Broadband in South Africa : Implications of International Experience

1. Broadband in South Africa:Implications of International Experience J. Scott Marcus

2. Broadband in South Africa Introduction Broadband usage Coverage and capabilities Challenges to achieving full NGA coverage Balance is needed Fixed versus mobile/satellite Supply side versus demand side SMART Metrics Observations

3. Usage trends

4. Global Internet traffic trends “It's tough to make predictions, especially about the future.” - Yogi Berra (US baseball player and manager)

5. Global Internet traffic trends Source: Cisco VNI (2012), WIK calculations. Full report, page 16.

6. Average busy hour traffic per user Full report, page 20. Source: Cisco VNI 2011 data, WIK calculations.

7. Traffic per household Few HH need large bandwidth (whether average or peak hour). Source: Cisco VNI 2011, WIK

8. Traffic asymmetry Video plays a huge role in traffic. Meanwhile, peer to peer not growing in relative importance. Source: Cisco VNI (2012)

9. Usage correlates weakly with speed MB / subscriber / month in selected countries (2012) Source: Cisco VNI online database (2013), WIK calculations

10. Effective download speeds Source: Ookla/Speedtest netindex, WIK calculations

11. Coverage and adoption

12. Overall adoption • Coverage and adoption are not the same thing. Source: OECD Broadband Portal

13. The Digital Agenda for Europe (DAE) The European Union is committed to an ambitious Digital Agenda for Europe (DAE). The DAE includes full broadband availability in 2013, 100% availability of 30 Mbps in 2020, and 50% adoption of 100 Mbps by 2020. They never fully defined these goals. It is widely acknowledged that meeting these goals is challenging, however defined.

14. Coverage poses challengesPopulation density Full report, page 41. Source: Center for International Earth Science Information Network (CIESIN).

15. Coverage in Europe • Basic broadband • The older Member States have nearly full coverage. • Some of the newer Member States have sufficient gaps in rural coverage. • NGA coverage • Member States with cable already enjoy full deployment (the Netherlands, Belgium, Malta). • Some eastern Member States have already deployed substantial fibre. • Some eastern Member States have lots of cable.

16. Basic broadband coverage Source: Point Topic (2012)

17. NGA coverage Source: Point Topic (2012)

18. The “broadband gap” in the United States Incremental CAPEX and OPEX needed to achieve 4 Mbps download and 1 Mbps upload speed Full report, page 42. Source: FCC: “The Broadband Availability Gap”, April 2010.

19. The “broadband gap” in the United States Full report, page 43. FCC: “The National Broadband Plan”, March 2010.

20. The developing world Many countries in the developing world face problems analogous to those of the newer Member States, but far more severe. The fixed network never reached the full population, and is unlikely to do so any time soon. Mobile voice coverage is nearly universal. Mobile data coverage has limitations. Cable television may or may not be present.

21. The developing world In general, there is no crisp dividing line between broadband and ultra-fast broadband. In developing countries, relatively few people have the interest and the disposable income to support NGA as a purely commercial proposition. First priority needs to be on achieving widespread basic broadband.

22. The challenge of broadband in India “The broadband growth has not only been slow but also biased in favour of urban areas. More than 60% broadband subscribers are in the top ten metros … Just 5% of the broadband connections are in rural areas …” – TRAI, Nat’l Broadband Plan Fixed network coverage is meagre Population of 1.2 billion 34.9 million fixed lines – just 3% of the population! Only 8.9 million rural lines Mobile subscriptions reach 77% in 2011, expected to reach 91% in 2012. – Merrill-Lynch How to achieve full broadband coverage???

23. The challenge of broadband in India Cable television is widespread, and could represent a key part of the answer. Between 80 and 140 million households subscribe to cable television. Penetration of pay TV is 78%, some 70% of this is cable. Cable in India suffers from serious handicaps. Very little cable has been upgraded to support digital transmission, addressability, and broadband data transmission. Fragmentation into some 60,000 cable operators. Many cable operators lack the scale, the technical competence, and the business and entrepreneurial skills to successfully deploy broadband data services.

24. The challenge of broadband in India • “The broadband growth has not only been slow but also biased in favour of urban areas. More than 60% broadband subscribers are in the top ten metros … Just 5% of the broadband connections are in rural areas …” – TRAI, Nat’l Broadband Plan • Fixed network coverage is meagre • Population of 1.2 billion • 34.9 million fixed lines – just 3% of the population! • Only 8.9 million rural lines • Mobile subscriptions reach 77% in 2011, expected to reach 91% in 2012. – Merrill-Lynch • How to achieve full broadband coverage???

25. Coverage poses challengesPopulation density Source: SEDAC/CIESIN Columbia University, Gridded Population of the World (GPW).

26. The situation in India A substantial fraction of India‘s population lives in villages, rather than cities or metros. There are vast differences in population density. Source: Census of India as reported in TRAI’s National Broadband Report (2010)

27. General broadband initiatives Source: Marcus/Jain, based on the TRAI National Broadband Plan (2010)

28. Wireless broadband in India Wireless solutions cannot serve as a replacement for wired solution in metros and in large cities (because spectrum is unlikely to be insufficient in the near to medium term), but will instead serve as an economic complement; Wireless solutions are likely to be adequate in villages, and may prove adequate as a full alternative to wired solutions in some towns as well; Moreover, no practical alternative to wireless solution exists in most towns and nearly all villages.

29. Challenges to achieving full NGA coverage

30. Migration to fibre WIK report on Next Generation Access for ECTA (2008). Sophisticated models of fibre roll-outs in France, Germany, Italy, Netherlands, Portugal, Spain. No country likely to achieve full coverage without public stimulus/subsidy. Only limited prospect of replicating infrastructure.

31. Migration to fibre

32. Migration to fibre

33. Challenges of achieving NGA deployment Cost and ARPU per customer per month for FTTH P2P Ethernet (Germany, 70% penetration) Full report, page 45. Source: WIK.

34. Challenges of achieving NGA deployment Investment subsidy per customer required for FTTH P2P Ethernet (Germany) Source: WIK Full report, page 46.

35. Fixed versus Mobile

36. Fixed versus mobile Globally, we have tended to fixate on fibre. Fibre is an essential element of all high speed solutions, including mobile and cable. For developing countries, FTTH/FTTP will be difficult and expensive except in dense areas. Mobile may be more realistic in rural areas. Fixed broadband is still needed for back-haul! For the sparsest areas, consider satellite.

37. Cost implications of different technologies Upgrades to FTTN/VDSL or to FTTH become more expensive on a per-subscriber basis as population density declines. LTE less expensive than fixed where population density below 3,000 inhabitants per Km2. Cable DOCSIS 3.0 upgrade costs do not depend on density. Annualized cost (€) per user to cover different geotypes in Spain Source: Feijoo / Gomez-Barroso (2010a). Full report, page 60.

38. A changing mix of devices

39. Changing characteristics of use Source: Cisco IBSG (2012)

40. A changing mix of applications Source: Informa / Mobidia (2013)

41. How much mobile off-load is taking place? Source: Informa/Mobidia 2013

42. How much mobile off-load is taking place in the EU? Data Sources: Cisco VNI (2012), Mobidia/Informa (2013), WIK calculations

43. Supply versus Demand

44. Supply versus demand A 2011 study on behalf of the IRG by Parcu et al. at the Florence School of Regulation found that, once broadband was well deployed, supply side measures could have greater effect than demand side. Consider carefully why Internet access is not more widespread. Survey data are crucial. It may not be just cost or availability.

45. Impediments to adoption in Europe “As Internet access becomes widely available, those who choose not to have Internet access are presumably more likely to have made an active decision to not do so.” “Among households without an Internet connection, the top reason given for not having one is that all household members are disinterested in accessing the Internet (63%). … Cost, as an explanation, has decreased in importance, 7% of respondents with no connection reported that they are not sure exactly what the Internet is. …”

46. Supply versus demand Source: Parcus et al. (2011)

47. Supply side measures Source: Parcus et al. (2011)

48. Demand side measures Source: Parcu et al. (2011)

49. SMART Metrics

50. SMART Metrics The Ministry has already noted a lack of good statistics for South Africa. If you cannot measure something, you cannot manage it. The European Commission just “declared victory” in regard to their 2014 coverage target thanks to satellite. By that measure, the goal was already satisfied before it was committed to paper.