Unit 1: Structural Separation and Divestiture

ICT Industry and Markets

Page 49 of 73 pages. Chapter: 5: OLD Unit 1: Fair Trade and Competition Policy

Structural Separation and Divestiture

Two other approaches, namely structural separation and divestiture, have been used by competition authorities and telecommunications regulators in cases of serious anti-competitive cross-subsidisation. Both approaches tend to be used only where there is evidence of significant anti-competitive conduct. This usually involved not only cross-subsidisation, but related conduct such as predatory pricing, anti-competitive use of information and discriminatory practices.

Structural separation generally refers to be the separation of different lines of business of a telecommunications operator into separate corporate entities.

As an example, a cellular business can be operated by a separate company from a wireline telephone business. Both may be owned by the same shareholders. However, existence of a separate cellular company makes it easier to ensure that the incumbent operator with which it is affiliated does not discriminate operator with which it is affiliated does not discriminate unfairly against cellular competitors as compared to its own cellular operations. Rules can be established to ensure that both cellular companies are treated the same, for example, with respect to interconnection charge. Other examples of telecommunications line of business that are frequently separated included ISPs and various types of mobile operators.

When structural separation is mandated by regulation, the different companies must typically be run on an "arm's length" basis. In that case, the companies must deal with each other on the same terms and conditions as they deal with third parties, such as competitors. The separate companies must normally not only have separate accounting records, but also separate management, offices, facilities, etc.

Regulatory conditions normally determine the degree of separation required in the companies' operations. Development of these conditions can pose challenges. Regulators must balance two competing objectives. One is to create sufficient separation to minimise the potential for cross-subsidisation, collusion or other anti-competitive actions between the separated companies. The other is to minimise the inefficiencies that will almost inevitably be created by structural separation.

For example, there may be efficiencies (economies of scale and scope) inherent in proving common administration services to both companies. On the other hand, the sharing of administrative services, such as accounting services, provides potential for anti-competitive conduct and for developing covert cross-subsidies. Similarly, sharing head office space can lead to efficiencies. On the other hand, it provides opportunities for collusive conduct between managers of the two companies. If structural separation is to be required, there should be a real separation of the two lines of business, including their management, premises, customer data bases, accounts and operations. Otherwise, the structural separation may be a sham.

The initial question, however, is not whether there should be structural separation between the companies, but whether the advantage of separation outweighs the disadvantages given the realities of a particular market. Other disadvantages or structural separation include high transaction costs (the costs of creating the separate companies) and the distraction for employees and customers as they work through the separation. Despite those disadvantages, structural separation may be the only way to ensure a level playing field for competition in some markets.

Structurally separate companies can often continue to operate under common ownership. Divestiture refers to a situation where a company, such as an incumbent, not only runs a particular line of business through a separate company, but divests (i.e. sells) some or all of the ownership of that separate company to independent parties.

Some competition advocates argue that only divestiture of ownership can ensure that a separate company is run in the interest of its separate shareholders, rather than merely as an operating arm of is parent company (e.g. the incumbent). Without divestiture, it is argued, a great deal of regulatory effort will be expanded to detect anti competitive dealings between affiliated companies. Once there are separate shareholders, the management of the separate companies must act in the interests of those shareholders. It will be safer to assume that the companies are actually run on an arm-length basis.

Structural Separation: The EU Cable Directive'

An example of structural separation directive can be found in the EU's 1999 Cable Ownership Directive. This Directive requires dominant telecommunications operators to place their cable televisions operations in a structurally separate company. The Directive builds on the EU's ONP Directive and other efforts to implement a competitive framework for telecommunications. It is intended to address specific problems which the EU Commission has concluded result from the joint operation of cable television networks and conventional telecommunication networks.

The Cable Ownership Directive makes is clear that the Commission views structural separation as the minimum corrective measure required at this time, and that it may impose further measures, including divestiture of cable interests to third parties, in specific cases. The Commission also appears to be adopting a practice of requiring dominant companies to divest their cable interests as a precondition to securing Commission approval for new mergers among telephone companies.

Divestiture: The AT & T Model

The most famous example of a telecommunications divestiture involved the separation of AT&T from the Regional Bell Operating Companies (RBOCs) in the United States in 1984. Not only were the local operations of AT&T structurally separated from its long distance and international operations, but ownership of the two groups of companies was separated by means of a share swap. The divestiture was, by most accounts, a great success.

With their ownership separate from AT&T, the RBOCs no longer had an incentive to favour AT&T over its long distance competitors, such as MCI and Sprint. Therefore, all long distance competitors obtained access to local telecommunications services from RBOCs on similar, non-discriminatory terms. More relevant to this section, the divestiture eliminated concerns about anti-competitive cross-subsidies between AT&T's local distance operations.

Divestiture is generally viewed as an extreme remedy that is only appropriate in cases of overwhelming dominance by very large operators in large economies such as the US. Policy-makers in other countries have been reluctant to consider dismembering incumbent, which are often seen as "national champions".

Vertical Price Squeezing

Vertical price squeezing is a particular type of anti-competitive conduct that may be engaged in by incumbent operators. This form of conduct can occur if the incumbent provides services in two or more "vertical" markets. Vertical markets are sometimes labelled "upstream" and "downstream" markets. For example, the oil production market is upstream of the oil refining market, which in turn is upstream of the gasoline sales market. Instead of upstream and downstream, the terms "wholesale" and "retail" are often used.

Vertical price squeezing can occur when an operator with market power controls certain services that are key inputs for competitors in downstream markets, and where those same key inputs are used by the operator or its affiliates to compete in the same downstream market.

To take an example, in telecommunications markets, incumbent often control local access and switching services. Consider one such service - the provision of dedicated local circuits from customer premises to local exchanges. Dedicated local circuits can be viewed as "Upstream" services. These services are used as input by the incumbents in providing "downstream" services, such as dedicated internet access services. Dedicated local circuits are also a key input for competitors who provide dedicated internet access services. In other words, both the incumbent and other suppliers compete in the downstream market for dedicated internet access services.

If the incumbent decided to engage in vertical price squeezing, it could increase the price to competitors for the upstream input (i.e. dedicated local circuit rates) - while leaving its downstream prices the same (i.e. prices for its dedicated internet access services). The effect would be to reduce or eliminate the profits (or "margins") of competitors. Their margins would be "squeezed". To increase the squeezing effect, the incumbent could also reduce its downstream prices for internet access. This would be a "two-way" or margin squeeze.

Put another way, an incumbent can often squeeze the margins of competitors by raising wholesale prices paid by competitors, while at the same time lowering retail prices on competitor services.

A simplified numerical example of a vertical price squeeze is included in Box 1-8.

Box 1- 8:
Example of Vertical Price Squeeze by Incumbent Operator

Cost to incumbent of upstream facility (e.g. dedicated loop)	$90
Price charged by incumbent to competitor for loop	$120
Cost of providing retail services to end users (e.g. dedicated Internet access service) in addition to loop cost (e.g. marketing, billing, etc.)	$20
Price charged by incumbent to end users for dedicated Internet access services	$130

Wholesale Cost Imputation Requirement

To prevent vertical price squeezing, a telecommunications regular may impose a wholesale cost imputation requirement, along the lines set out in Box 1-9.

Box 1- 9:
Basic Elements of Wholesale Cost Imputation Requirement

Conditions for Application

Applies to a monopoly or dominant provider of “wholesale services”
Where the dominant provider also competes in market for “retail services” that require the wholesale services at inputs.

Basic Rules

Dominant provider must provide evidence to the regulator that its retail prices are no lower than the sum of the following:

The price it is charging competitors for the wholesale services that form part of the retail services (this price is said to be “imputed” in the cost of the dominant provider whether it actually incurs this cost or not)
The actual incremental costs (above the imputed wholesale costs) that are incurred by the dominant supplier in providing the retail services. For example, marketing, billing, etc. costs

Variations on this type of imputation approach have been used by various regulators and competition authorities. It is relatively simple to use (compared to detailed accounting separations or cost allocations). To return to the margin squeezing examples in Box 1-8, it does not matter whether the actual cost of the wholesale service is $90, $120 or some other number. What the imputation requirement assures is that the same cost for essential wholesale services is imputed to the dominant operator's retail services as is passed on its competitors.

Imputation: A Canadian Example

A form of the wholesale cost imputation requirement has been applied by the Canadian regulator in response to complaints of targeted retail price discounting by incumbent operators. The CRTC's approach was tailored to the rather unique circumstances of the Canadian market. In that market, the CRTC established a universal service program in the form of subsidy for the access deficit incurred by operators in higher cost area.

All long distance operators, including new entrants, are required to make "contribution" payment to subsidise the deficit described above. However, as noted in our detailed discussion of the Canadian example, Incumbent local operators continue to receive the vast majority of contribution payments. Initially, the CRTX did not specifically require the incumbent operators to account for their own use of the local access network in providing competitive services. That is, it did not require incumbents to make contribution payments to themselves. This led to the potential for vertical price squeezing by incumbents. The CRTC's Response to this situation is described in Box 1-10.

Box 1- 10 :
Case Study: The CRTC Imputation Test

In1994 (Decision 94-13), the CRTC described the targeted price responses of incumbent operators to new entrants as follows:

Under a scenario of unrestrained target pricing by the telephone companies, competitors could be faced with the situation in which they must compete against telephone company prices that embody a contribution amount that is lower than the competitor contribution cost in that market segment. The Commission considers that, due too their previous status as monopoly toll providers, the telephone companies have an established and generally predominant share in all market segments. As a result, their traffic mix, the presence of barriers to entry and the existence of customer inertia would permit them, on a sustained basis, to recover contribution from the most highly contested market segments at a level below the contribution amount (payable by competitors).

As a result of these concerns, the CRTC implemented an “imputed test” to ensure that incumbents’ prices in competitive networks were subject to similar cost recovery requirements as competitors. This imputation test, as modified in a later CRTC decision (Telecom Decision CRTC 94-19), has the following requirements:

Revenues for each service offered by an incumbent must equal or exceed the sum of:

The costs for “bottleneck services” used by the company in the provision of the question, using tarried rates of those bottlenecks services (the Operator Access Tariff”)
The casual costs specifically attributed to the services, which are additional to the costs covered in above
Any applicable contribution payments

This imputation test is similar to the one described in Box 1-9. The main difference is that the CRTC imputes "contribution" subsidies, as well as wholesale facilities costs, as costs that must be covered in the incumbents' retail prices. The CRTC took the position that so long as a service recovers these inputted costs, plus the direct causal costs of the retail service, targeted pricing would not be anti-competitive.