India, Pakistan and Indonesia have improved Telecom Regulatory Environments since 2008, according to stakeholders.

Posted on July 15, 2011  /  14 Comments


According the LIRNEasia’s 2011 Telecom Regulatory Environment (TRE) survey, stakeholders in India, Pakistan and Indonesia have identified the telecom regulatory environments in their countries as improved since 2008, the last time the survey was carried out.   In contrast, Bangladesh, Sri Lanka, the Philippines have seen the regulatory environments decline in effectiveness, while Thailandremains more-or-less the same.

The TRE Survey asks senior level stakeholders to evaluate the effectiveness of the telecom regulatory environment in the fixed, mobile and broadband subsectors along a Lickert scale of 1 to 5 (1 being highly ineffective and 5 being highly effective, with the mid-point of 3 being considered average performance).  Seven different dimensions of regulation (market entry, tariff regulation, interconnection, universal service, anti-competitive-practices, quality of service) are evaluated by the stakeholders.    This year, 349 responded participated in the 7 countries.

Within a country, scores for each of the dimensions reflect specific issues: for example, in India, the lowest score (of 2.3 out of 5) was received by the Access to Scarce Resources dimension in the mobile-sub-sector.  This is perhaps not surprising given the 2G scandals in India.  However, India did finally get around to allocating 3G spectrum in 2010, and did so by having its first ever spectrum auctions. Perhaps because of this, or because stakeholders believe that that the 2G scandal has finally paved way for transparency in allocation, the score of 2.3 this year is still an improvement, though marginal,  over the 2008 score of 2.2.   India’s USD 4 billion+ undisbursed Universal Service Fund and related policies are responsible for its biggest TRE score decline: the TRE for USO drops from 3.1 in 2008 to 2.4 this year.  In contrast, tariff regulation in the mobile sub-sector continues to be the top performer with a score of 3.9 out of 5.0, indicating stakeholder satisfaction at TRAI’s policy of forbearance which has enabled Indian consumers to enjoy extremely low prices thanks to competitive forces.

Pakistan saw an increased in almost all dimensions, with the exception of 3 (fixed market entry, mobile access to scarce resources and mobile interconnection) which showed minor declines.   In contrast, Bangladesh saw scores in all seven fixed-subsector dimensions decline, in some cases by as much as 1 point.  The scores are perhaps reflective of the issues related to the cancellation of several fixed licenses.  Overall only seven dimensions showed improvements in Bangladesh, and even these were marginal.   Thailand, whose overall performance is unchanged, has however seen significant declines in its Market Entry scores due the uncertainties caused by the concession contracts granted to the mobile operators and what their status would be when they expire starting next year.

Detailed analysis of the scores is available in the draft country reports for Pakistan, India, Thailand, Philippines, Sri Lanka, Indonesia, and Bangladesh.

A comparative country analysis is also available online.

All reports and analysis is in draft form at present. We hope you will comment on the country reports and cross country comparisons so that we may improve our analysis.


  1. Chanuka Wattegama

    I have few comments on the Sri Lanka paper.

    Firstly, it is indeed good research. Congrats. My students will use it as reference material. Having said, that I do not want to spend time telling why it is good. Instead I will give few suggestions for improvements.

    Secondly, I am bit surprised about its length. It is 87 pages, while I was instructed to cut the Indonesia paper to 30 – 40 pages. (which I didn’t – I think a paper should be judged by not its length, but content) Anyway, this is not my concern. More the content, more useful it will be.

    Okay, here are my points:

    1. Page 11- Table 1 – CBSL Annual report for 2009 cannot have information on 2010. Since 2010 annual report is out why not take a little trouble to update all the data from it? Far better quoting CBSL than using data from LBO, FT etc. (I know you did the paper before March 2011)

    2. Page 12 – For ICT literacy please refer only to Department of Census and Statistics. Please don’t use data cooked up for promotional purposes. I explain this issue in detail in

    3. Page 17 – Table 7 – Annual Cost of a 256kbps Broadband Business Connection is USD 168 in Sri Lanka? I pay only Rs. 890 (USD 8) for a 1 Mbps link. You may refer to dedicated lease line rentals – not broadband.

    4. Pages 23 & 23 – TRC give the number of mobile broadband subscribers as 294,000 for Dec 2010 which contradicts with the data in your paper. Again 280,000 is not fixed broadband connections. That can and I think it does include the dial-up connections too. (Not sure you may clarify with TRC) Again, I do not think the market share data you provide are accurate. The round figures are not good for the credibility of the figures. (also in Figure 6 , 40%, 40% and 10% does not add up to 100%) Guestimations are okay, as long as you say so.

    5. Pages 23 – About the so called ‘price wars’ I have a fundamental difference about the argument, but don’t try to push my point. You love price controls. I love markets. We both can agree to disagree. Anyway, ‘Upahara’ is not the only or even main reason for 2009 dip in telco profits. That was a bad year for so many at global level.

    6. Page 47 – [Quote] Operators point out that under the earlier system they could set off the VAT incurred on their expenses by the VAT charged from consumers. The new tax regime has exempted operator revenue from VAT which leaves them with nothing to set off operator VAT against, which in turn increases their cost by 12 per cent. [Unquote]

    This is operators’ version. The fact is earlier they charged VAT to consumers but didn’t pay fully to the government. So the actual telecom charges were 12% higher than what had been advertised. Now, under the simplified tax structure operators cannot play that game. So if they can’t meet the operational costs, let them increase prices. It is the market. Who prevents them?

    Finally (and with direct reference to my last point) we normally work on the presumption only the government is inefficient, corrupted and selfish. We assume the operators are efficient, smart, not corrupted and always honest. To say the least, this assumption is wrong. Period.

  2. Chanuka Wattegama

    I have a question (to Sri Lanka researchers) on the following para on page 58.

    [Quote] As seen in Figure 25 below the results for regulation of anti‐competitive practices are the low in the fixed and mobile sub‐sectors. This perception reflects the arguments set out in Section 6 such as the partially government‐owned fixed operator SLT and mobile operator Mobitel and the perceptions of cross subsidizations existing between them. [Unquote]

    May I know who subsidises whom? (at least the perception)

    Is it SLT subsidising Mobitel or vice versa?

    1. Buddhika Brahmanage

      @ Chanuka: We were referring to perceived cross subsidization of Mobitel by SLT.

      1. Chanuka Wattegama

        Thanks, Buddhika. Okay, now let us look this perception with the real numbers.

        According to your own data SLT has 58% of fixed lines. That is 2,088,000. Mobitel has 24% of mobile SIMs. That is 4,320,000. Please note the latter is more than twice of the former.

        So if SLT were to cross subsidise Mobitel – which has roughly a twice bigger market, the former should be earning super-natural profits. Please note unlike a mobile connection a fixed connection is used almost always for voice calls. So the revenue is limited. I would really like to do this calculations with ARPUs but I do not have figures.

        Can SLT earn super-natural profits (a) in a market in which it no more has a monopoly and (b) when fixed is almost a dying market?

  3. @Chanuka: Thanks for your comments. All but two have been already given as feedback to the LK SPR team so hopefully will be reflected in the udpated version. The exceptions are number 6, and your last para about corruption/efficiency of operators vs. government.

    As for yoru question about x-subsidising, I’ve informed the IPS team and asked the team to respond.


  4. Chanuka Wattegama

    Final para is not a stand alone point and it is the explanation of No. 6. Sorry, I cannot and do not want to write everything I know here.

  5. Chanuka Wattegama

    The Sri Lanka SPR/TRE paper quotes Rohan Samarajiva Ph D, saying that TRC is not a profit making body but a regulatory body, and that the profits generated at the cost of revenues from the industry could have been employed for the development of the sector.

    Interestingly Provincial Council member of the main opposition party Shiral Lakthilaka brings a diametrically opposite argument accusing the TRC for not generating the profits it easily could (

    Demonstrating the archetypal immaturity common to most UNP politicians Lakthilaka compares India telecom market with Sri Lanka’s, which is more a pumpkin to a peanut than an apple to an orange. Still he got a point. TRC could have certainly earned more if it went for auctions.

    It would be too naïve to assume TRC not knowing the price of 3G licenses. A more realistic hypothesis is the difference between the real and nominal prices of the licenses have ended elesewhere.

    So Lakthilaka, in a way adds to the last para of the first comment I made. Any transaction is not complete without two parties, a taker and a giver. :-)

  6. My comments about the TRC not being a revenue-earning entity should not be interpreted as a blanket opposition to auctions of licenses.

    The decision to conduct auctions for licenses is a policy decision. In almost all cases, the revenues from auctions go to the government. These are one-off revenues that should not go to the regulator.

    The TRC, is by law, required to implement policy directions given by the government. In 2003, it conducted an auction for GSM 1800 band slots that yielded approximately USD 2 million. These funds were used to clear the band by compensating the previous occupants. This is a rare case of spectrum refarming in the developing world and has been written about.

    It is important to distinguish between the conduct of auctions and the imposition of additional costs in an ad hoc manner. Pakistan conducted auctions for the 4th and 5th GSM licenses in 2003-04 which yielded USD 291 million apiece. This was announced as the fee that would have to be paid by the previous licensees at renewal time: the auction in the end yielded USD 291 million x 5.

    That is a serious amount of money. Yet, given the reduction of regulatory uncertainty overall, Pakistan saw massive increases in FDI and as a result improvements in overall sector performance in terms of increased connectivity, low prices and increased choice.

    But I want to emphasize that support for transparent auctions for licenses does not mean that auctions should be held every few weeks for every input an operator requires. That would make business planning impossible.

    What is required is a balance between transparent allocation of scarce resources and licenses on one hand and stability and certainty on the other. Ad hoc revenue raising creates regulatory uncertainty. Not conducting auctions for operator licenses does not reduce costs to the operators; it just reduces revenues to the government.

  7. /In almost all cases, the revenues from auctions go to the government./

    The revenues collected from license renewal charges, spectrum fees, and a CESS of 2 per cent too goes to the government. Correct me if I am wrong, but as far as I know the regulator does not keep them. (The pathetic salaries some of my friends receive at TRC do not indicate so.)

    So what is the difference? Why one is wrong and the other is right?

  8. Sorry about distracting from the main point which is the balance between transparency and certainty.

    In Sri Lanka all the fees that are collected go to the TRC Fund. They are moved in bulk to Consolidated Fund at various times. This issue is relevant in certain settings such as Bangladesh where the regulator has discretion to use the funds directly as bonuses and thus incentives of organization and govt differ. Not to Sri Lanka.

    My main argument stands independently of the 2nd para.