Wireless & development in the Asia Pacific: Institutions matter Rohan Samarajiva, with contributions from LIRNEasia team High-level workshop, Annenberg

Wireless & development in the Asia Pacific: Institutions

matterRohan Samarajiva, with contributions from LIRNEasia

team

High-level workshop, Annenberg Research Network on International Communication, Oct 7-8, 2005

www.lirneasia.netwww.lirneasia.net

Agenda Wireless and development Wireless in the Asia Pacific

Backbone Mobile & “fixed” Mobile data WiFi

The strange case of Indonesia Regulatory environment

Spectrum management incl. refarming Importance of investment


How do South Asians on less than USD 100/mo. communicate?

Fixed(49%)

Mobile(19%)

‘Public’ access (66%)

21%2%

11%

3%23%

3%

37%

Base: 3199


Survey (3200 sample, 6 languages, 11 locations)Surveys in India (2099) & Sri Lanka (1100)

Not representative of India and Sri Lanka as wholes

‘Users’those who have used a phone in the last three months

‘Financially constrained’ users:Monthly household income < USD100

Socio-Economic Classification (SEC) groups B,C,D & E

Mix of urban (37.5%) and rural (62.5%) respondents


Tech

Development: Not by communication (wireless) alone

Development

Enabling/complementary conditions

Communication CoordinationKnowledge

Other communication

inputs

ICTs/Tech enabling

sync interaction/Info seeking/publ’n

Capital

WirelessOtherinputs


Asia Pacific: Wireless in the network Wireless in the backbone (digital microwave &

satellites) Mobile telephony is most visible manifestation

of wireless in the access network A-P is now the largest mobile market in the

world, overtaking N. America in 2003 Growth rate of 31% (v. 13% in N. America)

But mobile/100 was only 16 compared to 52 in Europe & 35 in the Americas Signifying potential for more growth

Mobile > fixed in many A-P countries (Afghanistan to Taiwan) More than ITU data shows, e.g., India


0.00

20.00

40.00

60.00

80.00

100.00

120.00

140.00

0.0

20.0

40.0

60.0

80.0

100.0

120.0

140.0

Mobile/100 CAGR

Mobile/100 & CAGRs high-mobile Asia Pacific (end 2003)


India, fixed & mobile growth, 1991-2005

0

10

20

30

40

50

60

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Fixed Mobile


Mobile/100 & mobile as % of total in high-mobile Asia Pacific, end 2003

0.00

20.00

40.00

60.00

80.00

100.00

120.00

Mobile/100 % of total subscribers


Wireless in the access network Much of current “fixed” growth driven by

wireless CDMA 800/1900 for voice CDMA 450 & other standards for data overlay

networks Claim that 77% of world’s mobile data users

are in AP (may be more with recent 3G launches in Taiwan, etc.)

Claim of 21,000+ WiFi hotspots in AP in 2003 Strange case of WiFi in Indonesia


Typical ISP Network

Infrastructure Network

Access Network

Twisted copper pair

ADSL

Dialup

Fiber optic

Link to higher tier ISP

ISP A

ISP B

ISP C

Coax cable

Cable modem


Network Reality-Indonesia

Infrastructure NetworkAccess Network

Neighbourhood Network $35/pm

Corporate Customer $200/pm

UTP Cable

ISP A

WiFi 2.4

IIX

Ethernet Ring

Microwave

ISP B

ISP C

Ethernet

School B

School C

Wifi 5.8WiFi

5.8

HouseSchool A$4000/pm (Internet link+ international bandwidth)

ADSL

UTP Cable


Strange case of Indonesian WiFi

WiFi deployment Not inside home; not available for free Blurring of access and infrastructure network; used

as backbone; up to the curb WiFi, last mile aerial cable

Many tiered retailing of Internet service…. WHY? To recover high input costs

In addition to “last mile,” need to recover domestic & int’l leased line and interconnection costs


What gave rise to strange network configuration in Indonesia?

Regulatory environment Non-independent regulator

Two regulatory bodies: DG POSTEL & BRTI DG POSTEL is unit of Ministry of Communication & IT BRTI under-staffed, powers under transition, chairman is DG

of DG POSTEL Exclusivity clauses extending historical monopolies

Indonesian govt owns 51% share in PT Telkom & 15% in Indosat plus “golden share”

Licenses prevent ISPs from deploying infrastructure No local loop unbundling

Exclusivity until 2015 No regulation of leased lines

Few suppliers, refusal to deal, high prices, quality


Market environment

Lack of competition in infrastructure sector

Resulting in high leased line prices

High international backbone prices

Proliferation of unlicensed “reseller-ISPs”

Telecom services

Telecom operations

Fixed wireline local

Exclusive right 1996-2010 PT Telkom

Fixed domestic LD

Exclusive right 1996-2005 PT Telkom

Fixed wireless local

Limited competition (Satelindo)

Fixed international

Duopoly 1995-2004 (Indosat, Satelindo)

Mobile Competitive (Satelindo, Excelkomindo, Telkomsel etc.)

Internet service provision

Competitive Currently 124 ISPs official, 54 unlicensed

No competition

Limited competition

Competitive


Annual leased line prices: 2Mbps link

Data compiled from Lokanathan, lirneasia.net, EU 10th report, interview with Indonesian ISP & Network Service Provider

Ratios India EU1:47.9 1:3.8

RatiosIndia EU1:5.9 1:4.9

2Mbps link 2km 200km

Indonesia US$ 18,000 US$ 45,000

India US$ 376 US$ 7,603

EUBenchmark (Denmark)

US$ 4,802 US$ 9,219


Institutional aspects of wireless Current quasi-property rights regime

Bundle of rights, less right to alienate Except by selling the licensee firm

Use highly constrained (e.g., specific standards, power, polarity)

Therefore significant role for effective spectrum management by government

Government responsible for refarming of frequencies Quasi-property rights require consent

of/compensation for displaced users


Hypothetical refarming processStep Main policy actions Parallel policy actions

1 Government sets overall policy and authorizes negotiations with seven operators (O1 – O7)

2 System and frequency license modifications negotiated (Modifications include removal of technology restrictions from O1, O2, etc.; and may include extending license term of O4 (which will gain no benefits but has to yield frequencies

3 & 3A O4 and O5 release GSM 900 frequencies; O1, O2, and O3 will also be requested to agree to phased release of frequencies to enable overall ordering of the bands

1800 MHz Tender Board releases funds for band clearing (some 1800 MHz frequencies have been auctioned to GSM operators)

4 & 4A O6 assigned GSM 900 frequencies & releases CDMA 800 frequencies

1800 GSM and 1900 CDMA bands fully cleared

5 & 5A O1, O2, and O3 assigned CDMA 800 frequencies

Auction frequency slots that may be used for CDMA 1900 or GSM 1800 to current operators but possibly also to newcomers


Difficulties with refarming for unlicensing In many countries, 2.4 GHz and 5 GHz

bands have occupants who require coordination/relocation In India, EESS (active) and SRS (active) services

in 5250–5570 MHz band In Sri Lanka, high-powered MMDS broadcasts on

either side of 2.4MHz band which is also used by data licensees for 10+ years

How to find the money to pay off users who are to be moved? Beneficiaries of unlicensing cannot be

asked to pay


Spectrum management is not enough . . . For WiFi to be effective in the access network,

backbone must exist & be offered on non-discriminatory basis at reasonable prices

Data or voice communication is a chain As fast as the speed of the weakest link; if link is

broken, no communication In these markets, sustainable prices determined

by input costs ISPs require access to backbone

In some countries only access regime needs improving

In others, need to create incentives for building as well


Spectrum management is not enough . . .

In addition Market entry Interconnection and access Effective regulation of competition

Investment is what connects people


Investment is necessary condition for improved access

Figure 3: Sri Lanka telecom growth 1991-2004

0

500,000

1,000,000

1,500,000

2,000,000

2,500,000

3,000,000

3,500,000

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004

Mobile

Fixed entrants

Incumbent

Wireless


Investment is necessary condition for improved access

WirelessAccess +Backbone

Fiber + fixed access


Wireless investmentWireless-based investment in Sri Lanka, 1993-2002

0

20

40

60

80

100

120

1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

US

D m

illio

ns

Mobile Entrants Fixed Entrants


Without effective policy/regulation, inadequate/skewed investment . . .


Where the money has gone . . . Private investment in telecom 1990-2003


What is needed . . . Market entry permitted

Case of Bhutan Environment for investment created

Regulatory risk reduced Participation by multiple suppliers

enabled Level playing field


In sum . . . WiFi vision of developed countries is

possible only because of fully developed backbone and access network Enabled by environment conducive to

investment, including effective regulation Without institutional conditions, little/no

WiFi outside developed enclaves

Technology matters; but not without appropriate institutional conditions

Documents

Wireless & development in the Asia Pacific: Institutions matter Rohan Samarajiva, with contributions from LIRNEasia team High-level workshop, Annenberg