities	Avoidance of duplicative mapping costs	Benefit estimates for sample projects (eg Sizewell C, road building, housing estate construction)
	Taxpayers	Time savings
	Infrastructure companies	Environment benefits	Questionnaire to identify nature of benefits recognised by major Ordnance Survey users

techniques to value, for example, time savings and the avoidance of premature death and non-fatal injuries. A preliminary framework is included as table 6.1.

In passing, we also note that Ordnance Survey delivers these benefits currently at a net cost to central government of around £15 million per annum whilst its paying customers contribute almost four times that amount in annual revenues. Although we have not undertaken or seen any quantitative analysis, £15 million seems a relatively small annual sum in relation to the likely size of the "external" benefits involved. There are, of course, also substantial sunk costs invested in the underlying datasets.

In practice, then, pricing policies have to be determined by economic and accounting analyses supported by reasoned argument and judgement. The neat solutions of the theoretical models and diagrams highlighted earlier in this section are usually not practicable because of a lack of information.

Regulation

All of the government information providers discussed earlier are in some sense already regulated. The form of this regulation varies somewhat but all answer to Parliament through Ministers. In some cases this is very direct: Ordnance Survey, for instance, has a Minister and its targets are approved by that Minister and promulgated in Parliament. The Public Accounts Committee and the Select Committee on the Environment make periodic inquiries into Ordnance Survey and the National Audit Office probes OS' accounts and value for money to the taxpayer. Members of the public, the eight Ordnance Survey Consultative Committees and, indeed, all other organisations are free to lobby the Minister to change policy. In the 1980s, for example, there was a major inquiry and public policy statement made in response to a challenge made by commercial publishers. In some other cases, this form of regulation is more indirect: the British Geological Survey for instance answers to the Council of the Natural Environment Research Council, members of which are appointed by the Secretary of State for the Environment.

In the last 15 years, however, government policy with regard to ownership of the provision of certain goods and services has changed dramatically, with moves towards deregulation and privatisation of many markets, including natural monopoly businesses. There has, been recognition of a continuing need for economic regulation in many of these markets. It is against this background that a considerable body of literature has developed on how regulation can provide appropriate incentives to encourage allocative and productive efficiency and, hence, help ensure overall economic efficiency.

The six main issues in economic regulation are:

• unnatural monopolies where a single producer is not more efficient than potential competitors;
• predation where a monopolist reduces prices to below marginal cost in order to deter or prevent new entry;
• unfair or unequal access where a monopolist might refuse or seek an excessive price for access to its infrastructure, network or database, in order to restrict or prevent competition;
• allocative inefficiency where output and prices do not mirror the outcome in a competitive market;
• organisational or productive inefficiency where resources are not used efficiently by organisations with market power; and
• inefficiencies where a monopolist does not take into account "external" costs and benefits.

These issues reflect the fundamental concepts of economic welfare explained at the start of this section and the themes highlighted on pricing rules. This is not surprising as pricing behaviour is a key concern of economic regulators. Furthermore, this linkage highlights that similar issues need to be tackled irrespective of the "ownership" of the producer.

A range of instruments is used in economic regulation including:

• restrictions on pricing policy to prevent predatory or exclusionary behaviour; quality of service standards; obligations to supply certain products or customers; and, possibly, direct regulation of profits or prices.

In institutional terms, there are number of possible regulatory models to implement these instruments ranging from:

• regulation by general competition law; through to
• specific industry regulation.

The first approach is usually the least interventionist and the least cost option. The Office of Fair Trading (OFT) and Monopolies and Mergers Commission (MMC) only become involved in the event of complaints. The bus industry is a high profile example of this regulatory model in the UK. In this instance, however, there is now pressure for a dedicated regulator given the volume of cases and the complexity of the economic issues involved.

A more interventionist variant of this option is to introduce special provisions giving the competition authorities additional powers and responsibilities to regulate. These might include, for example:

• an obligation to notify competition authorities of pricing structures; or
• a requirement on the competition authorities to monitor compliance with any licence obligations regarding pricing and quality of service.

The case for moving further to a separate industry regulator is stronger where:

• the business is profitable;
• a producer has a very strong market position that is likely to be sustained;
• there are potentially high economic and social costs associated with not regulating the market;
• products and markets are complex so special sector knowledge is required for effective regulation; and
• buyers are not well informed or have little buying power.

A dedicated regulator typically operates a system of price or rate of return regulation. The utilities provide an example. Rate of return regulation is used in the United States for regulating the utilities and involves a cap on the rate of return on capital that can be earned. Rate of return regulation, however, provides little incentive to improve efficiency as all benefits are, in theory, passed immediately through to the consumer. In practice, rate of return regulation in the US has become highly legalistic.

In the UK, by contrast, the utilities are subject to price regulation in the form of a limit on price increases or a requirement to reduce prices by a minimum amount each year. This is expressed, in real terms, using a formula of RPI plus or minus a factor. The key aim with this system is to provide an incentive to producers to improve efficiency. To be successful, it is vital that quality standards can be measured and monitored. It also works best where the output measure is stable, such as a unit of energy (e.g. a kilowatt hour).

The price cap formula is usually set for a five or ten year period and tends to incorporate a target improvement in efficiency. To the extent that additional efficiency savings are achieved, additional profits are thus, in theory, retained by the utility until the next review of the formula. There are some provisions for intervention within this period, but only in particular circumstances. The Monopolies and Mergers Commission acts as a "court of appeal" in the event of a dispute between the economic regulator and the regulated business.

New Zealand provides an interesting case study because, unlike in many major economies, its utilities are not regulated directly, but are subject simply to general competition law combined with an obligation to provide the general public with key accounting and other types of information. The view taken is that the mere threat of stronger economic regulation is a sufficient constraint on potential monopolistic behaviour.

Geospatial information providers

We have rehearsed above the theoretical background relating to a consideration of pricing and regulatory issues. Against this background, the key economic characteristics and principles that are applicable to geospatial information providers are summarised below.

As noted in Section II, geospatial information provision has several important economic characteristics:

• high fixed, collection costs and low variable, dissemination costs;
• non-rivalry in consumption (ie multiple users can consume the same information);
• there are many different uses and, hence, markets and the same information may have different values in those markets (ie the structure of demand is highly complex);
• there are additional "external" benefits from consistent dataset standards;
• a need to invest in regular collection activities to maintain the value of the dataset;
• key changes in the underlying business economics partly driven by developments in and access to electronic technology.
• the difficulty of defining a unit of information, especially where data integration and value-adding is practised

Furthermore, as we note in Section II, these characteristics imply that geospatial information is, therefore, a public good. It can be argued that simply the development and maintenance of a geospatial dataset comes close to being a "pure" public good and to meeting all the conditions of zero marginal costs, non-rivalry in consumption and "non-excludibility".

In practical terms, however, geospatial information provision is a slightly weaker form of public good because there are, albeit relatively small, real costs associated with geospatial information dissemination and individuals can be restricted from using products or services derived from the underlying dataset.

These characteristics highlight why governments throughout the world have not left geospatial data provision purely to the market. Section II explains in more depth the links between public goods, market failures, and the provision of these goods.

The US solution has been to try and separate the maintenance of the public good, a geospatial dataset, from its use by implementing a policy of pricing at dissemination cost. Figure 6.5 illustrates how an organisation with high fixed costs and low marginal costs (ie a declining cost curve) will make losses if it prices at short or even at long run marginal cost.

In effect, no attempt is being made in the US to recover the costs of the processes to support geospatial dataset maintenance. These costs are funded directly by federal government. Figure 6.3 shows how this type of pricing policy can help to maximise economic welfare as there are "external" benefits to users from the availability of consistent datasets.

This approach to policy, however, relies largely on government to determine the level of provision of the geospatial dataset, in terms of scale, quality and maintenance procedures, without any economic signals from users. There are, with geospatial datasets, important choices to be made because the underlying information has so many different uses and values and, hence, the structure of demand is highly complex. Many dimensions of the 'product' are important.

In the UK, there has been pressure on cost recovery levels. Using the example of Ordnance Survey, consultation with key user groups has shaped the nature and provision its geospatial datasets and through more commercial pricing policies, these user groups have contributed to the core information collection costs. These pricing policies have been sustainable because of copyright protection.

Although the costs of geospatial data providers like Ordnance Survey still exceed revenues and, hence, they generate operating losses, competitors and regulators may still be interested in the underlying pricing policies. This is because the combination of the cost structure of information provision (which has natural monopoly characteristics for any specified region) and copyright protection mean that the organisation has some monopoly power. Purely commercial pricing policies may, therefore:

• reduce output levels below the optimal (the monopoly prices and outputs on Figures 6.4 and 6.5) where price differentiation is difficult;
• make it possible to cross-subsidise "non-commercial" market segments through monopoly pricing in other segments; or
• make it easier for the organisation to operate in downstream markets with an unfair advantage.

Hence, economic regulation may be important, especially to downstream resellers, competitors and policy-makers even though excessive profit is not being generated overall. If, despite the difficulties of doing so, geospatial information providers were to be regulated more formally, key issues would include:

• the most appropriate measures to ensure that value added resellers have fair and equal access to the underlying geospatial databases and are not subject to predatory pricing;
• how to assure the continuing quality of the underlying geospatial databases;
• whether and how to take into account any "external" benefits, possibly through National Interest Mapping Contracts;
• whether to regulate any prices directly and, if so, which prices; and
• whether to set any limits on differential pricing; how to achieve a consistent and stable definition for a unit of information

In the case of information, there are in principle a number of institutional options for regulation, including:

• continued reliance on existing models;
• an Office of Information Regulation covering current government information providers;
• individual dedicated regulation for each provider;
• a more general office of Economic Regulation covering, say, both the utilities and information providers; and
• the threat of appeals to the competition authorities.

Given the relatively small size of the geospatial information providers compared with the utilities, the "weaker" monopoly positions that exist in some markets, and the reasonably well-informed user groups, there is a strong case for relatively light regulation either by a government department, the competition authorities or a general office of economic regulation. Furthermore, this regulation would have to be enlightened and expert in the particular fields of activity being regulated given the complexities of information products (which also vary somewhat between different fields).

If, however, a regulator were to be given responsibility for geospatial information providers using an RPI-X formulae then there would be a number of special practical issues to consider:

• a suitable method would be needed for measuring output, assessing quality, and computing future demand for geospatial data provision (output measurement is much more complex than for utility businesses and demand patterns are probably less stable);
• a justification for the underlying structure of controlled prices would be required (many very different products and services can be derived from access to the underlying geospatial datasets);
• the step function in maintenance costs often associated with geospatial databases would need to be dealt with;
• an approach would be needed to define the scope for productivity improvements in markets where customer needs are changing regularly; and
• methods of assessing any potential cross-subsidy between the regulated and unregulated businesses would need to be developed.

Overall, dedicated price control regulation is an expensive option in terms of the input both by the regulator and by the businesses themselves. It may be justified for the utilities which are very large and important sectors in the economy; for smaller sectors, including geospatial information providers, some of the regulatory issues are more complex and, hence, the cost of economic regulation on this model compared with the benefits are likely to make it unattractive. It should not, however, be ignored as a future possibility.

Conclusions

The key conclusions from this section on pricing by public bodies are:

• under conditions of perfect competition, price will equate with long run marginal cost which, in turn, will ensure economic efficiency (this is the theoretical justification for the US Geological Survey policy of pricing at dissemination cost);
• if there are "external" benefits from the provision of a good then lower prices will improve economic efficiency;
• the "classic" monopolist will set output below the level which would be obtained in a competitive market and charge higher prices in a manner which reduces economic efficiency;
• when there is a natural monopoly, two-part tariffs can be useful to establish a price structure that recovers full costs without distorting output levels too significantly;
• a monopolist who can set differential prices will generate an economically efficient outcome although there will be a major transfer of economic benefit from consumers to the producer compared with the competitive outcome;
• this transfer may dull the monopolist's incentives to use resources efficiently and, hence, lead to productive inefficiency; and
• whilst economic theory provides a useful background for considering pricing by publicly-owned bodies, there are, in practice, highly significant practical considerations.

The key conclusions on economic regulation are:

• it is concerned with similar pricing issues and aspects of economic efficiency as the public enterprise pricing literature;
• the need for regulation arises where there is significant and sustainable monopoly power creating a conflict between profit maximisation and user needs resulting in economic inefficiency;
• economic regulators will be concerned about aspects of price differentiation, especially if a monopolist is seeking to protect his position through predatory pricing; and
• the model of regulation used will depend, in part, on the perceived costs and benefits from regulation.

The key conclusions with regard to geospatial information provision are:

• the pricing policies of Ordnance Survey and some other government information providers mean users contribute to core collection costs and have, therefore, influence on the specification of geospatial datasets;
• although revenues of many of these bodies do not exceed its costs, competitors and regulators may still be interested in the underlying pricing policies;
• options for economic regulation in the UK for geospatial information providers range from reliance on the competition authorities and the threat of stronger economic regulation through to the appointment of a specialist regulator; and
• if a specialist regulator were to be appointed, price control based on an RPI-X formula is the most likely model but a strong case would be needed to justify the costs for what is both a relatively small sector of the economy and complex to regulate compared with the utilities.