APPLICATIONS OF A TELECOM SECTOR MODEL FOR
ESTABLISHING AND VALIDATING TELECOM POLICIES AND
STRATEGIES
Boris Ramos
Escuela Superior Politécnica del Litoral,
Campus Gustavo Galindo, Km. 30.5 Via Perimetral, Guayaquil, Ecuador.
E-mail address: bramos@ espol.edu.ec
Phone: 011 593 9 1456098
Abstract
This article describes applications of a system dynamics model of the telecom sector to
the definition and validation of telecom policies and strategies in developing countries.
This describes several cases where this model has been used to solve problematic and
complex behavior found in the telecom sector, such as saving telecom operators from
bankruptcy, validating a new structure for three telecom operators, defining and
validating universal service policies, and validating tax policies for the telecom sector.
Introduction
This paper shows different applications of a system dynamics model to the resolution of
problematic and complex behavior still found in the telecom sector of developing
countries. The telecommunications sector has several positive externalities and feedback
loops that trigger the diffusion of its services, the adoption of subscribers, and the
collection of revenues (Ramos et al. 2010). However, in spite of that, there are still
several operators within this profitable sector that face bankruptcy or economic loses.
This article shows the application of system dynamics to the definition and validation of
strategies in order to improve the performance of two government-owned telecom
operators in developing countries. In these cases, the poor planning previously applied
had led to the failure of their operational and financial performances. The one case
corresponds to a fixed telecom operator, which had been experiencing loses over the
past years of this study, and the other a mobile operator that had been declared in
bankruptcy after several years of not seeing profits since its creation.
The economic crises that face many countries worldwide obligate their governments to
find new sources of funding to balance their respective budgets. One common
mechanism is to increase tax revenues by imposing new taxes, especially to sectors
considered sumptuary and that generate very high levels of revenues.
Telecommunications is commonly considered a very profitable sector, but in some
cases, it has been mistakenly considered a sumptuary one.
This paper also describes the validation of a tax rate applied to the telecom sector,
specifically investigates the impact of applying a tax rate to the cellular telephone
traffic. It has been observed that telecommunications generate a significant fraction of
the GDP in most countries, which makes it attractive for new taxes and fees. On the
other hand, telecommunications are actually used as a luxury in some few cases, but in
most cases, these are utilized for increasing the productivity and the welfare of the
population (Hudson 1984; The ITU-D Focus Group 7 2000).
Several unsuccessful universal service policies have been applied in developing
countries in past years (Ramos et al. 2010). Universal service can be defined by the
provision of “universal” availability of connections to individual households from
public telecommunication networks, with non-discriminatory and affordable prices
(Hank and McCarthy 2000). An alternative to universal service is “universal access”,
which is defined by a situation where every person has a reasonable means of access to
a public telecom network (Cain and Macdonald 1991).
This work is also used to validate the design of a new proposed universal service policy.
This policy defines a variable universal service rate that impose a higher universal
service fee to fixed telephone operators with low capacity in rural areas and a lower
universal service fee to operators with higher capacity in rural areas. It was found that
this policy fails on achieving the goal of reducing the urban-rural telecom gap.
However, it was shown by this investigation that this policy might be effective on
reducing cream skimming in fixed telephony.
The simulations carried by this model helped identifying counterproductive behavior,
incomplete policy definitions, and the causes of failure of the proposed telecom policies
and strategies. For instance, counterproductive behavior was found after simulating a
tax policy for the telecom sector, because it reduced the total collected telecom taxes in
the long term. The incomplete policies showed the lack of complementary strategies to
increase the probability of implementation success. These deficient policies include the
procedure to merge three telecom operators, and the new proposed universal service
policy.
Finally, a telecom strategy transformed a cellular company into a mobile virtual
network operator (MVNO), which allowed leasing telecom capacity to other mobile
operators. This implementation was supposed to save the operator from bankruptcy by
improving the number of subscribers and the economic resources in the short term. The
underlying causes of the failure of this policy were also found through the simulations
of the model.
A System Dynamics Model of the Telecom Sector
The model sectors and their interactions are shown in Figure 1. These sectors are
grouped according to the functions they perform (Ramos et al. 2010).
Telephone Demand represents the subscribers with telephone lines and the unmet
demand, which is composed of subscribers in the waiting list and potential subscribers
willing to pay the monthly rental fee (Warren 2002). The subscribers in the waiting list
are connected to a rate, which depends on the telephone capacity (Jensen et al. 2002;
Ros and Banerjee 2000).
Telephone Deployment represents the process of deployment of new telephone lines,
which is a function of the economic resources available, the number of subscribers in
the waiting list, the current telephone capacity, and the number of telephone lines being
deployed (Lyneis 1980).
Telephone Traffic represents the level of usage of the telephone network as a function
of the price, income per capita of the population, the current telephone density, and the
telecom penetration in the region (Kayani and Dymond 1997; Shapiro and Varian
1999).
The Financial Resources sector accumulates the economic resources used in the
expansion of telecom infrastructure. This sector depends on the number of subscribers,
the telephone traffic, the price of the service, the operating costs, and the amount of
money invested in telecom infrastructure (Kayani and Dymond 1997).
Rural Telephone
Demand
+
W,
Rural Telephone +
Deployment
Rural Traffic
+4
. ar
+
Financial
esources
JT }
+,
Urban Telephone
Deployment
Urban Traffic
+
+
4
Urban Telephone
Demand
Figure 1. A simplified sector map of the Telecom Sector Model (Ramos etal.
2010)
Real Applications of the Telecom Model
The basic model described in the previous section has been extended and modified to
validate and define different policies and strategies applied to the telecom sector of
developing countries. The five telecom policies and strategies being tested and analyzed
are described as follows.
1. Defining Strategies to Save a Mobile Operator from Bankruptcy.
The mobile operator under analysis was government-owned and the smallest one of
a developing country. This operator has been losing money since its creation, and
the growth of mobile subscribers in the recent months has been lower than before, in
spite of that the company was recently transformed into a mobile virtual network
operator (MVNO). It is important to note that this operator was also the last one to
enter into the mobile competing market.
The main determinants of growth or stagnation were attributed to the network
extemality of the competing mobile operators, the installed capacity of the network,
the distribution channels, the advertising of the operator, and the operating costs.
The causal structure of the growth of the referenced mobile operator is described in
Figure 2.
As observed in positive feedback loop 1 of Figure 2, the economic resources and the
MVNO implementation increase the installed capacity of the mobile operator, which
improves the coverage of the network. The better coverage of the network improves
the demand of the mobile service, increasing the number of subscribers. The higher
number of subscribers increases the economic resources of the mobile operator.
Finally, the subscribers adopted by the operator are also a function of the
distribution channels and advertising.
On the other hand, as it can be seen in negative feedback loop 4 in Figure 2, the
increase of the installed capacity raises the operating costs, which reduces the
economic resources of the company. In addition, the positive feedback loop 2 in the
same figure indicates that a higher number of subscribers increase the network
externality of the mobile operator. This tends to reduce the subscribers from other
mobile operators. On the other hand, the higher number of subscribers from other
operators increases their network externality, reducing the subscribers from our
mobile operator.
The following strategies were tested in the model sequentially and in an aggregated
manner:
1. Price reduction
2. Increase of distribution channels
3. Higher investment on advertising.
4. Reduction of operating costs.
The base case is observed in line 1 of Figures 3 and 4. These strategies were
implemented in month 54.
Distribution
MNO, Coverage. Channels , Network Externality
nA from Mobile Operator
+
Installee—> (nn (9) oY
Capacity 13) osts Ne Subscribers from
Other Operators
Oh seo Extemnality “x
‘Resources from Other Operators. +
+ Advertising
Figure 2. The causal structure of the growth of the mobile operator
The price reduction was a common strategy implemented by the mobile operator in
order to improve its market share. However, this price strategy was shown to
improve the demand and the number of subscribers, as observed in line 2 of Figure
3, in a lower proportion than the reduction of the price. This led to the reduction of
the economic resources in the short to middle term, as observed in line 2 of Figure
4, which is a counterproductive result.
One year before this investigation, the mobile operator became a MVNO without
observing since then, a significant improvement on its financial, operational, and
market performance. The main reason for the failure of the MVNO implementation
that was found was the lack of distribution channels and a low level of investment in
advertising, as shown in feedback loop 1 of Figure 2. It was also proved through the
simulations that in spite that the MVNO implementation increased the installed
capacity of the operator, it was not appropriately accompanied with an increase on
the number of distribution channels and more investment on advertising.
The increase of the distribution channels represented in line 3 and the improvement
of advertising in line 4 of Figures 3 and 4 were strongly recommended because they
improved the performance of the system. It was shown by the simulations that they
improved the number of subscribers and the economic resources of the mobile
operator as shown in Figures 3 and 4 respectively.
The reduction of operating costs by increasing efficiency tends to improve the
economic resources of the mobile operator, as shown in feedback loop 4 of Figure 2.
This situation increases the advertising even more, raising the number of subscribers
as shown in line 5 of Figure 3. The higher number of subscribers tends to improve
the economic resources of the operator. However, more investment on advertising
end reducing a little bit the economic resources in spite of more subscribers, as
shown in line 5 of Figure 4.
200 foc
Months
0 14 34.8 52.2 69.6 87
Months
Figure 4. Economic Resources for Different Strategies of the Mobile
Operator
2. Defining Strategies to Improve the Performance of a Fixed
Telecom Operator.
There were two main government-owned fixed telecom operators serving a different
region of the same developing country. The company under analysis was the second
largest and has been losing money in most of the previous years, in spite of a
significant number of urban subscribers. It is important to note that the average
revenue per user (ARPU) has been reducing through the years, however its value is
still considered high. In addition, the revenue from international traffic is
considerably lower than previous years.
The main determinants of growth or stagnation were attributed to the network
externality of the operator, the installed capacity of the network, and the operating
costs. The causal structure of the system is described in Figure 5. As observed in
positive feedback loop 5 in this figure, the economic resources increase the installed
capacity of the operator, which improves the connection rate and the number of
subscribers. The higher number of subscribers raises the economic resources of the
company.
On the other hand, the negative feedback loop 7 in Figure 5 shows that the increase
of economic resources improves later the installed capacity of the operator raising
its operating costs, which in tum reduces the economic resources of the company.
Finally, the higher number of subscribers improves the network externality of the
operator, which increases the demand, the connection rate, and the number of
subscribers, as seen in positive feedback loop 6 in the same figure.
The following strategies were tested sequentially, in an aggregated manner, in the
model:
1. The use of wireless access technology
2. Reengineering of capacity implementation process
3. Reduction of installation prices
4. Reduction of operating costs
The base case is observed in line 1 of Figures 6, 7, and 8. These strategies were
implemented in month 78.
Efficiency Connection Rate
Demand
P somll, co
Installe ing operatizy costs A+3)
Capacity
+ 4-7) Es Network
. Extemality
Economic 4 Subscribers
‘Resources
+
a +
a
Inoome from “ARPU
International Traffic
Cellular Traffic
Bypass IP Telephony
Figure 5. The causal structure of the growth of the fixed telecom operator
It is already known that wireless technology reduces the costs of telecom
infrastructure and the implementation delays, which tends to improve the installed
capacity, as observed in positive feedback loop 5 in Figure 5. This situation was also
shown by the simulations, since wireless technology implementation considerably
improved the number of subscribers and the economic resources of the operator, as
observed in line 2 of Figures 6 and 7 respectively.
The reengineering of the capacity implementation process allows reducing the
delays incurred in the causal relation between the economic resources and installed
capacity, as shown in positive feedback loop 5 of Figure 5. It was seen through the
simulations that this strategy was able to improve the number of subscribers, the
economic resources, and the installed capacity of the fixed telecom operator, as seen
in line 3 of Figures 6, 7 and 8.
The reduction of installation prices was intended to improve the demand and the
number of subscribers of the service. It was observed through the simulations, that
the improvement of demand produced by this strategy was not able to increase the
connection rate and the number of subscribers, as observed in line 4 of Figure 6. It
happened because the connection rate also depends on the installed capacity and the
economic resources of the operator, as observed in positive feedback loop 5 of
Figure 5, which practically remained constant as can be seen in line 4 of Figures 8
and 7 respectively.
The reduction of operating costs considerably improved the economic resources and
the installed capacity of the telecom company in the short term, as observed in line 5
of Figures 7 and 8. This improvement on economic resources was able to
considerably increase the installed capacity and the number of subscribers, as
indicated in positive feedback loop 5 of Figure 5. The substantial improvement in
the number of subscribers achieved by this strategy in the simulations is shown in
line 5 of Figure 6.
730 4
550 7 1 : i
7660 4
-44680
Figure 7. Economic Resources for Different Strategies of the Fixed
Operator
Installed Capacity ,
egy aise
Ce Coenen
683 4 1 1
0.00 27.00 54.00 81.00 108.00
Months
Figure 8. Installed Capacity for Different Strategies of the Fixed
Operator
3. Validating the Merging Process of Three Telecom Operators.
The final strategy proposed to improve the performance of the previous
government-owned mobile and fixed telecom operators was to merge these two
companies with the remaining government-owned fixed telecom operator, which
was the only one generating economic profits.
In addition, the national fixed telephone density had not improved in the previous
years, and was relatively low compared with the mean values of the countries in the
region. In addition, the urban-rural fixed telephone gap has not been reduced in spite
of the existence of universal service policies in place in the country.
The purpose for merging the telecom companies was to guarantee universal service
under economic and social profitability, with affordable prices and national
coverage. The causal structure of the mobile and fixed telecom operators being
merged are similar to those shown in Figures 2 and 5 respectively.
The base case of the simulations is observed in line 1 of Figures 9, 10 and 11. It was
also considered that the merging process was implemented in month 78. The direct
impacts identified after merging the three telecom operators were achieving scale
economies for infrastructure implementation, increasing the number of distribution
channels for a merged service, and a better investment distribution, when using
aggregated economic resources. The performance achieved by the merged telecom
operator after the simulations was a considerable increase of the fixed and mobile
telecom subscribers, and the improvement of the aggregated economic resources, as
observed in line 2 of Figures 9, 10 and 11 respectively.
10
In addition to validate the merging process of the three telecom operators, some
additional recommendations were made in order to successfully implement this
process. Among these recommendations was the reduction of the operating costs by
improving the operational efficiency. This strategy was able to improve the
economic resources of the merged operator and the number of fixed subscribers, as
shown in line 3 of Figures 11 and 9 respectively. The mobile subscribers were not
improved since the mobile service remained under the same MVNO scheme.
The use of new business models for integrated wired and wireless services was
another recommendation made to successfully implement the merging process. This
strategy consisted on providing a more appealing and integrated wireless and wired
services to the fixed telephone subscribers with the main purpose of increasing the
number mobile clients of the merged company. It was shown through the
simulations that this strategy was able to improve the fixed and mobile subscribers
and the economic resources of the merged operator, as observed in line 4 of Figures
9, 10 and 11 respectively.
The use of wireless technologies in rural areas was the final recommendation made
and was also implemented in the model. The simulations showed that this strategy
was able to improve the number of fixed subscribers and the economic resources of
the merged operator, as can be seen in the line 5 of Figures 9 and 11.
Fixed Telephone Subscribers ‘-2-°-4-°-
25000 ]
18715 5
Figure 9. Fixed Subscribers for Different Strategies of the Merged
Operator
11
Mobile Subscribers: '-2-2-4-5-
30000
15000 4 oo
ra
0.00 30.00
Figure 10. Mobile Subscribers for Different Strategies of the Merged
Operator
Economic Resources : ' - 2-2 -
50000 7
-150000 +
-350000
0.
Figure 11. Economic Resources for Different Strategies of the
Merged Operator
12
4, Validating a New Universal Service Policy.
The proposed policy under analysis consisted of a variable rate for the universal
service fee, which should be charged to the fixed telecom operators of the country.
The two main purposes behind this variable percentage were to increase the
universal service fund and to reduce cream skimming of profitable subscribers in
urban areas. The universal service fee contributed by each fixed operator was
calculated as the product between the revenue from its service operation, and this
universal service rate represented as a percentage.
Under this new universal service policy, the fixed telecom operators that were
focusing mostly in high-income urban areas clients should be charged a higher
universal service rate, and operators that were investing considerable resources to
the expansion of fixed telephone services in rural areas should be charged a lower
universal service rate. Then, this universal service fee was defined as a decreasing
linear function of the rural and urban-low income telephone density. The rural and
urban-low income telephone density was calculated as the proportion of the operator
total capacity implemented in rural and urban-low income areas.
The causal structure of the system is shown in Figure 12. In this figure is observed
two fixed telecom operators, operator 1 and operator 2. The operator 1 invests on the
rural and low-income urban network besides the urban profitable network, and the
operator 2 only on profitable urban areas. The operator 1 and operator 2 contribute
to the universal service fund, which is invested on telecenters and infrastructure to
deliver Internet connectivity for public elementary schools.
In Figure 12, it is also observed the cream skimming process in profitable urban
areas. The operator 2 is doing cream skimming because only invests in high-income
urban areas, as shown in positive feedback loop 11, reducing the number of
profitable urban subscribers from operator 1.
In addition, this causal structure shows the causal relationships behind the
implementation of this variable universal service rate. The higher the urban
telephone density, the higher the universal service rate, as shown in negative
feedback loops 9 and 12 of Figure 12, and the lower the rural and low-income urban
telephone density, the higher the universal service rate, as shown in positive
feedback loop 10 in the same figure.
In addition, the feedback loops 9, 10 and 12 show how the universal service fee
reduces the economic resources of the telecom operator, which are used to increase
telecom capacity and the number of urban and rural subscribers of the operators.
13
Operator 1
Economic Operator 1
ore dl —* Resources Universal Service:
Wier +10 Fee
* ’ he
Operator 1 Rural al Operator 1
Low-Income Urbar| —> Universal Service
Subscribers 7 %
Mobile and Carri
+8 Operators Univer
\' 4 Service Fee
j» Operator 1 Urban +
Subscribers Uniserval
Service Fund
Total Rural and +
Low-Income Urban +
Subscribers Telecenters and
fs School Internet
Connectivity
Operator 2
Operator 2_—> Economic Operator 2
+ Resources Universal Service
Revenue
4h ht Fee
+
Operator 2
—> Universal Service
(arf
et
Operator 2 Rural ahd
Figure 12. The causal structure of the impact of a universal service variable
rate
The model was simulated combining this universal service variable rate with other
universal service policies. The base case of the simulations is observed in line 1 of
Figures 13 and 14. It was also considered that these policies were implemented in
month 108. The implementation of the universal service variable rate alone was able
to considerably raise the universal service fund collected from the fixed telecom
operators, as can be seen in the causal structure of Figure 12. However, this policy
did not improve the rural telecom capacity, as observed in line 2 of Figure 13, and
did not close the urban-rural telecom gap.
On the other hand, this policy was able to reduce the number of profitable urban
subscribers from operators focused mostly on these types of clients, and transferred
them to the operators that have been also investing in rural areas, reducing the so-
called cream skimming. An important aspect of this policy implementation is that
the total number of urban subscribers did not reduce, as can be seen in line 2 of
Figure 14.
14
The combination of this universal service variable rate with two different universal
service obligations is also simulated using the model. The first universal service
obligation implemented by the model mandated the fixed telecom operators to
invest four percent of its new telecom capacity in rural and low-income urban areas.
This implementation reduced the number of rural subscribers and improved the
number of urban subscribers, as observed in line 3 of Figures 13 and 14
respectively, which is a counterproductive result. This happened basically because
this level of investment in rural areas is lower than the investment being actually
made by the type of operator number one in Figure 12.
The second universal service obligation combined with this universal service
variable rate, mandated the fixed telecom operators to invest nine percent of its new
telecom capacity in rural and low-income urban areas. This implementation
improved a little bit the number of rural subscribers while at the same keeps the
same number of urban subscribers, as shown in line 4 of Figures 13 and 14
respectively. This improvement is achieved because the level of investment in rural
areas is equivalent to the one being actually made by the type of operator number
one in Figure 12.
Finally, if the last combination of universal service variable rate with universal
service obligation also considers investing fifty percent of the total universal service
fund in rural and low-income urban areas, a considerable improvement in the
number of rural subscribers is achieved, as can be seen in line 5 of Figure 13.
However, the total number of urban subscribers still remains the same, as observed
in line 5 of Figure 14.
Rural Subscribers 1-2-5-4-5-
3000 7 *
2000 4
1000 ‘ ‘ ‘ ‘
0.00 42.00 84.00 126.00 168.00
Figure 13. Rural Subscribers for Different Implementations of the
Universal Service Variable Rate
15
Usban Subscribers © '-2-2-4-5-
30000 7
20000 +
10000 u T
0.00 42.00 84.00 126.00 168.00
Months
Figure 14. Urban Subscribers for Different Implementations of the
Universal Service Variable Rate
5. Validating a Tax Policy for the Telecom Sector.
The economic crisis worldwide affected the economic resources of the country, and
a tax rate to the cellular phone calls was proposed in order to raise its fiscal revenue
and investment budget. This cellular tax was intended to be applied to the generated
cellular traffic, and was calculated as the product of this proposed cellular traffic tax
rate with the revenue generated from the cellular traffic.
The causal structure of the impact of a cellular traffic tax is shown in Figure 15. It
can be seen in the negative feedback loop 14, that the cellular traffic tax rate
increases the cellular traffic tax and the total collected cellular taxes. On the other
hand, the total cellular taxes reduce the cellular operator profit, which is used to
improve telecom capacity and later the revenue of the mobile operator. Additionally,
it is observed in negative feedback loop 16 that the reduction of the operator profit
also tends to reduce the income tax and the total collected taxes from the operator.
It can also be seen in negative feedback loop 15 of Figure 15 that total collected
taxes are also increased by the value added tax, which depends on the cellular
revenue and cellular capacity. Finally, the cellular revenue tends to improve the
operator profit as observed in positive feedback loop 13.
It was shown by the simulations that the application of a cellular traffic tax rate
increases the collected cellular traffic tax, and the total collected cellular taxes at the
beginning. However, this reduces the cellular operator profit and the operator
income tax, which reduces the total collected cellular taxes in the long term, in spite
of an increase on the cellular traffic tax. In other words, it was found by this
16
investigation that the implementation of a cellular traffic tax rate yields
counterproductive results.
-14
Cellular Traffic +
is Value ‘ie
+ + Total Cellular
Taxes
Cellular Traffic
Tax Rate ~ ellular
Operator AG
+ Income Tax
Cellular
Revenue +13 +
a Cellular
Capacity
Figure 15. The causal structure of the impact of a cellular traffic tax
Conclusions
This telecom model was extended to test, design, and validate several telecom policies
and strategies in a country. This article shows real applications of the model and its
effectiveness to find counterproductive behavior, deficient policies, and actual causes of
policy implementation failure in the telecom sector.
It was concluded after the simulations that implementing a cellular traffic tax rate
produces counterproductive results, and implementing a MVNO strategy could also be
counterproductive, if this is not accompanied by appropriate complementary strategies.
The strategy of merging three telecom operators and a proposed universal service policy
were found to be deficient, through the simulations of the model.
17
References
Cain P, Macdonald JM. 1991. Telephone Pricing Structures: The Effects on Universal
Service. Journal of Regulatory Economics 3: 293-308.
Hank I, McCarthy T. 2000. Telecommunications Regulation Handbook, InfoDev,The
World Bank: Washington DC.
Hudson HE. When telephones reach the village. The role of telecommunications in
rural development. Norwood, New Jersey: Ablex Publishing Corporation; 1984.
Jensen K, Trill S, Brady A. Dynamic modelling of demand and supply in
telecommunications. The J ournal of The Communications Network 2002; 1(3): 52-8.
Kayani R, Dymond A. Options for rural telecommunications development. In: World
Bank Technical Paper, No. 359. Washington, DC: The World Bank; 1997.
Lyneis JM. Corporate planning and policy design. Cambridge, MA: The MIT Press;
1980.
Ramos B, Saeed K, Pavlov O. The impact of Universal Service Obligations and
Intemational Cross-subsidies on the dispersion of telephone services in developing
countries. Socio-Economic Planning Sciences 2010; 44: 57-52.
Ros AJ, Banerjee A. Telecommunications privatization and tariff rebalancing: evidence
from Latin America. Telecommunications Policy 2000; 24(1):233-52.
Shapiro C, Varian HR. Information rules. A strategic guide to the network economy.
Boston, MA: Harvard Business School Press; 1999.
The ITU-D Focus Group 7. New technologies for rural applications. Geneva,
Switzerland: International Telecommunication Union; 2000.
Warren K. Competitive Strategy Dynamics. West Sussex, England: John Wiley & Sons
Ltd; 2002.
18
