1

Towards an understanding of ICT access and use in AfricaCRASA Workshop on NGN Interconnection, Lilongwe, March 2013

Dr. Alison Gillwald Executive Director: Research ICT Africa

Adjunct Professor - University of Cape Town,

GSB, Management of Infrastructure Reform and Regulation Programme

Network a response to the research

vacuum on the continent in relation to ICT policy

and regulation and dearth of capacity to respond to redress it.

3

Context! Demand side indicators as policy outcomes! Market structure and the institutional ! arrangements! Users response to services/applications ! offered by/on networks at particular price and ! quality

Household & Individual ICT survey‣ Lack of data - decision relevant data for

ICT policy making and regulation

‣ PARTNERSHIP ON MEASURING ICT FOR DEVELOPMENT: delivers all indicators required by the Partnership for household, individuals, and businesses

‣ COST EFFECTIVE: Using Enumerator Areas (EA) of national census sample frames and samples households, small business simultaneously minimizes costs.

‣ SCOPE :Apart from delivering ICT indicators required by international bodies the survey delivers data and analysis for several regulatory functions such as pricing regulation, number portability and universal access.

‣ LINKAGES:explains interactions between households, individuals and informal and small businesses on ICT access and usage.

5

Household & individual ICT survey methodology

• Step 2: EAs will be sampled for each stratum using probability proportional to size (PPS).

• Step 3: For each EA two listings will be compiled, one for house-holds and one for businesses. The listings serve as sample frame for the simple random sections.

• Step 4: 24 Households and 10 businesses will be sampled using simple random sample for each selected EA.

• Step 5: From all household members 15 years or older or visitor staying the night at the house one will be randomly selected based on simple random sampling.

HH SampleHH Sample

TotalHH Sample Business Sample

Total

Botswana 900 400 1,300

Cameroon 1,200 500 1,700

Ethiopia 1,600 600 2,200

Ghana 1,200 500 1,700

Kenya 1,200 500 1,700

Mozambique 1,200 500 1,700

Namibia 900 400 1,300

Nigeria 1,600 600 2,200

Rwanda 1,200 500 1,700

South Africa 1,600 600 2,200

Tanzania 1,200 500 1,700

Uganda 1,200 500 1,700

Tunisia 1,200 500 1,700

Total 15,300 6,200 21,500

Sample SizeThe desired level of accuracy for the survey was set to a confidence level of 95% and an absolute precision (relative margin of error) of 5%. The population proportion P was set conservatively to 0.5 which yields the largest sample size (Lwanga & Lemeshow, 1991). The minimum sample size was determined by the following equation (Rea & Parker, 1997):

n =Z

ap(1 ! p)

Cp

"#$

%&'

2

=1.96 0.5(1 ! 0.5)

0.05

"#$

%&'

2

= 384

Inserting the parameters for the survey yields the minimum sample size for simple random sampling. Due to the sampling method chosen for the survey the minimum sample size has to be multiplied by the design effect variable (Lwanga & Lemeshow, 1991). In the absence of empirical data from previous surveys that would have suggested a differed value, the default value of two was chosen for the design effect (UNSD, 2005). This yields then a minimum sample size of 768 per country for households and individuals. The actual sample size for countries is slightly larger than the minimum requirement to compensate for clustering effects and have a wide enough spread of EAs through out a country. For the businesses a design effect of 1 is assumed leading to a minimum sample of 384 businesses for each country.

Survey Characteris-tics

Household & Indi-viduals

Businesses

Target PopulationAll households

Individuals 15 years or older.

all businesses

Domains 1 = national level1 = national levelTabulation groups Urban, Rural NationalOversampling Urban 60% Rural 40%Urban 60% Rural 40%Clustering Enumerator Areas (EA) national Census Enumerator Areas (EA) national Census None Response Random substitutionRandom substitutionSample Frame Census sample from from NSOCensus sample from from NSOConfidence Level 95% 95%Design Factor 2 1Absolute precision 5% 5%Population Proportion 0.5, for maximum sample size0.5, for maximum sample sizeMinimum Sample Size 768 384

WeightingFour weights will be constructed, for households, individuals, small businesses and public institutions. The weights are based on the inverse selection probabilities1 and gross up the data to national level when applied.

Household weight: HHw = DW1

PHH *PEA

Individual weight: INDw = DW1

PHH *PEA *PI

Business Weight: Busw = DW 1PBus *PEAI

Household Selection Probability: PHHw =n

HHEA

EA Selection Probability: PEAw = mHHEA

HHSTRATA

RESEARCH ICT AFRICA

2! 2011 Brief Survey Methodology

1 See UNSD (2005) page 119 for a detailed discussion on sampling weights.

2012 COUNTRIESBotswanaCameroonEthiopiaGhanaMozambiqueNamibiaNigeriaSouth AfricaRwandaTanzaniaUganda

Census and RIA national ICT surveyCensus DataCensus Data RIA Survey DataRIA Survey Data

2006 2011 2007 2011

Households with Fixed Line 18,5% 14,5% 18,2% 18,0%

Households with Computer 15,6% 21,4% 14,8% 24,5%

Household with Radio 76,5% 67,5% 77,7% 62,3%

Households with Television 65,5% 74,5% 71,1% 78,2%

Households with Internet 35,2% 4.8% (Household)15.0% (Individual)

19.7% (Household)33.7% (Individual)

Cellphone Ownership 72,7% 88,9% 62,1% 84,2%

Africa’s Digital DivideHousehold data analysis

ICT Access and Usage 2011 Survey

General sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individualGeneral sample statistics of randomly selected individual

!""#$%&'$(

)*$+&,$"-./-*-/0&'"-.12%$"34"5)*$+&,$"-./-*-/0&'"-.12%$"34"5)*$+&,$"-./-*-/0&'"-.12%$"34"5

)*$+&,$"-.12%$"345"666)*$+&,$"-.12%$"345"666)*$+&,$"-.12%$"345"666)*$+&,$"&,$

!"7-89"&":&.;"&1120.8!"7-89"&":&.;"&1120.8!"7-89"&":&.;"&1120.8

!""#$%&'$(

)*$+&,$"-./-*-/0&'"-.12%$"34"5)*$+&,$"-./-*-/0&'"-.12%$"34"5)*$+&,$"-./-*-/0&'"-.12%$"34"5

)*$+&,$"-.12%$"345"666)*$+&,$"-.12%$"345"666)*$+&,$"-.12%$"345"666)*$+&,$"&,$

)'' ♂ ♀!""#$%&'$(

)'' ♂ ♀ )'' ♂ ♀

)*$+&,$"&,$

)'' ♂ ♀

Botswana 59% 270 340 222 460 579 378 34 48,4 52,4 45,6

Cameroon 52% 72 94 52 145 189 104 33 10,9 10,8 10,9

Ethiopia 45% 27 39 12 69 101 30 34 3,7 4,3 3,0

Ghana 55% 87 117 63 183 244 134 34 29,4 35,5 24,5

Kenya 62% 85 119 64 154 214 116 28 44,5 57,6 36,4

Namibia 57% 194 279 130 270 387 181 40 56,3 51,1 60,3

Nigeria 47% 102 151 47 171 252 78 34 30,5 39,8 20,0

Rwanda 50% 28 36 21 57 72 42 30 16,3 17,4 15,2

South Africa 54% 402 617 221 595 914 328 36 58,9 62,7 55,7

Tanzania 54% 35 45 26 89 115 68 34 6,2 7,4 5,1

Uganda 44% 52 59 42 126 144 102 31 15,2 18,7 10,7

!"#"$

%&#&$

'#&$

!%#"$

%(#($

%'#%$

%)#"$

%*#*$

%"#+$

%&#"$

%!#!$

%&#%$

'#&$

%&#&$

!!#($

!+#!$

(#)$

%&#)$

!,#'$

)#+$

(#($

(#+$

'#'$

)#'$

%,#%$

%!#*$

%,#+$

%(#)$

(#+$

%%#+$

)#&$

&#,$

"%#%$

"*#'$

!)#)$

%,#+$

"(#+$

"#"$

%&#!$

(#%$

%*#%$

*#($

%(#)$

!&#,$

!)#($

()#+$

"'#%$

!*#&$

*+#&$

!!#&$

()#"$

(*#+$

('#)$

(%#&$

'#%$

)#)$

*#&$

-./01/$

2304/$

5/06/07/$

89/01/$

:;<7=>7/$

?</0/$

@/A3B==0$

C7.3B7/$

C/A7D7/$

E=F;<$GHB7I/$

J=;K9/0/$

!"#$%&'(&)#*+&#,-.*/0&%+1#1%2&*+333&K;F130;$L$>F>7M$ F0>/71$9=BN$L$<=FK397H3$ B3OB31$ F03A>M=431$ 17K/DM31$/01$F0/DM3$;=$9=BN$ 3A>M=431$ K3MHP3A>M=431$

Table 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statisticsTable 2 – Gender disaggregated educational sample statistics

Highest Education: TertiaryHighest Education: TertiaryHighest Education: Tertiary Highest Education: SecondaryHighest Education: SecondaryHighest Education: Secondary Highest Education: PrimaryHighest Education: PrimaryHighest Education: Primary

!"" ♂ ♀ !"" ♂ ♀ !"" ♂ ♀

Botswana20,5% 21,9% 19,4% 53,9% 53,9% 54,0% 18,7% 19,3% 18,2%

Cameroon 7,4% 8,6% 6,2% 22,8% 19,2% 26,2% 30,6% 30,7% 30,6%

Ethiopia 2,1% 2,4% 1,8% 1,8% 1,3% 2,4% 13,1% 16,4% 8,9%

Ghana 10,5% 15,8% 6,2% 36,6% 38,9% 34,8% 27,3% 25,3% 28,9%

Kenya 26,2% 32,7% 22,3% 41,4% 41,1% 41,7% 27,4% 22,8% 30,2%

Namibia 7,1% 8,4% 6,1% 27,8% 24,3% 30,4% 45,2% 42,4% 47,4%

Nigeria 14,8% 19,5% 9,6% 37,8% 40,3% 34,9% 18,7% 18,1% 19,3%

Rwanda 1,2% 1,7% 0,7% 15,3% 16,8% 13,7% 58,4% 59,4% 57,4%

South Africa 13,3% 18,0% 9,1% 65,3% 65,8% 64,8% 17,0% 13,2% 20,2%

Tanzania 1,4% 1,5% 1,2% 11,1% 14,9% 7,8% 72,0% 73,3% 70,9%

Uganda 9,1% 11,2% 6,3% 29,9% 33,3% 25,6% 44,2% 44,6% 43,7%

Percentage of households with electricity still very low in many African countries, some even saw a decline

Uga

nda

Rwan

da

Ethi

opia

Tanz

ania

Nam

ibia

Nig

eria

Bots

wan

a

Keny

a

Cam

eroo

n

Gha

na

Sout

h Af

rica

89%

73%

65%60%60%58%

42%

19%18%16%13%

77%

63%57%

47%48%45%

13%

5%10%

2007/8 2011/12

46,6% average

Radio still main source of informationTV luxury good in several countries

Kenya

Uganda

Rwanda

Namibia

Ghana

Nigeria

Botswana

Tanzania

South Africa

Ethiopia

Cameroon 34%

41%

62%

63%

66%

70%

72%

72%

72%

77%

81%

Households with Radio

South Africa

Botswana

Kenya

Ghana

Nigeria

Cameroon

Namibia

Tanzania

Uganda

Ethiopia

Rwanda 9%

10%

13%

18%

41%

44%

53%

54%

54%

59%

78%

Households with TV

ga

Share of households with fixed-lines

South Africa

Namibia

Botswana

Ethiopia

Ghana

Kenya

Cameroon

Tanzania

Uganda

Rwanda

Nigeria 0,3%

0,2%

1,5%

0,4%

2,2%

0,6%

1,8%

4,0%

15,0%

11,5%

18,0%

0,1%

0,3%

0,9%

1,8%

2,3%

2,6%

7,6%

11,0%

17,4%

18,2%

2007/8 2011/12

Fixed-lines in decline except Botswana,

Cameroon, Uganda and Rwanda

High MTR = active contribution to fixed-mobile substitution

ga

South AfricaNamibiaSenegal

BotswanaEthiopia

Côte d'IvoireBurkina Faso

BeninGhana

Zambia*Kenya

CameroonMozambique

TanzaniaUgandaRwanda 0,1

0,30,9

1,71,82,32,42,6

4,64,74,8

7,611

11,717,4

18,2

High MTR = active contribution to fixed-mobile substitution

% of households with a working Fixed-line TELEPHONE at home

Share of households with a working computer

South AfricaBotswana

NamibiaKenya

CameroonGhanaNigeria

UgandaRwandaTanzaniaEthiopia 0,7%

1,6%

2,0%

2,2%

6,6%

8,5%

8,6%

12,7%

14,7%

15,7%

24,5%

Share of households with a working Internet connection

South AfricaKenya

NamibiaBotswana

NigeriaGhana

CameroonUganda

TanzaniaRwandaEthiopia 0,5%

0,7%

0,8%

0,9%

1,3%

2,7%

3,4%

8,6%

11,5%

12,7%

19,7%

Less than a quarter of households have a computer and even fewer Internet access

ga

Individual Access and Usage

South Africa

Kenya

Namibia

Botswana

Ghana

Nigeria

Tanzania

Rwanda

Uganda

Cameroon

Ethiopia 7%

15%

15%

19%

19%

23%

29%

30%

31%

32%

51%

15+ Own a mobile that is capable of browsing the

Internet15+ Own a mobile

South Africa

Botswana

Kenya

Nigeria

Ghana

Namibia

Uganda

Cameroon

Tanzania

Rwanda

Ethiopia 18%

24%

36%

45%

47%

56%

60%

66%

74%

80%

84%

Up to a two line subtitle, generally used to describe the takeaway for the slide

Means of sending and receiving money that the business usesMeans of sending and receiving money that the business usesMeans of sending and receiving money that the business usesMeans of sending and receiving money that the business usesMeans of sending and receiving money that the business usesMeans of sending and receiving money that the business uses

Mobile Money

Post Office Western Union etc

Banks send cash with someone

UgandaTanzaniaRwandaEthiopiaGhanaCameroonNigeriaNamibiaBotswana

16% 1% 2% 17% 81%14% 0% 0% 5% 93%

8% 0% 1% 10% 70%0% 0% 0% 5% 55%0% 1% 1% 12% 54%0% 1% 26% 4% 75%0% 0% 0% 11% 77%1% 25% 1% 41% 86%2% 16% 3% 27% 73%

sending cash with someone preferred way of sending money

Some mobile money use in East Africa19

Internet Access & Usage

20

!"!!

!#!!

!$%!!

!$#!!

!&%!!

!&#!!

!'%!!

$(()! $((*! $((+! $(((! &%%%! &%%$! &%%&! &%%'! &%%,! &%%#! &%%)! &%%*! &%%+! &%%(! &%$%!

-./01232! 42567..3!8/9:.;:2! <9232!=63>2! ?.@25A:BC6!D25:A:2! D:E67:2!F123G2! H.C/9!IJ7:K2!L23@23:2! ME23G2!

!"#$%&'&($)*+,*-'&.*$/0$1-%121%3&45$351-.$'6*$!-'*+-*'

Ethiopia

Tanzania

Rwanda

Uganda

Ghana

Cameroon

Namibia

Nigeria

Kenya

Botswana

South Africa 15%

6%

15%

9%

13%

6%

2%

2%

2%

1% Internet use (15+)

more than doubled within 4

years

34%

29%

26%

18%

16%

14%

13%

8%

6%

4%

3%

2007/8 2011/12

ga

South AfricaKenya

NamibiaBotswana

NigeriaRwanda

GhanaCameroon

UgandaTanzaniaEthiopia 5%

5%8%8%

13%15%16%

23%24%25%

28%

Using mobile to browse the Internet

How is this different from ICT access and

usage in mature economies?

South AfricaKenya

BotswanaNamibiaNigeria

RwandaGhana

CameroonUganda

TanzaniaEthiopia 10%

5%6%

4%10%

13%15%

12%17%

20%17%

Using mobile for emailingUsing mobile for Facebook etc.South Africa

KenyaBotswana

NamibiaNigeria

RwandaGhana

CameroonUganda

TanzaniaEthiopia 2%

5%7%8%

11%14%

16%17%18%

25%25%

Daily Internet use increased dramatically in past 4 years

South Africa

Kenya

Namibia

Ghana

Botswana

Tanzania

Ethiopia

Uganda

Nigeria

Rwanda

Cameroon 19%

57%

34%

28%

47%

52%

55%

43%

59%

53%

64%

11%

11%

13%

15%

15%

19%

31%

32%

35%

41%

56%

2007/82011/12

ga

Internet access:2007/08 VS 2011/12

Ethiopia

Tanzania

Rwanda

Uganda

Ghana

Cameroon

Namibia

Nigeria

Kenya

Botswana

South Africa 33,7%

29,0%

26,3%

18,4%

16,2%

14,1%

12,7%

7,9%

6,0%

3,5%

2,7%

15,0%

5,8%

15,0%

8,8%

13,0%

5,6%

2,4%

2,0%

2,2%

0,7%

2007/8 2011/12

Internet access double in three years

ga

Frequency of Internet daily use:2007/08 VS 2011/12

South Africa

Kenya

Namibia

Ghana

Botswana

Tanzania

Ethiopia

Uganda

Nigeria

Rwanda

Cameroon 19%

57%

34%

28%

47%

52%

55%

43%

59%

53%

64%

11%

11%

13%

15%

15%

19%

31%

32%

35%

41%

56%

2007/8 2011/12

ga

Where was the Internet used first?

Cameroon

Rwanda

Botswana

Ghana

Kenya

South Africa

Namibia

Tanzania

Nigeria

Ethiopia

Uganda 71,8%

66,7%

54,8%

54,2%

49,9%

34,9%

31,1%

29,5%

29,4%

29,2%

17,9%

28,2%

33,3%

45,2%

45,8%

50,1%

65,1%

68,9%

70,5%

70,6%

70,8%

82,1%

Computer Mobile phone

ga

Where the Internet was used in past 12 months

Cam

eroo

n

Gha

na

Bots

wan

a

Sout

h Af

rica

Rwan

da

Tanz

ania

Nig

eria

Keny

a

Ethi

opia

Uga

nda

Nam

ibia

23%

74%

42%

72%

45%

63%50%

33%

58%85%

80%

36%51%

21%39%

20%24%31%

21%32%51%

20%

48%55%17%31%29%45%52%36%51%

35%

10% 87%81%81%78%75%75%71%71%64%61%30%

Mobile phone Work Place of education Internet cafe

ga

Internet Access ModelsOld Internet New Internet

Hardware

Billing

Skills requirement

Electricity

Location

Computer / Laptop MobilePostpaid (monthly Internet

subscription) Prepaid

High (Windows + Internet explorer +

Viruses)Low

electricity mostly required at location of Internet use no required at home

Work, school, Internet cafe Anywhere

Implications for interconnection

!"#!$%

!&#'$%

'!#($%

')#($%

*"#+$%

&(#'$%

&(#)$%

!!#)$%

!'#,$%

!"#-$%

+*#,$%

*'#)$%

&&#'$%

&"#($%

./012345672834%9:02;<%

=>63;>01/1%0??;2@08>3<%

A7>6<234%BC/%D3B/73/B%

E/3123457/@/2F234%:>3/G%

H;0G234%40:/<%

E/3123457/@/2F234%B/IB%%

J2<</1%K0;;5H;/0</%K0;;%J/%%

!"#$%&'"()$(%*)&%+(#$,%-.'%!)/#0'%1.)$'222%L/:0;/% J0;/%

!"#$%

&'"&$%

&'"'$%

(&")$%

#!"!$%

*&"+$%

*",$%

&'"+$%

&-")$%

&*"*$%

##"*$%

*+"!$%

./%01234%56%782943%

:23;<%1=>4%=3/234%<2%?=@@%

A2%?2>48=B4%C1484%D%@5>4%

./%01234%B2<%6<2@43%

A2%4@4?<85?5</%=<%12E4%<2%?1=8B4%%

F=332<%=G28H%5<%

!"#$%&'(&$%#!')!&*"+&,)-,.,-/#0!&-'&)'1&"#.%&#&2'3,0%&4"')%555&I4E=@4% .=@4%

Conclusions‣ The mobile is closing the voice and the data gap in Africa

‣ First wave of Internet access through PCs and fixed-line /modem dial-up. Mostly through work, school or public access (Internet cafes)

‣ Second wave is through mobile phones- Easier to use- Cheaper equipment compared to computers- Prepaid (modem dial-up was postpaid)- No electricity at home needed

‣ Internet enabled mobile phones, low bandwidth applications, and social networking are the key drivers

‣ Mobile Internet reduces the cost of communication: Facebook Zero, whatsapp, Mixit

‣32

International bandwidth no longer a problem... electricity and backhaul national networks are....

South Africa

Ghana

Cameroon

Botswana

Kenya

Nigeria

Namibia

Tanzania

Ethiopia

Rwanda

Uganda 13%

16%

18%

19%

42%

58%

60%

60%

65%

73%

89%

HH connected to main electricity grid

South Africa

Botswana

Namibia

Ethiopia

Cameroon

Ghana

Uganda

Kenya

Tanzania

Nigeria

Rwanda 0,2%

0,3%

0,4%

0,6%

1,5%

1,8%

2,2%

4,0%

11,5%

15,0%

18,0%

HH with landline

Country ISP TechnologyProduct

nameDownstream bandwidth

Usage cap

Monthly cost (US$) Notes

India MTNL ADSL TriB 49 2 Mbps 200 MB 0,88 Speed reduces to 512Kbps after exceeding the usage cap + Additional charge INR1.00 per MB after exceeding usage cap

Sri Lanka SLT ADSL Entrée 2 Mbps 2.5 GB 3,78

Mexico Cablevision CableIntense 3.0

Mbps3 Mbps 10,56 Installation charge for the internet only package is taken as the connection

charge

South Africa MWEB ADSL Capped ADSL 384 Kbps 1 GB 17,55 Modem cost from ZAR 369 onwards, excludes voice line rentals.

Kenya TelkomKenya ADSL Surf and Talk 256 Kbps 34,99 The cost of Livebox+Panasonic Handset is taken as the modem cost

Cameroon Ringo Fibre Fibre Ringo 1 Mbps 47,29 XAF95000 is the charge for equipment and installation cost.

Uganda Uganda Telecom ADSL 64 Kbps 90

34

19

Figure 6: Conceptual framework for business modeling

Apple relies on its own hardware and operating system as well as its own application store. A credit card is required to register for iTunes, which disqualifies this configuration for the BoP. The price of an iPhone includes the operating system. Operating upgrades are free, unlike those for computers and laptops. Apple takes a 30 percent cut of applications sold in its App Store.

Google has a different combination. It offers an open-source operating system, leaving hardware development, production and sales to companies like HTC and Samsung. Some apps can be downloaded simply by registering (i.e., without requiring a credit card as iTunes does), which makes the Google configuration a potential candidate for the BoP. Some applications require payment through Google wallet, which can be loaded via credit card but also may be loaded by other forms of payment. Google has several revenue sources: paid applications, advertising and in-app purchases (Google charges app developers 30 percent of the revenues from in-app purchases, like Apple).

Facebook, like Mxit, has a platform that sits on top of operating systems such as iOS and Android. Unlike Mxit, however, Facebook is only available on computers and smart phones. (Facebook does have a product called Facebook Zero that works on feature phones, but this scaled-down version of Facebook does not offer app purchases). In South Africa, Facebook allows app purchases by charging a user’s prepaid airtime account. This is available with Cell C, Vodacom and MTN.

The Mxit model is also different in that it is available across all hardware and operating system combinations, from basic phones (via Unstructured Supplementary Service Data, or USSD) through smart phones (via an Android and an iOS App). Mxit also offers its own currency called Moola, which users can purchase with airtime or through their accounts with First National Bank or Standard Bank. Though Mxit has more users at the BoP at present, Facebook is relatively popular among the BoP, and many who cannot access it aspire to do so.

iOS

Up to a two line subtitle, generally used to describe the takeaway for the slide

See www.researchICTafrica.net

This research is made possible with generous

support of

35