This colloquium was conducted by Dimuthu Ratnadiwakara on a forecasting methodology for mobile penetration for years up to 2012 for Sri Lanka and Pakistan that is being developed. This method employs a simple mathematical process with less computational consumption.

The analysis was performed using historical teledensity figures together with T@BOP data and demographic data for each country. A logistic curve was fit for each country’s data after estimating an upper bound for the mobile teledensity.

The question was raised as to whether it possible to assume that people at the BOP own more than 1 SIM? The figures show 14% of the BOP having more than 1 SIM in Pakistan.
Harsha de Silva replies that this occurs because they use it as a cost cutting strategy.
Rohan Samarajiva noted that in Sri Lanka too this phenomenon is present.
Harsha de Silva noted that this percentage could fluctuate with policy changes.
Therefore it is important to state this as an underlying assumption. Helani Galpaya noted that it also depends on price/marketing action. eg giving away of free SIMs.

Professor Jhunjhunwala noted that an important factor when deciding to buy a mobile phone was coverage. As the BOP survey shows, within the rural areas coverage is one of the main factors.

Rohan Samarajiva noted that the an assumption needs to be made that coverage does exist. Also instead of having a precise figure, we should aim to forecast a band which consists of the lowest possible figure and the highest.
Professor Jhunjhunwala noted that telcos are constantly trying to get the people to buy connections. However, the infarstructure needs to be in place. But why is the saturation level so important?Rohan Samarajiva stated that if we have the worst case scenario predicted we can make an argument for certain policy processes/ strategies to help achieve the higher predictions.

Is market share of each operator a more important factor for measuring success?
Harsha de Silva replied saying that we do not have this data from the potential owners, but we do have this for existant owners. However, coverage does determine what operator a potential user will use.

Dimuthu Ratnadiwakara noted that the sampling error has to be atleast below 5%.

Dimuthu Ratnadiwakara stated that the forecast will also change depending on how many young people exist within the population.
Harsha de Silva commented that this affects the growth because it either increases or decreases the purchasing pool.

Rohan Samajiva noted that two assumption curves should be made using different assumptions.

The PTA data shows that the mobile penetration numbers are increasing at an accelerating rate. This seems to be consistant with the forecast.

Rohan Samarajiva noted that the multiple/ duplication of sims is a growing phenomenon. Harsha de Silva and Chanuka Wattegama noted that some countries are expectations. eg- Philippines has a high duplication rate but this maybe due to the tourist industry. With more tourists purchasing local SIMs for use when touring the country.

Harsha de Silva noted that a proxy should be adopted when trying to explain the affordability gap. Because the non poor people will be adopting the technology faster than the poor. And hence this will make the forecast higher.

But how can we take this study further?

Harsha de Silva noted that we can use it to study the assumptions that have a policy impact. eg- coverage.

Does fixed phones have a future?

Harsha de Silva replied saying yes, as affordability issues still exist.

Is this gonna be more to supplement ADSL; and the demand for higher speeds?

Professor Jhunjhunwala noted that wireless internet still cannot provide internet services at 1Mbps or greater, which is necessary for new technologies.

Harsha de Silva noted that this cannot be the end of the road for a copper connection.

Helani Galpaya noted that fixed lines maybe used for other services like ADSL. The demand for this service for making voice calls may decrease but the demand for a fixed line will exist.

Professor Jhunjhunwala noted that fixed, wireless will be able to compete to copper wire services. This is especially pertinent to the developing world that does not have the copper infrastructure built into the system. However, this fixed: wireless technology will be adopted only in the event where the copper wire does not exist.