The concept of an unfair burden first arose in the context of funding social obligations in the communications sector. In particular, markets have now been deregulated, but former incumbents such as BT in the UK, and eircom in Ireland, continue to take sole responsibility for social commitments, including quality of service targets, and universal access to the network. Under European law, incumbent operators can recover the (net) cost of providing these universal services, but only if the services represent an unfair burden.
The law does not clearly define ‘unfair burden’. In the majority of EU countries, the provision of universal services has not been deemed an unfair burden. In most cases, this has been assumed rather than explicitly stated. As discussed below, it often reflects the fact that much of the burden arises in parts of the market where there is residual market power, which implies that any ‘burden’ will be recovered through regulated prices.
However, this is not always the case. In Ireland, for example, the incumbent communications provider, eircom, has applied for funding of the net cost of the universal service. The analysis of the net cost was that it represented approximately €5m in 2009/10, at which point eircom was found to have a market share of marginally above 50%. In this case, the conclusion was that, on balance, this net cost did not represent an unfair burden. This article considers the meaning of ‘unfair burden’, how an unfair burden might arise in communications, and how the concept might be applied more generally in infrastructure sectors with social obligations.
Step 1: what is an economic burden?
In economic terms, the concept of a burden being imposed on one operator or a group of operators is well understood. A burden arises when an operator is required by some form of external intervention (regulation, legislation, or less formal intervention such as voluntary undertakings) to offer a level of service that is different from that which it would offer in a competitive market.
A simple example is quality of service obligations. In some markets, one or all operators can be required to meet minimum quality of service obligations that require a greater service performance from the individual operator or from the market as a whole than might be offered for purely commercial reasons. Figure 1 illustrates that such obligations may make products more attractive, but they are also likely to increase prices.
Figure 1 Impacts on service quality regulation in unregulated markets
On the assumption that, in the unregulated market, the operator set prices and quality levels in order to maximise profits, the operator will presumably be worse off if it faces an additional quality of service obligation. This is illustrated in the box below.
If an obligation, such as quality of service requirements, results in higher prices, is it such a bad thing? Not necessarily. The implication of imposing a properly designed obligation is that, while the regulated company may be worse off, society could be better off. In other words, the value placed by society on the benefits from the regulation could outweigh any economic loss that may occur as a result.
In the context of a universal service obligation (USO) in the communications sector, for example, this value would reflect the benefits in terms of social cohesion from all citizens having the same access to core communication services. Table 1 gives examples of such obligations.
Table 1 Examples of obligations in different industries
Source: Oxera, based on OECD (2005), ‘OECD Communications Outlook’; Ofcom (2012), ‘4G Coverage Obligation Notice of Compliance Verification Methodology: LTE’; ECORYS (2005), ‘Development of competition in the European postal sector’; ECORYS (2011), ‘Main Developments in the Postal Sector’; European Council (1998), ‘COUNCIL DIRECTIVE 98/83/EC’; and Webb Henderson and Strategy and Policy Consultants Network (2014), ‘The Future of Regulation: An Analysis of Developments in EU Telecoms Markets and the Implications for the European Commission’s Review of Relevant Markets’.
Step 2: introduction of the concept of an unfair burden
The concept of an unfair burden was introduced in the telecoms sector by European Directive 97/33/EC, as later amended by Directives 2002/22/EC and 2009/136/EC, which further developed the definition of universal service. The principle of an unfair burden was to allow for the potentially competing objectives of European communications legislation, where national monopoly operators were to be required to open their markets to competition, but at the same time to retain various obligations, including an obligation to provide universal access to the fixed network, including in non-commercial (often rural) areas. For example, in the UK, BT (and only BT) must offer services on request, and meet minimum quality of service standards for those services, including fault repair targets.
There are examples of the USO being found to be an unfair burden for the universal service provider (USP). In Portugal, in 2009–10 the communications regulator, ANACOM, found that the former incumbent, Portugal Telecom (PTC), had an unfair burden. This was in the context of PTC’s market share and profitability falling sharply, with market share below 80%, and negative profitability in 2009.
In the UK, the providers BT and KCOM have to meet USO conditions. In 2005, the communications regulator, Ofcom, estimated that the costs of the USO for BT were £50m–£70m.
However, despite BT now having only around 38% market share of the fixed lines in the UK and about 30% of fixed broadband provision, which are well below the levels of PTC in Portugal, Ofcom did not consider that it was unfair for BT to shoulder these obligations. In fact, the last time Ofcom considered whether there was an unfair burden was ten years ago. Why?
Step 3: the alternative concept: a ‘fair burden’
There is one legal test of what might make a burden ‘fair’: the Base case. In this case, the European Court of Justice (ECJ) concluded that the fairness of financing the USO related to a number of criteria, described in the box below. These include ‘the ability of the undertaking to bear’ the net cost. In other words, whether a burden is unfair is affected by the financial position of the company that bears the burden.
Equally importantly, the ECJ recognised that there was no one-size-fits-all solution, that a net cost does not necessarily imply an unfair burden, and that it should be for national authorities to determine how this would work in their respective markets.
Figure 1 in this article shows that a direct effect of imposing additional costs on operators will generally be to reduce profits. Therefore, if such costs are imposed on only one operator in a market, this is a form of market distortion. However, this will not necessarily mean that it is an unfair burden, as the impact on competition of imposing quality obligations, such as a USO, will depend on the market structure.
Table 2 compares a range of market structures, and what the imposition of a USO or quality of service obligation might mean in those markets.
Table 2 Impact of service obligations in different markets
Figure 2 illustrates the circumstances under which a burden may be ‘fair’—i.e. where it may affect competition but does not result in a net competitive disadvantage, as there are other effects within the market which outweigh its impact.
Figure 2 Impact of the USO in markets with developing competition
Figure 2 illustrates how the case-by-case analysis envisaged by the ECJ’s judgment in the Base case could work in practice, and how markets at different stages of competitive development may have different optimal outcomes. Another way of looking at this is by considering the impact on profitability, as in the following examples.
BT—Ofcom concluded that the primary effect was that the net cost of the USO, if positive, would have a small effect on BT’s profitability. As BT remains by far the largest and most profitable fixed-line operator in the UK, Ofcom has taken no action. This approach is common for quality obligations with a relatively small economic effect.
PTC—in Portugal, ANACOM identified that the net cost was positive (and material relative to the costs of a sharing mechanism) and could be considered an unfair burden, as PTC had lost significant market share and seen a material impact on profitability.
eircom—in Ireland, the electronic communications and postal regulator, ComReg, has implemented a formal process for assessing net cost applications by eircom. As discussed in the box below, ComReg concluded that the net cost for 2009/10, while material to the costs of a sharing mechanism, did not prevent eircom earning a fair return on capital, and therefore it could not be considered unfair.
What is the future of the USO?
The challenges of how to fund new infrastructure investment are arguably getting bigger. For example, in the UK, BT may be profitable, but it has been able to roll out new technology commercially only in certain areas. In some rural or remote areas, it may not be profitable to invest in new technology. Given trends in Internet usage, and with some public services moving towards Internet-only provision, a universal service based on postal services and voice calls might not be sufficient in the future.
Therefore, if governments want to continue the concept of universal provision into new technologies, they may either need to fund that investment directly, or the scope of the USO, and therefore the net cost, may need to increase. However, in communications markets, the funding of the USO through a retained dominant position is likely to be increasingly difficult over time. In fixed telecoms in particular there is increasing competition and falling market shares of former incumbents, as shown in Figure 3.
Figure 3 Change in incumbent market share in selected EU countries between 2005 and 2010
Source: Oxera analysis based on Ofcom (2011), ‘International Communications Market Report 2011’.
This ongoing reduction in market share for the previous incumbent operators is likely to continue to reduce the ability of such operators to fund the USO through cross-subsidy. Therefore, if a government wants its incumbents to offer non-commercial services, then, arguably, some intervention is likely to be required. There are three potential models:
subsidy (including tendering);
a sharing mechanism;
funding from the incumbent’s residual competitive advantage (including regulation).
As described above, a sharing mechanism may, in practice, either increase or decrease market distortions, depending on the market structure. In instances where a subsidy is an option, there will be a need to have regard to the Commission’s rules on state aid more generally, as described in the box below.
As competition continues to increase, and if governments want to continue the USO concept into new technologies, it would appear realistic for a sharing mechanism to become the economically optimal approach for funding an unfair burden. If so, rather than declining, there may be a future for a larger USO based on access to new technologies for all. In that case, the use of a sharing mechanism should be compared to the approaches of subsidy and regulation in assessing which approach is most relevant in competitive markets.