Connecting Africa
Mobile phones, mobile banking and economic growth, a close link in Africa?
Anouk van der Have
Supervisor: Prof. Dr. B.W. Lensink
Second Supervisor: Dr. P.P.M. Smid
October 2009
University of Groningen
Faculty of Economics and Business
Connecting Africa
Mobile phones, mobile banking and economic growth, a close link in Africa?
ANOUK VAN DER HAVE1
ABSTRACT
In this research, the relationship between mobile telephone penetration and economic growth will empirically be analysed for the groups ‘all countries’, ‘the rest of the world (excluding Africa)’, ‘Africa’ and ‘other developing countries’. Subsequently the mobile influence between these groups will be compared, to find out whether the effect of mobile telephony is the biggest in Africa. Furthermore, this study discusses the effect of mobile banking on national economic growth in Africa. In the dataset are 190 countries included for the period 2000-2006. Significant impacts are found for mobile penetration in Africa and the rest of the world , the mobile impact in Africa can be seen as the biggest worldwide. Furthermore, mobile banking is found to have a positive impact on economic growth in Africa.
Keywords: mobile communication, mobile banking, economic growth, reverse causality, Africa JEL classification: O47, O55, L96, H54
1
‘We put technology in the hands of people’ as Iqbal Quadir explains mobile impact on economic growth,
the man behind the creation of the success story GrameenPhone in Bangladesh. ‘Whatever people do with the phones, it’s good for them, good for the country. Phones bring people together in collaboration, cooperation.’ The general idea was that poverty can be challenged wit h private investment which was essential to create sustainable growth (Sullivan, 2007). Growth does not always translate directly into poverty reduction of course, as gains are never evenly distributed. But studies (Sullivan (2007); Adams (2003)) indicate that on average 1% of GDP growth reduces poverty more or less by 2%. The GrameenPhone in Bangladesh is a famous example of the impressive impact of mobile telephony and mobile banking on GDP (Sullivan, 2007). Since Africa is confronted with similar poverty, the success story of GrameenPhone is the inspiration behind this research about the mobile (banking) impact on economic development. Therefore this study will attempt to support the positive effect of mobile telephony in Bangladesh for Africa and the rest of the world. Furthermore this research will investigate,
whether the impact of mobile telephony on economic development in Africa differs from the rest of the world and other developing countries and whether mobile banking enhan ces national economic growth in Africa.
undertake financial transactions. The combined impact of the fast growing mobile penetration and markets with extremely low access to financial services, makes Africa possibly of particular interest.
Although much is written about the likely impact of mobile telephony worldwide and in developing countries, there are just a few scientific papers written about the effect (Roller and Waverman, 2001; Waverman et al., 2005; Sridhar and Sridhar, 2007). The contribution of this study is the particular focus on Africa, this research empirically examines whether the mobile impact in Africa is a special case when compared to the rest of the world . Furthermore this paper addresses a new subject, many recent articles discuss the likely impact of mobile banking and its convincing opportunities to reduce poverty. However, at this moment no empirical research has been published about mobile banking. To the best of my knowledge this is the first study that attempts to examine the effect of mobile bankin g empirically, it will investigate the possible impact of mobile banking on economic growth.
II. An overview of Mobile Telephony
A. The Expansion of Mobile Telephony
Mobile phones are often the only way of communication for a large number of people in less developing countries. The number of mobile phone services in developing economies has gone beyond the number of fixed-line phones (World Development Indicators, 2009), more than 800 million mobile phones were sold in developing countries in the past three years (GSM Association, 2006). Some developing countries have even experienced a decline in fixed line penetration (Kathuria, et al, 2009). Previous research (Jha et al., 1999) has found that developing countries with an extremely low fixed telephone penetration rate will almost skip the investment in fixed telephone infrastructure. The cost of installation of wireless and mobile networks is much cheaper and an even better way to build a broad network. As a consequence all telecom investment will go to the mobile network and this will lead to enormous growth of mobile subscribers. Especially developing countries have increased their share of mobile subscribers from 30 percent in 2000 to 50 percent in 2004 and to even 70 percent in 2007 (World Development Indicators, 2009). At the end of 2007 there were already three times more mobile phones than fixed phones worldwide. The broad introduction of mobile phones has led to more competition and improvements and expansion in the telecommunication sector (World Development Indicators, 2009).
B. The State of Mobile Telephony in Africa
telephone lines are situated in Africa. In addition, in some African countries, like Burundi, Central African Republic, Eritrea and Ethiopia there are less than 5 persons with a mobile phone per 100 inhabitants. In Figure 1 the mobile situation of Africa is compared with other continents. Nevertheless, in the last couple of years Africa relatively experienced the highest growth worldwide in the amount of mobile subscribers (ITU database, 2007). Between 2001 and 2005 the growth rate of the number of mobile subscribers in Africa was 46.2% (ITU database, 2007). The mobile market penetration in Africa at the end of 2008 was 40% higher than in the two previous years (Leishman, 2009). The mobile explosion in Africa will continue while in other world regions the growth seems to slow down probably due to almost saturated markets like in Europe and the Americas. Africa has the highest growth potential compared to other developing countries due to the huge lack of other communication and information infrastructure (Leishman, 2009). Parke r (2005) found that the discrepancy of digital development between African countries and the developed world is getting smaller. There were more new mobile subscribers in Sub-Saharan Africa than in North-America (Parker, 2005). However, this discrepancy is still much larger than the difference in income between developing and developed world (Wong, 2002).
Figure 1
Comparison of the average number of mobile phones per 100 inhabitants between continents
0 20 40 60 80 100 120 1996 1998 2000 2002 2004 2006
Mobile Phones per 100 inhabitants
Year Average Continental Mobile Phone
As a consequence ‘Connect Africa’ is an important goal of the International Telecommunication Union (ITU) and the United Nations (UN). Although Africa has a share of around 14% of the world’s population, it accounted for little over two percent of the world’s Gross Domestic Product in 2006 (ITU World Telecommunication/ ICT Indicators (WTI) database, 2007). Also between African countries GDP is not equally spread, South Africa has almost a share of 25% of Africa’s total GDP. In most developing countries, mobile operators are the most valued and trusted organizations, whereas banks are often distrusted (Mas and Kumar, 2008). That is why privatization of operators is growing very rapidly, since 2007 there are more private than state-owned fixed line operators in Africa (ITU database, 2007). Mobile operators are experiencing the highest growth in African economies. That is why investors are more interested in investment opportunities in the mobile market, which ha s become highly competitive. 4% of the global telecom investment is invested in Africa, which is the only region where investments are higher than the share of revenue. As a result, investors do not refrain from spending their money in African business.
C. The Development of Mobile Banking
benefit from mobile banking, transferring money by phone is also attractive for banked customers, it lowers the chances of loss and theft.
A key determinant of mobile money growth in Africa has been the rapid rise in mobile penetration, driven by low installation costs, expanded coverage and cheap phones. With mobile banking, people do not have to walk far to a bank or to take risk by carrying their money. It is also much cheaper to
do a mobile banking transaction than to use an Automatic Teller Machine2. Banking via mobile phones
offers features of automatic teller machines, internet kiosks and point-of-service devices like debit cards. Mobile operators introduced this cost-effective payment system in many African countries, the list of countries can be found in Appendix A. The network of mobile operators provides an exclusive reliable and efficient instrument for organizations to give out and collect money around the country, and around the world (The Gen, 2007). Customers can now pay bills, including electricity bills, school fees, public transport, drinks at a local pub and postpaid cell-phone bills. Furthermore, the mobile network can be used to repay microfinance loans , to buy micro insurance and to make salary payments to the poor, to distribute cash prizes in competitions and to pay out social expenses for example (Hughes and Lonie, 2007). It could also change the character of banking contacts themselves (Tiwari et al, 2007). With mobile banking it will become easier for people to transact money across borders and for micro banks to extend emergency loans to their clients. In addition, Hernandez-Coss et al. (2006) found that the cost of remittances between industrial and developing countries appears to be highest to Africa compared to other regions. Finance via mobiles can fill the gap by providing cost-effective, secure and fast remittance
services (Kapoor et al., 2007).There is also a trend towards the extension of individual loans with clear
2
responsibility for repayment which opens possibilities for microfinance with mobiles due to the personalized and individualized mode. These many possible services attract existing market players to replicate the success in other African markets. New market players have begun organizing mobile money services and aid organizations are increasingly supporting mobile money projects (GSMA, 2009).
III. Impact of mobile telephony: Literature review
A. Telecommunications
financial efficient markets. These consequences of telecommunication infrastructure are important spillovers and create externalities. The more users, the more value is derived by those users (Roller and Waverman, 2001). This has lead to an enormous explosion in the use of mobile telephony (Madden and Savage, 1998; Datta and Agarwal, 2004) and its importance for economic growth has attracted even more investors.
One of the earliest studies that investigated the relationship between telecoms penetration and economic growth, is the research of Hardy (1980). In this study data from 15 developed and 45 developing countries from the years 1960 through 1973 about telecom penetration and the number of radios are investigated. The results point out that the greater accessibility of telephones has a significant positive effect on GDP, while the access to radios did not. The different outcome was caused by the influence of network externalities of telephones from his point of view. Norton (1992) also found a significant and positive impact of telecommunication on growth. However, his exp lanation is the decline of transaction costs, the increase of effectiveness of investment markets and the increase in investment levels. Bayes et al. (1999) also emphasized the importance of telephones for economic activity; his results show that 50 percent of all telephone calls have economic purposes, like remittances, prices and opportunities. Subsequently, according to Leff (1984) telephones can realize economies of scale and scope in areas with more separated economic activity.
Waverman (1996) introduced a new structural economic model to prove the ‘endogeneity’ of the two-way relationship between GDP and the demand for telecoms. In 2001 Roller and Waverman found that about one-third of the economic growth of a cross-section of 21 OECD countries over the same period was stimulated by growth in telecommunications infrastructure.
B. Mobile telephony
increases the small network size and adds to the economy’s growth potential (Waverman, Meschi, Fuss, 2005). This is in line with economic theory of marginal diminishing returns, the marginal effect of one extra phone is the biggest where there is the least number of phones. Above a certain number of phones the marginal effect of one extra mobile telephone will become smaller and the effect can even become negative. The marginal product of each mobile phone will decline as the amount mobile telephones increases holding all other inputs constant. Mobile markets can get saturated and an overflow of mobile phones can be caused. Mobile phones can have such an impact because other types of communication such as postal systems, roads and fixed line networks are often poor in developing countries. Furthermore mobile phones have different externalities compared with fixed phones, they change people’s lifestyle and they are used to increase productivity (Waverman, Meschi, Fuss, 2005). Mobile telephony makes communication more flexible, reliable and cheap (Kapoor et al., 2007). The economic impact of mobile telephones is valuable and positive when it lowers information costs (Jensen, 2007).
C. The economic impact of mobile banking
markets more efficient. Getting cash into the hands of persons who can use it, is a priority. Since the creation of money, the capability to move it from A to B or the ‘velocity of money’ has been an essential cornerstone of economic activity (Hughes and Lonie, 2007). The ability to move mo ney around is rather difficult in underdeveloped regions or countries. The use of new technology, like mobile phones, is therefore an important tool to achieve more activity and financial access. Microfinance via mobile banking holds the potential to generate financial products that better fit the needs and it can create more flexible financial products to finance small businesses (Kapoor et al., 2007). Microfinance with the use of mobile banking can expand the nature of microfinance and can increase the reach of micro banks (Kapoor et al., 2007). The effectiveness of banks and other financial institutions can be highly improved by developing the information infrastructure (Honohan and Beck, 2007). Financing via mobile phones can therefore be a good opportunity to establish a more reliable and flexible information system for banks and their customers. Any financial service, like payment or savings, which enhance economic activity, requires scale, allocation and confidence, mobile phones are therefore useful. Especially in Africa, where corruption is high and where many people do not ha ve access to bank accounts, loans and other financial services.
to recent technological and financial innovations costs can be lowered (Cracknell, 2004; Truen et al.,
2005)3. Wright (2005) found that already in 2003, two-thirds of loans outstanding were through the use of
mobile banking. The growth of overall mobile penetration and mobile money will probably have a positive impact on social and economic conditions, also for those living on less than 2 dollar per day (Mas and Kumar, 2008). The findings of Ivatury and Pickens (2006) found that mobile banking providers must however build greater awareness of their services and must find the right balance between human interaction and technology to reach more low -income customers. Their study reveals that some people who are unemployed and earn no personal income seem to believe that they do not need banking services or cannot afford them.
So all in all, evidence (Cronin et al. (1991;1993a;1993b); Roller and Waverman (1996)) for the positive reverse causality of telecommunication and economic growth, mobile telephony and GDP (Roller and Waverman, 2001) and more specifically the positive reverse causality in developing countries (Waverman et al. (2005); Sridhar and Sridhar (2007)), make it interesting to test the first hypothesis. In addition, literature (Kapoor et al., 2007; Hughes and Lonie (2007); Mas and Kumar(2008); Cracknell (2004)) indicates a positive relationship between economic growth and mobile banking due to the extensive opportunities of mobile banking for economic activity, therefore the second hypothesis is tested.
D. Hypotheses
This research tries in the first place to prove the reverse causality of mobile phones and economic growth, subsequently it attempts to answer the question whether the effect of mobile phones in Africa
3
differs from the rest of the world and other developing countries. In table 1 an overview of empirical studies regarding the impact of (mobile) telecommunication on economic growth can be found; only these studies have used the same empirical model for their investigation as is used in this research. Roller and Waverman (2001) invented the model to estimate the influence of telephony on economic growth. The research of Waverman et al. (2005) further found that the mobile impact is positive and higher in developing countries. In addition, Sridhar and Sridhar (2007) found a bigger influence of mobile telephony than fixed telephony in developing countries. Due to the situation in Africa, the question rises whether the mobile impact in Africa is an exception compared to other developing countries. Previous research (Cracknell, 2004; Honohan and Beck, 2007) indicates that Africa is an exception in the area of development, technology and wealth. Africa has still the lowest total GDP in relation with the amount of inhabitants worldwide (ITU database, 2007) and is still far behind in the adoption of telecommunication infrastructure compared to other countries (World Development Indicators, 2009; ITU database 2007). According to many articles (Leishman 2009; ITU database; Jensen 2007; Jian and Sridhar 2003; Kathuria et al. 2009), the African mobile telephony is also of special interest due to the huge expansion and investments of mobile telephony throughout the continent. These conditions could make a difference for the influence of mobile telephony. As a result the main focus is the effect of mobile telephony in Africa.
The following hypothesis (H11) is formed: mobile telephony is related to economic growth and this
relationship is the strongest in Africa compared to the rest of the world and other developing countries.
The null hypothesis (H01) is that there is no mobile effect on economic growth and the mobile impact is
Table I
Study Period Region Tested Variable Coefficient
Roller and Waverman (2001) 1970-1990 OECD countries Fixed telephony 0.154***
Waverman, Meschi and Fuss (2005) 1996-2003 Developed and developing countries Mobile telephony in developed vs. developing countries 0.0003*** vs. 0.0006***
Sridhar and Sridhar (2007) 1990-2001 Developing countries Fixed vs. mobile telephony 0.14*** vs. 0.007***
Overview of empirical studies regarding the impact of telecommunications on economic growth
***Statistically significant at a level of 1 percent
Besides the impact of mobile telecommunication, this study will investigate the impact of mobile banking. This study is the first extensive attempt that looks at mobile banking empirically. Many articles discuss the enormous growth of mobile banking and its theoretical influence on economic growth in
Africa, but to the best of my knowledge no empirical articles have been written about mobile banking4.
This study tries to deal with the presence of mobile banking in some African countries due to its growing importance and the many interesting market opportunities nowadays . Since African economies modernize and need to grow, mobile banking can be crucial for doing business and creating efficient financial
markets. For the estimation of the impact of mobile banking, the following hypothesis (H12) is built : the
presence of a national mobile banking network lead s to economic growth in Africa. The null hypothesis
(H02) is that there is no relationship between mobile banking and economic growth. Mobile telephony and
mobile banking will be independent variables. H11 will be accepted in case mobile telephony has an effect
on economic growth and this relationship is the strongest in Africa. H12 will be accepted when the national
mobile banking network leads to more economic growth. Next to these variables, some macroeconomic and telecommunication control variables are added to the system of equations.
4
IV. Methodology
A. System of equations
Roller and Waverman (2001) invented a system of equations which consists of four different equations to determine the impact of telecoms infrastructure on economic growth (Roller and Waverman, 2001). Th is model captures the two-way causality of the investment of telecommunications and economic growth, it endogenizes telecom investment and mobile penetration. This model links telecom penetration and growth, factors that determine the supply and demand of telecom, to those that influence the change of telecom penetration. Table II includes the abbreviations of the variables, also the four equations are discussed.
Table II
GDP Gross Domestic Product
c constant
K Share of Capital Investment
LF Labor Force
MOB Mobile Penetration
TTP Total Telephone Penetration
MOBBK Mobile Banking
T Timetrend
GDPCAP GDP per Capita
RMC Revenue from Mobile Comunication
RTTC Revenu from Total Telephone Communication
WL Waiting list
TTI Total Telecom Investment
CHGMOB Change of Mobile Penetration
Abbreviations of the estimation ouput
The first equation is the production function or the output equation. This equation models the level of output (GDP) as a function of the share of capital investment, the total labor force and the total telephone (or mobile telephone) penetration rate.
related to a constant (c), the Capital Investment in US dollars (K), labor force (LF), the mobile phone penetration per 100 inhabitants (MOB) which indicates the one-way causal relationship from mobile infrastructure to economic development, a time trend (T) and an error term (e). In addition the equation will be adjusted for mobile banking, an extra dummy variable (MOBBK) is added to the equation. This will test the impact of a mobile banking network in a specific country. However to test for mobile banking equation (1.2) can also be used, mobile penetration is here excluded due to possible multicollinearity. All countries which have launched a mobile banking network in the past years will be included in the estimation. Moreover the variable for mobile banking is not transformed into a logarithm, because it is a dummy variable.
Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(3)*(T) + a(4)*(MOBBK) + e (1.2)
Furthermore the demand equation models the level of mobile telecoms penetration as a function of income (the level of GDP per capita) and mobile price (revenue per mobile subscriber).
log(MOB) = bC + b(1)*log(GDPCAP) + b(2)*log(RMC) + e (2)
The dependent variable is Total Telecom Investment (TTI), which is determined by the revenue of mobile communication (RMC) divided by the penetration rate (MOB). Ultimately, the fourth and last equation will estimate the change of growth of mobile penetration (CHGMOB) which is determined by Total Telecom Investment (TTI). This fourth equation indicates the relationship between the investment of telecom infrastructure and the change of mobile infrastructure. The last equation makes the model complete, adjustment for the reverse causality between the equations can now be done.
log(CHGMOB) = dC + d*log(TTI) + e (4)
The equations (2), (3) and (4) are necessary to endogenize mobile infrastructure and the investment of telecom infrastructure. Equations (2) and (3) represent the demand and supply of the mobile infrastructure, equation (4) shows the change of mobile infrastructure which is defined by the telecom investment. Subsequently, telecom investment is defined by equation (3), where mobile infrastructure is included and defines the size of telecom investment on its turn. This causal model is structured by a system of equations which identifies the two-way relationship of mobile penetration and economic growth.
B. Three Stages Least Squares
‘autonomously’ are labeled as instrumental variables (Greene, 2002). The exogenous or instrumental variables are timetrend, labor force, capital investment, mobile banking and the revenue of mobile telecom services. These variables should correlate with the dependent variables, but not correlate with the errors (Brooks, 2008). The rest is labeled as disturbances to get an econometric model. All equations are needed to estimate the effect of mobile phone penetration on economic growth (Greene, 2002). The regression has transformed the data into logarithms (Brooks, 2008).
To estimate the coefficients of this simultaneous -equations model, Three Stages Least Squares (3SLS) is used. 3SLS forms weights and reestimates the model using the estimated weighted matrix. The 3SLS is comparable with the Two Stages Least Squares (2SLS), which is the most common method to estimate simultaneous-equations models. 3SLS goes one step further by using 2SLS to estimate all coefficients of the entire system simultaneously (Zellner and Theil, 1962). In this extra step the OLS residuals of the equations are used to obtain a consistent estimate of the cross-equation covariance matrix. This will lead to a more efficient estimation (Woolridge, 2002), because 3SLS allows for non-zero covariances between the error terms. Deger (1986b) and Kusi (1994) have some concerns over the use of inter-country cross-section analysis, because it can only discover the average effect of a variable across countries. Therefore the database is split up in developing countries, Africa and the rest of the world.
V. Data
A. Databases
Telecommunication Indicators database 2008 contains time-series data from 1975-2006 annually for more than 200 economies. In most developing countries mobile phones are used less than a decade. Therefore panel data for the years 2000-2006 are included in the dataset (especially in Africa mobile phones are only been used since more or less 2000). The telecom data will be used to see to what extent mobile phones are in use around the world and particularly in Africa. The most important telecommunication variables are mobile telephone penetration, revenue of mobile telephone services and total telecom investment. Besides the comparison of the presence of mobile phones, this research will mainly investigate the different impact of mobile phones in Africa, other developing countries and the rest of the world. Therefore the World Development Indicators database and the new database on Financial Development and Structure of the Worldbank are included in the dataset to take demographic and macroeconomic data into account. The macroeconomic variables are GDP, GDP per capita, Labor Force and Capital Investment. These databases contain panel data from many economies including developed and developing countries for 1980 to 2007. Ultimately, the constructed dataset contains 190 countries from all over the world and from all continents; in appendix B the list of countries included in the dataset can be found. Before estimating the parameters of these variables, the nominal prices and numbers are converted to real prices (ITU, 2008) by using the Consumer Price Index for the prices and revenues of the telecom services and the Implicit Price Deflator for GDP and Total telecom investment.
table V contains information about other developing countries. Table III GDP K LF MOB P RMC TTI (*100000) (*100000) (*100000) (*100000) (*100000) (*100000) Mean 2.870.000 606.000 238 30,3 502 26.800 20.200 Median 186.000 39.900 47,3 18,5 103 1.630 1.620 Maximum 100.000.000 19.800.000 7.680 106,8 13.100 691.000 742.000 Minimum 244 0,72 0,92 0 2,42 0,00 0,66 Std. Dev. 11.000.000 2.230.000 855 30,3 1.610 88.600 71.100 Observations 286 286 286 286 286 286 286
Descriptive Statistics of all countries (incl. Africa) 2000-2006
Table IV GDP K LF MOB P RMC TTI (*100000) (*100000) (*100000) (*100000) (*100000) (*100000) Mean 3.990.000 866.000 2.160 81,3 478 37.000 29.200 Median 582.000 133.000 57,4 66,5 102 5.920 4.730 Maximum 100.000.000 19.800.000 128.000 267 13.100 691.000 742.000 Minimum 244 31,22 0,69 3,9 2,42 2,87 0,85 Std. Dev. 12.900.000 2.660.000 14.700 2,9 1.590 103.000 86.100 Observations 208 208 208 208 208 208 208
Descriptive Statistics the rest of the world 2000-2006
Table V GDP K LF MOB P RMC TTI (*100000) (*100000) (*100000) (*100000) (*100000) (*100000) Mean 645.000 225.000 604 7,6 1.120 17.300 9.110 Median 37.900 7.530 64,2 5,9 171 618 658 Maximum 8.640.000 3.520.000 7.680 25,6 13.100 267.000 128.000 Minimum 1.550 84,52 1,53 0 4,72 2,55 0,69 Std. Dev. 1.940.000 758.000 1.860 6,9 3.170 60.200 30.200 Observations 47 47 47 47 47 47 47
Descriptive statistics of other developing countries 2000-2006
operators is extracted from their websites: Vodacom, MTN, Glo Mobile, Djezzy, Mobinil, Vodafone,
Mobilis and Celtel, Safaricom, Telkom and Zain5. This information is used to include a dummy variable
for mobile banking into the system of equations. Note that information about mobile banking is only collected for Africa. Table VI contains information about the descriptive statistics of Africa.
Table VI
GDP K LF MOB P RMC TTI MOBBK
(*100000) (*100000) (*100000) (*100000) (*100000) (*100000) Mean 123.000 21.300 67 4,9 175 1.890 1.290 0,58 Median 22.800 5.090 47 2,1 113 219 212 1,00 Maximum 1.450.000 224.000 446 29,4 1.230 21.900 21.200 1,00 Minimum 1.400 0,72 1,41 0 4,40 0,00 0,64 0 Std. Dev. 282.000 44.500 75 6,5 209 4.650 3.440 0,50 Observations 79 79 79 79 79 79 79 79
Descriptive statistics of Africa 2000-2006
B. Descriptive Analysis
To analyze the structure of the dataset, some descriptive analyses are done. In Appendix C the growth of mobile telephony in Africa, other developing countries and the rest of the world are compared. The development of mobile phones has increased enormously during the last decade; Appendix C shows the difference in development of mobile telephony between Africa and other continents. As can be seen, Africa experiences the highest growth overall. This could be due to the enormous expansion of mobile phones in Africa, the more saturated mobile telecommunications markets in Europe and America or the relative small growth of mobile phones in other developing countries like in South-America and Asia.
Besides the progress of the mobile phone growth, the relationship between GDP and mobile growth must be investigated. In Figure 2 this relationship is shown: as can be seen the movements of mobile phone growth and GDP progress more parallel in Africa compared to the rest of the world. This
5 5
compared to the rest of the world.
Figure 2
The progress of GDP and Mobile Growth for Africa and the rest of the world
0 % 200% 400% 600% 800% 1000% 1200% 1400% 95% 1 0 5 % 1 1 5 % 1 2 5 % 1 3 5 % 1 4 5 % 1 5 5 % 1 6 5 % 1 7 5 % 2000 2001 2002 2003 2004 2005
% Mobile Phone Growth vs 1999
% GDP Growth vs 1999
Year
GDP & Mobile Growth
Africa GDP
Rest of the World GDP
Africa Mobile growth
Rest of the World GDP
LF K MOB MOBBK RMC TTI T GDP LF 1,00 0,41 0,00 0,54 0,46 0,37 0,17 0,41 K 0,41 1,00 0,53 0,27 0,97 0,93 0,13 0,99 MOB 0,00 0,53 1,00 -0,19 0,59 0,47 0,05 0,51 MOBBK 0,54 0,27 -0,19 1,00 0,29 0,29 0,23 0,31 RMC 0,46 0,97 0,59 0,29 1,00 0,91 0,15 0,98 TTI 0,37 0,93 0,47 0,29 0,91 1,00 0,16 0,94 T 0,17 0,13 0,05 0,23 0,15 0,16 1,00 0,12 GDP 0,41 0,99 0,51 0,31 0,98 0,94 0,12 1,00
Correlationmatrix of mobile telephony in Africa 2000-2006
C. Variables
Since the model is a system of equations that are interdependent, the variables need to be indicated as exogenous or endogenous. The most important dependent and also endogenous variable is GDP; the gross domestic product in US dollars. It is the market value of all final goods and services made within the borders of a nation in a year. Another endogenous variable is total mobile penetration, which measures the number of mobile phones per 100 inhabitants in a country in a year and is calculated by dividing the number of mobile subscribers by the population and multiplying by 100. Total telecom investment is the third endogenous variable also referred to as annual capital expenditure: the gross annual investment in telecom (including fixed, mobile and other services) for acquiring property and network. The last and fourth endogenous variable is the change of growth of telecom penetration for mobile phones. The change of growth in telecom penetration is measured by subtracting the growth of the previous year from the growth of this year.
such as cellular and radios, and excludes revenues from non-telecommunication services. Revenue consists of mobile communication services earnings during the financial year under review. At last an exogenous dummy variable will be included to test the impact of mobile banking in African countries. Note that not all countries have the same extensive mobile banking network and have launched a mobile banking application at the same time. Therefore the mobile banking variable is not linked to certain years.
VI Results
In this chapter the outcomes of the system of equations and more specifically the variables regarding the impact of mobile telephony and mobile banking on GDP will be discussed. The impact of mobile telephony will be measured and compared between four different groups; all countries, the rest of the world (excluding Africa), other developing countries and Africa. The abbreviations of the estimation output can be found in Table II.
A. Mobile telephony
these countries. As literature (Leishman, 2009) confirms, mobile telephony has expanded extremely, particularly in Africa. However, there are just a few other developing countries (Bangladesh and Philippines for example) where mobile penetration has exploded like in Africa. As a result the mobile influence is less clear, also information about mobile penetration is missing for several developing countries. In Appendix H a comparison for the mobile impact between developing countries and developed countries is made. Theory (Waverman et al., 2005) suggests that the impact in developing countries is twice as large compared to developed countries in a positive sense. However developing countries and developed countries have both insignificant and negative results, the impact in developing countries is more positive compared to the effect in developed countries.
Table VIII
Independent variables Coefficient Std. error Coefficient Std. error Coefficient Std. error Coefficient Std. error
Production Function : Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(4)*log(MOB) +a(5)*(T) Dependent Variable: GDP
c 1,60 0.80** 1,68 0.75** 4,44 1.39*** 3,38 1.15*** K 1,33 0.23*** 1,32 0.14*** 0,59 0,40 0,66 0.09*** LF -0,42 0,27 -0,40 0.14*** 0,36 0,42 0,31 0.10*** MOB -0,35 0,31 -0,33 0.18* -0,08 0,14 0,36 0.13*** T 0,00026 0,00065 -0,00042 0,00047 0,00120 0,00270 0,00162 0.00072** Nr. of Observations R ²
Rest of the World Other Developing Countries Africa
Overview of estimation output of mobile penetration 2000 - 2006
0,88 0,94 0,96 0,82
317 124 56 113
All Countries
Statistically significant at a level of 10 percent*, 5 percent** or 1 percent***
However the group ‘the rest of the world’ (excluding Africa) shows a positive and significant sign at a level of 1% for Capital Investment (K) which is consistent with theory (Sridhar et al., 2007). The parameters of Labor Force (LF) and Mobile Penetration (MOB) show also a significant, but unexpected negative sign -0.40 and -0.33, this indicates that Labor Force and Mobile Penetration decrease GDP or economic growth. This is contrary to theory; Labor Force and Mobile Penetration should enhance economic growth. An explanation for this negative sign of mobile penetration could be again the saturation of the mobile markets in the developed world. After a certain point, growth reduces and can even lead to a decrease of GDP due to an overflow or technological innovation. Figure 2 supports this outcome, here it can be seen that mobile growth in the rest of the world does not behave pa rallel with GDP growth.
Moreover Table VIII contains the estimation output of Africa. The outcomes for the parameters of Africa are all positive and significant at a level of 1%. This supports the suggested expectation about the high mobile impact in Africa. Labor Force (LF), Capital Investment (K) and Mobile Penetration (MOB) all stimulate economic growth. The coefficient of mobile penetration is 0.36, this indicates that mobile telephony enhances economic growth. This is consistent with Waverman et al. (2005), they consider that mobile phone should influence GDP positively due the reduction of transaction costs, including decision making regarding the production of goods and services. This is in line with the estimations of Sridhar and Sridhar (2007), the effect of mobile penetration is more positive than the impact of total telephone penetration. One extra mobile telephone per 100 inhabitants leads to 0.36 more GDP. R squared is also relatively high; the explanatory variables do explain much about GDP. In Figure 2 this positive effect is confirmed, economic growth and mobile growth behave parallel.
Until now, the outcomes of the mobile impact are analyzed separately for each group. To test the
null hypothesis (H01), the estimation output of these four groups must be compared. The null hypothesis
(H01) states that the mobile impact in Africa has no different effect compared to the groups ‘all countries’,
without any statistical testing or interaction term. The test-statistic assumes that two sample sizes are
unequal and the variances are assumed to be different6. So the calculation of the test-statistic is done for
Africa compared with all countries, the rest of the world and subsequently with other developing countries. The test-statistic is larger than the t derived from its critical value for every comparison, as a
result the null hypothesis (H01) can be rejected
7
. There is enough statistical evidence to accept that mobile telephony has the biggest impact on economic growth in Africa compared to all countries, the rest of the world and other developing countries.
B. Mobile banking
In Table IX the estimation output of mobile banking can be found, again Table IX contains only the first equation of the system due to its importance. The rest of the system which is also used to estimate
the outcomes, can be found in Appendix J. H02 is analyzed and will be rejected if mobile banking
enhances national economic growth in Africa. In Table IX there are two different groups; mobile banking with mobile penetration and mobile banking without mobile penetration. Both estimations can be found to double check the possibility of multicollinearity between the variable mobile penetration and mobile banking, although Table VII does not show any multicollinearity which was expected. These collinear variables could lead to insignificance, because both variables are related to the number of mobile phones. Nevertheless econometricians (Greene, 2002; Brooks, 2008) argue that this brings more problems than they solve. However one of the collinear variables can be dropped to get a better estimation , therefore both estimations are included in Table IX.
6
Hypothesis testing of paramaters of different groups can be done with the following formula:
where
7
The null hypothesis (H01 ) will be rejected if T(all countries; rest of the world; other developing countries) > t (a,8) , here
Independent variables Coefficient Std. error Coefficient Std. error
Production Function : Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(4)*log(MOB) + a(5)*(T) + a(6)*(MOBBK) Dependent Variable: GDP
c 4,28 1.51*** 3,55 1.56** K 0,79 0.10*** 0,91 0.07*** LF 0,08 0,15 -0,02 0,14 MOB 0,24 0.14* - -MOBBK 0,61 0.26*** 0,62 0.28** T 0,000024 0,0007 0,0002 0,0077 Nr. of Observations R ²
Overview of mobile banking estimation output in Africa 2000 - 2006
0,82 0,80
8 5 8 5
Mobile Banking & Penetration Mobile Banking (function excludes a(4)*log(MOB) )
Statistically significant at a level of 10 percent*, 5 percent** or 1 percent***
As can be seen, the estimated parameter of Mobile Penetration (MOB) with mobile banking is significant at a level of 10% and positive (0.24) as expected. Furthermore Capital Investment (K) and Mobile Banking (MOBBK) are positive and significant at a level of 1%. The estimated parameter for Mobile Banking (MOBBK) is 0.61. Additionally R squared is considerably high, so the equation does explain quite a lot of the economic development. As a result a mobile banking network stimulates
economic development, so H02 can be rejected.
rejected with certainty, Mobile Banking (MOBBK) creates economic development.
To summarize, mobile penetration and mobile banking both have a positive and significant estimated parameter (0.36 and 0.62) in Africa. The remaining interesting question is whether mobile penetration or mobile banking has a bigger impact on economic development. It is expected that there is no difference between the effect of mobile banking and mobile penetration. Again a test-statistic can be calculated to compare the estimated outcomes. The test-statistic is larger than the t derived from its critical
value, therefore the expectation can not be rejected8. There is not enough or no statistical evidence to
reject that mobile banking and mobile penetration have a comparable effect on GDP.
VII. Conclusion
Previous research (Waverman et al., 2005) has found that mobile impact on GDP is twice as large in developing countries compared to developed countries. Moreover, 10 mobile phones per 100 people in a typical developing country will lead to an extra half percent increase of GDP (Waverman et al., 2005). This research suggests that the mobile impact in Africa is positive and strongest compared to developed countries and other developing countries. The mobile impact in developing countries is large due to the lack of communication and information system, the easy way to buy and use a mobile phone and the huge investments that stimulate economic growth and the expansion of mobile phones. Besides these arguments, an extra tool of mobile phones, mobile banking, will speed up the mobile and economic growth even more. Especially in developing countries, like in Africa, mobile banking will make access to financial services for a large number of unbanked people possible. The mobile impact is the biggest in countries where the lack of infrastructure and financial system is the largest (Jensen et al., 2007). The use
long distances for example. The opportunities for mobile banking are immense.
In this study two hypotheses are empirically tested to answer the two following questions. Has mobile telephony a bigger impact on economic growth in Africa than in the rest of the world and other developing countries? Secondly, does mobile banking enhance national economic growth in Africa? These questions are tested for all countries, the rest of the world (excluding Africa), Africa and other developing countries for the perio d 2000 to 2006. On the first question, regarding the mobile impact on economic growth in Africa compared to the rest of the world, the following results are found. The groups ‘the rest of the world’ and ‘Africa’ show a significant estimated parameter for mobile penetration. The estimated coefficient for the rest of the world shows a negative relationship with economic growth, this could be due to the saturation of mobile markets in the developed world. The estimated parameter for Africa is positive, which is in line with theory (Waverman et al., 2005; Roller and Waverman 2001). Consequently, mobile telephony enhances economic growth in Africa. However, to answer the question whether the effect of mobile telephony is larger in Africa compared to all countries, the rest of the world and other developing countries, a test-statistic is used. The suggestive bigger impact in Africa can be accepted, so the null hypothesis is rejected.
The second question was regarding the effect of mobile banking on economic growth in Africa. The estimation output shows positive and significant results for the mobile money impact in Africa, therefore the null hypothesis regarding mobile banking can be rejected These outcomes support the expectations, as Iqbal Quadir said ‘Whatever people do with the phones, it’s good for them, good for the country.’ Mobile banking enhances economic activity and thus influences national economic growth in Africa positively , even in case the mobile penetration is included in the system of equations.
regarding mobile telephony. In addition, this research could not obtain reliable information regarding mobile banking from established databases, like the ones of the Worldbank and the International Telecommunication Union, due to the recent implementation of mobile banking. Data concerning mobile banking are at this moment too expensive to get access to due to the high competition element. Since mobile operators, aid organizations and new market players are trying to get a foot on the ground in many African countries, the competitive mobile money market is too volatile and variable.
The results that are included in this research leave enough space to investigate some interesting aspects of mobile telephony and mobile banking in particular. Especially mobile banking should be investigated extensively in the near future. Since mobile banking creates access to financial services, it would be interesting to investigate the relationship between financial access and mobile banking. If this relationship can be established, it will force banks to adjust their strategy and to venture in branchless banking. Such a development would drastically change the traditional banking system, especially in developing countries.
REFERENCES
Adams, Richard H. Jr., 2003, Economic growth, inequality, and poverty: findings from a new data set, Policy Research Working Paper Series 2972, Worldbank
Beck, Thorsten, Asli Demirgüç -Kunt and Maria S. Martinez Peria, 2007b, Reaching out: Access to and Use of Banking Services across Countries, Journal of Financial Economics 85(1), 234-266 Brooks, Chris, 2008, Introductory Econometrics for Finance, Cambridge University Press, London Cracknell, David , 2004, Electronic Banking for the Poor: Panacea, Potential and Pitfalls, MicroSave,
Growth, Telecommunications Policy 17, 677-696
Cronin, Francis J., Edwin B. Parker, E.K. Colleran and Mark A. Gold, 1991, Telecommunications infrastructure and economic growth: an analysis of causality, Telecommunications Policy 15, 529-534 Cronin, Francis J., Edwin B. Parker, E.K. Colleran and Mark A. Gold, 1993b, Telecommunications
Infrastructure Investment and Economic Development, Telecommunications Policy 17, 415-430 Datta, Anusua and Sumit Agarwal, 2004, Telecommunications and economic growth: a panel data
approach, Applied Economics 36(15), 1649-1654
Deger, Saadet, 1986b, Military Expenditure and Growth in less developed countries, Journal of Conflict
Resolution 27(2), 335-353
Demirgüç-Kunt, Asli, Thorsten Beck and Patrick Honohan, 2007, Finance for all? Policies and Pitfalls in
expanding access, A Worldbank Policy Research Report, Washington D.C., November
Eggleston, Karen, Robert Jensen and Richard Zeckhauser, 2002, Information and communication
technologies, markets and economic development, Department of Economics Working Paper, Tufts University
Fama, Eugene, 1970, Efficient Capital Markets: A Review of Theory and Empirical Work, Journal of
Finance 25(2), 383–417
Fama, Eugene, 1976, Foundations of Finance, New York
Greene, William H., 2002, Econometric Analysis, New jersey: Prentice-Hall
Grossman, Gene M. and Elhanan Helpman, 1994, Endogenous Innovation in the Theory of Growth,
Journal of Economic Perspectives 8(1), 23-44
GSM Association, 2006, GSM Hits Two Billion Milestone, June 16 GSM Association, 2009, Mobile Money for the Unbanked, March
Hardy, Andrew, 1980, The role of the telephone in economic development, Telecommunications Policy 4(4), 278-286
Honohan, Patrick, 2006, Household Financial Assets in the Process of Development, Policy Research Working Paper 3965, World Bank, Washington D.C.
Honohan, Patrick and Thorsten Beck, 2007, Making Finance Work in Africa, World Bank,Washington DC Hughes, Nick and Susie Lonie, 2007, M-Pesa: Mobile Money for the ‘Unbanked’, Innovations (1-2),
Winter/Spring
International Telecommunication Union, 2007, Telecommunication/ICT markets and trends in Africa Ivatury, Gautam, 2006, Using technology to Build Inclusive Financial Systems , Focus Note 32,
Worldbank and Consultative Group to Assist the Poor, Washington DC
Ivatury, Gautam and Ignacio Mas, 2008, The Early Experience with Branchless Banking, Consultative Group to Assist the Poor, Washington D.C.
Ivatury, Gautam and Mark Pickens, 2006, Mobile Phone Banking and Low-Income Customers -Evidence
From South Africa, Consultative Group to Assist the Poor, WorldBank and United Nations Foundation
Jain, P., and Varadharajan Sridhar, 2003, Analysis of Competition and Market Structure Of Basic Telecommunication Services in India, Communications & Strategies 52(4), 271-293
Jha, Raghbendra and Sumit K. Majumdar, 1999, A matter of connections: OECD Telecommunications Sector Productivity and the Role of Cellular Technology Diffusion, Information Economics and
Policy 11, 243-269
Kapoor, Mudit , Jonathan Morduch and Shamika Ravi, 2007, From Microfinance to M-Finance,
Innovations 2(1-2), 97-106
Kathuria, Rajat, Mahesh Uppal and Mamta, 2009, An econometric analysis of the impact of mobile, The
Policy Paper Series, January
Kusi, Newman K., 1994, Economic growth and defence spending in developing countries, Journal of
Conflict Resolution 38, 152-159
Change 32(2), 255-276
Leishman, Paul, 2009, Financially connecting Africa trough Mobile, Mobile Money for the Unbanked , March
Leishman, Paul, 2009, Mobile Money: a $5 Billion Market Opportunity, Mobile Money for the Unbanked, GSMA, March
Madden, Gary and Scott J. Savage, 1998, CEE telecommunications investment and economic growth,
Information Economics and Policy 10, 173-195
Mas, Ignacio and Kabir Kumar, 2008, Banking on Mobiles: Why, How, for Whom?, Consultative Group
to Assist the Poor, June
Norton, Seth W., 1992, Transaction Costs, Telecommunications, and the Microeconomics of Macroeconomic Growth, Economic Development and Cultural Change 41(1), 175-196
O’Sullivan, Arthur and Steven Sheffrin , 1996, Economics: Principles in action , Pearson Prentice Hall, Upper Saddle River, New Jersey
Parker, Stephen, 2005, W ireless Technology helping to shrink digital divide, Voice of America, May Peachey, Stephen and Alan Roe, 2006, Access to Finance: Measuring the Contribution of Saving Banks,
World Savings Banks Institute, Brussels, Belgium
Roller, Lars-Hendrik. And Leonard Waverman, 1996, Telecommunications Infrastructure and Economic
Development: A Simultaneous Approach, Social Science Research Centre Berlin, Discussion Paper
Roller, Lars-Hendrik and Leonard Waverman, 2001, Telecommunications Infrastructure and Economic Development: A simultaneous Approach, The American Economic Review 4, 909-923
Sridhar, Kala S. and Varadharajan Sridhar, 2007, Telecommunication Infrastructure and economic growth: evidence from developing countries, Applied Econometrics and International Development 7(2), 37-61
Sullivan, Nicholas P., 2007, You can hear me now, Jossey-Bass, San Francisco
Role of Innovative Business Solutions in Generating Competitive Advantage, Innovation and
Knowledge Management, 886–894.
Truen, Sarah, Richard Ketle y, Hennie Bester, Ben Davis, Hugh- David Hutcheson, Kofi Kwakwa and Sydney Mogapi, 2005, Supporting Remittances in Southern Africa: Estimating Market Potential and
Assessing Regulatory Obstacles, Consultative Group to Assist the Poor and FinMark Trust,
Washington D.C.
Waverman, Leonard, Meloria Meschi and Melvyn Fuss, 2005, The impact of Telecoms on Economic Growth in Developing Countries, London Business School, LECG, University of Toronto
Wong, Poh-Kam, 2002, ICT Production and Diffusion in Asia: Digital Dividends or Digital divide?,
Information Economics and Policy 14, 167-187
Wood, Christina and Sean de Cleene, 2004, Sustainability Banking in Africa, Worldbank,Washington DC Wooldridge, Jeffrey M., 2002, Econometric Analysis of Cross Section and Panel Data, Cambridge MA
MIT Press
World Development Indicators, 2009, States and Markets, 265 – 269
Wright, Graham, 2005, Designing Innovative Products, Processes, and Channels for the Promotion of
Microfinance, Conference Paper, MicroSave, Nairobi, Kenya
Zellner, Arnold and Theil, H., 1962, Three-Stage Least Squares: Simultaneous Estimation of
Appendix A
List of African countries with mobile banking networks
Africa Burkina Faso Egypt Madagascar Senegal Tanzania
Chad Gabon Mali Seychelles Uganda
Congo Ghana Niger Sierra Leone Zambia
Cote d'Ivoire Kenya Sao Tome and Principe Sudan Zimbabwe
Appendix B
List of countries included in the dataset
Africa Angola Congo Guinea Mozambique Sudan
Algeria Cote d'Ivoire Guinea-Bissau Namibia Swaziland
Benin Cote d'Ivoire Kenya Niger Tanzania
Botswana Djibouti Lesotho Nigeria Togo
Burkina Faso Egypt Liberia Rwanda Tunisia
Burundi Equatorial Guinea Libya Sao Tome and Principe Uganda
Cameroon Eritrea Madagascar Senegal Zambia
Cape Verde Ethiopia Mali Seychelles Zimbabwe
Central African Rep. Gabon Mauritania Sierra Leone
Chad Gambia Mauritius Somalia
Comoros Ghana Morocco South Africa
Asia Afghanistan Georgia Kazakhstan Nepal Thailand
Armenia Hong Kong, China Korea (Rep. of) Oman Turkey
Azerbaijan India Kyrgyzstan Pakistan Turkmenistan
Bahrain Indonesia Lao P.D.R. Philippines United Arab Emirates
Bangladesh Iran (Islamic Rep. of) Lebanon Qatar Uzbekistan
Bhutan Iraq Macao, China Saudi Arabia Viet Nam
Brunei Darussalam Israel Malaysia Singapore Yemen
Cambodia Kuwait Maldives Sri Lanka
China Japan Mongolia Syria
Cyprus Jordan Myanmar Tajikistan
Europa Albania Denmark Ireland Netherlands Spain
Andorra Estonia Italy Norway Sweden
Austria Finland Latvia Poland Switzerland
Belarus France Liechtenstein Portugal T.F.Y.R. Macedonia
Belgium Germany Lithuania Romania Ukraine
Bosnia and Herzegovina Greece Luxembourg Russia United Kingdom
Bulgaria Hungary Malta Slovak Republic
Czech Republic Iceland Moldova Slovenia
Antigua and Barbuda Bermuda Dominican Rep. Haiti Panama
Aruba Canada El Salvador Honduras Puerto Rico
Bahamas Costa Rica Greenland Jamaica St. Vincent and the Grenadines
Barbados Cuba Grenada Mexico United States
Belize Dominica Guatemala Nicaragua Virgin Islands (U.S.)
Oceania American Samoa French Polynesia Marshall Islands Papua New Guinea Tonga
Australia Guam New Caledonia Samoa Vanuatu
Fiji Kiribati New Zealand Solomon Islands Uruguay
Argentina Chile Guyana Peru Venezuela
Bolivia Colombia Netherlands Antilles Suriname
Brazil Ecuador Paraguay Trinidad and Tobago
North America
Comparison of mobile growth between continents from 1999-2006 9 0 % 110% 130% 150% 170% 190% 210% 230% 250% 1999 2000 2001 2002 2003 2004 2005 2006
% Growth versus previous year
Year Continental mobile phone growth
Africa Asia Europe North America Oceania South America Appendix D GDP K LF MOB RMC TTI GDP 1,00 0,99 0,23 0,19 0,92 0,94 K 0,99 1,00 0,29 0,20 0,95 0,96 LF 0,23 0,29 1,00 -0,07 0,39 0,31 MOB 0,19 0,20 -0,07 1,00 0,22 0,19 RMC 0,92 0,95 0,39 0,22 1,00 0,88 TTI 0,94 0,96 0,31 0,19 0,88 1,00
Correlationmatrix of mobile telephony in all countries 2000-2006
Appendix E GDP K LF MOB RMC TTI GDP 1,00 0,98 0,13 0,12 0,93 0,93 K 0,98 1,00 0,26 0,26 0,96 0,95 LF 0,13 0,26 1,00 0,99 0,28 0,33 RMC 0,93 0,96 0,28 0,28 1,00 0,89 TTI 0,93 0,95 0,33 0,32 0,89 1,00
GDP K LF MOB RMC TTI GDP 1,00 1,00 0,99 0,52 0,99 0,99 K 1,00 1,00 0,99 0,52 1,00 0,99 LF 0,99 0,99 1,00 0,50 0,99 1,00 MOB 0,52 0,52 0,50 1,00 0,54 0,51 RMC 0,99 1,00 0,99 0,54 1,00 0,99 TTI 0,99 0,99 1,00 0,51 0,99 1,00
Correlationmatrix of mobile telephony in other developing countries 2000-2006
Appendix G
Independent variables Coefficient Std. error Coefficient Std. error Coefficient Std. error Coefficient Std. error
Production Function : Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(4)*log(MOB) +a(5)*(T) Dependent Variable: GDP
c 1,60 0.80** 1,68 0.75** 4,44 1.39*** 3,38 1.15*** K 1,33 0.23*** 1,32 0.14*** 0,59 0,40 0,66 0.09*** LF -0,42 0,27 -0,40 0.14*** 0,36 0,42 0,31 0.10*** MOB -0,35 0,31 -0,33 0.18* -0,08 0,14 0,36 0.13*** T 0,00026 0,00065 -0,00042 0,00047 0,00120 0,00270 0,00162 0.00072** Nr. of Observations R ²
Demand Function : log(MOB) = c + b(1)*log(GDPCAP) + b(2)*log(RMC) Dependent Variable: MOB
c -4,45 1.02*** -3,65 1.19*** -8,97 2.32*** -5,86 1.31***
GDPCAP 0,82 0.14*** 0,67 0.12*** 1,70 0.42*** 0,42 0.14***
RMC 0,039 0,079 0,074 0,077 0,040 0,139 0,230 0,083
Nr. of Observations R ²
Supply Function : log (TTI) = c + c(1)*(log (RMC) / log(MOB)) Dependent Variable: TTI
c 0,36 2,01 1,01 2,59 1,60 3,20 3,04 2,19
RMC / MOB 1,103 0.12*** 1,083 0.16*** 0,964 0.19*** 0,833 0.132***
Nr. of Observations R ²
Growth Function : log(CHGMOB) = c + d(1)*log(TTI) Dependent Variable: CHGMOB
c -2,26 1,50 0,57 1,81 -2,88 2,07 -2,71 2,07
TTI 0,241 0.078*** 0,110 0,093 0,237 0.112** 0,203 0.120*
Nr. of Observations R ²
Significant at a level of 10 percent*, 5 percent** or 1 percent***
Rest of the World Other Developing Countries Africa
171 62 25 0,21 0,14 0,29 0,04 290 110 48 95 57 0,35 0,40 0,49 0,32 317 124 56 112 0,54 0,60 -0,36 0,43
Overview of estimation output of mobile penetration 2000 - 2006
0,88 0,94 0,96 0,82
317 124 56 113
Coefficient Std. Error Coefficient Std. Error
Production Function : Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(4)*log(MOB) +a(5)*(T) Dependent Variable: GDP
c 4,29 2,09** 3,99 0,71*** K 1,42 0,31*** 0,57 0,10*** LF -0,41 0,31 0,40 0,13*** MOB -1,61 1,14 -0,04 0,11 T -0,001 0,001 0,0005 0,0004 Nr. of Observations R ²
Demand Function : log(MOB) = c + b(1)*log(GDPCAP) + b(2)*log(RMC) Dependent Variable: GDP
c 1,05 0,82 -3,95 1,68***
GDPCAP 0,30 0,06*** 0,53 0,28*
RMC 0,00 0,04 0,09 0,11
Nr. of Observations R ²
Supply Function : log (TTI) = c + c(1)*(log (RMC) / log(MOB)) Dependent Variable: GDP
c 4,44 1,45*** -1,52 2,67
RMC / MOB 0,93 0,09*** 1,12 0,16***
Nr. of Observations R ²
Growth Function : log(CHGMOB) = c + d(1)*log(TTI) Dependent Variable: GDP
c 4,57 1,18*** -1,41 1,74
TTI -0,13 0,06** 0,08 0,10
Nr. of Observations R ²
Significant at a level of 10 percent*, 5 percent** or 1 percent***
-0,14 0,11 0,62 0,35 111 89 0,33 0,26 163 127 0,84 0,92 173 144
Overview of estimation output for mobile telephony 2000-2006
Developed countries Developing countries
Coefficient Std. Error Coefficient Std. Error Coefficient Std. Error Coefficient Std. Error Production Function : Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(4)*log(MOB) +a(5)*(T) Dependent Variable: GDP
c 3,39 0,88*** 2,55 0,69*** 4,24 1,62*** 1,88 1,28 K 0,85 0,24*** 1,04 0,22*** 0,68 0,43 0,79 0,12*** LF 0,04 0,24 -0,13 0,21 0,26 0,45 0,25 0,11*** TTP 0,40 0,50 0,04 0,40 -0,18 0,22 0,16 0,18 T -0,001 0,001 -0,001 0,001 0,002 0,003 0,001 0,001 Nr. of Observations R ²
Demand Function : log(MOB) = c + b(1)*log(GDPCAP) + b(2)*log(RMC) Dependent Variable: GDP
c 13,94 2,40*** 15,46 4,46*** 2,02 5,59 -2,23 3,67
GDPCAP -1,01 0,21*** -1,48 0,38*** 0,32 0,91 0,31 0,26
RTTC 0,12 0,17 0,21 0,30 0,35 0,42 0,56 0,22***
Nr. of Observations R ²
Supply Function : log (TTI) = c + c(1)*(log (RMC) / log(MOB)) Dependent Variable: GDP
c 2,24 4,18 8,73 4,73** 2,94 4,28 -1,68 3,61
WL 1,64 0,47*** 0,35 0,28 0,20 0,27 0,63 0,18***
RMC / MOB -0,08 0,37 0,34 0,37 0,72 0,27*** 0,73 0,23***
Nr. of Observations R ²
Growth Function : log(CHGMOB) = c + d(1)*log(TTI) Dependent Variable: GDP
c -1,73 0,97* -1,05 2,04 -4,86 1,91*** -5,21 1,54***
TTI 0,16 0,05*** 0,22 0,11** 0,37 0,10*** 0,33 0,09***
Nr. of Observations R ²
Significant at a level of 10 percent*, 5 percent** or 1 percent***
Overview of estimation output of total telephony 2000-2006
All countries Rest of the world Other developing countries Africa
363 135 62 94 0.93 0,97 0,96 0,84 251 84 41 67 0.32 0,24 0,01 -0,07 181 60 36 60 -1,90 0,02 0,35 0,14 226 70 31 54 0,21 0,13 0,20 0,43
Independent variables Coefficient Std. error Coefficient Std. error
Production Function : Log (GDP) = c + a(1)*log (K) + a(2)*log (LF) + a(4)*log(MOB) + a(5)*(T) + a(6)*(MOBBK) Dependent Variable: GDP
c 4,28 1.51*** 3,55 1.56** K 0,79 0.10*** 0,91 0.07*** LF 0,08 0,15 -0,02 0,14 MOB 0,24 0.14* - -MOBBK 0,61 0.26*** 0,62 0.28** T 0,000024 0,0007 0,0002 0,0077 Nr. of Observations R ²
Demand Function : log(MOB) = c + b(1)*log(GDPCAP) + b(2)*log(RMC) Dependent Variable: MOB
c -4,12 1.47*** -4,11 1.47***
GDPCAP 0,51 0.13*** 0,51 0.13***
RMC 0,110 0,090 0,110 0,090
Nr. of Observations R ²
Supply Function : log (TTI) = c + c(1)*(log (RMC) / log(MOB)) Dependent Variable: TTI
c 3,44 2.43* 3,46 2,43
RMC / MOB 0,810 0.15*** 0,810 0.15***
Nr. of Observations R ²
Growth Function : log(CHGMOB) = c + d(1)*log(TTI) Dependent Variable: CHGMOB
c -4,08 1.51*** -4,08 1.51***
TTI 0,250 0.09*** 0,250 0.09***
Nr. of Observations R ²
Significant at a level of 10 percent*, 5 percent** or 1 percent***
5 0 0,37
Overview of mobile banking estimation output in Africa 2000 - 2006
0,82 0,80
8 5 8 5
Mobile Banking & Penetration
0,37 5 0 0,43 8 5 8 5 0,43
Mobile Banking (function excludes a(4)*log(MOB) )
0,30 7 0 7 0
