Achieving Price Transparency in Bundles A Report on behalf of Hutchison Telecommunications
Joshua Gans and Stephen King
The analysis here represents the views of CoRE Research Pty Ltd (ACN 096 869 760) and should not be construed as those of Hutchison Telecommunications.
21st October, 2002
Contents
July, 2001
Page
1
Background ............................................................................. 2
2
Competitive concerns about bundling .............................. 3
3
Transparency to Competitors and Authorities................. 5
4
Customer-Level Transparency............................................. 9
5
Conclusion............................................................................. 11
6
Appendix ............................................................................... 13
i
Section 1
1
Background
Background On the 19th July, 2002, Telstra and Telstra Pay TV filed a notice with the ACCC regarding its intention to use a bundled price when reselling the pay television services of Foxtel. That price would offer to customers who have pre-selected a Telstra fixed line service a further discount of 5 percent on eligible services if they also subscribed to one of Telstra internet, mobile or pay TV services and 10 percent if they subscribed to any two of these. This report focuses on one aspect of Telstra’s proposed pricing policy – the issue of transparency. Section 14 of their submission deals with this issue. Telstra write: Telstra is aware that one of the primary concerns with the practice of bundling is that, where a single price is offered for the package, there is potential for disguising the prices of individual components of the package. The economist, Professor Stephen King has suggested that: “The problem is not that Telstra can sell a bundle of pay-TV, high-speed internet and telephony services, but rather that it can set a single price for the bundle of services.”1 The Telstra Rewards Options programme will not offer a single price for the bundle of packaged services. Neither the 5% nor the 10% discount would operate to disguise the prices of the individual elements of the package, such as the prices of mobile telephony services. The Telstra Rewards Options programme does not, and will not, allow Telstra to disguise the price of its telephony services. It will still be possible to compare Telstra’s telephony prices with those of its competitors. There is total transparency of pricing and no potential for Telstra to disguise the pricing of different elements of the Rewards Options programme. For this reason, the proposed conduct will not have the anti-competitive effect feared by Professor Stephen King. (p.27)
1 Stephen King, “Why this bundle should worry the ACCC,” Australian Financial Review, 16 April 2002, p.67.
2
Section 2
Competitive concerns about bundling
Hutchison Telecommunications has asked CoRE Research to comment on the issue of the transparency of Telstra’s proposed pricing policy. We begin by briefly recapping the potential anticompetitive problems that can arise with price bundling. We then consider the Telstra proposed pricing policy. Specifically we consider whether it is transparent to competitors and regulators and whether or not it could create competitive concerns. Finally, we comment on the issue of transparency from the customers’ perspective.
2
Competitive bundling
concerns
about
Gans and King briefly reviewed the economics literature on bundling.2 We noted two broad anti-competitive concerns about bundling that have been highlighted by economic research. First, bundling can be used to disguise predatory pricing. The firm selling the bundle can cross-subsidise one market from another market for the purpose of driving competition from the subsidised market. Second, bundling can be used to create a barrier to entry in a potentially competitive market. King discusses the first of these concerns in more detail.3 In particular, he focuses on the relationship between bundling and a vertical price squeeze. “If the incumbent is the monopoly provider of access to an essential facility, then anti-competitive bundling can involve a vertical price squeeze. In this situation, the access provider bundles retail products involving the essential input together (and possibly with other products where it also has market power) and sells the retail bundle at a price that cannot be profitable for a competitor who buys access. The competitors are ‘squeezed’ between the access price and the retail price” (p.1). King also notes that this form of anti-competitive bundling is similar to predatory pricing. Just as with standard predatory pricing, a vertical price squeeze instituted through a bundled product can benefit some customers in the short-term, but is aimed at reducing competition and raising profits in the longer term.
2 See Gans, J. and King, S. (2002) “Potential anti-competitive effects of bundling: A report on behalf of Hutchison Telecommunciations”, CoRE Research. 3 See King, S. (2002) “Bundling and imputation rules under accounting separation: A report on behalf of Hutchison Telecommunciations”, CoRE Research.
3
Section 2
Competitive concerns about bundling
In the quote presented above, Telstra argues that King’s concerns about anti-competitive bundling are irrelevant for the Telstra proposed pricing policy because there are individual product prices and “total transparency of pricing”. This claim is misleading. A vertical price squeeze and anti-competitive bundling involves the relationship between retail and wholesale prices and costs. The fact that various prices are ‘listed’ does not mean that there is no risk of a vertical price squeeze. Rather, a vertical price squeeze and anticompetitive bundling can arise when the relationship between prices is distorted by a firm with market power. A simple example can illustrate this point. Suppose an essentialfacility access provider has a wholesale access price of $10 per unit and the efficient retail-level costs of transforming a unit of wholesale input into a unit of retail output are $5. If the access provider sets a retail price of $14, then this is ‘transparent’ in the sense that the price is well known and observable. But it also involves a vertical price squeeze because, given the (transparent) wholesale access price, the retail price cannot be matched by an efficient competitor.4 Contrary to the claim made by Telstra, their proposed pricing rules may be used to hide anti-competitive conduct. This is explained in more detail in section 3. However, there are two underlying drivers of any potential anti-competitive conduct. First, the discounts in the Telstra proposed pricing policy are not explicitly related to any cost savings. There might be legitimate cost savings realised by Telstra through bundling, but such savings are not formally presented or measured by Telstra. Second, the discounts are ad valorem reductions in a customer’s payments over all relevant services. Such ‘total bundle’ discounts can lead to individual products being priced below incremental cost. In fact, as shown below, the Telstra proposed pricing policy might lead some customers to face negative prices for some components of the Telstra bundle. Such below cost pricing could clearly be used to undermine legitimate competition.
4 This is analysed in detail in King, S. and Maddock, R. (2002) “Imputation rules and a vertical price squeeze”, Australian Business Law Review, 30, 43-60.
4
Section 3
3
Transparency to Competitors and Authorities
Transparency to Competitors and Authorities One element of pricing transparency relates to the signals sent by prices to competitors and the information available for regulatory authorities. If there is no bundling, then the posted price for say Telstra’s mobile service is the price competitors know they have to beat in order to attract customers. Regulatory authorities can easily cross-check such prices against the relevant competition laws. Telstra’s bundling does not allow full price transparency to either competitors or regulatory authorities. From the perspective of competitors, they do not know the average price customers are actually paying for a Telstra service. In particular, competitors need to know what price the greatest proportion of Telstra mobile customers are paying to know what price to beat. However, without detailed information regarding Telstra’s subscribers to fixed line, Pay TV and internet, it cannot easily assess this. In this respect, bundling makes pricing less transparent. As noted above, the lack of transparency in bundled prices can make disguised predation more likely – allowing a dominant firm to set a price for a particular service below cost under the radar of both competitors and regulators alike. Telstra’s proposed bundling carries with it a risk to overall competition as price signals cannot be sent to competitors to Telstra or to regulatory authorities. To avoid a firm with market power using bundling to disguise predatory behaviour, it may be desirable to require the firm that provides the bundle to also sell the individual products at ‘consistent’ prices. The bundled product can be discounted to reflect legitimate cost savings. This approach allows both the producer and customers to enjoy benefits of bundling where they exist while avoiding the anticompetitive potential of bundling. The potential for predatory pricing under Telstra’s proposed bundled product pricing arises because the proposal does not guarantee consistency between prices of bundled products and the costs faced by Telstra for serving a customer. Telstra’s current bundling proposal can easily lead to incremental prices for customers that fall below incremental cost. Such prices will be prima facie inconsistent with standard competitive behaviour. To see this, consider the simple example of the first customer in the Appendix. The incremental price to the customer of Pay TV is only
5
Section 3
Transparency to Competitors and Authorities
$8.50. This is the extra revenue that Telstra earns from the customer when they subscribe to Pay TV given that they subscribe to all the other services listed in the example. To see this, note that the customer would pay $199.50 to Telstra if they did not subscribe to Pay TV. This involves a pre-Reward bill of $210 less a 5% discount. With Pay TV the customer’s bill is $207. Thus the customer only pays an extra $8.50 for the Pay TV subscription. While we do not have information on the incremental cost to Telstra of supplying Pay TV services to a customer, if these exceed $8.50 then Telstra is losing profits by selling Pay TV services to this customer. Even if the incremental costs of each of the services from Telstra’s perspective were close to $0, Telstra’s proposed bundling may lead to implicit product prices below marginal cost. In fact, a customer might pay a negative price for an additional product under Telstra’s bundling proposal. To see this suppose that a customer pays on average $250 per month in fixed line calls (local calls, long distance calls and fixedline calls to mobiles) using Telstra. The customer also buys Pay TV from Telstra, including say movie and sport packages, paying say $70 per month. The customer will pay $304 per month after the 5% discount. Now suppose that the customer subscribes to Telstra mobile. The customer has a relatively low mobile bill of only $15 per month on average before any bundling discount. The customer’s total bill before the Reward discount is $335 per month. After the 10% discount the customer pays $301.50. In other words, Telstra is paying the customer $2.50 per month to have the customer use a Telstra mobile phone. Telstra is charging a negative incremental price for mobile services to this customer. While this example is extreme, it is not impossible or unrealistic. Further, if pricing under incremental cost is not only feasible but wide-spread under Telstra’s bundled pricing, then Telstra’s competitors will face a substantial barrier to effective competition. In the above example, suppose that the customer was originally with another mobile carrier. Even if that carrier was significantly more efficient than Telstra, it would not be able to retain its customer – it could not compete with the negative price associated with mobile services under the Telstra bundle. More generally, the Telstra proposal does not guarantee that there will not be significant pricing below incremental cost on some products or for some groups of customers. If this occurs, efficient competitors will be driven from the market by the Telstra bundled pricing. Implicit pricing below incremental cost can clearly lead to competitive concerns. The potential anti-competitive effects of Telstra’s proposed pricing are exacerbated because they apply to products where Telstra’s competitors must buy access to wholesale services provided by Telstra in order to compete with Telstra at the retail level. For
6
Section 3
Transparency to Competitors and Authorities
example, the Telstra bundled products include fixed-line telephone services. In this situation, Telstra might be able to institute a vertical price squeeze even if the price to a customer for the retail service is above the incremental cost to Telstra of that service. If Telstra’s wholesale access price exceeds incremental cost then Telstra’s retail competitors will not be able to match Telstra’s pricing. In summary, Telstra’s proposed bundled pricing does not guarantee that Telstra will not engage in predatory conduct. In fact, the pricing proposal, by allowing customers to gain an increased percentage discount over their entire bill when they subscribe to an incremental service, will most likely lead to pricing below incremental cost for at least some customers and some services. It provides an excellent strategic cover for predation or a vertical price squeeze. We do not have enough information to predict the extent of either below-cost pricing or vertical price squeeze under the Telstra proposal. If such pricing is potentially wide-spread or will apply to important classes of customers, then Telstra’s proposed bundled pricing could severely retard efficient competition. At a minimum, to reduce anti-competitive concerns about disguised predatory pricing, Telstra’s bundling should proceed only on the basis that Telstra makes certain commitments to the ACCC regarding information and bundled prices. Telstra needs to assure the Commission that the incremental prices set for any customer or group of customers exceed incremental costs for the relevant services. Telstra needs to provide any relevant information to the Commission to allow the Commission to reassure itself that the bundled pricing will not be used to disguise predatory conduct. This information will include customer purchasing profiles and incremental cost information. For example, Telstra and the ACCC (possibly in consultation with other interested parties) could identify relevant classes of customers. Telstra would then provide the Commission with aggregate data relating to each customer group. This data could include average customer profiles as well as the profiles of the lowest ten percent (in terms of expenditure) for each bundled product. Telstra would also need to provide incremental cost data to the Commission for each customer class (and relevant subclasses) so that the Commission would be able to determine if any class or subclass of customers were receiving implicit product prices that are below incremental cost. The Commission would need to independently verify the information provided by Telstra. If the Commission determines that there is or will be a significant class of customers who receive a service or group of services under the Telstra bundle at an incremental price that is below incremental cost, then the Commission should, at a minimum,
7
Section 3
Transparency to Competitors and Authorities
require that Telstra alter its bundled pricing to eliminate this predatory behaviour. There are obviously considerable difficulties with this approach. The choice of relevant customer classes would be contentious. Even if the Commission were able to gain relevant information from Telstra, it would be difficult to verify that all relevant classes of customers were being competitively priced. Experience in telecommunications has shown the difficulty of determining incremental costs for a fixed-line telephone system. The relevant cost information, generated by computer models, is contentious and is sensitive to a variety of modelling assumptions. As a result, in practice, it is unlikely that the Commission could ever be certain that the Telstra proposed pricing policy was not being used as a cover for illegitimately undermining competition. Further, the checking of Telstra’s bundled pricing would not be a once-off exercise. The relevant information and the potential for predation to emerge needs to be regularly checked by the Commission. For example, the Commission might need to review Telstra’s bundled options every eighteen months. Alternatively, firms that are potentially being harmed by below-cost pricing could apply to the Commission for a formal investigation if there is a reasonable possibility that some incremental prices have fallen below incremental cost. Again, the checking would be information intensive and contentious. An alternative approach would involve the Commission requiring that Telstra justify its bundled discounts. This alternative approach would require that Telstra can only discount a bundled offering to the extent that they can show legitimate cost-savings from bundling. For example, bundling might lower the costs of billing for a customer relative to the situation where that same customer bought the relevant individual products. Telstra should be allowed to discount its retail price to the customer to reflect such cost savings. But Telstra would not be allowed to discount a bundle in excess of the legitimate cost savings. Under this alternative approach, the Commission would require relatively little information from Telstra. The Commission would only require the individual prices of each retail product and the relevant cost savings that Telstra generates through bundling. These cost savings are likely to be fixed dollar amounts rather than percentage savings. The Commission could require that Telstra provide any relevant information about these savings before it approved a bundled discount. In other words, the onus would be on Telstra to show the legitimate cost savings before it would be able to include these discounts in its retail prices.
8
Section 4
Customer-Level Transparency
This alternative approach was summarised in the report by King. 5 The pricing of Telstra’s bundled products “must be consistent with (1) the individual retail prices set for the relevant bundled products by Telstra and (2) any cost savings associated with the bundle that can be explicitly shown to exist to the satisfaction of the ACCC”. The report by King also noted the information required by the Commission to ensure that there was no price squeeze occurring on a single product. In this sense, the Telstra bundled pricing proposal is simply one example of a general class of pricing concerns that face the Commission.
4
Customer-Level Transparency The Telstra proposal is to set its prices for all of its services and then to offer discounts to customers who select a broader range of their services. To see what this means from the customer’s perspective, consider two services, A and B supplied by a single firm. Let pA and pB be the prices offered to customers for the stand-alone services and pAB be the price paid by customers subscribing to both services. Bundling occurs in the pricing of these services if pAB < pA + pB. In the Telstra case, there is bundling as pAB = α.(pA + pB); where α = 0.95 or 0.9 as the case may be. However, if it is more difficult for Telstra to change α than to change pA and pB, there is a sense in which Telstra might be more constrained in its bundling pricing options than a textbook case of pricing bundles. In terms of transparency for consumers, for a sophisticated customer, these prices are relatively transparent. What concerns us more is the impact of bundling on less sophisticated customers. In particular, what does bundling do to those customers’ ability to compare price offerings from Telstra with its rivals; in particular, those rivals who do not offer the full range of services Telstra does? In this regard, we believe that Telstra has an opportunity both to provide customers with greater information regarding their prices as well as an indication of the benefits customers are receiving by having their services through Telstra. Our proposal would also give competitors the opportunity to readily compare their price offerings
5 See King, S. (2002) “Bundling and imputation rules under accounting separation: A report on behalf of Hutchison Telecommunciations”, CoRE Research.
9
Section 4
Customer-Level Transparency
with Telstra in a way that allows customers to easily compare competing offers – whether they are bundled or not. Our proposal is for Telstra to offer on their bill a breakdown of the Rewards benefits to customers. That would take the form of a table that described – beyond fixed line pricing – what are the savings the customer received each month from subscribing to a particular service. An example of the information that would be provided is in the Appendix. The idea would be that a simple reporting of the contribution of a service to the reduction in the customer’s bill for other services could be listed at the end of the bill. The Appendix does this under the heading “Rewards Breakdown.” So, for example, the first customer who received a total discount of $23 would be told that their bill was reduced by $10.50 because they subscribed to Telstra’s mobile service and by $8.50 because they subscribed to Telstra’s pay TV service.6 These amounts would then allow customers to readily compare any savings they would receive, in terms of lower prices, by switching to say another mobile carrier with the increase in their Telstra bill that this would cause. Thus, by providing this information Telstra would allow their customers to more easily compare the benefits of being a Telstra subscriber (in terms of reduction in their overall bill) to those offered by other carriers who might not have a bundle of services to offer. Even less sophisticated customers would have the benefits of increased transparency in terms of comparing the options available from different providers. Practically, in calculating this additional information, no further data is required than amounts already on the Telstra bill – used to calculate the Rewards discount. Thus, the practicalities of doing this would appear to be very modest.
6 Note that these do not add up to $23 as they are incremental discounts from each product.
10
Section 5
5
Conclusion
Conclusion In this paper we have briefly considered the transparency of Telstra’s proposed bundled prices. We consider the transparency for competitors, customers and regulatory authorities. From the customers’ perspective, Telstra’s bundled discounts should be made more explicit. This can easily be achieved by a simple addition to the customer bill. This increased transparency will aid customers when choosing between Telstra and alternative service providers for particular products. From the perspective of competitors and regulatory authorities, Telstra’s bundled options might be used to hide predatory prices. In this sense, the prices are non-transparent from a regulatory perspective. Bundled prices might involve incremental services supplied to customers below incremental cost or even at a negative cost. Such pricing is inconsistent with standard competitive behaviour and may significantly harm competition. Telstra’s bundled pricing might also be used to disguise a price squeeze for products such as fixed-line telephone services where Telstra’s retail competitors are dependent on Telstra for wholesale essential facility access. We consider two alternative approaches to the bundled pricing that could be adopted by the Commission. First, given the competitive concerns, the Telstra bundled pricing might only be given regulatory approval if the regulator (1) is satisfied, given current customer purchasing profiles, that no significant group of customers will receive any incremental services form Telstra under a bundle at a price that is less than incremental cost; and (2) has on-going procedures in place to check the Telstra’s bundled pricing to ensure that no elements of the bundle are priced below incremental cost in the future. As we noted, however, this approach requires high levels of information and may be impractical. Alternatively, the Commission could require that Telstra’s bundled pricing falls within the bounds of standard imputation rules. In particular, Telstra would only be able to offer bundled product discounts to the degree that these discounts reflected legitimate cost savings. Telstra would be required to verify these savings to the satisfaction of the regulator before it could offer the relevant retail discounts to customers. This approach considerably reduces the information burden on the regulator. It requires that Telstra provide information on relevant savings. However, this information should be readily available, as Telstra would need this information internally in order to determine its own bundled pricing policies. Further, this
11
Section 5
Conclusion
alternative approach ensures that there is alignment between cost savings and retail customer discounts. Such alignment is consistent with standard competitive market conduct.
12
Section 6
6
Appendix
Appendix Consider the following Telstra bill for a customer who subscribes to two additional services: Local Call Costs: Long Distance Calls: Fixed-to-Mobile: Mobile Phone: Pay TV:
$20 $100 $30 $60 $20
Total Reward Discount:
$230 $23
Amount Owing:
$207
Rewards Breakdown: this month you saved the following amounts by subscribing to the following Telstra services. Mobile Service: $8.507 Pay TV: $10.508 Or, alternatively, consider a customer who only subscribes to one additional service: Local Call Costs: Long Distance Calls: Fixed-to-Mobile: Pay TV:
$20 $100 $30 $20
Total Reward Discount:
$170 $8.50
Amount Owing:
$161.50
Rewards Breakdown: this month you saved the following amounts by subscribing to the following Telstra services. Pay TV: $8.509
7 This is the total bill without the mobile service = $170 (= $230 - $60) multiplied by 0.05 (= 0.1 – 0.05). 8 This is the total bill without the Pay TV service = $210 (= $230 - $20) multiplied by 0.05 (= 0.1 – 0.05). 9
This is the total bill = $170 multiplied by 0.05.
13