Universal service in Telecom: Regional experience

1
Universal service in Telecom Regional experience

• Rohan Samarajiva
• SAFIR workshop on Universal Service Obligation
• New Delhi, 14 July 2007

2
Agenda

• What does research say about regional experience
  on USO?
• Findings on regulatory environment on USOs
• Explaining the elative positions of India and Sri
  Lanka
• General conclusions
• Studies on teleuse at the Bottom of the Pyramid
  (by those who are the intended beneficiaries of
  USO)
• Characteristics of current owner-users
• Barriers facing non-owner users from owning and
  what can be done by governments, operators and
  equipment manufacturers

3
Telecom Regulatory Environment (TRE) study, 2006

• Panel study in six countries (reduced to five
  during analysis)
• India, Indonesia, Pakistan, Philippines, Sri
  Lanka, (Thailand)
• 40-101 respondents per country in four
  categories, statistically normalized
• Likert scale scoring across six dimensions of
  internationally accepted regulatory performance
• Diagnostic tool for assessing regulatory
  performance and communication

4
Universal service obligation
Highest Mobile
Highest Fixed
5
Universal service

• Every country is below the threshold of
  satisfactory performance of 3.0/5 (midpoint)
• Not so on all other dimensions
• India has worlds second largest USO Fund and has
  done several disbursements
• Yet, worst performance among the dimensions and
  within the group
• Sri Lanka which had done almost nothing re
  universal service and does not collect USO
  contributions from operators (only from
  international traffic) gets high scores
• Less is more?

6
Interrogating the Indian score

• India
• Does quite well overall (tied for 2nd place in
  fixed and 3rd in mobile), and
• Is head-and-shoulders above others in tariff
  regulation (3.7/5 in fixed and 3.5/5 in mobile)
• Compared with 1.9/5 in fixed and mobile for
  universal service (lowest of six dimensions)

7
Possible reasons

• USO levy in India (one of the highest) is 5 of
  adjusted gross revenues of all operators except
  pure value-added service providers
• Transparent means of disbursing the funds, but
• Until 2006, collected from all, but given only to
  fixed
• Most funds went to incumbent, due to design
  biases
• The latest round of least-cost subsidy auctions
  was the best designed attracted low to zero bids
• Getting low bids in least-cost subsidy auctions
  should not be a cause of unhappiness
• But now DoTs problem is getting rid of funds!

8
Possible reasons

• Worldwide, USD 6b collected (mostly in Brazil
  India) but USD 4.4b unspent (also in Brazil
  India)
• Assuming levy percentage unchanged on an
  expanding base, at least another USD 3.8b will be
  collected by 2010 (GSMA)
• By March 2005, INR 72.4b (USD 1.6b) collected in
  India, but only 25 spent (41 by end 2005-06)

9
USO collections disbursements in India,
2002-2010
Source Malik, Payal (2007). Forthcoming book
chapter.
10
What of Sri Lanka?

• Generally a laggard in TRE performance (4th out
  of 5 in fixed and last in mobile) but
• Shares first place for mobile in USO dimension
  and second by a pip for fixed
• Still below the midpoint (3/5)

11
Why?

• Has NO general USO
• 1997 privatization agreement precludes the levy
  of USOs from the incumbent
• By extension, its competitors cannot be subject
  to USO
• Small universal service initiatives such as
  connecting rural post offices and payphone
  subsidies funded directly from general taxation
  or regulatory fees
• Only source of USO funds is levy on international
  calls resulting from 2002-03 reforms
• First tranche of LKR 7b (USD 70m) disbursed in
  2007 April, after the TRE study was conducted

12
Possible conclusion from LK-IN relative positions

• Country that does little to nothing on USOs
  scores higher than the country most active in
  USOs, with the largest fund and the highest levy
  
• Sri Lanka rewarded for doing the least harm
• Indictment of USO policies and implementation

13
General conclusions

• Public Finance theory tells us that the source of
  US funds that distorts the least is general
  taxation
• India NO Sri Lanka YES
• Taxing a sector (India YES) causes distortions,
  so best to tax as little as possible (India NO)
• Sri Lankas tax primarily affects incoming
  international calls and causes minimal
  distortions and protects the most vulnerable
  customers
• If original design of decreasing percentages
  (down to zero by 2008) was kept, even better

14
General conclusions

• Once the tax is collected, it should be spent as
  quickly as possible
• Both India and Sri Lanka fall short, but the
  magnitude of the Indian failure is larger
• When the tax is spent, it should be spent fairly
• In the early rounds, major flaws in the Indian
  auction design resulted in most of the money
  going to incumbent
• Privileged wireline fixed access over wireless

15
Consequences of high USO levies

• USO levies dont come from companies they come
  from the existing customers and non-owning users,
  who include the rich, but increasingly the poor
• Teleuse _at_ the Bottom of the Pyramid
  representative sample survey conducted by
  LIRNEasia/ACNielsen in 2006 illustrates the
  problem

16
Bottom of the Pyramid defined

• Many definitions of poverty, but this study uses
  SEC D and E between ages 18-60
• SEC does not take into account income, but it is
  closely related to income levels

excluding FANA/FATA Tribal Areas excluding
NE Provinces
17
Overall access is very high

• Most have used a phone in the last 3 months

18
But many BOP users dont own phones
19
Current owner users
20
BOP makes/receives around 1 call per day in South
Asia
Source Diary
21
Calls are short
Source Diary
22
Mostly local calls at the BOP
Source Diary
23
Non-owner users

• Primary beneficiaries of USO

24
Key barrier to ownership is affordability
25
The expected cost of getting connected
26
compared to what can be afforded
Indian BOP expects initial cost to be as high as
USD25 but many can only afford USD10
27
Expectations affordability are more in-line re
monthly charges
28
(No Transcript)
29
Adding up to the next billion?
30
What is to be done?
31
Mostly, reduce entry costs for the BOP

• Policy makers/regulators
• Sustain competitive conditions
• India has lowest HHI concentration ratios
• Improve regulatory environment on all fronts
• Universal service and spectrum management seen as
  most problematic by stakeholders responding to
  TRE questionnaire
• Rethink tax regimes to remove barriers ownership
  and use
• Step down USO tax to zero ASAP
• Move away from taxing handsets

32
Mostly, reduce entry costs for the BOP

• Policy makers/regulators
• Accelerate USO Fund disbursements and
  infrastructure sharing to increase rural coverage
• Encourage well-functioning second-hand market
  (e.g., Pakistan)
• Exclude stolen sets by mandating equipment
  registration

33
Mostly, reduce entry costs at the BOP

• Telecom operators
• Leverage benefits of direct access security,
  keeping in touch
• Improve coverage
• Infrastructure sharing is the right way to go
• Solve the connection/use charge problem
  installment plans?
• Collaborate on meaningful content and
  applications for BOP
• e.g., SMS-based remittances
• Handset manufacturers
• Ultra low-cost phones with warranty
• Affordable, functional handsets (e.g., with local
  language SMS capability)

34
In sum

• Research tells us that present approaches to USO
  in the region are not working
• Need to understand the needs of the intended
  beneficiaries of USO policies and what is
  preventing them from having the market meet their
  needs
• At least in telecom, strong case for phasing out
  universal service levies

35
samarajiva_at_lirne.net
36
Method

• Panel studies using as small a number of
  questions as possible
• Effective perception studies are conducted using
  one question alone
• Need to make minimal demands on time of senior
  respondents such as CEOs
• Use a large number of respondents to balance out
  the biases also use statistical methods to
  reduce bias caused by group representation
• Use traditional social-science perception
  instrument 5-point Likert scale
• Focus on telecom regulatory environment not on
  regulatory agency per se
• In India, three dimensions reflect TRAI
  performance three reflect DoT performance
• In Pakistan, PTA controls five

37
Method

• Five dimensions from GATS regulatory reference
  paper (document with greatest international
  legitimacy) price regulation
• Others under consideration
• Service-quality regulation
• Consumer protection
• TRE studies being conducted in Latin America and
  the Caribbean and Africa
• Asian six-country study conducted in
  August-October 2006
• Building on pilot conducted in 2003
  http//www.regulateonline.org/content/view/207/65/
  

38
Panel composition categories

• Category 1 Operators, industry associations,
  equipment suppliers
• Category 2 Financial institutions, private
  investment houses, banks and credit rating
  agencies
• Category 3 Educational/research organizations,
  telecom consultants, law firms
• Category 4 Journalists, telecom user groups,
  civil society, former members of regulatory/other
  government agencies

39
Panel Composition
40
Detailed analysis, across each of six dimensions

• Countries arranged in order of per capita GDP
  (Purchasing Power Parity), with Pakistan (PK)
  lowest and the Philippines (PH) highest
• Numbers reported are average scores, rounded to
  one decimal
• Midpoint (3) as threshold of adequate performance
• Analysis by each of the six dimensions can be a
  useful diagnostic tool
• Does the score reflect the perceptions of the
  regulatory agency?
• If not, perhaps more should be done to
  communicate its actions

41
Market Entry
Highest Mobile
Highest Fixed
42
Access to scarce resources
Highest Mobile
Highest Fixed
43
Interconnection
Highest Mobile
Highest Fixed
44
Tariff regulation
Highest Fixed
Highest Mobile
45
Regulation of anti-competitive practices
Highest Mobile
Highest Fixed
46
fixedOverall TRE scores
Maximum score
Mid-point
Minimum score
47
MobileOverall TRE scores
Maximum score
Mid-point
Minimum score