That the ACCC’s price setting has resulted in regulated access prices being set below attributed costs is readily illustrated using the Local Call Service (LCS). That service is a ‘resale’ version of a retail service — local calling — that is directly price capped. Specifically, under the price-capping arrangements, the retail price for an un-timed local call is capped at 22 cents, excluding Goods and Services Tax (GST), with such an untimed service to be available nationally at an essentially uniform price.[11]
The capped retail price for untimed local calls has generally been regarded both by Telstra and by the ACCC as being below the average cost of a local call, where that average cost is determined by attributing to local calls some responsibility for the joint and common costs of the network. Given that, the ACCC has had two options:
It could set wholesale (that is, access) charges on the basis of attributed costs, so that the wholesale charge might exceed the capped retail price; or alternatively,
It could set the wholesale charges on the basis of the capped retail price. To the extent to which this would result in those charges being below a relevant benchmark of costs, the issue of where the cost shortfall would be recovered by the hypothetical operator would need to be addressed so as to retain consistency with the underlying costing approach.
Some regulators, faced with this choice, have chosen the former approach — on the basis that this places access seekers in a position that is competitively neutral with the access provider, as both incur a cost shortfall in supplying the price-capped services. The ACCC, however, has opted for the latter, which it has implemented by setting the wholesale price on the basis of subtracting from the capped retail price the cost of those activities that a wholesale-only provider would avoid (a price determination methodology generally referred to as ‘retail minus’). As a result, to the extent to which there is a shortfall between the capped price and costs, that shortfall will remain and need to be recovered elsewhere.
The extent of the shortfall can be estimated by reviewing the ACCC’s setting of LCS prices for 1999/00 and 2000/01, when it still disclosed detailed information on its cost estimates and methodology. The relevant estimates for those years are set out in Table 2, noting that the data required to make similar calculations for more recent years has not been disclosed by the ACCC.
|
1999/00 |
2000/01 (GST exclusive) |
|
|---|---|---|
|
Total costs allocated to local calls |
21.54 cents |
21.21 cents |
|
Wholesale local-call price |
19.26 cents |
17.51 cents |
|
Under-recovery per call |
2.28 cents |
3.7 cents |
|
Number of local calls |
11,566 million |
11,987 million |
|
Under-recovery of costs (based on ACCC cost estimates) |
$264 million |
$443 million |
Three points help explain the Table.
First, the ACCC, when it calculates the cost of PSTN services, allocates costs to all types of PSTN traffic, including local calls.[12] For the years here at issue, the Commission allocated PSTN costs of 21.54 cents to each local call in 1999/00 and 21.21 cents to each local call in 2000/01 (see Table 2, first row).
Second, despite that cost allocation, the ACCC, in setting the price for LCS, used a retail-minus approach which involved starting with the retail price of local calls and then deducting the ‘avoidable’ local-call retail costs. For 1999/00 through to 2002 the ACCC used an estimate of 2.74 cents per call for retail costs and adjusted this to 2.49 cents when the GST was introduced on 1 July 2000. As a result, in 1999/00 the LCS price was 19.26 cents per call and in 2000/01 was 17.51 cents per call (see Table 2, second row).
Third, there was consequently an under-recovery of costs for each local call, that under-recovery being of 2.28 cents per local call in 1999/00 and of 3.70 cents per local call in 2000/01 (see Table 2, third row). The aggregate under-recovery was of $264 million in 1999/00 and of $443 million in 2000/01. Additionally, the quantum of that under-recovery would have continued to increase up to 2002/03, as local-call minutes increased more rapidly than all call minutes until then.
That shortfall is essentially a common cost that the hypothetical builder of the new, wholesale-only network would have had to recover from other services. Consistency with the thought experiment therefore suggests that the shortfall should have been allocated to other access services, notably PSTN Originating and Terminating Access (OTA), and within that, largely to terminating switched access.[13] Had that been done, total revenues from the hypothetical network would indeed have equalled total costs, both being evaluated using total call minutes. But the ACCC failed to make any such allocation — rather, it simply ignored the inconsistency in its price-setting.[14]
As a result, a cumulative shortfall that (capitalised to the present) probably amounts to close to $1 billion, was simply placed upon Telstra, which was left with access charges plainly below the relevant costs as determined by the ACCC’s costing methodology.
[11] Telstra Carrier Charges — Price Control Arrangements, Notification and Disallowance, Determination No. 1 of 2005, later amended by Telstra Carrier Charges — Price Control Arrangements, Notification and Disallowance, Determination No. 1 of 2005 (Amendment No 1 of 2006), clause 16 ‘Untimed local calls’. The effect of the requirements set out in the price determination in ensuring that call charges for an untimed local call are the same in all areas is discussed in ACCC (2005a, p. 111ff.).
[12] The sum of those allocated costs equals total costs, so that total cost recovery in the relevant thought experiment requires that each service be priced in such a way as to recover the share of costs allocated to it.
[13] OTA involves the origination and termination of switched calls over Telstra’s PSTN (for example, from a calling customer preselected to Optus to a called customer who is preselected to Telstra, where both are connected to Telstra’s Customer Access Network). In general, it is efficient to place higher mark-ups on terminating than on originating access, as it is more difficult for terminating access to be bypassed.
[14] It is important to note that had the ACCC allocated the shortfall to other wholesale services, this would not mean that the shortfall would have been largely or entirely borne by access seekers. Rather, because the shortfall would have been allocated over all minutes of use, each user of the network, including Telstra, would have faced the same unit charge per end-minute-of-use. In contrast, under the ACCC’s approach (of simply ignoring the shortfall), the deficit would only fall on Telstra. This is the natural consequence of the definition of the relevant increment. In the ACCC’s approach, that increment is the total traffic carried over the PSTN. As a result, common costs are unitised over that total. Analytically, the shortfall is merely a common cost to the PSTN as a whole and hence would be spread over all the traffic making use of that network.