Chapter 3 Previous reports on international mobile roaming
3.1
In the previous chapter, the Committee described how the administrative
and technical framework of roaming contributes to the cost of this service. Even
with the technical nature of roaming, charges are considered high, with both
the Consumers’ Telecommunications Network (CTN) and the Australian Competition
and Consumer Commission (ACCC) making this point.[1]
3.2
Two reports on Australian international mobile roaming charges have been
published in recent years:
n the ACCC’s 2005 Mobile services review: International inter-carrier roaming (hereafter called
the ACCC report); and
n the Department of
Broadband, Communications and the Digital Economy’s (DBCDE’s) 2008 Report of
findings on: International mobile roaming charges, prepared for the
Department by KPMG (hereafter referred to as the KPMG report).
These reports provide an insight
into the underlying price components of international mobile roaming services.
3.3
Both reports investigated a range of issues including:
n the uptake and usage
of roaming services;
n the level of
competition in the market;
n the cost of
international roaming services compared to non-roaming services; and
n what proportion of
the end user charges is attributable to the wholesale price, and what
proportion is attributable to the retail mark-up applied by home network providers. [2]
Report findings
3.4
Both reports contain a number of similar findings about how the person
who subscribed to the roaming service (described as an ‘end user’) is charged.
Wholesale charge
3.5
The wholesale part of the end user price reflects the Inter Operator
Tariff (IOT) charge as negotiated between the home network operator and the
visited network operator.[3]
3.6
Optus provided the Committee with some insight as to how these wholesale
arrangements are agreed upon and passed on to consumers. Optus asserted that
carriers negotiate the best wholesale price they can with the visited network
operator, then reflect this wholesale price in their end prices to consumers.[4]
3.7
The Committee also heard evidence as to how these wholesale transactions
occur. When a call is made on a visited network, the billing personnel of the
visited network compile a billing file with the details of the call and the
wholesale charges associated with it. This billing file is then sent to a
clearing house where the file is distributed to the home network operator. The
home network operator then pays the wholesale charges as recorded in the
billing file.[5]
3.8
These wholesale costs are incurred by the home network operator whenever
one of their subscribers makes a call on a visited network. The home network
operator then attempts to recover these wholesale costs by including them in
the retail roaming bill to the subscriber who made the call.
3.9
The Australian Mobile Telecommunications Association’s (AMTA’s) submission to the inquiry highlighted that, due to this wholesale arrangement, home
network providers shoulder some amount of risk to facilitate the roaming
service. The submission stated:
… IOT charges are paid to the visited network by the home
network irrespective of whether the home network recovers any fees from its
customer. The home network operator therefore takes on all bad debt risk (i.e.
the risk of the non-recovery of charges from the end customer).[6]
3.10
The Committee heard evidence that this wholesale billing method can
cause delays to the billing of international mobile roaming charges to the end
user. The CTN stated that sometimes providers are unable to provide current
balances of international roaming charges to their customers because of delays
in receiving billing information from visited network providers.[7]
However, Vodafone Australia argued that these delays are an exception to the
rule.[8]
Home network operator’s mark-up
3.11
The ACCC determined that the mark-up component of roaming retail charges
is not governed by any common set of principles. Rather, each home network
operator is free to determine the size of the mark-up component of the retail
price. The ACCC suggested that usually this mark-up is determined by adding a
percentage of the IOT onto the wholesale charge as negotiated in the
party-to-party agreement.[9]
3.12
Telstra advised that the mark-up varies depending on both the carrier
and the destination country in which the subscriber uses international roaming.[10]
3.13
A range of evidence was provided to the Committee as to how the size of
these retail mark-ups are determined by carriers.
Operational costs
3.14
First, many home network providers asserted that there are significant
operational costs that must be recovered by providers through the retail
mark-up. For example, Telstra pointed out that retail mark-ups must cover the
operational costs of the home network operator including front-of-house
customer service, customer support and customer billing costs.[11]
3.15
AMTA argued that the mark-up component of the retail price covers the
costs of negotiating and administrating party-to-party roaming agreements,
marketing, customer support as well as cost associated with the maintenance and
construction of the network operator’s infrastructure.[12]
3.16
Vodafone Australia added the operational costs associated with
facilitating wholesale transactions is included in the retail mark-up to the
list. According to Vodafone Australia, the operation of the clearing houses,
where the wholesale billing information is transacted can be quite costly. The
cost of operating of these clearing houses must be recouped by the home network
provider by including the cost in the retail mark-up.[13]
Bundling
3.17
Another factor contributing to the home network providers’ mark-up was
the bundling of roaming with other mobile phone services.
3.18
Vodafone Australia stated that its international mobile roaming service
comes bundled with a range of other mobile services, such as domestic voice and
SMS. Thus, revenues to Vodafone from the retail mark-up on roaming may allow it
to reduce the mark-up applied to other high-demand mobile services within the
bundle. This makes the bundle more attractive to a wider range of consumers.[14]
Premium service
3.19
Finally, there was a consistent view amongst the providers that
international roaming was a premium service and that this may be a factor
considered by providers when determining the size of the mark-up on the
service.[15] In other words,
international mobile roaming is considered to be a privilege type of service,
attracting a commensurate cost.
ACCC report
3.20
The size of the IOT tariff is negotiated between providers and can be
expected to vary depending on the home network operator’s call volume, customer
expenditure, call volume growth and destination of calls as well as the number
of providers in the foreign country.[16]
3.21
In order to investigate what proportion of the final consumer price of
roaming is attributable to the wholesale price and what proportion is
attributable to the retail mark-up charged by home network providers, the ACCC
used publicly available information from Telstra stating that the retail
mark-up for their outbound international roaming services is 30 percent. Using
the Telstra figures, the ACCC extrapolated a general figure for the Australian
market of a 25 percent markup.[17]
3.22
Given this 25 percent retail mark-up, the report infers that wholesale
charges make up 75 percent of the final price charged to consumers.[18]
This conclusion is illustrated below.
Figure 3.1 – ACCC’s conclusion regarding wholesale and mark-up
components of final consumer price
Source: ACCC, Mobile services review: International inter-carrier roaming, 2005, p. 22.
3.23
According to the ACCC, the size of the wholesale charge is based on the profitability
of an Australian operator’s customer base and the nature of the visited network
providers.[19]
3.24
The BDCDE suggested that network providers who have a relatively small
share of the global international roaming market, such as many Australian providers,
are inevitably price takers when it comes to party-to-party agreements:[20]
…[Australian network operators] tend to be price takers
rather than price setters. They are often confronted with negotiating roaming
agreements with, in some cases, a limited number of alternatives and they are
often negotiating with existing alliances of international carriers, so they
are confronted with existing pricing arrangements. [21]
3.25
According to the ACCC, another factor affecting IOT pricing is the
number of providers in a country. Where there are the least number of mobile providers
in a country, IOTs are likely to be highest.[22]
3.26
Industry representatives generally agreed with this assessment. Vodafone
Australia supported the arguments put forward by the DBCDE and the ACCC, advising
the Committee that the scope, scale and volume of the Australian international
roaming market puts most Australian providers at a disadvantage when
negotiating IOTs with foreign providers.[23]
3.27
However the Committee also heard evidence that some network providers
are in a position to limit the impact Australia’s small volume has on their
negotiating power. Vodafone Australia, a subsidiary of the Vodafone Group which
has a presence in twenty six countries,[24] is sometimes able to
take advantage of this global presence when negotiating prices.[25]
3.28
Another possibility for limiting the impact of Australia’s small volume is for Australian providers to become members of inter-carrier
alliances. The ACCC’s submission to the inquiry notes the emergence of
inter-carrier alliances that allow network providers from different countries
to form a coalition to negotiate IOTs with other larger network providers,
increasing the negotiating power of the providers.[26]
The DBCDE was also of the view that some Australian providers are no longer
price takers due to participation in such alliances.[27]
3.29
There are two weaknesses in the ACCC’s analysis. The first is its
extrapolation from publicly available Telstra figures on the retail mark-up to
all Australian providers. This weakness is quite difficult to overcome because
of the commercial in confidence nature of this information. In other words, the
ACCC could not obtain the same information from other providers.
3.30
The second weakness is the age of the report. The report was published
in 2005 and was based on information obtained some years earlier. In the mobile
phone market, that is enough time for substantial changes to have occurred.
KPMG report
3.31
KPMG’s investigation employed a different method to the ACCC’s analyses.
KPMG used publicly available international benchmark data, published by the
Technical University of Denmark, to estimate the actual per-minute costs of
providing an international mobile roaming service and the average retail cost
per-minute to consumers of this service. These costs were then converted to
Australian dollars. These estimates include the actual costs associated with
termination rates, international call transit rates and roaming specific costs.[28]
3.32
To determine the approximate mark up applied to a roamed call by the
overseas and Australian providers, the report deducted the total estimated actual
cost from the average end user cost.[29] The figures for this
analysis are illustrated in the table below.
Figure 3.2 – KPMG’s analysis of wholesale and mark-up price
components
Charge element
|
Charge per minute (AUD$)
|
Estimated total actual cost
|
0.46
|
Average retail cost to consumers
|
2.75
|
Approximate retail margin on the actual cost
|
2.29
|
Source: Department
of Broadband, Communications and the Digital Economy, Report of findings on international
mobile roaming charges, 2008, p. 23.
3.33
Figure two shows that KPMG determined that where a consumer pays $2.75
per minute for an international mobile roaming call, $0.46 of this per minute
charge is accounted for by the actual cost and $2.29 by the mark up applied by
the overseas and home network providers.
3.34
Figure three demonstrates KPMG’s conclusion as a percentage.
Figure 3.3 – KPMG’s conclusion regarding actual cost and
mark-up components of final consumer price
Source: Department of
Broadband, Communications and the Digital Economy, Report of findings on international
mobile roaming charges, 2008, p. 23.
3.35
The disadvantage of this approach is that the actual cost of the roamed
call is not what Australian providers are charged. As has already been
discussed, the Australian market has some specific peculiarities, such as being
a small market with little bargaining power in international negotiations over IOT tariffs. This could mean that the Australian situation is very different.
Different approaches by the ACCC and KPMG
3.36
Whilst both the ACCC and KPMG analyses are validly based, the Committee
notes that the ACCC and KPMG reports each adopt a different approach to
investigating roaming, leading to findings that emphasise different aspects of
the roaming market. The ACCC’s approach focuses attention on the role played by
the party-to-party agreements in determining the end user cost. KPMG’s approach
directs attention to the discrepancy between the actual cost of roamed calls
and the end user cost.
3.37
In doing this, the KPMG report relies on international benchmark cost information
and then assumes that Australian providers are charged this actual cost for
roamed calls by overseas providers. The ACCC, on the other hand, has relied on
publicly available information directly from a single Australian carrier and
extrapolated this to all Australian service providers.
3.38
It should also be noted that the ACCC has investigative powers which
provide it with information resources, and an understanding of Australian
markets, largely unavailable to private-sector consultancy firms.
3.39
The Committee asked the ACCC to comment on KPMG’s findings. The ACCC
suggested that:
...the KPMG 2008 Report appears to correctly identify the
actual component costs of providing a roamed call … as being quite small
compared to the charges faced by the end-user. However, no account appears to
have been made of the wholesale charges levied by [visited network operators].[30]
3.40
The ACCC also suggested that the benchmark cost information provided by
the Technical University of Denmark, and used in the KPMG report, may
underestimate the component costs of transmitting a mobile call to and from Australia. The transmission of mobile calls between northern hemisphere countries, which
make up the bulk of international call traffic, cover much smaller distances,
and may use significantly less resources, than transmitting a call between
Australia and Europe.[31]
3.41
When carriers were asked to comment on the reports, they suggested that
the ACCC’s analysis was a better reflection of the reality of roaming
arrangements. Telstra, for example, suggested that the conclusion of the ACCC was
accurate when compared to their cost data.[32]
3.42
Vodafone Australia concurred, stating that the wholesale price paid by
home network providers to visited network providers constitutes the biggest
component of the end user price to consumers.[33]
3.43
The Committee is of the view that, while both the KPMG and ACCC reports
are based on valid sources, neither entirely reflects the pricing situation as
both rely on extrapolating conclusions rather than direct data, and that as a
consequence, both are flawed.
Market distortions
3.44
In the case of roaming services offered by Australian providers, the
Committee considers that the market does not operate effectively because the
size of the Australian population reduces competition. Australian providers
cannot offer enough customers to providers in other countries to make
negotiations over price competitive.
3.45
In addition, Australian customers do not generally choose their provider
on the cost of international mobile roaming, but on the domestic charges.
3.46
The result of this distortion is that the price of roaming in Australia is high, and the product is considered by Australian providers to be a premium service.
3.47
In the next chapter, the Committee considers what can be done to ameliorate
this distortion.