Telecom access rankings in South Asia

Posted on October 24, 2009  /  18 Comments

According to the ITU ICTeye, which is now carrying 2008 data, Pakistan’s surge to overtake Sri Lanka has petered out, leaving the Maldives (143 active SIMs/100 people) as the undisputed leader in mobile connectivity (apparently all adult Maldivians carry two active SIMs; there are only two operators in the Maldives), and Sri Lanka second with 52 SIMs per 100 people.

On the fixed side, assisted by CDMA phones that are counted as fixed, Sri Lanka is the leader (17 connection per 100 people), followed by Maldives (15 per 100).

Like in cricket, the middle of the rankings are the most interesting. Both Pakistan (50/100) and Bhutan (37/100) are ahead of India (29/100) in mobile. This shows that India cannot afford to let up the pace of 10 million connections a month for some time. If it does, it might be overtaken by Afghanistan (29/100) and even Bangladesh (28/100).

Of course, the fact that Afghanistan is ahead of Bangladesh in mobile penetration should cause all sorts of palpitations in government offices in Dhaka. Bangladesh was one of the earliest in South Asia to adopt mobile and is the most densely populated country in the world. How they were overtaken by Afghanistan, a war-torn country with difficult terrain, should cause serious re-examination of policies such as the BDT 800 SIM tax. The fact that Afghanistan’s CAGR for 2003-08 is 109%, higher than Bangladesh’s 2003-08 CAGR of 101%, suggests that the gap between the two countries is more likely to increase than decrease.

In the fixed rankings, we find Afghanistan occupying the cellar (0.37/100 people) behind Bangladesh (0.84/100). Pakistan (2.5/100) is behind Nepal (2.8/100). This is very surprising given the apparent superiority of the Pakistan policy and regulatory framework. Both use CDMA on the fixed access side, so that cannot be the explanation. Comments from Pakistani colleagues would be most welcome.

India is the only country showing negative growth in fixed over the 2003-08 period (-2%), but this simply because India is more honest in its reporting, counting CDMA on the mobile side instead of on the fixed side. For example Sri Lanka is experiencing negative growth in wireline, that is masked by CDMA growth.

From 2003 to 2008, the number of active SIMs has increased by over 12 times, while the number of fixed connections has decreased marginally, the negative growth in India wiping out all the gains in the rest of South Asia. South Asia is clearly the territory of the mobile.


  1. Congratulations Afghanistan (29/100) for beating Bangladesh (28/100). “The Afghans increase teledensity without adding any phone. They just shrink the population through mass killing,” – that was how this country was ridiculed under the Talibans. Bangladesh was far ahead of Afghanistan at that time. That is history. Now the Afghans are on their long march to make history. Beating Bangladesh in telecom access ranking is one example.

    Afghanistan introduced mobile phone in early 2002 – six years after Bangladesh did. The Afghan government has been supportive to the industry, while Bangladesh government has been erecting roadblocks. Infamous SIM tax is one of the worst examples. The industry has been subsidizing it at the cost of aggressive network expansion and customer acquisition. Annual growth of mobile in Bangladesh has come down to double digit.

    Unlike Afghanistan, the mobile covers over 95% territory and population in Bangladesh. Let Dhaka sleepwalk backwards. Kabul knows its job.

  2. WTH is this BD has 50 million Mobile connection, with a population of 150 Million. That’s around 33/100. itu is an organization that ALWAYS uses laggard data to depict BD in a certain way. What an utter BullSh*t.

  3. @Amar

    I understand your unhappiness, but this is a necessary feature of using comparative data. The ITU does not have 2009 data for all the countries. Its latest comprehensive data are for 2008, so that is what is used for comparison.

    India is adding 10 million SIMs a month, so I imagine Indians will be even more frustrated than you are.

    However, this does not make the comparison valueless. At a particular point in time, all these countries were measured (the data were reported by the respective governments, not collected independently by the ITU). At that point, end of 2008, Bangladesh was behind Afghanistan and India. That is a fact and it should be cause for concern.

    Bangladesh may have added 6 million SIMs after the data was reported, but so have India and Afghanistan; perhaps more. The SIM tax has slowed new connections in Bangladesh considerably, down to single digits perhaps, so I think it is unlikely that it will catch up with India and Afghanistan.

  4. I am venting my frustration because the data that they use are so out of whack. The data for total population and active/total cell subscribers are not clear. BTRC started maybe a year ago to count ONLY active sim so how does this relate to the data from other countries. Thus at this point it like comparing apples and oranges.

  5. Case in point: Vietnam, the ITU reported 70 million cell subscribers by the end of 2008. And as you mentioned was provided by the Govt. Now by their own admission, the Ministry of communication has just said “Vietnam Says Half of Official Subscribers May Be Inactive” Now when the ITU compares countries like Bangladesh, India, Vietnam all in the same benchmark, in the same context, they are basically fooling people. Investors (Journalists) all around the world look at these stats to make investments and form opinion about a particular nations. And these falsified stat just bring negative connotation to BD. Maybe the BTRC should also be embellishing/inflating the mobile subscriber data to fit into the Govt’s agenda of “Better Image” to the world.

    1. @Amar, Again, no dispute about the need for standardized reporting. See our unhappiness about Pakistan numbers that were overstated by 10 million:

      But as I said, ITU gets the numbers from the governments. Governments can tell ITU to use standardized definitions, not us. Whether you or I like the numbers or not, people all over the world use ITU numbers. There is no alternative to them.

  6. This is hard to believe that Afghanistan is ahead of Bangladesh in terms of, well, anything that is positive. Except Kabul, do they have any stable government? How they install those base necessary for 29% penetration? How come those base stations remain safe amidst all those machine gun and mortar fire?

    Surely, Rohan you have got all the answers. Just show me the RIGHT link where I can read about it.

    Off topic: I saw you in Dhaka at the South Asian Mobile Summit. I was one the delegates from Grameenphone.

  7. @Hadi,

    I am with you in being surprised. But we all knew that there was considerable growth in mobile, driven by the systematic and transparent licensing policies.

    Despite a few instances of towers being blown up, for the most part the Taliban allows the mobile networks to function. As Roshan CEO Khoja says, they operate in all of Afghanistan, unlike the Karzai government that operates in “Kabulistan.”

    Also, you must not make too much of headlines. Sri Lanka achieved high growth in the middle of the 30 year civil war and a tsunami. People live and work in Afghanistan every day, despite a few bombs here and there.

    Again, I want to emphasize that our comment is on the ITU data, as supplied by the respective governments, not data that LIRNEasia has collected.

    1. Thanks Rohan for your reply. I have got all my answers.

      Its the GOVERNMENT which provides all these data……….that basically settles it.

  8. So now we can agree that the BD Govt should inflate the numbers to make us look good. Can’t believe this has come down to this.

    1. I think the number for Bangladesh is just OK. Only we have to consider the 30% multi-SIM users.

    2. “We agree” seems to include people other than yourself. Sorry, that may be your conclusion, but not mine.

      Governments collect data from operators, compile them and submit to ITU. There may be errors due to lack of standardization of definitions, but these are common across all countries and operators. Governments periodically correct these errors, as has been shown in Pakistan and Vietnam. This does not support the conclusion that the Afghan government is purposely inflating the numbers.

  9. Nirmali Sivapragasam

    Official statistics published by each country’s telecom regulatory authority/governmental websites (MCIT in Afghanistan and BTRC in Bangladesh) also reveal similar findings. The no. of GSM/cellular subscribers as of june 2009 are as follows:

    Afghanistan: 10.4 million (
    Bangladesh: 46.69 million (

    Using 2009 estimates of population numbers reported by the IMF ( , the no. of GSM/cellular subscribers per 100 comes to 36.0 and 28.9 for Afghanistan and Bangladesh, respectively.

    Mobile subscriber numbers for Bangladesh have since increased (see September data:, but per 100 figures are still below that of Afghan data for June 2009)

  10. @ Hadi

    Does Grameenphone consider the 30% multi-SIM users while it reports to BTRC every quarter?

    1. Well, you got it all wrong. This is not about Grameenphone. It is an ESTIMATED figure based on the entire Bangladeshi mobile phone market. I found it in some article (can’t remember right now).

      I think its not possible for any operator to know for sure whether their customers are using other operators’ connections.

  11. @Hadi: even war lords (including the Taliban) included want to be connected. we see this in Africa also. so don’t be surprised that telecom networks are “allowed” to operate.

    Also, don’t underestimated the over-reporting of SIMs by Grameenphone. It’s the largest operator. So 30% of that is huge. You are right – an operator doesn’t know if it’s subscribers are using SIMs of competitors. But they CAN and do know if their own SIMs have been used (for a fee generating service) in the past X months (x being 3 months, or 6 months or what ever time frame that the regulator should deem as “active” for a SIM).