regulatory models for internet growth: what way forward for south africa?

1

Regulatory models for Internet growth:

What way forward for South Africa?

J. Scott Marcus, Director and Department Manager

Cape Town, 24 May 2010

2 Regulatory models for Internet growth: Cape Town, 24 May 2010

What way forward for South Africa?

• The Internet is widely recognised as contributing to societal welfare.

- Increasing the overall performance of the economy by making markets more efficient.

- Enhancing the access of the disadvantaged to information.

- … and far, far more.

• What can regulation and public policy do to foster healthy growth of the Internet?

• What can South Africa learn from international best practice?


Agenda

• Introduction

• Regulation and public policy in the EU and US- Authorisation / licensing- Access- Interconnection- Spectrum policy- Universal service: basic service for all- Industrial policy: ultra-fast service for many

• Relevance to South Africa today

• Conclusions: What way forward?


Introduction:What drives Internet growth?

• Physical availability to all.

• Affordable prices.

• Applications and content of interest.

• Prerequisites:- Suitable devices: PCs, smart phones, whatever- Educated consumers

• Consumer privacy, security, and trust



• Competition is crucial- Widespread availability- Consumer choice- Affordable prices

• Promote widespread availability of fast services

• Avoid bottlenecks to applications and content

• Ensure that users have access to suitable devices, and know how to use them to access the Internet

• Promote a culture of security and privacy

• … but how????



• As conventional networks migrate to IP-based NGNs, the technological basis is essentially the same as that of the Internet.

• Substantial practical differences remain between (closed) NGNs and the open Internet.

• All in all, it is increasingly clear that the health of the Internet is closely linked to that of the electronic communications sector overall.

• If the sector at a whole does not enjoy effective competition, Internet innovation will be impacted.


Regulation and public policy:Key principles

• Let the market operate unimpeded wherever it is likely to generate appropriate results.

• Intervene only to the extent needed to address likely market failures:

- Market power- “Public goods”, universal service, and related challenges- Management of scarce public resources (spectrum, numbers)

• Prefer wholesale remedies over retail.

• Prefer ex post competition law over ex ante regulation in those cases where ex post would likely be effective.


Regulation and public policy

• Authorisation / licensing

• Access

• Interconnection

• Universal service: basic service for all

• Industrial policy: ultra-fast service for many

• Spectrum policy


Regulation and public policy

• The US and the EU represent two increasingly divergent approaches.

• Through the nineties, liberalised US approaches were widely admired and emulated.

• Since 2001, pro-business US regulators radically deregulated, with mediocre results.

• The EU system put in place in 2002 should be viewed as representing best practice today.


Regulation and public policy:Europe

• In the past, nearly every European country had a government-owned telecoms operator (PTT).

• Fixed, mobile, and in many cases cable television were all a single government monopoly.

• Comparisons with (especially) the US convinced most European experts that these government monopolies were inherently inefficient, and were impeding technological innovation.

• A period of privatisation and liberalisation followed, culminating in a European framework in 2002-2003.


Regulation and public policy:Authorisation / licensing

• Intense licensing regimes are often put in place in order, ostensibly, to protect consumers.

• There is a cost! They impede competitive entry.

• European practice:- Set low thresholds for the maximum burdens that

national regulators (NRAs) can impose.- The ECS can be required to notify the NRA.- If, however, the NRA fails to quickly respond, the ECS

can proceed as if a licence had been granted.


Regulation and public policy:Access

• The wired last mile is a competitive bottleneck in nearly all countries.

• In the absence of regulation, last mile market power leads to:

- Inflated prices- Lack of consumer choice



• Mitigating factors need to be considered.

• Portions of the national territory might support some facilities-based telecoms competition.

- High density of subscribers.- High disposable income.

• Cable television and wireless may, where present, provide an alternative means of access.

• The degree to which these represent meaningful competition needs to be carefully and objectively assessed by means of competition economics.



• Mitigating factors need to be considered.

• Portions of the national territory might support some telecoms competition.

- High density of subscribers.- High disposable income.

• Cable television and wireless may, where present, provide an alternative means of access.

• The degree to which these represent meaningful competition needs to be carefully and objectively assessed by means of competition economics.



• Europe is characterised by a systematic approach where the European Commission initially identifies markets that are potentially problematic.

• National Regulatory Authorities then analyse:- The market definitions in their national context- Whether any market players have Significant Market

Power (SMP)- What remedies should be applied to those with SMP

• The Commission then reviews the results.

• The process is public and very transparent.



• Last mile fixed network access has been a central focus in Europe.

• A range of remedies, enabling competitive entry with different levels of investment and different risk/reward profiles, comprise a “ladder of investment”.

- Simple resale- Bitstream access (ATM or IP)- Shared access (high frequency only)- Full Local Loop Unbundling (LLU)

• As you move downwards, greater investment is required, but there is greater opportunity as well.


Regulation and public policy:Bitstream Access

• Bitstream access is similar to Telkom wholesale service.

• With bitstream, the incumbent always operates the DSLAM.

Source: ERG Common Position, Bitstream Access, 2005


Regulation and public policy:LLU and Shared Access


Regulation and public policy:Forms of Access

Source: European Commission, 13th Implementation Report


Regulation and public policy:Access in France

Source: European Regulators’ Group (2006)

Broadband market competition report - French Case Study



France Broadband Adoption 7/2004

2,270,407 , 46%

13,066 , 0% 717,654 , 15%

854,205 , 17%

635,155 , 13%

425,000 , 9%

Incumbent DSL LinesFull ULLShared AccessBitstream AccessResale

Cable, FTTH, WLL, PLC, Other

Source data: European Commission, 10th Implementation Report



France Broadband Adoption January 2008

46%

22%

11%

15%

1% 5%

Incumbent DSL Lines

Full ULL

Shared Access

Bitstream Access

Resale

Cable

Source data: European Commission, 13th Implementation Report



Source: European Commission 14th Implementation Report


Regulation and public policy:Access in the United States

• The US was on a similar trajectory in the nineties, but has now taken a very different course.

• Historic recognition of market power, but little or no explicit market power analysis.

• No over-arching technological neutrality.

• Competition law mutually exclusive with regulation.

• Radical deregulation during the period 2001-2008.

• Increasing market concentration.

• Collapse of competitive network operators.



CLEC Percent of ADSL High-Speed Lines

0.0%

1.0%

2.0%

3.0%

4.0%

5.0%

6.0%

Dec-01

Jan-02

Feb-02

Mar-02

Apr-02

May-02

Jun-02

Jul-02

Aug-02

Sep-02

Oct-

02

Nov-02

Dec-02

Jan-03

Feb-03

Mar-03

Apr-03

May-03

Jun-03

Jul-03

Aug-03

Sep-03

Oct-03

Nov-03

Dec-03

Jan-04

Feb-04

Mar-04

Apr-04

May-04

Jun-04

Jul-04

Aug-04

Sep-04

Oct-04

Nov-04

Dec-04

Source: FCC reports based on Form 477 carrier data



Source: FCC reports based on Form 477 carrier data



Source: European Commission 14th Implementation Report

US 2%



• Apologists for US incumbents will say that the US is not doing all that badly.

• Possibly true but irrelevant. US performance is vastly inferior to what it could have been.

0

5

10

15

20

25

30

35

40

Nether

lands

Denmark

Norway

Switzer

land

Korea

Icelan

d

Sweden

Luxe

mbour

g

Finlan

d

Canad

a

German

y

Franc

e

United

King

dom

Belgium

United

Stat

es

Austra

liaJa

pan

New Z

ealan

d

Austria

Irelan

dSpa

inIta

ly

Czech

Rep

ublic

Portug

al

Greece

Hunga

ry

Slovak R

epub

lic

Poland

Turke

y

Mexico

Source: OECD

DSL Cable Fibre/LAN Other

OECD Broadband subscribers per 100 inhabitants, by technology, June 2009

OECD



• United States had an enormous head start on broadband deployment over everybody else.

• Ubiquitous cable television: A second pipe to nearly every home.

• High GDP, high disposable income, lots of PCs.

• The US arguably should have been among the top countries in the OECD in broadband adoption.



• Slower-than-expected roll-out and adoption of broadband.

• Loss of consumer choice.

• Higher retail prices?

• Network neutrality problems that are likely to necessitate highly intrusive re-regulation.

• Possibly some acceleration of fibre deployment by incumbents, but at the cost of greatly impacted deployment by competitors.


Regulation and public policy:US versus EU

• Network Neutrality only a minor concern in the EU.- The more robustly competitive environment discourages

anticompetitive discrimination.- Richer palette of regulatory tools.

• Most Europeans have access to multiple broadband providers (not all of which are fully facilities-based).

• EU regulatory reform seeks minor changes to ensure e.g. that consumers are informed, and can switch without cost if their network operator changes its policies.


Regulation and public policy:Interconnection

• Internet interconnection globally takes place primarily through variants of two main mechanisms:

- Peering: ISPs exchange traffic destined for their respective customers (or customers of their customers), often without explicit payment.

- Transit: An ISP carries another party’s traffic to third parties, possibly to the entire Internet, generally for pay.

• These arrangements typically do not depend on any regulation.

• Interconnection in the telephony network, by contrast, tends to be highly regulated and highly focused on voice minutes.



OriginatingNetwork

TerminatingNetwork

Call placed Call received

RetailCPP

Payment

Wholesale CPNP PaymentWholesale CPNP Payment

• Termination rates represent wholesale payments between network operators under the Calling Party’s Network Pays (CPNP) arrangements.



• A substantial economic literature tells us to expect high termination fees (from small operators as well as large) in the absence of regulation.

• These prices result from the termination monopoly.

• Termination rates for the fixed network have long been constrained by regulation in European Member States so not to exceed the terminating network operator’s marginal cost.

• Mobile termination rates were, however, unregulated in most Member States until 2003 or so.



• European experience supports the expectation of high MTRs in the absence of regulation.

- MTRs before regulation (2002): € 0,187.- MTRs after regulation (2008): € 0,086.

Source: 14th Implementation Report, Annex 2, 2009,


How low should they go?Mobile Termination Rates (MTRs):How low should they go?


Termination Rates

• A substantial economic literature argues that termination rates should be set at the level of the terminating network operator’s cost (however determined).

• An alternative school of thought argues that there should be no wholesale payments (Bill and Keep).

• There are some arguments for setting TRs lower than cost, but few if any have argued that TRs should be higher than cost.

• There is no perfect price.


Regulation and public policy:Interconnection: Effects of MTRs

• The MTR affects MNOs in two very different ways:- For calls Mobile-to-Mobile (M2M) calls, a lower MTR

represents a reduced wholesale cost for the originating MNO.• In a competitive market, a reduced cost should lead to a reduced

price.• A reduced unit price will tend to lead to increased consumption.• The impact on ARPU depends on the relative magnitude of these

effects, since they push in opposite directions.- For calls to the mobile network from either fixed or mobile, a

lower MTR tends to mean reduced wholesale income.• However, as noted above, it is also likely to result in reduced retail

unit price, both for F2M and for M2M.• Again, reduced unit price for calls to the mobile network should

result in increased call volumes.• The increase in call volume pushes ARPU in the opposite direction

as the reduction in MTRs, such that the combined effect is not easy to predict.


Regulation and public policy: Interconnection: Effects of MTRs

• Historical experience is that overall European unit prices for mobile voice service move in parallel with MTRs.

Service-Based Revenue per MoU vs MTRs in Europe

$0.00

$0.05

$0.10

$0.15

$0.20

$0.25

$0.30

2004 2005 2006 2007 2008

$ (U

S) SBR/MOUMTR (PPP corrected)

Source: ERG (for MTR data), Merrill-Lynch



• One would expect high unit price to be associated with low demand and vice versa (price elasticity of demand).

Source: WIK. based on Merrill Lynch 3Q2008 data.

y = -1720,3x + 534,54R2 = 0,5691

-

100

200

300

400

500

600

700

800

900

- 0,05 0,10 0,15 0,20 0,25 0,30 0,35

USA

IndienHong-Kong

Singapur

Österreich

Deutschland

Finnland

Kanada

Schweiz

FrankreichSchweden

IrlandNorwegen

NiederlandeSpanien

UK DänemarkBelgienGriechenland

Italien Portugal

ARPU USA

ARPU Deutschland

ARPU ÖsterreichSouth Africa



Voice Revenues

0

500

1.000

1.500

2.000

2.500

3.000

3.500

2Q2006 3Q2006 4Q2006 1Q2007 2Q2007 3Q2007 4Q2007 1Q2008 2Q2008 3Q2008 4Q2008 1Q2009 2Q2009 3Q2009

Mill

ion

euro Other (Mio.)

M2F (Mio.)M2M off-net (Mio.)M2M on-net (Mio.)

Source: Spanish CMT Data



€ 0,110€ 0,100

€ 0,080€ 0,070

€ 0,057

€ 0,195 € 0,195 € 0,191

€ 0,178

€ 0,190€ 0,180 € 0,182

€ 0,177€ 0,170 € 0,168 € 0,166 € 0,164 € 0,163 0,16

€ 0,006 € 0,005 € 0,005 € 0,008 € 0,008 € 0,009 € 0,010 € 0,010 € 0,011 € 0,011 € 0,013 € 0,013 € 0,015 € 0,014

€ 0,189 € 0,190€ 0,186

€ 0,170

€ 0,181€ 0,171 € 0,171

€ 0,166€ 0,159 € 0,157

€ 0,153 € 0,151 € 0,148 € 0,147

€ 0,110

€ 0,100€ 0,100€ 0,100

€ 0,057€ 0,060€ 0,060

€ 0,070€ 0,080

€ 0,000

€ 0,050

€ 0,100

€ 0,150

€ 0,200

€ 0,250

2Q2006 3Q2006 4Q2006 1Q2007 2Q2007 3Q2007 4Q2007 1Q2008 2Q2008 3Q2008 4Q2008 1Q2009 2Q2009 3Q2009

MTR Voice SBR / orig MOU Subscription fees / orig MOU Traffic SBR / orig MOU

Source: Spanish CMT Data



• Our data for Europe show the following relationships to MTR:

- Retail price per minute: +0.7- Minutes of use per month: -0.5 to -0.6

• The instrumental variable used to represent retail price (Merrill Lynch Service Based Revenue per Minute of Use) is about 85% retail and 15% wholesale revenue.



• Most studies (not all) find that high MTRs encourage more rapid mobile adoption.

- Lower initial fees.- Higher handset subsidies.- Lower monthly fees.- All result in a lower cost to acquire and retain service.- Cost to use the service, however, can be higher.

• Does penetration greater than 100% represent a benefit to public welfare?

- Are multiple subscriptions a response to different roles and responsibilities (work versus leisure)?

- Or are they a response to charging anomalies (on-net off-net price discrimination, roaming charges)?



Source: 14th Implementation Report, Annex 2, 2009,



Source: Eurbarometer June 2008, data from Nov-Dec 2007



• Mobile penetration in South Africa was about 101% as of 3Q2009 (Merrill Lynch), about 81% of them pre-paid.

• Actual adoption of cell phones is highly variable from one province to another, ranging from 43.2% (Western Cape) to 80.8% (Mpumalanga). (Source: Statistics South Africa, 2009 Household Survey)

• Compare this also to roughly 10% fixed penetration (or 18% of households).



• The move to lower MTRs in Europe, or to eliminate them altogether, has to a significant degree been motivated by the migration to IP-based NGNs.

• Recovery of the cost of the network solely from the voice service makes no sense going forward. Voice represents a small fraction of total traffic and cost.

• With the emergence of over-the-top VoIP services, the service provider might not be the network operator in the first place.

• Distorted interconnection pricing inhibits eventual migration of the PSTN to IP interconnection.


Regulation and public policy:Driving widespread broadband

• Separate approaches are often needed to driving- Deployment- Adoption

• Two interrelated but distinct programmatic aspects- Universal service

• Ensuring that anyone who wants a basic service can get it.• Typically addressed as a regulatory matter.

- Ultra-fast broadband• Ensuring that high speed (e.g. fibre-based) broadband is

available to as much of the population as possible.• Generally treated as a matter of industrial policy.• Typically driven by a ministry, not by the regulator.



• Universal service recognises network effects – the more people are connected, the better for all.

• Twentieth Century universal service emphasised voice service over the fixed network.

• Twenty-first Century universal service must recognise:

- That in a country like South Africa, where the fixed network reaches only 18% of households, that mobile service is of enormous importance.

- That in an increasingly IP-based world, the meaningful wired universal service is a broadband access.



• Concepts of best practice emerge in studies by the ITU and by the World Bank (cf. Bjorn Wellenius)

• The national territory can be viewed as consisting of three kinds of areas:

- Those where commercial incentives are sufficient to ensure deployment and ongoing viability of services.

- Those that require subsidy indefinitely.- Those that could be self-sustaining once initially “jump

started”.

• An analogous categorisation into white, black and grey areas appears in European State Aid rules.



• Important to avoid needless subsidies to services that could sustain themselves. Not only is it wasteful, but it also distorts competition.

• “Reverse auctions” are a best practice means of providing no more subsidy than necessary.

• Reverse auctions are not trouble free:- The winner may be unwilling or unable to actually

complete the build-out at the agree-on price. Encourages “bid to win”.

- Does not automatically adjust to changing circumstances.



• 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.



Investment per home connected (in Euro), market share 50%, urban cluster, stand alone first mover **

VDSL

PON

P2P

** Based on the investment of the urban cluster and a market share of 50%. If other marekt shares are used, it is mentiond in brackets.

1,882 1,1602,111 (54%) 2,025 1,333 1,548

254 433

2,039 1,580 1,238 1,411 1,771 1,110

457 n.v. 352 218

Network Type

Country [in €]

DE FR SE PT ES IT



Viability of NGA roll-out for incumbents across countries and technologies

VDSL

PON

P2P

SE PT ES IT

71.5% n.r. 18.3% 39.0% 67.4% 100.0%

DE FR

25.1% 25.2% 18.3% 19.2%

13.7% 18.6% 18.3% 19.2%

12.2% 17.6%

12.2% 12.6%

CountryNetwork Type



Replicability of NGA roll-out for a second mover, 80 % access to existing ducts at current cost-based prices

VDSL

PON

P2P

n.v. 1.6%

0.0% 6.8% n.v. n.v. n.v. 0.2%

0.3% 6.8% n.v. n.v.

Network Type

CountryDE FR SE PT ES IT

18.5% n.r. n.v. 39.0% n.r. 17.6%



• Important initiatives to drive ultra-fast broadband are under way in a number of European countries, Singapore, Australia, New Zealand, and the United States.

• Australia’ National Broadband Network:- Connect 90-93% with high speed fibre at 100 Mbps.- Connect the remainder at 12 Mbps (peak) with some

combination of fixed or mobile wireless and satellite.- Cost initially estimated at $43 billion AUD; recently

revised to $26 billion AUD (about 171 billion ZAR) taking substantial advantage of aerial fibre.


Regulation and public policy:Spectrum policy

• An enormous area in its own right. In the interest of time, we will make only brief comments.

• A global consensus has emerged for commercial spectrum allocation and assignment:

- Allocations and assignments should have as few restrictions as possible, consistent with the need to avoid harmful interference.

- Technological and service neutrality are desirable.- Market mechanisms, exemplified by auctions, help to

ensure that spectrum is assigned to those who value it most (and thus are most likely to put it to good use).

- Secondary markets (trading, leasing) are a useful complement to auctions.


Relevance to South Africa

• Competitive access remedies

• Interconnection and MTRs

• Submarine cable, long-haul capacity

• Broadband for all

• Spectrum management

• Government ownership of telecoms

• Institutional arrangements



• The following observations, suggestions, and recommendations are based on an admittedly fragmentary understanding of your environment.

• Special thanks to Christoph Stork of researchICTafrica.net, who kindly made a number of unpublished manuscripts available to me.

• Other observations reflect publicly available data from sources such as Statistics South Africa, and from commercial sources such as the Merrill Lynch Quarterly Wireless Matrix 4Q2009.

• Any errors are my own.


Relevance to South Africa: Licensing

• Observation: Licences for competitive new fixed and mobile entrants have been delayed for years.

- Neotel- CellC- Infraco (government-owned)- Conversion of licences from VANs to ECNS

• Observation: European experience demonstrates that lengthy licensing procedures and/or onerous conditions and unnecessary and unproductive.


Relevance to South Africa: Licensing

• Recommendation: Licensing procedures should be simplified and streamlined, and maximum conditions firmly limited. ICASA needs to make a decision, up or down, within a short period of time.

• Question: I am told that ICASA granted and issued 288 individual Electronic Communications Network Services (I-ECNS) licences in January 2009. They granted another 256 I-ECNS licences without issuing them. Is this problem solidly on the mend? Is it fixed?


Relevance to South Africa:Competitive access

• Observation: Policies for ensuring that competitors gain access to bottleneck facilities seem to be ineffective. New rules have languished pending ICASA’s definition of relevant markets.

• Observation: There are large parts of the country that are not served by the fixed network; nonetheless, it is vital in key metropolitan areas.

• Recommendation: Detailed rules for a basic ladder of investment need to be put in place, using rules in a country where they are effective as a template.

• Recommendation: ICASA’s ability to promptly enforce such rules needs to be assessed.


Relevance to South Africa: Interconnection

• Observation: MTRs in South Africa continue to be high, despite commitments to lower them.

• Observation: High MTRs may be leading to high unit prices in South Africa, and clearly to low usage (MoU).

• Observation: High MTRs can also likely inhibit effective competitive entry (e.g. by on-net off-net price discrimination).

• Observation: MTRs in Europe are about to decline 80% or more (e.g. to € 0.006 in the UK in 2015).


Relevance to South Africa:Mobile Subscribers (000)

-

10,000

20,000

30,000

40,000

50,000

60,000

7-Mar 7-Jun 7-Sep 7-Dec 8-Mar 8-Jun 8-Sep 8-Dec 9-Mar 9-Jun 9-Sep

OtherCell CMTNVodacom

Source: Merrill Lynch Quarterly Wireless Matrix 2Q2009


Relevance to South Africa:Mobile Minutes of Use per Month


0

20

40

60

80

100

120


VodacomMTN


Relevance to South Africa:Mobile Revenue per MoU


ZAR 0.00

ZAR 0.20

ZAR 0.40

ZAR 0.60

ZAR 0.80

ZAR 1.00

ZAR 1.20

ZAR 1.40

ZAR 1.60

ZAR 1.80


VodacomMTN


Relevance to South Africa: MTRs

• Observation: Market shares here, in terms of subscribers, are about 52% / 33% / 13%.

• Observation: Relatively low levels of on-net off-net price discrimination, despite high MTRs, suggests a market where competition is weak.

• Conjecture: MNOs will not return all of the savings of lower MTRs to the public. In Europe, they return about 70% on average. Here, the proportion might be somewhat lower. Lowering MTRs is nonetheless in order.



• Observation: Changes need to implemented with care. High MTRs probably also lead to high penetration, and to widespread availability of pre-paid plans.

• Recommendation: MTRs urgently need to be brought to much lower levels, but perhaps not quite as low as the target rates in Europe.


Relevance to South Africa:Broadband

• Observation: Broadband deployment initiatives in South Africa need to take an integrated view of multiple potential bottlenecks:

- Submarine cable- Back-haul between metropolitan areas- Back-haul into the countryside- Last mile (or last air mile)


Relevance to South Africa: Submarine cable

• In many markets I have worked in (Jamaica, Australia, New Zealand), undersea cable capacity represents a crucial bottleneck.

• Low capacity can impede development of content and applications in country.

• Low capacity sometimes forces hosting to the US.• Observation: Submarine cable, at least, appears

to be in a promising state, including:- 2009 –SEACOM (commercially launched mid-2009)- 2010 –EASSy- 2010 –MainOne- 2011 –West African Cable System (WACS)


Relevance to South Africa: Backhaul

• Back-haul capacity between metropolitan areas, deeper into the countryside, and also to submarine cable landings, seems to continue to represent a crucial ongoing bottleneck.

• Experience in many countries suggests that lack of back-haul at competitive prices is likely to impede a wide array of other forms of competition.

• ISPs need back-haul to link major cities, reach cable landings, and serve their customers. Mobile operators need large capacity back-haul. High back-haul prices lead to many distortions.


Relevance to South Africa: Backhaul

• Observation: Infraco was supposed to address the shortfall in back-haul; however, Infraco encountered substantial delays in licensing and launch.

• Questions: Commercial parties are building out on their own. Will capacity be available to third parties? Will it be enough to close this gap?

• Recommendation: Back-haul appears to need much more focussed attention in terms of public policy initiatives.


Relevance to South Africa: Broadband for all

• Observation: Only 18% of households have access to the wired network; thus wireless broadband, whether fixed or mobile, has to play a critical role in any broadband strategy.

• Observation: In Europe, many bands have been made available under liberalised arrangements suitable for broadband (e.g. WAPECS). Digital Divided spectrum is being made available.

• Observation: An inquiry into spectrum allocation (especially for WiMAX), launched in 2006, published findings in 2008 but has not concluded.



• Observation: The Department of Communication (the SMA) has committed itself to an audit that it believes will take two years.

• Observation: ICASA has said that it intends to auction spectrum.

• Recommendation: Attention is urgently needed. Large blocks of spectrum suitable for wireless broadband need to be identified, and need to be made available through auction with as few restrictions as possible on mode of use.



• Observation: South Africa has a diverse population with a fairly wide dispersion of income (e.g. GINI coefficient).

• Observation: In metropolitan areas with high disposable income, the challenge is not to achieve coverage, but rather to lower price.

• Suggestion: For prosperous metropolitan areas, the focus should probably be on competitive access to bottleneck facilities, not on stimulating deployment.



• Observation: As previously noted, the tools used to achieve basic broadband to a larger segment of the population are different from those used to achieve ultra-fast broadband to those who can afford it.

• Suggestion: South Africa might be well advised to focus first on getting basic broadband rolled out to a larger proportion of the population. Otherwise, you risk exacerbating the Digital Divide.



• Observation: South Korea was an early leader in ADSL. They did not subsidise the last mile.

- They subsidised a fibre backbone linking all major metropolitan areas.

- They subsidised PCs for Korean families

• Observation: About 15% of South African homes have a PC, which poses an effective ceiling to the number of Internet connections.

• Question: What could be done to stimulate PC penetration in South Africa, thus stimulating demand for broadband services?


Relevance to South Africa: Spectrum

• A comprehensive program of spectrum liberalisation is urgently needed.

• Elements should include:- Auctions- Secondary markets (but restricted to simpler forms of

spectrum trading or leasing)- Liberalised allocations- In a second phase, consider implementing administrative

controls for spectrum for defence forces (as exemplified by the Netherlands)


Relevance to South Africa: Government ownership

• Observation: After the transactions of 2009, the South African government reportedly still owns 37.7% of Telkom, and retains a 14% direct shareholding in Vodacom. (Source: researchICTafrica.net)

• Observation: Experience in Europe and elsewhere strongly suggests that a share this large distorts the government’s own incentives.

• Recommendation: More of this share should be sold off.

• Recommendation: Any direct government investment should focus on segments where competitive entry would otherwise be unlikely.



• Observation: ICASA seems to have enormous difficulty in bringing proceedings to a definitive conclusion. MTRs are a case in point.

• Observation: Interminable delays cause uncertainty that harms businesses, and ultimately consumers.

• Suggestion: I would respectfully suggest that ICASA’s institutional arrangements need analysis and probably some serious re-thinking.

• Suggestion: ICASA might benefit from capacity building.


Conclusions: What way forward?

• Implement effective competitive access remedies

• Lower MTRs

• Address back-haul deficits

• Liberalise spectrum management

• Focus broadband initiatives on universal service

• Consider demand stimulation

• Further reduce government ownership of telecoms

• Bolster the institutional effectiveness of ICASA

wik-Consult GmbHPostfach 200053588 Bad HonnefTel 02224-9225-0Fax 02224-9225-68eMail [email protected]. wik-consult.com

regulatory models for internet growth: what way forward for south africa?

Documents