Tax Review 06 07

Global Mobile Tax Review
2006–2007
Tax regimes that recognise mobile phones
as a need not a luxury benefit all stakeholders
Full Report
Contents
1 Introduction ................................................................................................ 4
1.1 Tackling the digital divide........................................................................... 4
1.2 This report.................................................................................................. 5
2 Analysing taxation rates across the sample ............................................... 6
2.1 Tax as a percentage of the total cost of mobile ownership ........................ 6
2.3 Tax and the cost of mobile services......................................................... 12
2.4 Taxes on Mobile Handsets....................................................................... 14
2.5 Telecom specific taxes............................................................................. 16
2.6 Other tax related issues ........................................................................... 17
3 Modelling the impact of alternative tax policies........................................ 19
3.1 Overall modelling approach ..................................................................... 19
3.2 Usage and penetration elasticities ........................................................... 21
3.3 Mobile communications and GDP............................................................ 24
3.4 Jurisdictions included in the study............................................................ 26
4 Impact of tax changes on the mobile sector and Government revenues . 28
4.1 Overall approach...................................................................................... 28
4.2 Base forecasts ......................................................................................... 28
4.3 Removing all telecommunications specific taxes ..................................... 28
4.4 Reducing sales taxes............................................................................... 30
4.5 Reducing general customs taxes ............................................................. 34
5 Africa........................................................................................................ 37
5.1 Summarised market overview.................................................................. 37
5.2 Taxation regime ....................................................................................... 38
5.3 Reducing mobile specific taxes................................................................ 39
6 Magreb and the Middle East .................................................................... 42
6.1 Summarised market overview.................................................................. 42
6.2 Taxation Regime...................................................................................... 43
7 Latin America ........................................................................................... 46
7.1 Summary market overview....................................................................... 46
7.2 Taxation regime ....................................................................................... 47
8 Asia Pacific .............................................................................................. 51
Global Mobile Tax Review 2006-2007

2
8.2 Taxation regime ....................................................................................... 52
9 Central and Eastern Europe..................................................................... 56
9.2 Taxation regime ....................................................................................... 57
10 Conclusions and policy implications......................................................... 61

3
1 Introduction
In 2005, the GSM Association (GSMA) developed its first study on tax and the digital
divide, seeking to understand more fully the tax rates affecting telecommunications in
developing countries and the impact that cutting taxes may have on mobile handsets
and new services.
The study’s key findings showed that telecommunication taxes were
disproportionately high in many developing countries and that even small cuts in
taxes many attract significantly more mobile users.
In this second report, the analysis is extended to include a larger set of countries – in
particular adding the transitional Eastern European countries. The report also
investigates more fully the link between lower taxes and revenue opportunities for
governments in the long term, showing that cutting taxes may lead to increased
economic growth in the least developed countries.
1.1 Tackling the digital divide

How can governments tackle the digital divide and bring universal communications
access to businesses and communities? The continued technical advances in
developed countries including 3G, on-the-go digital devices and, in the near future,
mobile broadband, promise to ease still further access to information through the
internet, e-mail and telephony in the developed world. However, whilst benefiting a
privileged minority, these advances threaten to widen even further the economic gap
between rich and poor countries. Information access rates in developing countries
are often more than 20 years behind those in the developed world.
There is little doubt amongst academics, business leaders, governments and citizens
that access to reliable communication services aids in the economic development
and social welfare of countries. Increasing access to communications is an aim of
virtually all governments and is central to the philosophy of leading global
development organisations such as the World Bank.
Academic studies have traditionally sought to quantify the impact of fixed line
telephony investment on growth. However, as mobile technology has evolved and
shown itself to often be a more cost effective solution to universal access, particularly
in rural and isolated areas, attention has turned to the role of mobile communications
in promoting growth. Investment in mobile communications and increasing
penetration rates can promote economic growth, with the greatest impacts being in
those countries with the lowest levels of GDP.
Government has a key role in supporting communication advancements, as mobile
penetration varies dramatically across countries. Consumers, particularly in
developing countries, are price sensitive in relation to their uptake and usage.
Government actions, such as taxation levels and regulatory fees, are a key variable
cost impacting upon the total cost of mobile ownership and use. Indeed, this report
shows that an average of 17.4% of the total cost of mobile ownership is formed of
consumer and import taxes.
It is widely forecast that by 2010 the worldwide penetration of mobile telephony will
have doubled. This provides the potential for the mobile industry to reduce and
eventually eliminate the digital divide. In order to capitalise upon, or increase, this
growth the challenge to Governments and the mobile industry is to work together
more effectively to make mobile phones and usage affordable.
The mobile industry is already taking steps in this direction and the impact of such
actions is being felt. Mobile operators in emerging markets are extending their

4
coverage rapidly and over 80% of the world’s population now live in areas covered by
mobile technology. However, at present over 60% of the world’s population do not
have a mobile phone. The implication of this is that Governments, particularly in
developing countries, may choose to consider providing a greater focus to their
taxation strategies to increase economic development.
1.2 This report
The aim of this study is to investigate the impact of mobile taxation on consumer
behaviour and government revenues. This report:
• Provides an initial benchmarking and analysis of taxation rates across the 101
countries in our sample. The report highlights those countries whose taxation
policies may provide a positive example to others and also those where
taxation may be considered a hindrance to mobile penetration and, ultimately,
economic growth;
• Details the methodology adopted to calculate the impact upon consumer
behaviour and government revenues. This report develops last years
analyses:
o Taking into account the corporate tax impact of all changes to the
taxation of mobile operators; and
o Providing an initial quantitative analysis of knock on effects across the
economy.
• Provides the results of the study. The results of our quantitative investigation
of 101 countries is initially presented at continental level but then
disaggregated further in subsequent regional sections. A series of interviews
and case studies are provided focusing on countries of particular interest to
support the quantitative analysis.

5
2 Analysing taxation rates across the sample
This section of the report provides a consideration of the taxation structures that
apply to the cost of ownership and usage of mobile services. Our analysis considers
the position in 86 developing countries and 15 developed (EU) markets. This report
provides an explicit analysis of the:
• The role of tax relative to the total cost of mobile ownership (TCMO), the total
cost of mobile services (TCMS), and the cost of handsets;
• Telecom specific taxes, i.e. taxes where the telecoms sector is treated in an
inconsistent manner relative to the remainder of the economy; and
• Other tax related issues, including the treatment of mobile vis-à-vis fixed
networks and the role of corporation tax (excluded from above) and license
fees.
This analysis has been conducted by Deloitte based on taxation data collected by
Deloitte offices internationally and third party market data.
2.1 Tax as a percentage of the total cost of mobile ownership
The initial comparison is the percentage of the TCMO that is made up of taxes on
handsets, subscription and airtime1. In order to calculate the TCMO with taxes, VAT
and custom duties have been added to the handset cost, fixed subscription taxes
have been added to connection charges, VAT and other telecom specific taxes have
been added to rental and usage and fixed taxes have been added to the total2.
There appears to be no direct relationship between the percentage of tax in the
TCMO and penetration.
1
The proportion of tax in the TCMO was calculated by comparing the TCMO inclusive of tax with the TCMO excluding taxes,
i.e. (TCMO with tax/TCMO with no tax)-1.
2
Calculation used average handset prices and blended ARPU data collected in the 101 markets, and assume the life cycle of each
handset and subscription is three years. The ARPU was calculated from tariffs data provided by Tarifica for over 120 operators
and minutes of use data collected from Wireless Intelligence and Pyramid Research. Where data was not available for a particular
country then assumptions were made (see Section 3 for details). It is important to note that this analysis did not include those
taxes which are payable by the company rather than directly by the consumer.

6
Figure 1: Tax as a percentage of TCMO and penetration
0.5
0.4
Taxs as % of TCMO
0.3
0.2
0.1
0
0 100% 200%
Penetration
Maghreb and ME CEE
EU15 Latin America
Subsaharan Africa Asia Pacific
Source: Deloitte analysis based on Wireless Intelligence, Tarifica and Deloitte Tax data.
The following figures provide a further disaggregation of the tax burden, splitting
taxes between handsets, connection, rental and usage. These results suggest that
there is broad overall similarity in the role of taxation (vis-à-vis the TCMO) across:
• Pre-pay and post-pay platforms, and
• The different regions of the world.
Figure 2: Tax as proportion of the TCMO for post-pay services
Type of tax Africa Middle Asia Latin CEE EU15

East Pacific America
Handset 0.8% 0.68% 0.73% 0.4% 0.7% 0.5%
Connection 0.01% 0.12% 0.53% 0.01% 0.2% 0.01%
Rental 4.31% 2.6% 2.8% 4.85% 3% 4.7%
Usage 10.8% 10.2% 7.7% 13.4% 16.5% 14.4%
Total 16% 13.6% 12% 18.6% 20.4% 19.6%

Source: Deloitte

7
Figure 3: Tax as proportion of the TCMO for pre-pay services
Type of tax Africa Middle Asia Latin CEE EU15

East Pacific America
Handset 3.15% 1.6% 2.25% 1.4% 1.6% 1.6%
Connection 0.05% 0.3% 0.7% 0.04% 0.4% 0.02%
Usage 14.11% 12% 9.4% 17.3% 18.4% 18%
Total 17.3% 14.1% 12.4% 18.7% 20.6% 19.6%

Source: Deloitte
This point is further illustrated by the cumulative distribution of tax as a proportion of

TCMO for pre and post paid platforms. The lack of difference between taxation policy
adopted across the two platforms is somewhat surprising given the increasingly
differing roles undertaken by the two platforms across most developing countries:
• Pre-pay is now regarded as the provider of universal service across
developing countries 3, i.e. communications as a social good; and
• Post-pay provides a greater business focus and possibly more focus on
communications as a driver of economic activity.
Figure 4: Tax as proportion of TCMO for Pre-paid and post paid, cumulative
distribution
100.0%
90.0%
80.0%
70.0%
60.0%
% of countries
Pre Paid
50.0%
Post Paid
40.0%
30.0%
20.0%
10.0%
0.0%
0
4
02
04
06
08
12
14
16
18
22
24
26
28
32
34
36
38
42
44
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
0.
tax as % of TCMO
Source: Deloitte
Further disaggregation of tax charges for pre-paid platforms suggests that usage and
connection taxes are proportionately higher in those countries where tax is a higher
percentage of the TCMO. However, no similar pattern exists for handset taxes.
3
GSMA, Universal Access, 2006.

8
Figure 5: Tax charges: Pre paid
100%
90%
80%
70%
60%
usage
rental
50%
connection
handset
40%
30%
20%
10%
0%
tax<=10% of TCMO 10%<tax<15% of TCMO 15%<tax<20% of TCMO tax>20% of TCMO
Tax as proportion of TCMO
Source: Deloitte
For post-paid platforms:

• Usage taxes are proportionately higher in those countries where tax forms a
higher percentage of the TCMO; and
• Handsets taxes are proportionally higher in countries where tax forms a lower
proportion of TCMO.
Figure 6: Tax charges: Post-paid
100%
90%
80%
70%
60%
usage
50% rental
connection
handset
40%
30%
20%
10%
0%
tax<=10% of TCMO 10%<tax<15% of TCMO 15%<tax<20% of TCMO tax>20% of TCMO
Source: Deloitte

9
2.2 Country specific results
The figure below ranks the 101 jurisdictions in our sample by the percentage of
TCMO that is accounted for by tax and demonstrates differences across countries:
• The jurisdictions where tax forms the highest proportion of TCMO are Turkey,
Uganda, Tanzania, Ukraine, Dominican Republic, Zambia, Brazil, Ecuador,
Sweden, and Denmark;
• Three of the five jurisdictions (China, Malaysia and Bhutan) where tax forms
the lowest proportion of TCMO are in the Asia Pacific region; and
• On average, 17.4% of end user spend is kept by the Government in tax
revenues. This result is consistent with that found in the previous Tax and the
Digital Divide survey.

10
Figure 7: Tax as a share of total cost of mobile ownership
Turkey
Tanzania
Uganda
Brazil
Ukraine
Zambia
Dominican Republic
Ecuador
Greece
Argentina
Kenya
Denmark
Sweden
Nepal
Mozambique
Zimbabwe
Finland
Italy
Poland
Cameroon
Chad
Senegal
Rwanda
Portugal
Ireland
Dem Rep. Congo
Morocco
Colombia
Slovenia
Uzbekistan
Hungary
Austria
Bulgaria
Gambia The
France
Azerbaijan
Georgia
Chile
Peru
Burkina Faso
Jordan
Guinea
Romania
Albania
Czech Republic
Slovakia
Bangladesh
Gabon
The Netherlands
Cote d'Ivoire
Russian Federation
Lithuania
Latvia
Malta
Estonia
Madagascar
Tunisia
United Kingdom
Ghana
AVERAGE
Thailand
Trinidad and Tobago
Mexico
Spain
Ethiopia
Germany
Pakistan
Venezuela RB
Mauritius
Nicaragua
Luxembourg
Kazakhstan
Guinea-Bissau
Sri Lanka
Cyprus
Egypt Arab Rep.
South Africa
Samoa
Mauritania
Bolivia
Cambodi
Guatemal
a
a India
Philippines
Indonesia
Iran Islamic Rep.
Lao PDR
Vietnam
Botswana
Paraguay
Yemen
Sierra Leone
Papua New Guinea
Lesotho
Angola
Nigeria
Syrian
ArabMalaysia
China
Bhutan
Myanmar
Swaziland
0.00% 5.00% 10.00% 15.00% 20.00% 25.00% 30.00% 35.00% 40.00% 45.00% 50.00%
Source: Deloitte
Further analysis demonstrates that, when applied, telecom specific and special taxes
are the main contributors to the heavy burden of taxes on the TCMO.

11
Figure 8: Tax as proportion of TCMO, top 4 countries in the ranking
50%
45%
Fixed
40%
Other
35%
Custom
VAT
30%
25%
20%
15%
10%
5%
0%
Turkey Tanzania Uganda Brazil
Source: Deloitte analysis based on Tarifica, Wireless Intelligence and Deloitte tax data
Turkey stands out as the market with the highest cost of tax as a percentage of
TCMO; the tax burden of 44.6% is one and a half times as much as that of the
nearest country. Whilst Turkey imposes a VAT rate on telecommunications services
that is relatively similar to VAT applied in other countries in the study, it charges the
most extreme telecom specific taxes:
• A special communication tax of US$16.87;
• A wireless license fee of US $6.99 at subscription; and
• A wireless usage fee of US$6.99 annually on top of 25% special
communication tax (apart from 18% VAT).
2.3 Tax and the cost of mobile services
It may be argued that consumers are more sensitive to the cost of mobile services as
the cost of handsets and connection are often discounted when they receive their
service4. As such, this section provides a focus on the role of tax in relation to the
total cost of mobile services (TCMS).
Taxes on mobile services account for a substantial portion of the overall tax burden
on mobile phone ownership. These taxes include consumption taxes and any tax
charges on mobile rental charges and usage:
• All the countries bar two (Bhutan and Swaziland) impose some form of VAT,
general sales tax (GST) or similar taxes on mobile phone usage; and
• 74 countries of the countries considered impose rates between 10% and
20%, while 5 charge a VAT of 20% or above.
4
Handsets are not, typically, subsidised in the Middle East and Africa.

12
Figure 9: Tax as percentage of Total Mobile Service Cost
Tax as proportion of TMSC
Turkey
Uganda
Brazil
Dominican Republic
Zambia
Ukraine
Ecuador
Tanzania
Greece
Kenya
Sweden
Denmark
Argentina
Zimbabwe
Finland
Poland
Italy
Nepal
Portugal
Ireland
Morocco
Jordan
Colombia
Austria
Uzbekistan
Slovenia
Hungary
Bulgaria
Albania
France
Cameroon
Peru
Chile
The Netherlands
Slovakia
Romania
Czech Republic
Senegal
Rwanda
Madagascar
Guinea
Gambia The
Gabon
Dem Rep. Congo
Cote d'Ivoire
Chad
Burkina Faso
Tunisia
Russian Federation
Malta
Lithuania
Latvia
Georgia
Estonia
Azerbaijan
United Kingdom
Mozambique
Thailand
AVERAGE
Spain
Germany
Mauritius
Guinea-Bissau
Ethiopia
Egypt Arab Rep.
Trinidad and Tobago
Nicaragua
Mexico
Luxembourg
Kazakhstan
Cyprus
Sri Lanka
Pakistan
Bangladesh
Ghana
South Africa
Mauritania
Venezuela RB
Bolivia
Cambodia
Samoa
India
Guatemala
Philippines
Yemen
Botswana
Papua New Guinea
Lao PDR
Indonesia
Sierra Leone
Paraguay
Vietnam
Iran Islamic Rep.
Nigeria
Lesotho
Angola
Malaysia
Syrian Arab Republic
China
Swaziland
Myanmar
Bhutan
0.00% 5.00% 10.00% 15.00% 20.00% 25.00% 30.00% 35.00% 40.00% 45.00% 50.00%
As with TCMO, Turkey has the highest taxes as a proportion of TCMS, a position
unchanged from the 2005 report. Uganda, Brazil, Dominican Republic, Zambia and
Ukraine are additional markets which have overall service taxes above 25%.

13
At the low end of the range are Angola, Malaysia, Syria and China, which all have
VAT rates of 5% or less. The majority of Asia Pacific countries have VAT (or
equivalent) tax rates that are lower than 10% and the least taxed jurisdictions are
Myanmar and Bhutan.
The figure below shows the split of the burden of taxation as a percentage of the
TCMS between VAT and other tax charges on mobile rental charges and usage.
Figure 10: Tax as proportion of TMSC, top 4 countries in the ranking
50%
45%
40%
Other
35% VAT
30%
25%
20%
15%
10%
5%
0%
Turkey Uganda Brazil Dominican Republic
2.4 Taxes on Mobile Handsets

Taxes on mobile handsets are another significant component of the TCMO. Taxes on
handsets typically consist of:
• Import duties, included already into the retail price of a handset; plus
• The sum of sales and VAT taxes and handling taxes which are paid directly
by the consumer.
Investigation demonstrates that, for our sample:
• On average, tax accounts for 24.8% of total handset costs; and
• 45% of countries imposed import duties on handsets.

14
Figure 11: Tax as a percentage of handset cost
Tax as proportion of Handset cost
Syrian Arab
Iran Islamic Rep.
Cameroon
Chad
Rwanda
Brazil
Tanzania
Mexico
Mozambique
Trinidad and Tobago
Bhutan
Argentina
Ghana
Gambia The
Turkey
Senegal
Azerbaijan
Dem Rep. Congo
Samoa
Guinea
Burkina Faso
Georgia
China
Venezuela RB
Uganda
Gabon
Ethiopia
Cambodia
Cote d'Ivoire
Chile
Sweden
Denmark
Bolivia
AVERAGE
Russian Federation
Peru
Zambia
South Africa
Morocco
Zimbabwe
Finland
Poland
Colombia
Lesotho
Portugal
Ireland
Myanmar
Lao PDR
Bangladesh
Nepal
Italy
Austria
Uzbekistan
Ukraine
Slovenia
Hungary
Bulgaria
Albania
France
The Netherlands
Greece
Slovakia
Romania
Czech Republic
Vietnam
Indonesia
Madagascar
Tunisia
Malta
Lithuania
Latvia
Estonia
United Kingdom
Botswana
Kenya
Spain
Germany
Nigeria
Angola
Mauritius
Guinea-Bissau
Nicaragua
Luxembourg
Kazakhstan
Cyprus
Sri Lanka
Swaziland
Mauritania
Paraguay
Yemen
Guatemala
Ecuador
Philippines
Sierra Leone
Malaysia
Egypt Arab Rep.
Papua New Guinea
Thailand
India
Jordan
Dominican Republic
Pakistan
0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% 100.00%
The jurisdictions with the highest tax as a proportion of total handset cost are Syria,
Iran, Cameroon, Chad and Rwanda. Syria charges on average $24 in fixed taxes on
top of 20% VAT and 10% custom duty (reduced from the previous rate of 45.6%).
Iran’s custom duties are 60%, by far the highest custom duties on mobile handset of

15
all the countries in our sample, whilst Cameroon has also some of the highest
custom duty charges (32%), which are additional to a VAT rate of 19%.
Figure 12: Tax as a percentage of handset cost, selected countries
100%
90%
80%
Fixed
Custom
70% VAT
60%
50%
40%
30%
20%
10%
0%
Syrian Arab Republic Iran Islamic Rep. Cameroon Chad
At the low end of the taxation spectrum are 17 jurisdictions, whose governments do
not impose any non-recoverable taxes (such as customs duties) on the importation of
mobile handsets. The majority of European jurisdictions have a zero-rate, as do
countries such as Egypt, Madagascar, Pakistan, Thailand, Guatemala and Ukraine.
The jurisdictions at the very bottom of the ranking are Philippines, Sierra Leone,
Egypt and Papua New Guinea. None of these jurisdictions charge custom duties or
fixed taxes on handsets and charge VAT between 10% and 12%. It should be noted
that Kenya recently removed import taxes on handsets in June 2006.
2.5 Telecom specific taxes
Most countries do not impose specific taxes on mobile telephony, recognising that it
is important not to discriminate against a sector of the economy that is recognised as
a driver of economic growth. However, a minority of jurisdictions in our sample did
levy such taxes which place a greater taxation burden on the telecommunications
sector than the rest of the economy. Whilst mobile specific taxes are generally
established for social purposes, they are generally regressive in nature in developing
countries (proportionally more costly to poorer consumers) and hinder mobile take-up
amongst the poorer sections of society and so often harm exactly those people that
they are intended to aid. Examples of key mobile specific taxes are detailed below.

16
Figure 13: Telecoms specific taxes
Country Telecommunications specific taxes
Tunisia • 5% Telecom Royalties: mobile and fixed
• 5% Telecoms Royalties: calculated on cost of handset plus VAT
Iran • Fixed charges at subscription of: $4.44 for pre-pay / $135 for post-
paid / $0.33 on fixed line. This comprises of subscription tax for pre-
pay/post-paid and fixed line. In addition there is Rural Areas
Development tax for post-paid
Jordan • 4% special VAT due on telecoms services: charged on mobile and
fixed
Turkey • 25% Special communication tax - on overall usage and fees for
mobile phones. Fixed telephony has a 15% rate of Special
communication tax.
• Fixed Taxes at subscription: $23.86 Wireless usage and new
subscription special communication tax (mobile telephony only).
Bangladesh • $11.76 Paid at subscription: mobile and fixed
• $3 Fixed charge on import of mobile and fixed handsets
Nepal • 10% Telephone services charges - paid on usage only. Not
applicable to subscription and applies to both mobile and fixed
• $20.15 Telephone ownership fee: fixed and mobile
Pakistan • $8.30 Fixed activation charges: mobile only
Sri Lanka • 2.5% Mobile Subscription Levy paid on total value of monthly
charges: mobile and fixed
Dominican Republic • 2% Communications development tax - paid on usage and
subscription on mobile only
Venezuela • $1.56-6.25 Subscription Tax on mobiles and fixed
Albania • Post pay $59 / Prepay $7 / Fixed $146 : registration and subscription
charges
Greece • $1.92 - $5.75 Subscriber's special duty. Varies based on the monthly
invoice total
Italy • $6.82 Mobile specific stamp duty on bills issued. Applies only on
post-paid services
Kenya • 10% Telecommunications Tax on usage: fixed and mobile
Tanzania • 7% Telecommunications Tax on usage: fixed and mobile
Uganda • 12% Telecommunications Tax on usage: fixed and mobile
Source. Deloitte
2.6 Other tax related issues

2.6.1 Taxes on Mobile vs. Taxes on Fixed
While mobile is widely recognised as equivalent to fixed communications in terms of
providing connectivity, there appear to still be significant difference in some countries
in the tax treatment of these two communication platforms. 20 markets
demonstrated lower taxes on fixed telephony, the key differences being as follows:
• A lower VAT rate or no VAT rate on fixed telephony services e.g. Senegal,
Mauritania and Pakistan;

17
• Lower ‘other taxes’ due on fixed telephony, such as lower or no excise tax
due on services e.g. Thailand, Kenya, Tanzania, Uganda, Zambia and
Argentina; and
• Different levels of subscription / connection charges e.g. Albania.
The indirect tax burden with respect to mobile telephony appears to be less than that
in the fixed line business only in India (higher rate of VAT on fixed telephony handset
purchases) and Columbia (subscription tax for the installation of fixed telephony).
2.6.2 Corporate taxation
The focus of this study is on the impact of consumer taxes. However, governments
clearly receive telecoms related revenues from corporation tax. Whilst this project
has not sought to explicitly consider the impact of corporation tax, any reduction in
consumer taxes on mobile handsets / services can be expected to lead to higher
penetration / usage.
All countries in our study levy a corporate tax charge on telecoms profits. These
rates were collected by Deloitte offices worldwide and are used to model the
offsetting impact of corporation tax receipts. The following table provides a summary
of the rates of corporation tax in our sample by region of the world.
Figure 14: Corporation tax in our sample, average rates by region
Region Average corporate tax rate
Asia Pacific 31%
CEE 20%
EU15 28%
Latin America 24%
Maghreb and ME 29%
SubSaharan Africa 32%
Source. Deloitte
The lowest corporate tax rates are found, on average, in East European countries,
with the highest rates in Sub Saharan Africa and Asia Pacific. This has an important
and significant impact on the relative counterbalancing of the impacts of reductions in
consumer taxes.
2.6.3 Licence fees
License fees are charged in many countries as a proportion of the revenue created
under that licence. As such, it can have a similar counter balancing impact on total
Government revenues (albeit often paid to the regulatory authority rather than directly
to the Government) following a reduction in consumer taxation, i.e. increased
penetration / usage leading to increased revenue from operation of the license.
Information on the percentage of revenue that is paid as a licence fee is often
confidential and was not available for each country for this project. However, rates
were provided by a leading global mobile operator and based upon this we have
adopted an informed estimate for regulatory fees of:
• 2% in licence fees in EU15 countries; and
• 7.5% in developing countries.

18
3 Modelling the impact of alternative tax policies
3.1 Overall modelling approach

The previous section of this report demonstrated that both taxation rates and
structures vary considerably across countries, but that they form a substantial portion
of the cost of mobile ownership and usage.
The overall aim of this report is to quantitatively consider the impact of reducing /
removing consumer and import taxes on mobile penetration and mobile usage. To
achieve this we model:
• A reduction in the price charged to the end customer 5;
• The impact this change will have on mobile penetration and usage; and
• The subsequent impact on tax revenues and GDP.
The overall modelling approach adopted for this report is set out diagrammatically
below.
Figure 15: Approach to modelling the impact of taxation changes
Impactoftax
changeon
Country company
characteristics revenues
Impactoftax
changeonmobile
Regression penetration Corporationtax
Taxrates Licencefees
General
Telecom
GST
Demand Impactoftax Impactupon
Handset
equations changeonmobile total
Mobileprices (Penetrationand governmenttax
usage(minutes Taxon
Usage) andtexts) mobile revenues
services
Currentprices
Additional tax
Impactoftax revenues
changeon
demandfornew
handsets
ImpactonGDP
This model was constructed by Deloitte using taxation information collected by

Deloitte practices internationally alongside market and forecast information from third
parties. The economic modelling conducted as part of this project relies on a number
of assumptions as to the development of markets, population levels and the
continuation of existing economic relationships 6. As with all such models the results
5
Specifically, the following consumer and import taxes were included in our analysis: fixed taxes paid at the time
of subscription for mobile users; fixed taxes paid after subscription for mobile users; traditional sales and variable
taxes levied on mobile telephone use (e.g. VAT); telecom specific variable taxes levied on mobile usage; and fixed
and variable taxes due on the importation and sale of mobile handsets e.g. VAT or customs duties.
6
Our analysis depends significantly on data provided by third parties and that we have no responsibility for this,
we have not reviewed or audited it in any way and that errors in this data would affect the outcomes/conclusions.

19
should therefore be regarded as indicative rather than definitive. All assumptions
have been discussed with representatives of major mobile operators.
The model was estimated at the jurisdiction level for 101 jurisdictions across the
world and then aggregated for presentational purposes. In order to compare the
impact of various taxes:
• A base case scenario was created for all 101 jurisdictions, which projects for
5 years market development and tax revenue collection, assuming application
of the current taxation structure. The jurisdictions included in the study are
detailed below. 7. External market projections and an econometric model are
used to forecast penetration rates, population, GDP growth, mobile usage
and handset sales; and
• All changes in penetration, handset sales, mobile usage and Government tax
revenue collection are compared to a base case scenario and calculated for
each period.
The initial basis of our model is that the decision of an individual to purchase a
mobile handset and to demand mobile services is based upon the total cost of mobile
ownership (TCMO). As such, the annual cost of mobile ownership can be expressed
as:
TCMO = ( H / 3) + (C / 3) + TCMU (1)
Where:
TCMO Total Cost of Mobile Ownership
Cost of handset. It is assumed that handsets have a three year lifetime and are replaced at the end of
that lifetime. The handset cost is therefore spread across the three years equally. This may be a
H
conservative assumption, with the lifetime of a handset in many developed countries now believed to be
8
below 2 years
Connection fee. It is assumed that the connection fee mimics the lifetime of the handset and that the
connection fee is spread across the three years equally. This assumption may be conservative. For post
C pay customers in developed markets this appears appropriate. However, the assumption may under
reflect the churn in developing countries with a high pre pay market share although we have no basis for
an alternative assumption
Total cost of mobile usage. The TCMU is defined as twelve times the monthly rental, plus twelve times
TCMU the monthly minutes times the rate, plus twelve times the monthly texts times the rate per text. The issue
of free texts has been addressed as part of this construction
This framework is then used to consider the impact of changes to consumer taxes. A
number of key assumptions were required to model the “direct” impact upon
penetration, usage and government revenues from these consumer taxes, notably:
• Any reduction in the tax rate is fully reflected in prices faced by consumers 9;
• The reaction of the penetration and usage levels to a change in price is
determined by the value of the elasticity of demands for usage and
penetration (see below for details);
7
The tax rates were collected by Deloitte member firms from across the Deloitte network.
8
The analysis does not consider the impact on the black market due to a lack of available data. Discussions with
handset vendors suggest that a significant proportion of new handsets come from black market sources, suggesting
that this estimate would provide a potential overstatement of the total impact on handset sales. For example, in
many African countries with import taxes black market handsets are regarded as the norm by operators. However,
no reliable data was available on this issue.
9
This assumes note the competitive nature of international mobile markets and simplifies to the point where no
additional tax revenue is kept as producer surplus. Relaxation of this assumption would reduce the positive impact
of any taxation changes.

20
• Total penetration is capped at 130% and can never increase beyond this
point as this is taken as representing market saturation; and
• Total mobile usage is assumed to increase by the number of new subscribers
multiplied by the average number of minutes per subscriber. It is assumed
that average minutes of use will remain constant over the next 5 years in the
base case assumption10.
The change in penetration and usage levels drive the direct change in government
revenue that results from a change in consumer taxes. However, this reduction in
consumer taxes is, at least, partly offset by an increase in corporation tax and licence
fee revenues that result from the high penetration and usage. Modelling this impact
of this “indirect” impact required the following additional key assumptions:
• An assumed profit margin on new revenue was multiplied by the country
specific corporate tax rate to provide the increase in corporate tax 11; and
• Multiplying the additional revenue by the assumed license fee percentage
provided the increase in the licence fee rate12.
Finally, the relationship between mobile penetration, economic activity and hence
GDP is considered via an econometric calculation of the “growth” impact before
inputting it into the model. This is discussed below.
One final simplifying assumption was required to allow the comparison of different
scenarios: all tax change measures are introduced in early 200613.
3.2 Usage and penetration elasticities

Core to the approach outlined above is the need to estimate the impact of changes in
process caused by a reduction in taxation on the demand for mobile services. There
are two key relationships that, jointly, need consideration:
• Demand for new connections, i.e. how penetration is affected by price of
access both in terms of handsets and usage; and
• Demand for usage, i.e. how minutes of use is linked to the price charged for
the call.
The following key sources of data were used to model these relationships:
• Mobile usage prices. Tarifica provided call and text rates by destination,
connection and rental prices for nearly all operators in each jurisdiction. An
average cost per minute was calculated using average tariffs across call
plans for all operators in each jurisdiction. The calculated average cost per
minute required several further assumptions including the proportion of calls
by destination14. For countries for which data was not available a regional
average of each of a selection of tariffs (on-net, off-net, fixed and
10
This reflects two contrasting trends. Firstly, new subscribers are more likely to be prepaid and will therefore
drive usage down. At the same time, in a number of countries with higher penetration levels the minutes of use
have started to grow over the last two years due to steep price declines and fixed mobile substitution. Average
usage in these markets is expected to continue to grow during the forecast period.
11
The assumed profit before tax (PBT) was based on actual information from a leading worldwide mobile
operator. PBT is set at the following levels in each region of the world: Asia Pacific (26%); EU 15 (28%); CEE
(41%); Sub Saharan Africa (46%); Latin America (46%) and Margreb and ME (67%)
12
A simplifying assumption was required to model license and related fees as this information is not publicly
available for many markets. We assume that these fees are 2.5% of revenue in the EU 15, and 7.5% elsewhere.
These assumptions are based on the views of a leading worldwide mobile operator.
13
In reality different budgetary processes would clearly apply internationally.
14
Assumptions were required on the proportion of calls going to on-net, off-net, fixed and international numbers.
These assumptions have been made on a regional basis, on the basis of data provided by a leading worldwide
mobile operator

21
international, for both the pre and post pay platforms) was adopted. Finally,
an allowance was made to allow for the provision of free texts as part of a
bundle offered to consumers;
• Penetration. Wireless Intelligence provided data for the number of pre and
post pay connections by jurisdiction. The number of connections was divided
by population, taken from the World Bank database, to obtain penetration.
Forecasts were then developed for penetration until 2010 based on market
intelligence and a statistical forecasting model;
• Usage. Wireless Intelligence provided data for usage for pre and post paid,
both separately and as an average. Pyramid Research provided similar
average usage data for a number of additional countries. Average usage was
split between pre and post pay platforms and where no data was available a
regional average was used; and
• Handset prices. An international survey was conducted in 2004 and 2005.
Initially, a regional average change in price was calculated, based on those
countries for which data from both years was available. This change was then
applied to countries for which 2004 data was available. For countries where
no data was available, a regional average from 2005 data was used.
We sought to estimate subscription and usage elasticities at the market level, i.e.
through combining data for individual operators in each jurisdiction15.
The analysis indicated that there is a linear relationship between the price of handset
and the level of penetration in the different countries16.
Figure 16: Regression results for the level of penetration
Dependent variable: Number of connections divided by
population
Explanatory variables Coefficient t-statistic
Handsets adjusted for
-0.0006 3.22
purchasing power parity
Percentage of population that 0.0044 3.68

is urban
GDP per capita 00.0093 2.16
Source. Deloitte
In particular, the statistically significant linear relationship between handset prices,

adjusted for purchasing power parity, and mobile penetration. The estimated linear
relationship suggests that the elasticity will:
• Differ between jurisdictions;
• Increase with the increase in price of the handset; and
• Decrease with the growth in penetration.
The following figure uses the results for Ghana to demonstrate the procedure through
which elasticity estimates on the relationship between the price of handsets and
penetration are calculated.
Figure 16: Calculation of the relationship between the price of handsets and
penetration, example for Ghana
15
This approach appears consistent with that adopted in the previous GSMA “Tax and the Digital Divide” report
16
The impact was statistically significant at the 5% level and passed the standard econometric diagnostic tests. In
the 2006 report no significant elasticity was identified. For ease of presentation a statistically significant constant
and regional dummy variables are not reported

22
Factor Result
A Regression co-efficient for Ghana -0.00064
B Handset price for Ghana adjusted for PPP ratio 302
C Penetration rate for Ghana 0.126
D Eventual elasticity impact, calculated as (A*B) / C -1.53
Source. Deloitte
However, it was not possible to collect data for handset prices for every country and
as such regional averages were assumed for the countries for which data was not
available. Given this simplification, we believe that the data has greater integrity at
the regional level and therefore we have chosen to adopt regional averages of the
elasticities calculated. This provided the following elasticities, which whilst more
consistent still show some considerable variation across the regions of the world.
Figure 17: Penetration elasticity by region
Region Elasticity
Asia Pacific -1.42
CEE -0.46
EU15 -0.04
Latin America -0.16
Maghreb and ME -0.83
Sub Saharan Africa -2.44
Source. Deloitte
In most cases the results reflect a-priori expectations as to the relative elasticity,
except Latin America where the results are lower than may have been expected17.
For usage no statistically significant relationship with penetration was apparent within
our data. However, the data available was only available as a one year cross
section. It was concluded that the data was insufficient to allow a full investigation of
this impact which is better estimated on a time series. As such, the elasticity
estimated in the previous GSMA Digital Divide (2005) report of -0.5 was adopted.
This elasticity is held consistent for all countries.
Finally, the two elasticities are then combined based on a weighting calculated based
on the regional relative share of the TCMO.
For the model of minutes of use we identified an isoelastic relationship, i.e. a
relationship characterised by constant elasticity 18, between usage and the cost per
minute.
Figure 18: Regression results explaining minutes of use
Dependent variable: log (average minutes of use per year)
Explanatory variables Coefficient t-statistic
(logged)
GDP per capita 0.17 2.13
Total penetration 0.12 0.88
Average cost of usage per -0.72 6.14
minute
Source. Deloitte
17
Further investigation of the results for this region suggested the relatively low elasticity was a factor of low
handset costs reported and PPP values.
18
The impact was statistically significant at the 5% level and passed the standard econometric diagnostic tests.
For ease of presentation a significant positive constant is not reported

23
Finally, it is noted that in none of the above regressions was it possible to identify a
statistically significant difference in the coefficients for prepaid and post-paid. As
such, total penetration and average minutes of use across pre and post pay
platforms have been applied.
3.3 Mobile communications and GDP

The penetration of telecommunications has been found to have a significant positive
impact on growth19. Mobile penetration in particular, being easier and cheaper to
supply than fixed telephony, can be expected to play a crucial role in the economic
growth of African and other developing countries.
In order to consider this issue for our sample, the specification of Waverman et al
(2005) 20 was adopted to estimate a model in averages over 24 years 21 for two
groups:
• Developing countries (57 developing countries from our data set from
SubSahran Africa, Latin America, Maghreb and ME and Asia Pacific); and
• A fuller data set including higher GDP capita countries from the EU 15 and
CEE.
A statistically positive relation between average GDP per capita growth and mobile
penetration was found in both cases 22.
Figure 19: Regression results explaining average GDP growth
Dependent variable: average GDP growth
Developing country set Full data set
Explanatory variables Coefficient t-statistic Coefficient t-statistic
Average mobile penetration

0.0012 2.42 0.00069 2.62
per 1,000 people
Average investment as a
0.00208 5.78 0.00204 5.77
percentage of GDP
Literacy rate at the beginning
-0.00011 -0.96 -0.00160 1.79
of the period
GDP per capita at the
-0.0036 -2.15 -0.00215 1.52
beginning of the period
Source. Deloitte
The result of the developing country set was applied to developing countries
(SubSahran Africa, Latin America, Maghreb and ME and Asia Pacific). For this
19
Waverman L., Meschi M., Fuss M., (2005), “The Impact of Telecoms on Economic Growth in Developing
Countries”, Africa: The Impact of Mobile Phones, The Vodafone Policy Paper Series, Number 2, find a 10%
increase in penetration leads to a 0.6% increase in GDP. Further evidence is available from Sridhar K.S. and
Sridhar V., (2004), “Telecommunications Infrastructure and Economic Growth: Evidence from Developing
Countries”, available at http://ideas.repec.org
20
Following Waverman et al. (2005) we have estimated a growth model following the Endogenous Growth
approach. This is a cross-section estimation of the relation between average GDP per capita growth over a period
of time (1980 to 2003 in our case, as in Waverman at al. (2005)) and the initial level of GDP per capita, literacy
rate at the beginning of the period as proxy for initial human capital, average investment as a proportion of GDP
and average mobile phone penetration.
21
The choice of a model in averages is driven by the fact that mobile penetration is expected to have an impact on
long term economic growth, rather than year-to-year growth. However, similar results are found when the
regression is estimated using a panel data approach
22
The results reported pass standard econometric diagnostic tests. For ease of presentation, a significant constant
term is omitted.

24
group, a 10% increase in mobile penetration leads to a 1.2% increase in GDP. As
there was insufficient data to allow a separate developed country regression, the full
data set results were applied to the EU 15 and the CEE, i.e. a 10% increase in
mobile penetration leads to a 0.6% increase in GDP 23.
Whilst not considered in this report, a further positive impact of the mobile industry is
the additional value add created by the mobile communications industry across the
economy. This is, where measured, captured by multiplier that captures expenditure
in subsequent rounds. The following figure shows the values of multipliers that have
been calculated in other studies.
23
This may potentially overstate the results for developed countries and results should be interpreted considering
this point.

25
Figure 20: Various multiplier benchmark studies
Title of study Multiplier
The contribution of mobile phones to the UK economy, 02 for ONS 1.13
Ovum studies on economic impact of mobile telephony in Bangladesh and
USA based on review of various other studies* 1.6
Association Française des Opérateurs Mobiles * 1.7
Economic impact of spectrum use in the UK, Europe economics, based on
ONS 1.1
Sicrana, R., and de Bonis, R.: "The Multiplier Effects of Telecommunications
Investments on Economic Growth and Restructuring** 1.5
Radio authority, UK 1995, Economic impact of radio 1.4
Notes. * On employment, ** On GDP
These results suggest a range of 1.1 to 1.7 may exist, i.e. the value added created by
network operators and other parts of the value chain is multiplied by this number to
capture the total impact of this expenditure. It is not possible to include this factor for
this study as:
• No information has been collected on value added created by mobile
operators or their value chain in each jurisdiction; and
• Leakages from the economy in each jurisdiction will vary, i.e. differing levels
of imports.
As such, it is important to note that the estimates included in this study are likely to
be an underestimate of the total impact of the mobile sector.
3.4 Jurisdictions included in the study
The following table identifies all jurisdictions included within this study. Our country
list makes up about 83% of the global population. The final selection of countries was
determined by data availability.

26
Figure 21: Jurisdictions within the study
Sub Saharan Maghreb and
Africa ME Latin America Asia Pacific CEE EU15
Angola Egypt Arab Rep. Argentina Bangladesh Albania Austria
Botswana Iran Islamic Rep. Bolivia Bhutan Azerbaijan Denmark

Burkina Faso Jordan Brazil Cambodia Bulgaria Finland
Cameroon Mauritania Chile China Cyprus France
Chad Morocco Colombia India Czech Republic Germany
Syrian Arab
Cote d'Ivoire Republic Dominica Indonesia Estonia Greece
Dem Rep. Congo Tunisia Ecuador Lao PDR Georgia Ireland
Ethiopia Yemen Guatemala Malaysia Hungary Italy

Gabon Mexico Myanmar Kazakhstan Luxembourg
Gambia The Nicaragua Nepal Latvia Portugal

Ghana Paraguay Pakistan Lithuania Spain
Papua New
Guinea Peru Guinea Malta Sweden
Trinidad and
Guinea-Bissau Tobago Philippines Poland Netherlands
Kenya Venezuela RB Samoa Romania United Kingdom
Russian
Lesotho Sri Lanka Federation
Madagascar Thailand Slovakia
Mauritius Vietnam Slovenia

Mozambique Turkey
Nigeria Ukraine
Rwanda Uzbekistan
Senegal
Seychelles
Sierra Leone
South Africa
Sudan
Swaziland
Tanzania
Uganda
Zambia
Zimbabwe

27
4 Impact of tax changes on the mobile sector and
Government revenues
4.1 Overall approach

This section reports results from using the methodological framework discussed
above to investigate the impact of changing various taxes on the take up / usage of
mobile technology and thereafter Government revenues. In particular, the report
considers the impact of:
• Reducing telecommunications specific taxes, i.e. those which place a burden
on the telecommunications sector over and above the remainder of the
economy; and
• Reducing general VAT / GST (hereafter VAT) and customs / related taxes
which, typically, apply to telecommunications services in the same manner as
other goods across the economy. Whilst these taxes also apply to much of
the remainder of each jurisdiction’s economies, these scenarios only reduce
taxation on the mobile sector as it is a key driver of economic growth.
All tax reductions are based on a proportional reduction of 10% of the prior year tax
rate or level in order to allow effective comparison of impacts across jurisdictions.
The alternative would have been to reduce the total VAT rate by an arbitrary number,
say 1%24, which would have lead to dramatically different impacts countries / regions
with the differing rates. For each scenario the report:
• Estimates the impacts upon penetration, usage, handset sales and
Government tax receipts using the approach outlined above; and
• Compares the results to the base forecast for 2006-10.
4.2 Base forecasts
The sample of 101 jurisdictions has an aggregate population of 5.6 billion, which
represents 84% of the worldwide population. In order to compare the impact of
various taxes, a base case scenario was created for all 101 markets, which projects
market development and tax revenue collection for the next five years, assuming a
constant tax structure.
The base case scenario projects real market developments in all jurisdictions,
alongside a number of assumptions used to forecast mobile penetration rates, usage
and handset sales as documented above.
As noted above, these forecasts are indicative and based on Deloitte analysis using
the above methodology and data collected by Deloitte international practices, third
party data and assumptions. All assumptions have been discussed with
representatives of major mobile operators.
4.3 Removing all telecommunications specific taxes
This section considers a scenario where mobile services and equipment are treated,
from a taxation perspective, comparably to other goods and services across the
economies of our sample. In order to represent this, the report estimates the impacts
of removing all telecom specific taxes whilst leaving VAT and custom / related duties
24
As adopted in the GSMA (2005) Tax and the Digital Divide. For example, VAT is 3% in China and 17.5% in
the UK. A 10% reduction in the rate charged would lead to a 0.3% reduction in China and 1.75% in the UK. The
alternative approach of a 1% overall cut would lead to 33.3% reduction in VAT in China and a 5.7% cut in VAT in
the UK

28
unchanged. Of the sample, 16 jurisdictions had at least some telecoms specific
taxes, which are discussed above.
4.3.1 Impact of tax changes on penetration, handset sales and usage
Analysis suggests that changes to mobile specific taxation can have a significant
impact on penetration. The following figure shows the percentage increase in
subscribers in 2010 from the base forecast by reducing such mobile specific taxes to
90% of their prior year value.
Figure 21: Percentage change in subscribers from base case in 2010 following
the removal of telecoms specific taxes
Kenya
Uganda
Tanzania
Tunisia
Jordan
Iran Islamic Rep.

Venezuela RB
Dominican Republic
Italy
Greece
Turkey
Albania
Sri Lanka
Pakistan
Nepal
Bangladesh
0% 0.1% 0.2% 0.3% 0.4% 0.5% 0.6% 0.7% 0.8% 0.9% 1.0%
Change in subscribers relative to base forecasts
Source. Deloitte
The level of impact is driven by the size of the mobile specific taxation burden and
the impact upon the TCMO through the various elasticities discussed above. Turkey
has the highest impact, reflecting its reliance on mobile specific taxation. Also of
note are the high impacts on penetration in East Africa, where the countries impose
excise duties on mobile usage.
It is assumed that new subscribers will purchase a new handset in both the year of
purchase and every three years thereafter. As such, the new cohort in 2006 will also
purchase a new handset in 2009, and the new cohort in 2007 will purchase a new
handset in 2010. The model therefore suggests that the removal of all mobile specific
taxes will also have a significant effect on handset sales over the five year period,
and that this will be broadly double the percentage impact on penetration.
4.3.2 Impact of tax changes on Government revenues
The approach adopted to model the impact of tax changes on Government revenues
has been described in some detail above. The key conclusion from the investigation
is that in 14 of the 16 countries, the impact of reducing mobile specific taxes to 90%
of their prior level leads to a loss of taxation revenue relative to the base case
forecasts of less than 0.6% and in many cases considerably less. This result is
driven by the positive impacts of indirect impacts of higher penetration / usage

29
(notably additional corporation tax) and the growth caused to the remainder of the
economy as a result of higher penetration.
In only three countries would the model suggest a higher adverse impact of reducing
mobile specific taxation to be felt:
• Turkey, where the relative importance of mobile specific taxation is large
relative to other economy wide taxes (VAT and corporate tax). The net
impact of this is that economy wide taxes will never be able to counter
balance the lost direct taxation revenue25. It may be argued that taxation
models that are inherently biased towards particular sources of revenue may
have inefficient allocation properties and as such run the risk of misleading
investment decisions across the economy; and
• Greece, where there is limited increase in penetration to counter balance the
loss of direct taxation revenue. This is due to the model capping penetration
growth in response to reductions in taxation, as these markets are assumed
to have reached a saturation level before 2010.
Figure 22: Percentage change in government revenues from base case in 2010
following the removal of telecoms specific taxes
Kenya
Uganda
Tanzania
Tunisia
Jordan
Iran Islamic Rep.
Venezuela RB
Dominican Republic
Italy
Greece
Turkey
Albania
Sri Lanka
Pakistan
Nepal
Bangladesh
-2.50% -2.00% -1.50% -1.00% -0.50% 0.00% 0.50%
Change relative to base forecasts
Source. Deloitte
4.4 Reducing sales taxes

This section of the study considers a scenario where indirect taxation in the form of
VAT on mobile services and handsets are reduced. All jurisdictions within the
sample had imposed some form of VAT on mobile operators. There are nineteen
jurisdictions for which VAT on handset and services differ, a similar pattern exists for
VAT on handsets 26. The majority of jurisdictions within our sample have a VAT
between 15-20%, although some considerable variation is apparent.
25 It may be that additional latent demand and a higher economic impact exist for Turkey than can allowed for in a
multi national model
26 Six jurisdictions have a higher VAT rate on handsets than services and thirteen have a higher rate of services
than handsets.

30
Figure 23: Levels of VAT on services across sample, frequency by group
60
50
40
Number of countries
30
20
10
<5% 5%-10% 10% -15% 15% -20% 20% -25% 30%+
Source. Deloitte
Given this variation in rates investigation of the impact of reducing VAT is conducted
by reducing rates to 90% of their prior year value, i.e. a 10% reduction of the current
rate in each jurisdiction. This scenario also assumes that this reduction in VAT rates
is only made on mobile usage and handsets and that VAT rates for the remainder of
the economy remain constant.
The results suggest that reducing VAT to 90% of its prior year value in each
jurisdiction would have a significant impact on penetration relative to our base case,
but that this impact varies significantly across regions of the world. These results
primarily reflect the elasticity impacts discussed above and hence the fundamental
nature of the markets themselves 27.
Dennis Knowles - Partner and Indirect Tax TMT Leader, Deloitte & Touche LLP
In an ideal global market, VAT (which also encompasses sales tax and GST) should
not be a material cost component of roaming charges. Transactions between Telco’s
internationally should not attract VAT and should only be taxable where the
customer, private or business, belongs. This survey demonstrates that many
emerging economies charge VAT to Telco’s outside their own country which is often
not recoverable. This, therefore, increases the cost of roaming and arguably does not
mean that VAT accrues to the country of consumption.
The OECD has a technical working group looking at the future of international
services generally. It is likely to recommend that no VAT should be charged
internationally between businesses, unless the services fall with a limited number of
exceptions. It can only be hoped that once the OECD has finalised its deliberations
that countries follow the international trading model which should reduce the overall
27
Many of the EU 15 and several of the CEE jurisdictions reach our assumed level of market saturation and, as
such, no further penetration growth is allowed

31
cost of roaming and reducing much of the administration caused by charging VAT
internationally.
Figure 24: Percentage change in subscribers from base case in 2010 following
a 10% reduction in the value of VAT on usage and handset sales
1.5%
Change compared to base case forecast
1.0%
0.5%
0.0%
Maghreb Subsaharan Asia Pacific CEE Latin America EU15
and ME Africa
Source. Deloitte
In terms of impact on handset sales over the five year period, the same relationship
holds as discussed above. Average minutes of use vary, with the scale of increases
reflecting differences in VAT rates internationally, where Asia Pacific has the lowest
average rates.
Figure 25: Percentage change in minutes of use in 2010 following a 10%
reduction in the value of VAT on usage

32
1.60%
1.40%
Change compared to base case forecast
1.20%
1.00%
0.80%
0.60%
0.40%
0.20%
0.00%
Maghreb and Subsaharan Asia Pacific CEE, Latin America EU15
ME Africa
Source. Deloitte

The approach follows the same approach for direct, indirect and growth impacts as
noted above:
• The direct impact of reducing mobile specific taxes upon Government
revenues, i.e. the foregone revenue;
• Adding the indirect impact of greater usage and penetration on corporate
revenue / VAT and hence corporation tax and regulatory fees; and
• Adding the growth impact of greater economic activity on corporation tax
revenues.
The results suggest that reducing VAT on mobile usage and handset sales only
leads to limited reduction in Government revenue vis-à-vis the base case forecast.
The model suggests a total impact of between 1-3%, depending upon region of the
world, after allowing for the indirect and growth impacts. The indirect and growth
impacts have a substantial impact upon this result, counterbalancing a direct impact
of c.7% vis-a-vis Government revenues in the base case forecast.
There is some variance in the results obtained between regions of the world, mainly
reflecting the different levels of additional penetration created by reducing VAT.

33
following a 10% reduction in the value of VAT on usage and handset sales
Direct effect Uplift from indirect effect Uplift including growth effect
Maghreb SubSaharan Asia Pacific CEE Latin America EU15

Impact on tax revenues relative to base case
8% and ME Africa
6%
4%
2%
0%
-2%
-4%
-6%
-8%
-10%
Source. Deloitte
4.5 Reducing general customs taxes

This section of the study considers a scenario where general customs and related
taxes are reduced to 90% of their prior year value. Of the sample of 202
jurisdictions, 46 had a customs tax on handsets.
Figure 27: Levels of customs tax across sample, frequency by group
14
12
10
Number of countries
0
<5% <10% <15% <20% <25% <30% <35% >35%
Source. Deloitte

34
The customs tax rates levied varied considerably across jurisdictions and are
demonstrated in the above figure, which demonstrates a long tail in the taxation rate,
i.e. Iran has a customs rate of 60%.
As with VAT, it has been assumed that customs taxes are only charged on the
importation of mobile handsets and they remain unchanged in other sectors of the
economy.
The following figure considers the impact of reducing customs charges on handset
sales, which is proportionate to the change in penetration.
Figure 28: Percentage change in handset sales 2006-2010 following a 10%
reduction in the value of customs tax
0.20%
Change relative to base case forecasts
0.15%
0.10%
0.05%
0.00%
Maghreb and SubSaharan Asia Pacific CEE Latin America EU15
ME Africa
Source. Deloitte
It is noted that:
• The suggested impact of a reduction of 10% of the rate of customs taxes is
significantly lower than that recorded for VAT, as these taxes only apply to
handset costs and not services. Handset costs only form 8.6% of the TCMO
on average; and
• Regional differences are quite clear, reflecting the high levels of customs
taxes across Sub Sahran Africa and Latin America, and the lack of such taxes
in other regions such as the EU.
The impact on Government taxation revenues follows the same approach for direct,
indirect and growth impacts as noted above. The key conclusion suggested by the
model is that reducing customs taxes by 1% of their current value has a slightly
negative impact upon Government revenues. However, within this impact, there are
larger negative direct impacts counterbalanced in the main by the indirect and growth
impacts.

35
following a 10% reduction in the value of customs tax
Direct effect Uplift after indirect effect Uplift including growth effect
Maghreb and SubSaharan

Global ME Africa Asia Pacific CEE Latin America
2.5%
2.0%
1.5%
1.0%
0.5%
0.0%
-0.5%
-1.0%
-1.5%
-2.0%
-2.5%
-3.0%
Source. Deloitte

36
5 Africa
5.1 Summarised market overview

This study included 35 jurisdictions across Africa, who had a total population of about
600m accounting for more than 80% of the region’s total population.
The penetration rate in these jurisdictions is the lowest of all the regions analysed in
this study. However, penetration has increased dramatically in the past year and at
the end of 2005 the mobile subscriber base in these jurisdictions was c.84m. The
following figure shows the rapid increases that have been recorded and are forecast
for the near future.
Figure 30: Average penetration rates in African sample markets
0.3
0.25
0.2
Mobile penetration
0.15
0.1
0.05
0
03
04
05
01
01
01
01
02
02
02
02
Q 3
03
03
04
Q 4
04
05
05
Q 5
06
06
06
06
07
07
07
07
0
0
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
2
4
1
4
Q
Source: Wireless Intelligence
There are large variations in penetration rates across the region, with a high
correlation between GDP per capita and penetration rates. For example, penetration
in South Africa now stands at almost 60% whilst in Zambia penetration is less than
10%.

37
Figure 31: Penetration and GDP per capita for selected African markets, 200528
Source: Deloitte analysis based on Wireless Intelligence and the World Bank database
The average monthly minutes of use per person is only just over 60 minutes across
the region. However, this result varies considerably across the region with Kenya
having the highest usage and the Democratic Republic of Congo the lowest at c. 30
minutes of usage per user per month.
Mobile tariffs in the region are notable for two conclusions:
• Whilst tariffs are the second lowest in the sample, tariffs are lower in Asia
Pacific, despite the lower ability to pay in Africa; and
• The difference between pre-pay and post-pay call charges is far less
pronounced in Africa than in the other areas. This reflects the strong
proportion of prepaid customers in the sample, more than 90%, and the role
of prepaid mobile as the source of universal service in Africa.
5.2 Taxation regime

The 35 African jurisdictions in this sample have an average GDP of around $19b and
26 markets have GDP below the sample average and are reflected in the mobile
penetration across the region. However, these conditions provide the greatest
potential for a reduction in taxation. The elasticity of demand is estimated to be
higher in Africa than elsewhere, reflecting the potential for further marginal
consumers, and hence to increase penetration greatly by a reduction in the TCMO.
Any such increase in penetration is also magnified through the growth impact across
the economy.
Although the average tax burden on mobile ownership in the region is 17.2%, slightly
lower than the global average, the tax regime in Africa varies considerably from
country to country. Of the 16 jurisdictions in out sample that charge some sort of
28
The size of circles in this (and subsequent similar) figures represent the number of connections for a jurisdiction
divided by the average number of connections for the region

38
telecom specific taxes, 3 are in SubSaharan Africa: the East African jurisdictions
(Kenya, Tanzania and Uganda all impose an excise tax on usage at varying rates.
This is also reflected in the proportion of tax in the TCMO, where the East African
jurisdictions are at the top of this comparison.
Figure 32: Tax share in TCMO and penetration
0.35
Tanzania
0.3 Uganda
Kenya
0.25
tax as % of TCMO
Senegal
0.2
Congo
0.15
0.1
South Africa
0.05
Angola
Nigeria
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Penetration
Source: Deloitte analysis based on Wireless Intelligence and Deloitte Tax
There have been some changes in taxation levels that impact mobile operators since
the GSMA 2005 study on this issue, with taxation increases in:
• Tanzania: Excise duty on services / use has increased from 5% to 7%; and
• Ghana: An additional national insurance levy due on the import of handsets of
2.5% has been introduced.
There have, however, been some reductions in mobile specific taxes in the region
within the last year:
• Uganda: Customs duties due on handsets have fallen from 25% to 10%; and
• Kenya: After 3 years of lobbying by operators, the government gradually
reduced the 10% custom tax on handset and since June 2006 handsets have
been made exempt to import taxes. This decisions was made was in
recognition of the importance of the price of a handset in the mobile take-up
decision
5.3 Reducing mobile specific taxes
The following figure summarises the overall effects of the removal of all mobile
specific taxes on penetration and tax revenues in the 3countries that levy such taxes.
The results suggest that reducing such taxes would have a positive impact upon
subscribers and only a marginally negative impact upon overall Government taxation
revenues.

39
Figure 33: Overall impact by 2010 of reducing mobile specific taxes for
SubSahran African countries with those taxes to 90% of current level
0.80%
0.60%
0.40%
0.20%
Subscribers
0.00% Taxation
revenues
Tanzania Uganda Kenya
-0.20%
-0.40%
-0.60%
Source: Deloitte
Mr Phuthuma Nhleko, Group President and Chief Executive Officer - MTN

Group Limited
Are there particular taxes your company faces that differ from those paid by fixed line
operators, the mobile industry in other countries, or companies in other parts of the
economy?
For corporate taxes and transactional taxes like customs duties and VAT, generally
no, although the manner in which the laws governing these taxes is applied is not
always in line with internationally accepted best practice. The license fees in some
territories are imposed in the form of a revenue share, which is similar to a tax, as
these funds are paid to governments or their agencies. Some countries also impose
an excise tax on the sale of airtime.
To what extent do you believe that taxes lead to lower mobile ownership or reduce
mobile usage?
I do not believe that taxes on handsets generally lead to lower mobile ownership as
handsets tend to be subsidized in one form or another. What is clear, however, is
that the absence of a specific tax on the industry or its products will lead to higher
levels of usage by consumers, as consumer’s disposable income levels tend to be
limited. The more expensive calls are, the less calls will be made, the less revenue is
earned and the more future growth becomes limited.
In countries where taxation levels are above average, would setting them at
‘standard’ levels increase penetration and usage to such an extent that it would have
the potential to make the impact revenue neutral or positive for governments over
time?

40
Possibly, but this has to be balanced against other relief that the governments
sometimes give, such as tax holidays, customs duty relief and accelerated or
enhanced depreciation allowances.
Do the national governments of your operating jurisdictions understand the total
economic impact of the mobile industry on the national economy and factor this into
the design of any prohibitive taxation structures for the sector?
Other than the revenue share and excise duties imposed in some jurisdictions, we
are not aware of any prohibitive structure that is imposed on the industry in any of our
operating jurisdictions. Where the revenue share due to the local government is high,
it may be accompanied by a compensating measure like a lower corporate tax rate or
a tax holiday, or perhaps some customs duty relief.
Are there any particular issues in relation to taxation on mobile operators in general,
or your operations in particular, that you would like to highlight?
With respect to transactional taxes like VAT, governments should be encouraged to
follow internationally accepted best practice in their VAT legislation and enforcement.
It would also be encouraging if governments followed the OECD guidelines on
withholding tax on payments to non-residents – only withholding tax if the amount in
question is sourced in the country from which the payment is made, and not on a
blanket basis where all payments to non-residents are subject to a withholding tax.
It would also benefit the industry tremendously if foreign governments were rewarded
by external funding bodies (i.e. the IMF) to negotiate double taxation agreements.
These agreements clarify each government’s taxing right with respect to a cross
border transaction that may take place between two taxpayers situated in each
country.

41
6 Magreb and the Middle East
6.1 Summarised market overview

This study included 5 Middle Eastern jurisdictions and a further 3 from Magreb.
Together, these countries have a population of 238m. The average penetration rate
in the region is 38%, although this does vary within the region. The following figure
shows the increases in penetration in the region over recent years and a forecast or
continued steady growth.
Figure 34: Average penetration rates in Middle Eastern and Magreb sample
markets
0.6
0.5
0.4
Penetration rate
0.3
0.2
0.1
0
Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4 Q1 Q2 Q3 Q4
2001 2001 2001 2001 2002 2002 2002 2002 2003 2003 2003 2003 2004 2004 2004 2004 2005 2005 2005 2005 2006 2006 2006 2006 2007 2007 2007 2007
Penetration in the region, however, varies considerably. For example, in 2005

Jordan and Tunisia had a penetration rate of 53% and 67% respectively, while other
countries such as Yemen and Iran still display very low penetration rates of around
10%29. As is internationally the case, the relationship between penetration and GDP
per capita is very strong.
29
Forecasts for these markets, such as those from Wireless Intellegence, suggest that penetration is expected to
growth much faster in Jordan and Tunisia than in the other countries in the sample. For example, Jordan is
estimated to have reached a penetration rate of 85% by the end of 2006.

42
Figure 35: Penetration and GDP per capita for selected Maghreb and Middle
East markets, 2005
0.7
Tunisia
0.6
Jordan
0.5
Morocco
Penetration rates
0.4
0.3
Mauritania
0.2
Syria
Egypt
0.1
Yemen Iran
0
0 500 1000 1500 2000 2500 3000 3500
GDP per capita ($)
6.2 Taxation Regime

The structure of the taxation regime imposed on mobile services and handsets in the
countries in this region differs from the remainder of the sample.
Whilst none of the countries in the Maghreb and Middle East region have a high tax
as a proportion of TCMO, the two countries with the highest proportion of tax in the
cost of handset are:
• Syria: total tax burden of c.90% due to the fixed tax which is imposed on top
of VAT and custom duties; and
• Iran: total tax burden of c.60% due to the highest level of custom duties (60%)
within our total sample.
The following figure shows the relation between mobile penetration and tax burden
on the total cost of handset in this region, showing a strong negative relationship.

43
1.2
Syria
1
0.8
Iran
tax as % of handset cost
0.6
0.4
Tunisia
0.2
Egypt
0 Morocco
Jordan
-0.2
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7
Penetration rate
Syria and Iran are also the only two countries in this region which have changed their
mobile related taxation regime since the GSMA 2005 study on this issue:
• Iran: Customs duties on the import of handsets have increased from 4% to
60%; and
• Syria: Customs duties due on handsets have fallen from 45.59% to 10%.
The following figure summarises the overall effects of reducing mobile specific taxes
to 90% of their current level in the 3 jurisdictions that levy such taxes. The results
suggest that a reduction of such taxes would have a positive impact upon penetration
relative to the base case forecasts and that the loss in tax revenue is limited.

44
Figure 37: Overall impact by 2010 of reducing mobile specific taxes for Magreb
and Middle Eastern jurisdictions with those taxes to 90% of current level
0.3%
0.2%
0.1%
0.0% Change in
subscribers
Iran Islamic Rep. Jordan Tunisia
-0.1%
Change in
taxation
-0.2% revenues
-0.3%
-0.4%
-0.5%
Source: Deloitte

45
7 Latin America
7.1 Summary market overview

In this study we have analysed 14 countries from the Latin America region, with a
total population of 508m, representing over 90% of the total population of the region.
The subscriber base in this area is around 230m.
The average penetration for our sample is 30%, one of the lower regions in the study.
Forecasts from Wireless Intelligence suggest that penetration in the region may slow
down on average in the next year.
Figure 38: Average penetration rates in Latin American sample markets
0.7
0.6
0.5
penetration rates
0.4
0.3
0.2
0.1
0
01
01
01
01
02
02
02
02
03
03
03
03
04
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
1
4
Q
There is considerable variation in penetration across the sample, with Chile having a
penetration rate of over 60% and Nicaragua and Peru being nearer 15%.
Interestingly, there is little evidence of a relationship between GDP and penetration,
unlike with the most regions within the sample.

46
Figure 39: Penetration and GDP for selected Latin American markets, 2005
0.9
0.8
Chile
0.7
0.6
Argentina Brazil
penetration rates
0.5
Colombia Mexico
0.4
Venezuela
0.3
Guatemala
Peru
0.2
0.1
0
0 200 400 600 800
GDP ($b)
Average pre tax tariffs in this region are relatively high compared with other countries
with similar penetration and macroeconomic characteristics. Prepay dominates in all
Latin American markets although post pay is stronger in a majority of countries,
including Argentina and Chile.
7.2 Taxation regime
The average burden of taxation in the total cost of mobile ownership in the region is
19%, just above the global average. However, there is considerable variation in rates
across the region. For example, Brazil and Argentina are among the ten countries
with the highest tax as proportion of TCMO, while other jurisdictions, e.g. Guatemala,
rank well below the global average.
The results for handsets are particularly of note:
• The tax burden on handsets in Brazil and Mexico are among the ten countries
with the highest tax costs imposed on handsets; and
• Handset prices in Latin America are among the lowest in our sample of
countries.
These two findings combine to suggest that even though the proportion of tax in the
total cost of handset is considerable, the handset itself forms a smaller proportion of
the TCMO30. Interestingly, the Latin American countries in our sample do not
demonstrate any relationship between tax as a percentage of TCMO and
penetration, as per the majority of regions in this study.
30
This, together with the elasticity impact identified above, is one of the main determinants of the relatively low
impact on penetration of reducing taxation

47
0.35
0.3
Brazil
0.25 Argentina
tax as % of TCMO
0.2
Colombia Chile
Peru
Mexico
0.15
Venezuela
0.1
0.05
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Penetration rate
Several Latin American jurisdictions have lowered the mobile specific taxes since the
GSMA 2005 study on this issue:
• Venezuela: VAT on services has decreased from 15% to 14%;
• Brazil: Customs duties on import of handsets have decreased from 20% to
16%; and
• Mexico: “General Import duties" have fallen from 15% to 0% on the import of
handsets.
María del Rosario Guerra de la Espriella - Ministra de Comunicaciones,
Colombia
How successful has mobile communications been in Colombia, and how does
taxation affect the mobile industry?
Mobile penetration has been growing due to greater competition from the entrance of
a third mobile operator—Colombia Movil (OLA, now called TIGO) in January 2003.
This generated a very steep increase in subscriber growth between 2003 and 2006,
reaching almost 70% market penetration by the end of September 2006.
Mobile telephony in Colombia has not benefited from special reductions in the tax
system. On the contrary, mobile communications is considered a luxury good and
has a special VAT rate of 20%, compared to common goods and services which are
charged at 16%. These four percentage points are used to support sports and culture
in Colombia. As yet, the government has not introduced any specific tax-reducing
measure for the mobile industry.
What impact would you expect a reduction in mobile taxes to have on Government
taxation revenue over time?

48
Taking into account data for 2005, mobile telephony revenues have been US $2340
million based on 21.8 million users and an average ARPU of US $9 per month, with
total VAT collection of US $390 million (equivalent to 0.32% of the GDP). If we had
reduced VAT to 16% during 2006, keeping the same ARPU levels, we could have
seen VAT collection of US $431 million, equivalent to 0.33% of the GDP reported as
of September 2006.
In general terms, therefore, a VAT rate reduction could have kept constant the
sector’s level of collection as a % of the GDP.
Does the Government regard mobile penetration and usage as a driver of economic
development across the economy as a whole?
It is this administration’s vision to generate a massive mobile telephony expansion,
as stated in the IT Government Plan for 2006-2010 with the slogan “Every Colombian
connected and informed”.
In light of the fact that teledensity and infrastructure development levels are very
high, the Colombian Government considers that mobile telephony is a fundamental
driving force to access the Society of Information. To that end, the Ministry of
Communications of Colombia is planning to have 90% geographical mobile coverage
over the 1098 municipalities that are part of the national territory. This could be the
best alternative to connecting country villages that are far away from high-populated
areas and promote universal service policies.
Would you encourage other countries to reduce mobile specific taxations?
As mobile telephony reaches higher levels of market penetration, the service
becomes a very important instrument for spreading IT and consolidating the
Information Society. Once this is recognised by a government as a means to reduce
the digital divide, it could be appropriate to consider the reduction of certain specific
taxes that can favour access conditions to this service. Today in Colombia, because
of its high level of use and penetration, mobile telephony is now a basic consumer
good, but it is taxed as a luxury good.

Of the 16 countries in our sample that impose mobile specific taxation, 2 of them are
in Latin America. The following figure summarises the overall effects of reducing
mobile specific taxes to 90% of their current level. The suggested impact on
penetration is positive, but limited by the relatively low penetration on handsets (see
section 3), whilst the impact on overall government revenues is only mildly negative.

49
Figure 41: Overall impact by 2010 of reducing mobile specific taxes for Latin
American jurisdictions with those taxes to 90% of current level
0.10%
0.05%
Change in
0.00% subscribers
Dominican Republic Venezuela RB Change in
taxation
-0.05% revenues
-0.10%
-0.15%
Source: Deloitte

50
8 Asia Pacific

In this study we have analysed 17 countries from the Asia Pacific region, with an
overall population of 3.3 billion, equivalent to 50% of the world population. The
mobile subscriber base of our sample of countries is low at 14%, only 3% higher than
in Subsaharan Africa.
The figure below shows the historical and forecasted average penetration rate for the
region. The recent growth path for this region is more stable than other developing
country regions and is forecast to remain so in the near future.
Figure 42: Average penetration rates in Latin American sample markets
0.4
0.35
0.3
0.25
Penetration rates
0.2
0.15
0.1
0.05
0
01
01
01
01
02
02
02
02
03
03
03
03
04
04
04
04
05
05
05
05
06
06
06
06
07
07
07
07
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
1
4
1
4
1
1
2
4
Q
Q
Q
Q
Q
Q
Q
Similarly to most other regions a strong correlation exists between GDP and
penetration. As such, there is a significant difference in penetration levels across the
sample, e.g. Malaysia has a penetration rate of 57%, while countries such as Nepal
and Papua New Guinea have rates of around 1%.

51
Figure 43: GDP per capita and penetration - Asia Pacific
0.9
0.8
Malaysia
0.7
0.6
Thailand
Penetration rate
0.5
Philippines
0.4
China
0.3
0.2
Indonesia
0.1
India
0
0 1000 2000 3000 4000 5000 6000
GDP per capita ($)
Tariffs pre tax in this region are the lowest of all those considered in our sample: the
weighted average cost of a call per minute is estimated to be less than half than in
Subsaharan Africa.
8.2 Taxation regime

The average share of tax in total cost of mobile ownership in the region is 11.2%; the
lowest in our sample. Indeed for 15 of the 17 jurisdictions in the region the share of
tax of TCMO is below the average for all jurisdictions in the sample. The two
exceptions to this are Nepal and Bangladesh, which have at present a share of tax
equivalent to 23% and 19% respectively.
The following figure shows this, but does not demonstrate any relationship between
tax as percentage of TCMO and penetration.

52
Figure 44: Tax as a percentage of TCMO - Asia Pacific
0.3
0.25
Nepal
0.2
Tax as % of TCMO
Bangladesh
Samoa
0.15
Indonesia
0.1
Malaysia
India Vietnam
0.05
China
0
0 0.2 0.4 0.6 0.8 1
Penetration rate
Source: Deloitte analysis based on Wireless Intelligence and Deloitte tax
There have been considerable changes in the taxation regime since the GSMA 2005
study into this issue. The following jurisdictions have increased their taxation
structure as it impacts mobile:
• Bangladesh: there has been the introduction in 2006 of a supplementary duty
due on SIM cards of 35%;
• India: service tax on use has increased from 10.2% to 12.24%;
• Philippines: VAT on services and handsets has increased from 10% to 12%;
and
• Cambodia: new specific tax of 10% applied to handsets.
Conversely, there has been a decrease in taxes in the following jurisdictions:
• Bangladesh: the fixed subscription has decreased from US$13.80 to
US$11.76. Fixed customs duty has also been reduced from US$4.60 to
US$3;
• Cambodia: specific tax on services reduced from 10% to 3%; and
• Vietnam: customs duties have fallen from 15% to 7.5% on import of handsets.
Of the 17 countries in this region considered, 4 impose mobile specific taxes:
Bangladesh, Nepal, Pakistan and Sri Lanka. The following figure summarise the
suggested overall effects of reducing mobile specific taxes to 90% of their current
level. In aggregate this has a positive impact on penetration and a marginally
negative impact on overall Government revenues.

53
Figure 45: Overall impact by 2010 of reducing mobile specific taxes for Asian
Pacific jurisdictions with those taxes to 90% of current level
0.8%
0.6%
Change relative to base case
0.4%
0.2%
Change in
subscribers
0.0%
Change in
-0.2% Bangladesh Nepal Pakistan Sri Lanka taxation
revenues
-0.4%
-0.6%
-0.8%
Source: Deloitte
As suggested above, these results do not capture the full impact on the remainder of
the economy and the intangible benefits for users of mobile technology
General Malik, Chairman of Pakistan Telecoms Regulatory Authority
General Shazada Alam Malik, Chairman of the Pakistan Telecoms Authority, has
played an instrumental role in transforming the mobile communications sector in his
country in a very short period of time. He introduced much greater competition into
the mobile market, established clear rules distinguishing between mobile and
wireless local loop networks, and – perhaps most critical of all – oversaw a reduction
of import duties on mobile handsets to zero.
“Five years ago the Government used to charge approximately 2000 Pak rupees
($35) SIM tax activation on every handset sold, and the regulators charged a 4%
regulatory fee for operators, based on gross revenue. At the same time, I realised
that countries with a similar GDP per capita and income level had mobile penetration
ten times higher than Pakistan. I determined that this low mobile penetration rate in
Pakistan – around 1% - was mainly due to the high taxation levels. So I started to
work towards reducing the taxes in one form or another.”
Malik successfully convinced the Government to reduce the $35 SIM activation tax
by half to $17.5. He then managed to reduce this by 50% again, lowering it to
approximately $8. In addition, Malik cut the regulatory fee on operators to 1.5%.
“I believe that taxation can have a very negative impact on the growth of mobile. The
aim was to reduce the barriers to entry for people who could not afford the $35
access fee. This was too much – after all, we are not a tax collecting agency. We
reduced that entry barrier substantially.”
This work has had a huge effect on access to mobile communications in Pakistan.
The proportion of Pakistanis with a mobile phone has grown from under 1% in 2000
to over 20% in 2006, with mobile network coverage in the country having increased
from 40% 3 years ago to around 80% today. “Pakistan is a text book example,” says

54
Malik. “Reduce the taxes or the duties and there is tremendous growth. The
operators are now investing heavily because they think there is huge potential, and
the market now has confidence in the regulators.”
“I am a firm believer that taxes should be reduced. When the Government was
charging $35 SIM activation tax the total tax collected by the Government was much
less than it is today as so many more customers are now able to enjoy the benefits of
mobile communications and pay the $8 activation fee. In addition, the Government
collects 5% GST (General Sales Tax) on mobile operator revenue, so the more
people that are able to enjoy the benefits of mobile communications, the more
revenue the Government collects in this format. Governments will not be the loser in
the reduction of taxes – there will be many more people able to use mobile
technology and this will have a tremendous social and economic benefit for all. Lower
taxes have also reduced the threat from the black market as there is no incentive to
create such a situation.”
Looking ahead, Malik aims to cut taxes even further, targeting a SIM activation tax of
$4. “We are confident we can convince the budget financiers to reduce taxes next
June. This will surely increase the growth of mobile in Pakistan.”

55
9 Central and Eastern Europe

In this study we have included 20 jurisdictions, including the EU transition
economies, i.e. those outside of the EU 15, in a broad Central and Eastern Europe
(CEE) region. The total population of the countries in our sample in this region is
almost 470 million. The average penetration rate in this region is 77%. In particular,
a number of jurisdictions have a penetration rate above 100%, i.e. Lithuania, Estonia
and the Czech Republic.
The following figure demonstrates the historical and forecasted penetration rates for
the region. Given the level of saturation in a number of the more populous countries
within the region the growth in penetration is forecast to slow over the next year.
Figure 46: Average penetration rates in CEE sample markets
120%
100%
Penetration percentage
80%
60%
40%
20%
0%
01
01
01
01
02
02
02
02
03
03
03
03
04
04
04
04
05
05
05
05
06
06
06
06
07
07
07
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
1
4
Q
A positive relationship between GDP per capita and penetration exists and is the key
driver of remaining divergences in penetration rates across the region.

56
Figure 47: GDP per capita and penetration - Asia Pacific
1.8
1.6
Lithuania
1.4
Czech Republic
1.2 Russia Slovenia
penetration rates
0.8 Ukraine
0.6
Poland
0.4
Turkey
0.2
Uzbekistan
0
0 5000 10000 15000 20000
GDP per capita ($)
9.2 Taxation regime

The majority of countries in this region have high taxes as a proportion of the TCMO.
In particular, Turkey has the highest percentage across our sample, driven by a
series of very high mobile specific taxes.
Figure 48: Tax as a percentage of TCMO - CEE
0.52
Turkey
0.47
TCMO
0.42
a percentage of
0.37
asTCMO
0.32
Tax in
0.27
Poland
0.22 Romania
Lithuania
0.17
Russia
0.12
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
Penetration rate
Source: Deloitte analysis based on Wireless Intelligence and Deloitte tax

57
Binali YILDIRIM, Minister of Transport and Communications, Republic of
Turkey
What is the current state of the mobile market in Turkey and its levels of taxation?
“The number of mobile telephone subscribers is approximately 50 million. The
increasing trend in the number of GSM subscribers gained momentum particularly
after liberalisation of the sector in 1998 and this number has doubled over the last
three years. Penetration with respect to population is around 70 %. The number of
base stations naturally increased in parallel with this rise and efforts were ramped up
to ensure service quality.
The total rate of taxes on communication services in our country has reached 56
percent in GSM and 43-44 percent in fixed lines. In Uganda, which is the second
highest taxed country in mobile phone services, the rate is 30 percent, and it has
dropped to even 3 percent in some countries. Besides the excessive magnitude of
rates, there is also a sheer multitude of taxes. While there is only one tax in other
countries, there are five or six different types in Turkey. Given this situation, there
comes a point where investors feel reluctant to move ahead. The investors' path
must be cleared.”
Why does Turkey have such high taxation rates?
“These taxes were imposed for the purpose of overcoming the hardships
encountered in our country after the Marmara earthquake in August 1999. Everyone
agrees that taxes on communications are high in Turkey. It's also said that this state
of affairs encourages off-the-record business practices. High taxes certainly push
investment costs higher, which both slows down investments and affects the users in
the form of costly service, blocking the mechanism of competition. There is no doubt
that this is not a desirable situation.
Intensive efforts are being spent in our country in the fields of telecommunications,
information technologies, and communication services under the influence of
liberalisation, among other factors. The industry's liberalisation has been completed
by the handover of Türk Telekom shares. Consequently, competition conditions
became far more challenging. Also, prices continue to decrease. Campaigns and
promotions are being offered in both mobile and fixed telephones. These price cuts
eventually ended up in an impasse. They cannot go lower because the tax rates are
too high. It has therefore become necessary to lower taxes.
Providing access to information in an effective, faster, and more reliable manner at
reasonable prices and offering alternative infrastructure and services are the issues
we dwell upon most. Our goal is to make sure that more telecommunication services
are made available to individual or corporate users.”
What steps are you taking to reduce such rates?
“We have initiated efforts to bring about a gradual reduction of the tax burden in order
to ensure that the high tax rates on electronic communication services in Turkey do
not constitute an obstacle to our goal of transformation into an information society.
We gave a comprehensive presentation on this issue to the Council of Ministers
several months ago. We should continue this study in 2007 in coordination with the
Ministry of Finance in order to determine the discount rates and schedules. That is
what we are going to do in 2007. Our work will focus on both the rates and the
implementation. One thing worth mentioning is that if one tax is to be discontinued,
another source has to be found to replace it.

58
With regard to communication tax, the ideal figure in our minds is the Value Added
Tax (VAT). The world's developed countries in the information/communication
industry have only VAT. And its rate is low. Our final goal is to reach that point. It's an
ambitious goal but unfortunately we are unable to take large strides to match it due to
the circumstances prevailing in our country. We have to take small but firm steps.
Our government lowered the corporation tax so that entrepreneurs could invest more.
We dropped the VAT on food and textiles. We accomplished the same in the civil
aviation sector, where we removed taxes with the aim of developing the aviation
industry so that the state could collect more taxes from the resulting turnover. We
were proven right. There was a huge growth in turnover and in the number of air
passengers. The Finance Ministry (The Ministry of Finance) derived revenue in
aviation, from VAT. Corporation tax was generated. 15,000 to 20,000 people have
been employed. In conclusion, this was a profitable and successful implementation in
every respect. I'm sure the same will happen in mobile too.”

Of the 20 jurisdictions in this region considered, 2 impose mobile specific taxes:
Albania and Turkey. The results suggest that reducing such taxes to 90% of their
current level will only have a limited impact in Albania but is much more significant in
Turkey.
Figure 49: Overall impact of removing mobile specific taxes on Asian Pacific
countries with those taxes
1.50%
1.00%
Change relative to base case
0.50%
Change in
0.00% subscribers
Albania Turkey Change in

-0.50% taxation
reciepts
-1.00%
-1.50%
-2.00%
Source: Deloitte
The result for Turkey is unsurprising given the higher reliance on mobile specific
taxes in Turkey. The results suggest that the positive taxation levers (VAT and
corporate tax) are set at too low a rate to counterbalance the large direct losses from
the removal of the high level of mobile specific taxes. In a static analysis it could be
argued that this makes a case against tax reform. However, a more dynamic
conclusion is that this suggests the need for a larger general reform of taxation in
order to create appropriate and fair incentives for providers of goods and services
across the economy and hence encourage appropriate allocative efficiency. Clearly,

59
the taxation structure in Turkey differs from the remainder of our sample set and a
more detailed investigation of these issues would be required.

60
10 Conclusions and policy implications
Mobile phones have the potential to link together communities within and between
countries, driving social welfare, investment and growth. Where fixed line networks
have traditionally been considered the provider of universal service, developments in
mobile technology (particularly in terms of speed and bandwidth) mean that mobile
services now provide a cost effective method for providing general communications
access. In effect universal service is now provided by mobile technology in many
developing countries where mobile penetration dominates that of fixed line by 10:1.
Governments play a crucial role in determining both the speed of both mobile roll-out
and the take-up of services. The ongoing challenge is to continue working with the
mobile industry to increase this trend through reducing the total cost of owning and
using a mobile phone.
Previous studies have shown that mobile phones are beneficial in reducing the cost
of doing business and driving entrepreneurialism and growth. This is particularly the
case in the many developing countries that do not have an extensive fixed line
telecommunications infrastructure where mobile phones make conducting business
easier. Mobile phones also directly increase social welfare, connecting families and
providing simpler and easier access to educational and health resources in rural
communities. This report shows a positive link between mobile communications and
economic welfare, with a 10% increase in mobile penetration increasing GDP growth
by 1.2% in developing countries.
The mobile industry has made considerable strides towards driving down the cost of
mobile services to consumers. As noted in the 2005 report, the GSMA challenged the
manufacturing sector to respond to the cost of a mobile phone being a significant
barrier to entry and to produce an “ultra low cost” handset for developing markets.
Initially this was achieved at $40 cost, although though further innovation the cost
was reduced to $30. This has been instrumental in increasing take-up and driving
forward the benefits of increasing communications access.
However, whilst the global mobile penetration level now stands at 40% there remain
large differences between developed countries and many developing countries
where penetration lies below 25%. More still needs to be done to increase these low
rates, particularly in those countries where mobile, as opposed to fixed line, is the
main source of access technology. To achieve the maximum possible economic and
social growth, Governments and mobile operators need to work together to
determine the optimal tax mix and level for their particular countries.
High taxes on mobile services run counter to government’s commitments to
improving access to communications. At the World Summit on the Information
Society in 2003, 175 countries signed up to a commitment to give more than half the
world’s population access to information and communications technologies by 2015.
Taking the optimal approach to taxation is a clear route to achieving this goal and
recognising the huge benefits to developing countries and their people. This study
has provided support for this rationale. In particular, it finds that:
• Taxes are higher in many developing countries relative to developing
countries;
• 16 countries levy telecom specific taxes on top of standard sales and import
taxes on mobile phone users;
• Removing mobile specific taxes would increase penetration, usage and
handset sales in the poorest regions of the world, with its strongest impacts
felt in SubSahran Africa; and

61
• The direct impact of reducing mobile specific taxation is, in most cases,
almost fully counterbalanced through indirect taxation and growth impacts.
The relatively limited reduction in Government taxation revenues by 2010
need to be contrasted with the additional economic and social impacts that
can not be captured that can be obtained over and above those allowed for in
this study.

62
Global Mobile Tax Review 2006–2007
Executive summary
Acknowledgements
About Deloitte For further information, or if you would like to talk to one of our
Deloitte has the broadest and deepest range of skills of experts on Telecommunications or Tax matters please contact:
any business advisory organisation. We are renowned
Global – John Ruffolo T + 1 416 601 6684
for our innovation, collaboration, our solutions-led approach,
E [email protected]
and our outstanding quality of client service.
EMEA – Dennis Knowles T +44 20 7007 3822
The Deloitte Touche Tohmatsu (DTT) Technology, Media &
E [email protected]
Telecommunications (TMT) Industry Group consists of more
than 5,000 member firm partners, directors and senior managers Asia Pacific – Nigel Mellor T +65 6530 5523
supported by thousands of other professionals. E [email protected]
TMT has dedicated practices in 45 countries, centres of Latin America – Carlos Ianucci T +54 11 4320 2736
excellence in the Americas, EMEA and Asia Pacific and serves E [email protected]
over 90 percent of the TMT companies in the Fortune Global 500.
Economic Consulting – Chris Williams
Clients include some of the world’s top software companies, T +44 20 7007 1071
computer manufacturers, wireless operators, satellite broadcasters, E [email protected]
and advertising agencies – as well as leaders in publishing,
telecommunications and peripheral equipment manufacturing.
The tax data collection and economic analysis for this report was
conducted by Deloitte & Touche LLP. The data and analysis is
subject to assumptions and limitations which are highlighted in
various sections within the full report.
This information contained herein has been written in general terms and Deloitte refers to one or more of Deloitte Touche Tohmatsu, a Swiss Verein, its
therefore cannot be relied on to cover specific situations; application of the member firms, and their respective subsidiaries and affiliates. Deloitte Touche
principles set out will depend upon the particular circumstances involved and Tohmatsu is an organisation of member firms around the world devoted to
we recommend that you obtain professional advice before acting or refraining excellence in providing professional services and advice, focused on client
from acting on any of the contents of this publication. service through a global strategy executed in nearly 140 counties. With access
to the deep intellectual capital of approximately 135,000 people worldwide,
Deloitte Touche Tohmatsu would be pleased to advise readers on how to apply Deloitte delivers services in four professional areas – audit, tax, consulting and
the principles set out in this report to their specific circumstances. Deloitte Touche financial advisory services – and delivers more than 80 percent of the world’s
Tohmatsu accepts no duty of care or liability for any loss occasioned to any person largest companies, as well as large national enterprises, public institutions,
acting or refraining from action as a result of any material in this report. locally important clients, and successful, fast-growing global growth companies.
Deloitte & Touche LLP refers to the United Kingdom limited liability Services are not provided by the Deloitte Touche Tohmatsu Verein, and, for
partnership, a member firm of DTT. regulatory and other reasons, certain member firms do not provide services in
all four professional areas.
As a Swiss Verein (association), neither Deloitte Touche Tohmatsu nor any of
its member firms has any liability for each other’s acts or omissions. Each of the
member firms is a separate and independent legal entity operating under the
names “Deloitte”, “Deloitte & Touche”, “Deloitte Touche Tohmatsu”, or other
related names.
15 63
Executive summary
The GSM Association (GSMA) is the global trade association

representing more than 700 GSM mobile operator Members
across 218 countries and territories of the world. In addition,
over 190 Associate Members (manufacturers and suppliers)
support the Association’s initiatives as key partners.
Encompassing technical, commercial and public policy initiatives,
the GSMA focuses on ensuring wireless services work globally;
thereby enhancing the value of mobile services to individual
customers and national economies while creating new businesses
opportunities for operators and their suppliers.
For further information contact:
Mark Smith / David Pringle
GSM Association
Tel: +44 78 50 22 97 24 / +44 79 57 55 60 69
Email:[email protected]
16
64
Executive summary
Acknowledgements
Since 1976, Tarifica has been providing clear, up-to-date and Wireless Intelligence is a comprehensive database on the
accurate tariff information to network operators, regulators global mobile market. It covers all mobile technologies
and financial institutions. and includes over a million individual data points across
600 operators, 1,100 networks in 220 countries. Because of a
Our global, multilingual team of researchers constantly monitors
need for an up-to-date and accurate view of the global mobile
and reports on the ever-changing tariff environment whether for
markets, the GSM Association formed a unique partnership
voice, data, Internet or video services. Coverage is comprehensive
with Ovum, the leading industry analyst firm.
with Tarifica’s databases containing the Fixed and Mobile tariffs
for over 400 operators in 130 countries. Through the GSMA, the majority of the world’s operators have
access to their own data and, with over 40,000 database queries by
In-depth comparisons are easily obtained from quarterly pricing
members in 2006, this makes Wireless Intelligence one of the most
benchmarks of European, Middle-Eastern and Asia-Pacific
referenced sources of its kind in the world.
markets. The most important tariff developments worldwide are
analysed in a weekly newsletter which has become the must-read Wireless Intelligence provides operator data across operational
publication for service provider pricing specialists. and financial metrics and allows analysis at an operator, country,
Contact: John Lilley – Tel: +44 (0)207 692 5287, regional or global level. The metrics in the service cover subscriber
[email protected]. connections, growth rates, technology market shares as well as a
range of operational metrics such as churn, minutes of use and
financial metrics including revenue, capex and EBITDA margin.
17 65
For more information contact:
GSMA London Office GSMA Dublin Office
1st Floor Mid City Place Block 2
71 High Holborn Deansgrange Business Park
London WC1V 6EA Deansgrange
United Kingdom Co. Dublin Ireland
Tel: +44 (0)20 7759 2300 Tel: +353 (0)1 289 1800
Email: [email protected]
www.gsmworld.com

Tax Review 06 07

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tax Review 06 07

Uploaded by

Copyright:

Available Formats

Global Mobile Tax Review

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

1.1 Tackling the digital divide

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

Maghreb and ME CEE

EU15 Latin America

Subsaharan Africa Asia Pacific

Type of tax Africa Middle Asia Latin CEE EU15

Handset 0.8% 0.68% 0.73% 0.4% 0.7% 0.5%

Connection 0.01% 0.12% 0.53% 0.01% 0.2% 0.01%

Rental 4.31% 2.6% 2.8% 4.85% 3% 4.7%

Usage 10.8% 10.2% 7.7% 13.4% 16.5% 14.4%

Total 16% 13.6% 12% 18.6% 20.4% 19.6%

Global Mobile Tax Review 2006-2007

Type of tax Africa Middle Asia Latin CEE EU15

Handset 3.15% 1.6% 2.25% 1.4% 1.6% 1.6%

Connection 0.05% 0.3% 0.7% 0.04% 0.4% 0.02%

Usage 14.11% 12% 9.4% 17.3% 18.4% 18%

Total 17.3% 14.1% 12.4% 18.7% 20.6% 19.6%

This point is further illustrated by the cumulative distribution of tax as a proportion of

Global Mobile Tax Review 2006-2007

For post-paid platforms:

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

Tax as proportion of TCMO

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

Tax as proportion of TMSC

Global Mobile Tax Review 2006-2007

2.4 Taxes on Mobile Handsets

Global Mobile Tax Review 2006-2007

Tax as proportion of Handset cost

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

2.6 Other tax related issues

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

3.1 Overall modelling approach

This model was constructed by Deloitte using taxation information collected by

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

3.2 Usage and penetration elasticities

Global Mobile Tax Review 2006-2007

Percentage of population that 0.0044 3.68

In particular, the statistically significant linear relationship between handset prices,

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

3.3 Mobile communications and GDP

Developing country set Full data set

Explanatory variables Coefficient t-statistic Coefficient t-statistic

Average mobile penetration

Global Mobile Tax Review 2006-2007

Global Mobile Tax Review 2006-2007

Notes. * On employment, ** On GDP

Global Mobile Tax Review 2006-2007

Angola Egypt Arab Rep. Argentina Bangladesh Albania Austria

Botswana Iran Islamic Rep. Bolivia Bhutan Azerbaijan Denmark

Cameroon Mauritania Chile China Cyprus France

Chad Morocco Colombia India Czech Republic Germany

Ethiopia Yemen Guatemala Malaysia Hungary Italy