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).

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

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

Afghanistan: 10.4 million (http://www.mcit.gov.af/)
Bangladesh: 46.69 million (http://www.btrc.gov.bd/newsandevents/mobile_phone_subscribers/mobile_phone_subscribers_june_2009.php)

Using 2009 estimates of population numbers reported by the IMF (http://www.imf.org/external/pubs/ft/weo/2009/02/weodata/index.aspx) , 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: http://www.btrc.gov.bd/newsandevents/mobile_phone_subscribers/mobile_phone_subscribers_september_2009.php), but per 100 figures are still below that of Afghan data for June 2009)

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.

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

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.

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.

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.