The Access Deficit Contribution for PSTN

Interconnection Pricing
A Submission to the Australian Competition and Consumer
Commission

Joshua Gans and Stephen King

26th February, 2003

 Executive Summary
This submission evaluates the contention that Telstra ought to receive a contribution
from all PSTN users towards a so-called access deficit. That access deficit is the short-
fall between the line rental and subscription charges that Telstra may be able to earn
under current retail price controls and the costs associated with the customer access
network (CAN).

From an economic perspective, the construction of an access deficit is rather odd. The
CAN is used as an input for a wide variety of telecommunications services including
local and long distance calls, calls to mobiles, and internet access. Economically, there is
no reason to presume that an input that is used to provide a variety of different services
must have its cost recovered from a particular charge. Rather, if one input is used to
facilitate a large variety of retail products then any evaluation of any ‘deficit’ or ‘surplus’
associated with that input needs to be made by considering the entire operations of the
relevant firm. But the access deficit is calculated only with regards to customer access
charges, ignoring any economic profits that might accrue to Telstra from other
telecommunications products. As such, it is not clear that the access deficit has any
economic meaning.

There appear to be two potential arguments are made in favour of an access deficit
contribution (ADC). First, it is sometimes argued that the lack of an ADC will not
allow Telstra to earn a reasonable market rate-of-return on assets. Second, it is
sometimes argued that without an ADC, inefficient entry might occur. That is, entrants
may seek opportunities to make profits where Telstra is constrained to meet their price.

We evaluate each of these arguments. With regard to investment incentives, the nature
and operation of the universal service obligation fund in Australia essentially under-
writes any losses that may result from non-recovery of the ADC and effectively
guarantees Telstra a market rate of return on PSTN assets. The inefficient entry or
‘cream skimming’ argument neglects the fact that most PSTN dependent service
markets (such as long distance and fixed-to-mobile calls) involve competitive provision
and, consequently, entry will be efficient regardless of whether there is an ADC in the
access price or not. This is because in these markets there are no actual or implicit
constraints on Telstra’s ability to meet an entrant’s price. For these reasons, we do not
believe there is sound economic basis for the inclusion of the ADC in access
arrangements for the PSTN.

Finally, even if there was an economic argument for Telstra to receive an ADC,
building this ADC into access prices on the basis of calls or call minutes is an
inefficient means of recovering the costs of the CAN from access seekers. This is
because such pricing will discourage entry and soften price competition in
telecommunications markets. A better approach would be to recover the ADC in a
manner consistent with that of the USO.
Contents Page

1. Background ............................................................................. 3

2. What is the ‘Access Deficit’? ................................................ 4

3. Related Telecommunications Policies ............................... 5

4. The Access Deficit as an Economic Construct.................. 6

5. Should Telstra Receive an Access Deficit Contribution?9

5.1 The Investment Incentives Argument ............................. 9

5.2 The Cream Skimming Argument ................................... 12

5.3 Conclusion ......................................................................... 14

6. Methods of Recovering any “Access Deficit” ................ 16

6.1 The Telstra Approach....................................................... 16

6.2 The ACCC Approach ....................................................... 17

6.3 Evaluation of Call-Based Approaches ........................... 17

6.4 An Alternative................................................................... 17

6 Conclusion: Answers to the ACCC’s Questions ............ 19

6.1 Should access seekers still be required to make a
contribution to the AD? ................................................... 19

6.2 If there is to continue to be an AD, how should it be

defined and measured?.................................................... 19

6.3 If an ADC were to be retained, how should it be spread

over calls and minutes?.................................................... 20

Appendix A ....................................................................................... 21
A.1 The Universal Service Obligation Scheme .................... 21

July, 2001 i
Contents Page

A.2 Regulation of Telstra Prices............................................. 31

July, 2001 ii
Section 1 Background

1. Background

This submission is a response to the ACCC’s paper “The Need for an

ADC for PSTN Access Service Pricing” issued in February, 2003.
The ACCC calls for responses to a number of issues that arise in
relation to arguments made for an Access Deficit Contribution
(ADC).

The purpose of this paper is to evaluate the arguments made for an

ADC. We begin by reviewing what an “access deficit” actually is. In
particular, does the access deficit, as defined by the Commission, have
a sensible economic basis? We then turn to consider the various
arguments made for an ADC. Finally, we consider how an ADC
ought to be made if it is thought it is legitimate to do so.

To understand the access deficit, it is important to also understand

two related parts of Australian telecommunications policy – the USO
scheme and price caps on Telstra. We review these
telecommunications policies in the appendix of this submission.

3
Section 2 What is the ‘Access Deficit’?

2. What is the ‘Access Deficit’?

In its 1999 and 2000 Undertaking Reports, the ACCC recognised the
access deficit as a reasonable item that Telstra be allowed to recover
as a cost item in PSTN origination and termination. The ‘access
deficit’ was defined as the shortfall between the cost of providing
basic access and the revenues that Telstra is able to secure under the
price control regulations. Telstra’s basic access service is associated
with a customer being able to make and receive calls over the
customer access network (CAN).

The ACCC (July 2000) stated a formulaic definition of the access

deficit:

AD = Annual Line Costs + Retail Costs – USO funding - Connection

Revenues – Line Rental Revenue

Notice that, in calculating the access deficit, regard must be had to

two other telecommunications policies – retail price controls and
universal service obligation funding. We consider the impact of these
in the next section.

4
Section 3 Related Telecommunications Policies

3. Related Telecommunications
Policies

Two other telecommunications policies impact on the size of the

access deficit:

• USO Funding: The USO is the obligation placed on universal

service providers (USPs) to ensure that standard telephone
services, payphones and prescribed carriage services are
reasonably accessible to all people in Australia on an equitable
basis, wherever they reside or carry on business.1 Telstra,
currently the sole USP, is subsidised for providing this obligation.
The subsidy is funded by all licensed telecommunications carriers
in the telecommunications industry.2

Note that in the definition of the access deficit, it is recognised

that for some customers, the access deficit is covered by USO
funding received by Telstra. Hence this is treated as revenue
offsetting the difference between line rental and CAN costs.

• Retail price controls: The line rental revenue component is the

maximum line rental revenue Telstra can secure under the retail
price controls. That is, Telstra does not actually charge the
maximum line rental to all of its customers. However, the idea of
the access deficit is that it is a shortfall caused by retail price
controls and hence, an increment over actual revenues is used in
calculating the deficit.

The existence and nature of retail price control policies and USO
funding impact upon the size and rationale for an access deficit.
Appendix A provides a detailed description of these schemes and
considers how they impact upon the logic for an access deficit
contribution.

1
See Section 9 of the Telecommunications (Consumer Protection and Service Standards) Act
1999
2 See further at http://www.aca.gov.au/consumer/ uso/funding/funding.htm

5
Section 4 The Access Deficit as an Economic Construct

4. The Access Deficit as an Economic

Construct

From an economic perspective, it is far from clear that the access

deficit, as defined by the ACCC, has any useful meaning. To see this,
note that the services provided by the CAN are used as inputs for
other telecommunications services. Few if any customers value basic
access in its own right. Rather, customers value basic access because it
enables them to consume a variety of telecommunications products,
like local and long distance calls, that have the services provided by
the CAN as one input. In this sense, the services provided by the
CAN are not stand-alone products but rather they are services that,
when combined with other services, create valuable
telecommunications products.

The approach to the access deficit adopted by the ACCC treats basic
access as a stand-alone product. It treats basic access as an isolated
product and asks whether the direct revenues from that product
cover its costs. If the stand-alone product revenues do not exceed the
cost then there is a deficit. However, this calculation has no meaning
from an economic perspective. It is irrelevant whether there is a
surplus or a deficit in terms of direct contributions for one input that
is used in combination with other inputs to produce final services.

To see this, suppose that two inputs x and y are combined with each
other to form a final product z. One unit of final product z requires a
consumer to purchase exactly one unit of x and then to purchase one
unit of y. The consumer then puts these two components together to
form a unit of z. Further, the consumer can only form a working unit
of z if they purchase the unit of y from the same manufacturer as the
unit of x. For example, x might represent a razor-handle, y might
represent a blade, and z might represent a finished razor. The
consumer must purchase the razor handle and then purchase the
blade that ‘fits’ that handle. The manufacturer might help facilitate
this dual purchase by bundling the blade with the razor handle.
Assume that consumers only desire the final product z. Having a unit
of x by itself provides no benefit to a consumer. Similarly, having a
unit of y by itself provides no benefit to a consumer. Also assume that
consumers only care what they pay in total for the final product z.
Suppose that it costs $1 to produce a unit of x and $1 to produce a
unit of y. If z is sold at a competitive price then it will sell for $2. But
the seller might ‘break down’ the $2 price of a unit of z in numerous
ways. For example, the seller might sell product x for $1.40 and
product y for $0.60. From the consumers’ perspective they pay $2 for
the final product z. However, if the seller did this, then there would

6
Section 4 The Access Deficit as an Economic Construct

be a ‘product y deficit’ of $0.40 per unit – the ‘revenues’ for y are only
$0.60 per unit while the cost is $1 per unit. Alternatively, the seller
could hold the price of z fixed at $2 but set a price of x of $1.20 and a
price of y of $0.80. If the seller did this, then the ‘product y deficit’
would be $0.20 per unit. In fact, by altering the way they break-down
the $2 price of the final product, the seller could create a ‘product y
deficit’ anywhere between $1 per unit and -$1 per unit. However,
none of these deficit figures would have any economic meaning. They
would not alter consumers’ final product demand, the total sales of
the final product or the profits that are generated in total from the
final product.

The economic irrelevance of a ‘deficit’ that relates to one input of a

final product for a single producer is even more obvious if the
producer bundles the products. Suppose the producer bundled one
unit of x together with one unit of y and sold this bundle as one unit
of z to customers. Again suppose that the producer sets the
competitive price of $2 for the bundle. Then the producer could
immediately create any ‘product y deficit’ between $1 and -$1 that
they like by ‘breaking down’ the $2 price of the bundle between the
two inputs. For example, if the producer stated that of the $2
consumer price, $1.35 was a payment for x and $0.65 was a payment
for y, then the producer would create a ‘product y deficit’ of $0.35 per
unit. Again, this figure is meaningless as it does not affect the price
that consumers pay for the bundle and it does not affect the sales or
cost of production for the bundle. From an economic perspective the
‘deficit’ is an artificial figure.

There is no reason in economics why the provider of a product that

involves a number of inputs needs to receive revenues on each
individual input that exactly offset that input’s cost. For example,
mobile phone companies regularly ‘sell’ mobile phones below cost.
However, this does not reflect irrationality on the part of the sellers.
Rather, the sellers expect to make up any ‘phone deficit’ through call
charges and customer access revenues. The sellers care about total
profit, not the revenues and costs associated with each individual
input. Further, there is nothing ‘anti-competitive’ about such pricing.
Such pricing could arise even in highly competitive markets and often
reflects a benefit to customers. For example, in mobile phones, new
mobile phone users might be reluctant to pay a large sum for a phone
when they are unsure of how much they will use that phone. By
selling the phone at a low price that is below cost the mobile phone
company takes the risk away from the customer. If the customer
finds that they do not use the mobile phone very much, then they pay
little. Conversely, if the customer finds the mobile phone useful, then
they pay for the phone through the call charges. By taking the risk
away from the customers, more customers will use mobile phone

7
Section 4 The Access Deficit as an Economic Construct

companies that set low phone prices and these companies will be
more profitable.

From an economic perspective, the correct way to view the CAN is

as one of the common fixed costs necessary to provide a variety of
telecommunications products, including PSTN access. To evaluate
any deficit created for Telstra due to price constraints on the CAN, it
is necessary to consider all of Telstra’s costs and revenues from
providing services that involve customer access. If providing access
to a customer is profitable to Telstra, when all revenues and costs
associated with that customer are considered, then there is no
meaningful ‘access deficit’ for that customer.

A further problem with the access deficit calculation arises due to its
relationship with the Telstra price caps. If these price caps do not
bind then the access deficit is calculated using the maximum access
revenues that Telstra could have received under the price caps rather
than the revenues Telstra actually did receive. While this reduces the
access deficit, it also highlights that the deficit is an arbitrary construct
with little if any economic meaning. If a firm subject to a price cap
sets profit maximising prices that fall under the price cap, then that
price cap is not a binding constraint on the firm. Under standard
economic assumptions that the firm has a well-behaved profit
function and seeks to maximise profits, a non-binding price cap is
redundant. The price cap could be doubled, tripled, or removed
altogether without altering the behaviour of the firm. In this situation,
market conditions other than the price cap are constraining the firm’s
behaviour so that it is not profit maximising for the firm to price ‘up
to’ the cap. In this situation, to define an ‘access deficit’ on the basis
of a non-binding price cap has no economic meaning. The deficit
could be eliminated by simply raising the (non-binding) price cap. As
the cap is not binding, increasing the cap has no effect on any actual
market behaviour. But as the cap rises, the maximum ‘allowed’ line
rental would rise and the deficit would fall. The access deficit could
be arbitrarily raised or lowered by altering the non-binding price cap
without changing any actual behaviour.

In summary, the access deficit, as calculated by the ACCC, has little if

any economic meaning. It does not consider the CAN as an input to
a variety of telecommunications products but rather treats the CAN
as if it were a stand-alone product. The deficit calculation can be
based on fictitious revenues if the price cap for access charges is not
binding. A sensible approach to deal with any potential problems that
arise for Telstra due to its provision of the CAN and its price
regulations is to consider all the revenues and costs borne by Telstra
from services that are produced using the CAN. This is the approach
used for USO funding.

8
Section 5 Should Telstra Receive an Access Deficit Contribution?

5. Should Telstra Receive an Access

Deficit Contribution?

In section 4 we considered the economic basis for the access deficit,

as calculated by the ACCC. For the purpose of the current section we
will take that definition as given. In other words, we assume that
there is an ‘access deficit’ in the sense that Telstra’s line rental and
connection charges do not cover the stand alone costs of providing
the CAN.

If there is an access deficit, why should Telstra receive a contribution

from access seekers to the PSTN for this deficit? From our reading of
various documents articulating the case for an ADC, there are two
distinct arguments that have been put forward to support an access
deficit contribution. The ACCC emphasises issues of investment
incentives if an access deficit contribution is not made:
… to the extent that legislation restricts Telstra’s
ability to raise line charges to cost, then preventing
Telstra from seeking a contribution from charges for
call services would lead to under-recovery of fixed
line costs, thereby discouraging efficient levels of
infrastructure of investment. (July 2000, p.37)

It has also been argued that failure to have an ADC will lead to an
under-pricing of access to Telstra’s PSTN and hence, excessive entry
in related downstream markets. Thus, there is an investment incentives
argument and a cream skimming argument. The investment incentives
argument states why Telstra should receive an access deficit
contribution. The cream skimming argument is used to justify why
the ADC should come from access seekers to the PSTN.
Unfortunately, there is little elaboration of these arguments in the
Australian debate. Each, when presented, is stated as a contention. In
what follows we evaluate investment incentives and cream skimming
arguments and identify the facts that would have to be established to
give them some validity.

5.1 The Investment Incentives Argument

At a fundamental level, any losses Telstra makes on basic access as a

stand-alone business represent a fixed cost of being a ubiquitous

9
Section 5 Should Telstra Receive an Access Deficit Contribution?

provider of telecommunications services. As we noted above, it is

existence of related services that make basic access valuable and it is
the extent to which profits can be earned on those other services that
allows an access deficit to be financed. The access deficit is a fixed
cost that is recovered from the profits of other related activities.

This is the essence of the investment incentives argument. Suppose

that entrants did not have to bear any part of the fixed costs
associated with providing basic access. Then it is possible that such
entrants could compete away all of Telstra’s business where it was
generating profits to pay for the fixed costs of basic access. In this
dire scenario, Telstra would go bankrupt and would certainly have no
incentive to maintain and renew infrastructure for providing basic
access.

To see this argument more clearly consider the following example:3

We begin with a single telecommunications carrier who provides
three services to three different customers (or types of customers).
Service
Customer A B C Total Profit
1 -10 -2 4 -8
2 -4 2 8 6
3 -1 3 10 12
Total Profit -15 3 22 10

In this example, service A is the basic access service. Notice that the
firm makes a loss on this service alone. Service B is a less profitable
related service such as local calls while service C is more profitable.
These figures show the profits on each segment if the provider is a
monopolist. However, suppose that the firm is exposed to
competition. As services B and C are profitable, entry could occur in
those related services if access to service A was mandated. If such
competition was strong, the firm would be left with service A alone
and an overall loss.

For this reason, a provider of access would, at first glance, appear to

warrant some contribution to its fixed costs – including the access
deficit. Moreover, note that it is precisely because the access service is

3
This example comes from Mark Armstrong and Chris Doyle, “Social Obligations
and Access Pricing: Telecommunications and Railways in the U.K,” in Gabel,
David and Weiman, David F., eds. Opening networks to competition: The regulation and
pricing of access. Topics in Regulatory Economics and Policy Series, Boston; Dordrecht and
London: Kluwer Academic, 1998, pages 159-79.

10
Section 5 Should Telstra Receive an Access Deficit Contribution?

regulated that a contribution is potentially required. If it were not

regulated, the provider would be able to increase its line rentals as
related services became more valuable to customers; i.e., as
competition improved consumer surplus in those areas.

This simple argument is incomplete because Telstra receives a

payment from the government for losses in providing basic access.
The universal service obligation and its associated industry fund is
designed to cover situations where Telstra has a customer (or an
exchange area) on which it makes an overall loss -- including what it
earns in competition with others. Therefore, under the dire
competitive scenario, where Telstra is left with no profits on services
B and C, Telstra’s access deficit would be precisely covered.

To see this, consider what would happen if there is intense

competition faced by the provider in our example on services B and
C. Telstra is only left with the economic losses associated with service
A. In this case, all three customers would be loss making for the firm
and the USO fund would grow to $15.4 Most of the USO would be
paid by entrants ensuring that the provider did not make a loss.

As such, the existence of the USO rules out the dire scenario of
Telstra being made bankrupt and indeed any real possibility of entry
causing serious losses for Telstra. Indeed, the USO fund guarantees
that if Telstra were to only be a provider of basic access, it would earn
a market return on that service.

The existence of the USO fund means that Telstra’s fixed costs
associated with providing access will always be covered. What,
however, does this mean for its incentives to invest in the CAN?
Recall, that it has been argued that Telstra would have little if any
incentive to invest in the CAN if it did not receive an access deficit
contribution. However, Telstra’s universal service obligations require
such investments and compensate Telstra for them through the USO
fund. Indeed, as that fund is cost-based, if anything it may provide
too much rather than too little incentive to maintain and expand the
CAN. This is because Telstra is effectively reimbursed for such
expenses; subject, of course, to its use of best available technologies.

Consequently, the existence of the USO fund and the basis upon
which it is said to operate mean that the investment incentives
argument is unlikely to provide an appropriate justification for an
access deficit contribution. For this to be established would require

4 We can consider customers 1, 2 and 3 as individual customers or as groups of

customers such as a local exchange area.

11
Section 5 Should Telstra Receive an Access Deficit Contribution?

proof that the USO fund was not performing its stated functions and
that this was leading to competitive forces that are not allowing
Telstra to cover its fixed costs (including any access deficit).

5.2 The Cream Skimming Argument

The second argument used to justify the access deficit is based on

cream skimming. This argument suggests that if no ADC is allowed
then entrants will face more favourable conditions than the access
provider in related service markets. If this is the case, then it is
possible that entry could occur even when an entrant has higher
production costs or lower product quality than the access provider in
those related service markets.

To see this argument, suppose that in, say, long distance markets,
Telstra has a cost of C per subscriber and generates gross value of U
per subscriber. Suppose also that the price of Telstra’s services is P
per subscriber. In this situation, Telstra’s customers receive U – P.5

Now consider a potential entrant who offers a long distance product

of value u to subscribers and that costs for the entrant are c per
subscriber. Suppose that c includes the costs associated with access to
Telstra’s CAN but no ADC. That is, if the cost of the CAN to Telstra
is CA, the access charge is t and the long distance cost to the entrant is
cL, then c = cL + t = cL + CA. In this case, entry is desirable if total
surplus u – c exceeds that generated when the incumbent serves that
customer, that is, U – C. In other words, entry is socially efficient if C
≥ c + [U – u].

Given Telstra’s price of P, the entrant can in fact sign up a subscriber

so long as the price it charges, p, is such that u – p ≥ U – P. Thus, the
maximum price that could be charged by an entrant is p = P – [U – u]
and so entry will occur if this price covers the entrant’s costs, c. That
is, entry will occur if P ≥ c + [U – u]. Notice that this is different from
the condition for socially efficient entry, which occurs whenever P
does not equal C. If P exceeds C, there may be too much entry.

In this situation, if an additional payment was made by the entrant to

the incumbent, a = P – C, then for entry to be privately profitable, P -

5
This example is based on one from Mark Armstrong, “The Theory of Access
Pricing and Interconnection,” M. Cave, S. Majumdar and I. Volgelsang (eds.),
Handbook of Telecommunications Economics, Amsterdam: North-Holland, 2002.

12
Section 5 Should Telstra Receive an Access Deficit Contribution?

a ≥ c + [U – u] or C ≥ c + [U – u], which is the same as condition for

socially efficient entry.6

How does this argument relate to the ADC? Suppose that C is

decoupled into a CAN, CA, and long distance cost, CL. In this case,
the total access payment would be t = a + CA which would equal P –
CL. P exceeds CA + CL so that P – CL exceeds CA which exceeds any
price earned on basic access if there is an access deficit. Hence, the
entrant is implicitly required to make a contribution to that deficit.
Note that this is not to fund the provider’s losses on access, but to
correct for distortions in the long distance market because the
provider’s long distance price deviated from marginal cost.

This raises a critical question: why does the provider’s long distance
price differ from marginal cost? The above argument is predicated on
the assumption that long distance prices are regulated and fixed. In
other words, it assumes that Telstra cannot lower its price below P in
the presence of competition. However, in Australia, long distance
prices are deregulated and indeed, subject to considerable
competition. In this situation, the argument that there is a long
distance distortion that needs to be ‘cured’ by the use of an ADC
does not hold.

To see this, suppose that Telstra has a retail price for long distance
above a + c (where a is a given ADC). In this case, Telstra will receive
a contribution of a per unit from providing access. If, however,
Telstra lowers its price in response to entry, it will be able to charge
up to P = a + c + [U – u]. In this case, it will earn a profit of a + c +
[U – u] – C. This will not be profitable for the incumbent if a ≥ a + c
– C + [U – u] or C ≥ c + [U – u]. Hence, regardless of the choice of
ADC, a, entry will be socially efficient. As such, no ADC is required
to generate efficient entry conditions.

In this situation, whether an ADC is allowed or not does not change

the efficiency of entry. This occurs so long as the long distance (or
related service market) is fairly competitive (notwithstanding product
differentiation) and there is no constraint on the provider in terms of
lowering price to meet an entrant.

What might constrain Telstra’s ability to lower its price in response to

entry? At a first glance, Telstra may be unable to simply match a
competitor’s price to a customer without lowering prices to all of its
customers. This is the infra-marginal problem in pricing, which

6
This argument basically states that the access price should be driven by an ECPR
logic.

13
Section 5 Should Telstra Receive an Access Deficit Contribution?

reduces Telstra’s incentive to meet small-scale competitive pressure.

However, even subject to this constraint, Telstra is choosing not to
lower overall prices because it is still able to earn a positive margin on
other customers. To be sure, inefficient entry is possible due to
Telstra’s exploitation of market power, but equally true is that an
ADC is merely another way of deterring entry in general. This is
because such a charge reduces competitive pressure from efficient
entrants as well inefficient entrants. Because of this, access pricing
should include a factor that offsets any ADC.7

However, there are two reasons why this is unlikely to be a strong

constraint on price responsiveness. First, the nature of current retail
price regulations gives Telstra an additional incentive to respond to
entry. Prices for a basket of services are regulated, thus Telstra can
lower its price on one service and raise it on services that are not
subject to as intense competition. Moreover, to the extent that such
entry is with respect to customers in loss making areas (or where
entry creates a new loss making area), any profit reduction from the
incumbent is compensated for by increased USO payments. In this
case, the potential consequences of entry are borne by all non-USO
carriers who have increased USO payments as a result. For Telstra, a
cost associated with meeting competitive prices is removed and price
responsiveness is more likely.

5.3 Conclusion

There are two reasons given why Telstra needs to receive a

contribution for the access deficit. The first – based on investment
incentives – arises because of a concern that competition based on
access to the PSTN will eliminate Telstra’s ability to earn sufficient
profits on related telecommunications services to earn a market rate
of return on investments in the CAN. However, the existence of a
USO fund underwrites that investment to the extent that there is any
short-fall caused by competition in related service segments. Hence,
Telstra’s return on assets is assured; thereby not leading to a
reduction in incentives to investment in the CAN (let alone no
incentive as Telstra is contending).

The second argument – based on the productive inefficiencies that

might result from cream-skimming – is a legitimate concern where
competition is in markets where Telstra is unable to lower its price to
meet competitive pressure from entrants. When Telstra can compete

7
Armstrong, op.cit.

14
Section 5 Should Telstra Receive an Access Deficit Contribution?

effectively on price, entry will occur only if it is efficient, regardless of

the existence of an ADC. Consequently, the cream-skimming
argument does not provide a rationale for an ADC. Telstra does not
face any significant constraints on price responsiveness in related
telecommunications service markets and, indeed, has additional
incentives to compete resulting from USO funding measures and
retail price controls.

To be sure, these arguments are contingent upon the effective

operation of the USO fund and the continued application of the
existing methodology for calculating retail price caps. At present, we
have not seen any arguments suggesting that either of these key
related policies are operating in a manner significantly different than
is intended or are likely to change in the future.

15
Section 6 Methods of Recovering any “Access Deficit”

6. Methods of Recovering any “Access

Deficit”

In this section we suppose that there is a legitimate reason that would

justify Telstra receiving an access deficit contribution. We consider
how they ADC should then be recovered by first looking at the
methods proposed in the Australian debate. We consider an
alternative approach that is likely to generate more competitive and
efficient outcomes.

6.1 The Telstra Approach

It has been contended that the access deficit should be recovered

from all PSTN users. For example, Telstra has argued that the AD
should be allocated totally on all calls (flagfall) – their 100:0 rule.
Telstra argue that this allocation is optimal because the demand for
calls is less elastic than that for call minutes (ACCC, July 2000, p.29).

This approach has been justified by appealing to work done by Park,

Wetzel and Mitchell.8 This is a famous study but in many ways is not
applicable to the present day Australian context. In particular, (1) the
study looked at calls made in Central Illinois in 1975-77; hardly a
model for today’s environment and usage patterns; (2) the
specification of their demand model does not appear to be of a form
that could give rise to extrapolations that would justify a 100:0 rule;
(3) the study was for local calls, whereas the access issues here are for
other services whose demand patterns are unlikely to be comparable;
and (4) not all of the relevant coefficients in the Park, Wetzel and
Mitchell study were statistically significant although this factor may
change with a more comprehensive study tailored to a modern
telecommunications system.

8 “Price Elasticities for Local Telephone Calls,” Econometrica, 51 (5), 1983, pp.1699-

1730.

16
Section 6 Methods of Recovering any “Access Deficit”

6.2 The ACCC Approach

The ACCC has also considered allocating the ADC on the basis of
call ends and call minutes. In its initial undertaking reports, the ACCC
was concerned that by using call ends this would result “in a greater
amount of the access deficit being allocated to the declared PSTN
services than when call minutes [are] used to allocate the access
deficit.” (ACCC, July 2000, p.25) This is because average call times
were lower on declared services as opposed to the PSTN in general.
It appears that because of a lack of information upon which to
evaluate this distribution the ACCC chose a 50 percent weighting of
call ends to call minutes (July, 2000, p.38).

6.3 Evaluation of Call-Based Approaches

Both the Telstra and ACCC methods of allocating the access deficit
among access seekers and the provider are based on call volumes
(although weights on ends and minutes vary). The difficulty with this
approach is that when pricing its services, the marginal cost of an
additional call and call minute are increased above Telstra’s costs of
supplying PSTN services.

This has two effects. First, access seekers face a different marginal
cost than does Telstra. This is because the access deficit, as we argued
earlier, is a fixed cost of operating a PSTN network for Telstra.
Hence, it does not factor into their pricing as it would into the pricing
of entrants. As such, entrant’s prices are likely to be too high.

Second, this increased marginal cost on access seekers will tend to

reduce price competition between entrants and Telstra. As a
consequence, Telstra’s call prices will be higher than would be socially
desirable. By passing on access deficit costs through unit access
charges, rivals face higher costs at the margin than Telstra reducing
the pressure on Telstra’s prices. Indeed, it is potentially this impact
that motivated the ACCC to structure the weighting of call ends to
minutes so as to reduce the overall marginal impact of the ADC.

6.4 An Alternative

As a matter of principle, fixed network costs – like the access deficit –

should be recovered through fixed charges. One approach to
recovering fixed network costs that is currently used in Australia is

17
Section 6 Methods of Recovering any “Access Deficit”

the approach associated with Telstra’s fixed universal service

obligation costs. It seems sensible that, if an access deficit exists that
requires a contribution from access seekers, then this deficit should
be recovered in a manner similar to the USO costs. That is, the deficit
would be recovered according to a weighting based on relative carrier
revenues (net of access payments).9

Maintaining consistency between the attribution of the USO costs

and any access deficit would add transparency to the regulatory
system. This said, the USO funding mechanism is not a true ‘fixed
charge’ but may have competitive effects by altering a carrier’s
incentives to compete vigorously and raise revenue at the margin. To
our knowledge, an economic analysis comparing the competitive
effects of the current USO funding mechanism with alternative
mechanisms has not been undertaken in Australia. While it seems
likely that the de-coupling of marginal costs and payments under the
USO funding mechanism will be more pro-competitive than many
alternatives, this is an area that requires further economic analysis.

The advantage of recovering access deficits in a similar manner to the

USO is that it adds transparency and clarity to the system, decouples
payments for fixed costs from distortions to marginal prices and
potentially avoids distortions on competition that would otherwise
result. As the access deficit contribution would be determined in a
similar manner to the USO, it would require a minimal additional set
of information to implement.

9
This is the recommendation of the Centre for Research in Network Economics
and Communications at the University of Auckland in their paper “Estimating the
Cost of the KSO,” 1st June 2001 and also of the Productivity Commission in their
draft report.

18
Section 6 Conclusion: Answers to the ACCC’s Questions

6 Conclusion: Answers to the ACCC’s

Questions

Here we consider answers to the some of the ACCC’s questions in its

2003 issues paper.

6.1 Should access seekers still be required to

make a contribution to the AD?

There appear to be no broad efficiency arguments to justify an ADC.

We have demonstrated that the absence of an ADC will not
undermine Telstra’s investment incentives nor encourage inefficiency
entry. This is due to the existence of a USO fund in Australia.

It would have to be established that there was an overall concern

regarding Telstra’s financial viability coming from imperfections in
the USO fund rules. However, that viability would have to be
assessed over all markets which may receive a benefit from vertical
integration of Telstra with the CAN. If there is no benefit from
vertical integration then this would suggest (1) Telstra’s divisions
should be structurally separated from the CAN; and (2) that Telstra’s
CAN charges reflect an AD from all downstream firms. We are
sceptical of claims that Telstra’s internal transfer prices replicate a
vertically separated structure.

6.2 If there is to continue to be an AD, how

should it be defined and measured?

Even if an access deficit is to continue, the current approach to the

access deficit used by the Commission makes little if any economic
sense. A sensible measure of any access deficit needs to take into
account the revenues and costs Telstra faces over all its operations
that use the CAN as a basic input. Such a calculation would, of
course, be similar to the current USO approach.

19
Section 6 Conclusion: Answers to the ACCC’s Questions

6.3 If an ADC were to be retained, how should it

be spread over calls and minutes?

The answer is that it should not be spread over calls or minutes. Both
are distortionary and likely lead to consumer harm. Instead, a fixed
charge similar to the USO obligations should be imposed on
downstream telecommunications providers.

20
Section 0 Appendix A

Appendix A

This appendix describes in detail the Universal Service Obligation

(USO) funding regime and the retail price controls placed on Telstra.

A.1 The Universal Service Obligation Scheme

A.1.1 The Regime

The USO regime was introduced in the Telecommunications Act 1991

and subsequently contained in the Telecommunications Act 1997.
Regulatory oversight for the USO regime is the responsibility of the
Australian Communications Authority (ACA). The current universal
service arrangements are specified in the Telecommunications (Consumer
Protection and Service Standards) Act 1999 (“the TCPSS Act”). The
TCPSS Act allows for the specific regulation of the prices charged for
universal service, however by 2002 price regulation had only occurred
via the general price controls under the Act.10

The TCPSS Act also allows contestability in delivering the USO. On

21 June 2002 the ACA released guidelines designed to assist carriers
and carriage service providers prepare applications to become
approved competing USPs. To date, Telstra has been the only
declared provider of the Universal Service Obligation.11

USO subsidies are determined by the Minister, based on advice from

the ACA. These determinations may be made up to 3 years in
advance. Subsidies for 2002/03, 2003/04, 2004/05 have already been
set.

Before 2000 USPs reclaimed a Net Universal Service Cost (NUSC).

Though new legislation (discussed below) does not refer to the

10
For further details see Productivity Commission, Telecommunications Competition
Regulation, Final Report, December 2001 at Chapter 18
11
See further at http://www.aca.gov.au/consumer/uso/index.htm

21
Section 0 Appendix A

concept, the NUSC model is still applied by the ACA when

formulating its advice for the Minister.12

A.1.2 The Access Deficit and the USO

A 1999 ACCC paper on the Access Deficit charges notes that “a

proportion of the access deficit will be incurred in Universal Service
Obligation (USO) net cost areas. As the funding of the USO is
separately determined, this proportion of the access deficit should not
be recovered from PSTN services”13. The NUSC is subtracted from
the Access Deficit Charge, subject to a reduction for the net loss
from payphones and the proportion relating to provision of
wholesale PSTN services. In its 2000 PSTN undertaking (originally
submitted September 1999) Telstra sought to recover part of its USO
costs as part of the access deficit charges. However, this application
was rejected by the ACCC which subtracted costs recovered under
the USO regime from the access deficit loss. 14

A.1.3 Calculating the Net Universal Service Cost (NUSC)

The formula for calculating the overall size of the NUSC is the
amount by which the costs attributable to the provision services to
net cost areas (avoidable costs) exceed the revenues earned from the
provision of the services (revenues foregone). This formula was originally
contained in the relevant USO legislation. A model for calculating the
NUSC in accordance with the legislation was developed by Bellcore
International in 1996.15 The TCPSS Act (the current USO legislation)
was amended under the Telecommunications (Consumer Protection and
Service Standards) Amendment Act (No. 2) 2000 (commencing July 2000).
Under the amendments the avoidable costs minus revenues foregone
formula has been removed from the Act and no alternative
methodology is prescribed. The ACA Advice to the Minister on New
USO arrangements (prepared prior to passage of the legislation) notes
that:

12 Australian Communications Authority, Annual Report 2001-2002, October 2002

at p. 71
13
Australian Competition and Consumer Commission, Interconnection Charges and
Telstra's Access Deficit, Discussion Paper, September 1999 at p. 5
14
Australian Competition and Consumer Commission, A report on the assessment of
Telstra's undertaking for the Domestic PSTN Originating and Terminating Access services, July
2000 at p. 66
15
Australian Communications Authority, USO Costing and Assessment Arrangements,
May 1999

22
Section 0 Appendix A

Importantly, the Bill does not refer to NUSC, nor

use that concept in any way. In other words, the
policy inherent in the Act of enabling USPs to
recoup their losses or costs is not replicated in the
Bill... Therefore, there will be no inconsistency with
policy or proposed legislation if subsidies do or do
not recompense USPs for losses incurred.16

Notwithstanding, the Explanatory Memorandum to the Act states

that the use of the current methodology is a matter of judgement for
the Minister and ACA. In its Advice (prepared September 2000) the
ACA recommended the continued use of the avoidable costs,
revenues foregone methodology and the NUSC model.17

Advice provided to the Minister on subsidies for the 2002/03,

2003/04 and 2004/05 financial years was derived from the same
Bellcore methodology.18 The approach has three major parts:
determining the net cost areas (NCAs), determining the avoidable
costs and determining the revenues foregone.

A.1.4 Net Cost Areas

The first step in calculating the NUSC is to determine the regions of

Australia where the provision of telecommunications services is loss
making. In defining Net Cost Areas the aim is to identify areas where
it was possible that a profit-driven carrier may not provide services.19
In doing so, the ACA considers that it is not appropriate to assess the
profitability of a single service in isolation, even though there will
clearly be loss making services/customers even in urban areas (e.g. a
particularly difficult-to-access property). The Bellcore Deliverable # 1
sets out the rationale for costing the USO on a regional rather than
per-customer basis. The reasons are twofold:
The first is to ensure that universal service funding is
directed only to those areas for which it is intended -
low density, costly to serve rural and remote areas
and customers. … The second … is a practical one.
By eliminating the need to study cost structures and
revenue sources in urban areas, both the

16
Australian Communications Authority, Advice to the Minister New USO
Arrangements, September 2000 at p. 6
17
Australian Communications Authority (July 2000), op. cit., at p. 7
18 Australian Communications Authority (October 2002), op. cit., at p. 71.
19
Australian Communications Authority, Estimate of Net Universal Service Costs for
1998/99 and 1999/2000, January 2000

23
Section 0 Appendix A

development of appropriate cost functions and

overall model processing requirement can be
significantly reduced.20

Methodology

Under the Bellcore methodology, the first step in determining the

NCAs is a threshold determination of Potential Net Loss Areas
(PNLAs). The Universal Service Provider must lodge proposed
Potential Net Loss Areas (PNLAs) with the ACA within 60 days of
the beginning of the financial year. Not all PNLA will be loss-making;
the PNLA test is a 'first cut’ to identify those types of areas which
should be subject to analysis”.21 The ACA evaluates whether the
PNLAs provided are NCAs using the avoidable costs - revenues
foregone methodology.

The analysis of PNLAs begins at the Exchange Service Area (ESA)

level. There are several different regions that will be considered a
PNLA:
1) Small exchange areas (less than 150 SIOs)
2) ESAs with more than 150 SIOs. These are broken down
into two PNLAS: built-up areas (BUAs) (rural townships)
and non-built-up areas (NBUAs) (the areas surrounding
rural townships). There are a number of rules for
determining how to divide an ESA into a BUA and
NBUA.22 The aim here has been to capture the loss
making areas outside of rural townships that would not be
identified as loss-making if the region was taken to be a
single area.23
3) Customers receiving radio services
4) From 1998/99, customers receiving satellite services
5) Payphones

Small exchange areas are costed at the full exchange level. NBUAs
and BUAs are costed separately but on a regional basis. Customers
receiving radio or satellite services and payphone services are costed
individually.

20
Bellcore International, Net Loss Area Specifications, Net Universal Service Cost
Consultancy Agreement Phase 2, Deliverable #1, November 1996 at p. 1
21
Bellcore International (1996), op. cit.,, at p. 1
22
Bellcore International (1996), op. cit., at p. 3
23
Australian Communications Authority (January 2000), op. cit.

24
Section 0 Appendix A

The PNLA criteria was debated and discussed in 1997 by the

participating carriers at the time (Telstra, Vodafone and Optus)
during the development of the Bellcore NUSC methodology and is
contained in Deliverable #1.24 However, since this time the
methodology has been criticised by the other participating carriers. In
particular, the carriers have criticised the inclusion of radio services as
an independent NCA (Vodafone) and the separation of BUAs and
NBUAs (AAPT).25 The ACA continued with the Bellcore approach
in its declarations for 1998/99 and 1999/2000. However, while it
accepted the validity of separating BUAs and NBUAs it saw
problems with the inclusion of radio services as a separate service
arguing that:
Including a ‘radio services’ NCA in a NBUA NCA
may reduce the NUSC in the few situations where a
loss making radio service NCA is included in a
profitable NBUA and the NBUA remains profitable
after this inclusion.26

A.1.5 Avoidable Costs

The NUSC model developed by Bellcore International proposed a

methodology for calculation of avoidable costs. This methodology
was adopted by the ACA and incorporated by reference into the Net
Universal Service Avoidable Costs Determination 1998 (5).27 It continues to
be the methodology adopted for calculating avoidable costs and the
NUSC. Avoidable costs are calculated according to the long run
incremental costs of providing standard telephone service, payphones
and prescribed carriage services to customers in net cost areas.
Incremental costs are the costs that would be avoided if the services
to customers in net cost areas were no longer supplied. These are
calculated according to the costs a prospective carrier would incur in
providing the service using the most cost efficient technology and
production practices, rather than the historical costs actually incurred
by the USP.28 The long run cost concept takes into account the cost
of replacing fixed assets for the provision of the service.29 The cost of

24
Bellcore International (1996), op. cit
25
Australian Communications Authority (January 2000), op. cit
26
Australian Communications Authority (January 2000), op. cit.
27
This determination only applied to the 1997/98 but the same methodology has
been used for future NUSC assessments.
28
Australian Communications Authority (May 1999), op. cit.
29
Ibid.

25
Section 0 Appendix A

assets is valued on the basis that a new fully optimised network is

supplied each year.

To calculate total avoidable costs, first the installed cost of

technology, cost of capital and operating costs for a range of different
technologies (using best practices) are determined for each NCA.
Sampling processes are used in determining the cost for each NCA.30
Once the costs associated with each technology have been
determined an optimal mix of technologies for the provision of the
USO services is selected whereby each NCA is supplied using least
cost technology. Consequently, the major determinants of the size of
the NUSC are the:
• Technology installation and operating costs
• Cost of capital
• Choice of least-cost technologies

Technology Costs

Before the costs associated with each technology are calculated a

short-list of appropriate technologies is created based upon
availability, suitability to Australian conditions, capacity for
integration with existing telecommunications network, ability to meet
Australian regulations, codes and standards.31 Then the installed cost
of technology, depreciation rates and operating expenses are
calculated for each of the short-listed technologies.32

Installed Cost of Technology

These are the installed costs of switches, junctions, cable, wireless

local loop, satellite and any other technologies associated with each
type of telecommunications network in an NCA.33 The historical
costs of the universal service provider are not used as the basis for
deriving these costs. Rather, the costs are those that would be borne
by the most cost-efficient operator using best practices.34 A

30
Australian Communications Authority, Net Universal Service Cost Assessment for
1997-98, October 1999
31
Australian Communications Authority (October 1999), op. cit., at p. 52
32
ibid., p. 67
33
ibid.
34 ibid., p.12.

26
Section 0 Appendix A

‘greenfields approach’ is adopted for calculating installed costs.35 The

installed cost represents:
the fully engineered, furnished, and installed (EF&I)
costs of replacement facilities ... including
design/engineering time, first level supervision, travel
time, commissioning costs.36

The Bellcore NUSC methodology allows for sampling to derive

values rather than requiring analysis of a complete greenfields
network.37

Depreciation Rates

The Avoidable Costs Determination defines depreciation as “the

allocation of the initial investment in plant to periods of service
provided by the plant”.38 The formula for calculating depreciation
factor is:39
Depreciation Factor = (1-Net Salvage)/Asset Life

As the above formula implies, straight-line depreciation rather than an

economic rate is adopted.40 Asset lives are determined for each type
of asset based on known asset lives. Net Salvage is the proportion of
the net investment that is salvageable at the end of its useful life. This
is determined at the ESA level.

Operating Expenses

The Avoidable Costs Determination recommends a methodology for

determining avoidable operating expenses. Avoidable operating
expenses are defined as “any and all costs which are a direct result of
providing telecommunications service and are expensed (or ‘written
off’) for tax and/or book purposes in the year in which they occur”.41
Operating expenses are measured using an Activity Based Costing
approach. Avoidable operating expenses are categorised as follows:

35 ibid., p.14.
36 ibid., p.14.
37
ibid., p.14.
38
ibid., p.92.
39 ibid., p.92.
40
ibid., p.5.
41
ibid., p.101.

27
Section 0 Appendix A

• Direct depot costs (resulting from customer demand activity

occurring in the depot)
• Direct central costs (resulting from customer demand activity
at a central/regional location)
• Indirect depot costs (costs occurring at the depot not related to
customer demand)
• Indirect central costs (costs occurring at a central location not
directly related to customer demand).42

Opportunity Cost of Capital

This is calculated as a pre-tax weighted average cost of capital

(WACC). The Capital Asset Pricing Model is used to determine the
required rate of return on equity. According to the methodology set
out in the Avoidable Costs Determination asset costs are valued on
the basis that a new fully optimised network is supplied each year.
There is thus a concern that if a standard calculation of asset
depreciation were adopted this would overstate actual costs as the
rate of depreciation is higher during the first year of the life of an
asset.43 Consequently, it was recommended to the ACA by Austel and
the Allen Consulting Group that a levelised version of the WACC be
used.44 Levelisation works to make annual capital costs the “annuity
amount which would make the NPV related to the replacement cost
of an asset zero at the WACC over the expected economic life of the
replacement technology”.45

Least-Cost Technology Mix

Avoidable costs are calculated according to the prospective costs a

carrier would incur in providing the service using the most cost
efficient technology and production practices.46

The least-cost technology is the one with the lowest annual costs after
taking into account capital costs (WACC and depreciation) and
operating costs. This approach is favoured to an approach whereby

42
ibid., p.101.
43
ibid., p.46.
44
ibid., p.46.
45
ibid., p.46. The levelisation formula can be found at page 49.
46
Australian Communications Authority (May 1999), op. cit.

28
Section 0 Appendix A

the least-cost technology is that with the least cost in year one or with
the lowest NPV. 47

A.1.6 Revenues Foregone

Under the NUSC methodology, revenue foregone is the amount of

revenue earned by the USP during that financial year that the USP
would not have earned if it had not supplied NCAs. During the
development of the NUSC model Bellcore International derived a
basis for calculating revenues foregone that was agreed to by all
participating parties. This method has been adopted by the ACA as
an appropriate method for calculating revenues foregone.48

Revenues foregone are broken up in a series of categories:

• Local call revenue
• Long distance originating revenue
• Long distance terminating revenue
• Long distance between NCAs
• International
• Mobile-related
• Payphones
• Access Revenue
• Other (Operator-assistance, white and yellow pages directories,
ISP revenue)

Revenues foregone are calculated according to the amount actually

earned by the USP rather than the amount it could feasibly earn given
price caps in existence. From 1997/98 to 1999/2000 the NUSC was
calculated in arrears making it easier for the actual revenues that
Telstra earned in NCAs to be ascertained. The amount of revenue
foregone in 1998/99 and 1999/2000 was derived from the level of
revenue in 1997/98 subject to a revenue variation amount provided
by Telstra.49 The revenue variation amount was the variation
experienced in its overall operations for each of a series of
telecommunications service categories (eg local calls, STD, IDD,

47
Australian Communications Authority (October 1999), op. cit., pp. 105-106
48
ibid.
49 Australian Communications Authority, Estimate of Net Universal Service Costs for

1998/99 and 1999/2000, January 2000

29
Section 0 Appendix A

Fixed to Mobile). Some concern was expressed that the amount of

variation would not be constant between NCAs and non-NCAs.
Because revenues foregone are calculated using sampling processes
there is a statistical error associated with the estimate.

From 2000/01, the NUSC was calculated and paid in advance.

Consequently, an estimate of the amount of revenues foregone is
now used.50 Variances in key inputs, such as revenue, are made based
on advice from the Reserve Bank, industry consultants and major
industry participants.51

A.1.7 Funding the NUSC

Since 1997-98 the Universal Service Levy to be paid by each

telecommunications carrier has been calculated according to each
participating carrier's share of the total ‘eligible revenue’ earned in the
Australian telecommunications industry. Prior to 1997 the universal
service levy was shared amongst carriers in proportion to each
carrier's share of timed telecommunications traffic. “Eligible revenue”
was chosen by the government as the method for apportioning the
universal service levy on the basis that it :
Broadly spreads the burden of USO contributions
across the telecommunications industry, is
transparent, makes use of readily accessible data, is
administratively simple and competitively neutral,
both between carriers and between carriers and non
carriers with whom they compete.52

50
Australian Communications Authority (September 2000), op. cit., at p. 52
51 Australian Communications Authority (October 2002), op. cit., at p. 71.
52
Department for Communications, Information Technology and the Arts,
Explanatory Statement to the Telecommunications Universal Service Obligation (Eligible
Revenues) Regulations 1998, 1998

30
Section 0 Appendix A

The ACA prepared a new Determination, the Telecommunications

Universal Service Obligation (Eligible Revenue) (“the Determination”), on
27 June 2002 regarding the eligible revenue arrangements. In making
the Determination, the previous legal instruments that provided for
the calculation of eligible revenue were repealed.53

Participating Carriers

All licensed carriers are considered “participating persons” in the

scheme, unless exempted by regulations under section 20A of the
Telecommunications (Consumer Protection and Service Standards) Act 1999. At
the date of the Determination, only licensed carriers have been
required to pay the Universal Service Levy, however the Minister has
been written to allow for the possibility of carriage service providers
being made participating persons.54

Eligible Revenues

According to the Explanatory Statement accompanying the

Determination, “eligible revenue” determines how much each
participating person must contribute to the USO subsidy. The
Determination defines eligible revenue as gross telecommunications
sales revenue less certain deductions.55

A participating person’s sales revenue is taken to be telecommunications

revenue as described in its consolidated financial statements.
Revenue earned from an activity outside the telecommunications
industry may be deducted since the USO is a telecommunications
specific scheme.

Allowable deductions include “revenue earned for certain acts outside

Australia” and “inter-person input payments”.

A.2 Regulation of Telstra Prices

The second set of policies impacting upon decisions regarding access

pricing to the PSTN are the regulatory constraints on Telstra’s retail
prices. Here we describe those regulations in detail.

53 See further at http://www.aca.gov.au/consumer/ uso/funding/funding.htm

54 Department for Communications, Information Technology and the Arts, op. cit.
55 ibid.

31
Section 0 Appendix A

A.2.1 Current Regime

The majority of the regulation of Telstra's retail prices is found in the

Telstra Carrier Charges - Price Control Arrangements, Notification and
Disallowance Determination No. 1 of 2002 (“the Determination”) made
under the Telecommunications (Consumer Protection and Service Standards)
Act 1999 (“the Act”). Other relevant price controls fall under Part
XIC of the Trade Practices Act 1974 which regulates the prices Telstra
charge to other carriage service providers for access to its
telecommunications infrastructure. 56

The Determination specifies a series of price caps on selected baskets

of telecommunications services:
• CPI – 4.5% on a basket of three of Telstra’s services: local, trunk and
international calls (this means that the charge for these services as a
group must fall, in real terms, by at least 4.5 per cent each year).57
• CPI + 4.0% on a basket of line rentals (this means that the
charge for these services as a group may rise, in real terms, by
up to 4.0 per cent each year),58
• CPI - 0% on a basket of connection services.59

It also specifies further price caps of:

• 22 cents per call on the provision of untimed local calls,60
• 40 cent per call for local calls made from payphones.61

The revenue weighted average untimed local call price from

residential (as well as business) lines in non-metropolitan Australia in
2002/03 is not to exceed the revenue-weighted average local call price
in metropolitan Australia in 2001/2002 by more than 0.4 per cent.62
Also, where Telstra proposes to increase a line rental charge for
residential customers written consent from the ACCC is required.

56
Australian Competition and Consumer Commission, Review of Price Control
Arrangements, Final Report, February 2001., at p. 6. The history of regulation of
Telstra’s retail prices can be found at page 7.
57 Sub-clase 9(a)
58
Sub-clause 9(b)
59
Sub-clause 9(c)
60
Sub-clause 13(1)
61
Sub-clause 13(2)
62
Sub-clauses 13(4) to 13(7)

32
Section 0 Appendix A

Charges for directory assistance services are subject to notification to

and disallowance by the ACCC (section 158 of the Act).

Telstra is able to “carry forward” price changes where it does not

increase prices by the maximum allowable in any given year.63

A.2.2 ACCC Review of Price Control Arrangements

The ACCC undertook a review of the regulation of Telstra's retail

prices with its final report released in February 2001. In the report it
recommended that regulation of Telstra's retail prices continue, but
recommended a broad price cap applying to all price controlled
services and the removal of many of the existing sub-caps on prices.64
This was on the basis that a broad price-cap would more greatly drive
Telstra to adopt Ramsey pricing and increase efficiency in the
provision of telecommunications services.65

In particular, the ACCC recommended the removal of the sub-cap on

the price of line rental.66 There are significant efficiency losses
associated with the sub-cap on line rental charges. According to the
ACCC, the price cap prevents Telstra from covering the cost of
providing line rentals. The average cost of a line rental was estimated
to be $346 for 2000-01, but Telstra charged $166.20 per annum to
residential customers.67 The efficiency loss from the sub-cap on line
rentals alone was measured at between $25 and $45 million dollars
per annum.68

63
Clause 19
64
ACCC (February 2001), op. cit., at p. vii
65
ACCC (February 2001), op. cit., at p. 32
66
ACCC (February 2001), op. cit., at p. ix
67
ACCC (February 2001), op. cit., at p. 36
68
ACCC (February 2001), op. cit., at pp. 37-39

33