An interview with Parvez Iftikhar, Member PM’s Taskforce on IT & Telecom
A telecom-industry veteran, Parvez Iftikhar is the founding CEO of Pakistan’s Universal Service Fund (2007-2011), which is a public-private telecom development organisation. Previously, he also served as the country head of Siemens Telecom, Pakistan. After leaving USF, Parvez has been active in ICT policy advisory. He has been consulting for organisations like the International Telecommuni-cation Union, USAID, CTO and the World Bank in various countries of Asia and Africa. BR Research recently sat down with him in Islamabad to discuss issues facing the ICT industry in Pakistan. Selected excerpts are provided below:
BR Research: Let’s start with the official business. You are part of the PM’s IT & Telecoms Task Force and the Chairman of its telecom subgroup. What kind of issues has the task force taken up so far?
Parvez Iftikhar: Well, the task force has been reaching out to stakeholders, trying to take stock of barriers and roadblocks to ICT development, and formulating recommendations for concrete steps on how to remove those hurdles. Currently we are looking at various issues, like how to achieve massive fiberisation, the license renewal process for cellular operators, a long-term spectrum plan for the country, addressing lack of ICT-related HR capacity in the government, budgetary measures for telecom and IT industry, and so on.
Due to the coming budget, taxation is going to be the number one agenda item in the near future. We are giving proposals to rationalize the taxes on the ICT industry so that it is no longer the highest-taxed sector in the country. Specifically, we will recommend lower taxes on smartphones so the entry to internet can become affordable.
BRR: Why is there a holdup in the renewal of licenses for three operators this year and, for the future?
PI: The license renewal process should have started 24 months ago, as per laid-down rules. However, that was not to be. Last year was consumed by political transition. Also, a full-time Chairman PTA was not appointed for a year. Now with all things in place, and the deadline approaching, there is an urgency to get this issue resolved.
The holdup is mainly on the price tag. Should the price of renewal fee be $291 million, like originally in 2004? Should it be the last auction price of 2016-17, which was hardly an auction? Or should there be a new auction to establish the price? Different stakeholders have different opinions. There were a lot of open and closed debates. A committee was formed by the PM to reach a decision – but its legality was questioned. You know that there are always people ready to throw a spanner in the works whenever anything positive is going to happen in the country. Anyhow, the Cabinet is expected to decide this issue hopefully within this month.
BRR: In your view, what kind of a license-renewal mechanism suits this market?
PI: My own view is that given the market situation, the price of license renewal should be kept low, ideally the original price of $291 million. I question the very concept of using a license fee to bring a windfall for the government. Selling license is not a way to make money anymore. If you do it for short-term cash, you are closing your eyes to the big picture.
The more an operator pays for a license, or its renewal, the less it spends on expanding and upgrading its network, where the real economic benefits lie. Network expansions bring investment; create jobs, not just while building networks but also as a result of internet expansion and adoption. Moreover, if high renewal fee is charged, there will be low appetite for new spectrum because operators would have less to spend. Without new spectrum, the pathetic quality of broadband would never improve.
I would rather give the operators tough ambitious targets of quality and roll-out, instead of charging high license fees.
BRR: Well, let’s see how things settle on that front. Let’s turn to the mobile broadband market and the ICT ecosystem. Five years on, how far has the market for mobile broadband (3G and 4G) developed in Pakistan?
PI: The main positive result of the 2014 mega-auction was that Pakistan increased its broadband subscribers from 3 million back then, to around 60 million now. That was great. But after the first two years, the growth stalled because there were a few follow-up steps that were not pursued after the auction to increase broadband adoption.
For instance, the government should have tried to improve broadband’s affordability and to raise public’s awareness of broadband’s beneficial uses, for example by aggressively introducing e-Government services. That’s one of the reasons why the operators did not see profitability beyond the most lucrative areas of operations. Beyond the dense urban areas, the operators did not expand as much as they could have. I don’t have exact figures for mobile broadband coverage as of now. But even within cities, there are big holes where there is no 3G and 4G.
Then the government brought very little additional spectrum to the market, and that, too, at exorbitant floor prices. Consequently, only those operators went for it who were just too desperate.
BRR: What have been the barriers to higher broadband adoption?
PI: The main barrier has been the cost of the smartphone. A GSMA study names smartphone price as the biggest barrier against broadband adoption. The prices of broadband devices are high not just because they are imported and hence prone to currency depreciation, but also because of various kinds of taxes.
Thankfully, prices of broadband services are not a barrier because per megabit prices in Pakistan are one of the lowest in the world, due to vibrant competition. But, of course, there are barriers like awareness and perceived usefulness of broadband internet.
BRR: With Warid absorbed into Jazz, it’s a four-player market now. Do you think there is room for further consolidation?
PI: Right now, I don’t think we are going for any further consolidation in that space. There are bigger countries with just two players and there are smaller countries with five players. We can do with three operators, and we can also do with four – I think the current number of players is fine. And I don’t see a new operator coming in either. For that, the present lot has to be making good money on their investments.
BRR: The decade is about to turn and with it the next “G” is also in the offing. How premature is it to talk about 5G coming into Pakistan?
PI: I think it is a bit premature to talk about 5G when even 4G has not been adequately deployed in Pakistan. There are several issues that we need to fix before some degree of readiness can be achieved for 5G.
The number one issue in 5G deployment is going to be that of spectrum. You need to have a lot more spectrum to operate 5G than is currently made available to the market. Currently, we do not have enough spectrums even for 4G, which brings us back to the issue of spectrum pricing. With low ARPUs (average revenue per user) that we have, if the spectrum made available is priced high by the government, the service providers will go only for the bare minimum required, and 4G quality will remain pathetic.
Secondly, we have terrible lack of fiber-optic connectivity. 5G will require a lot of fiber for its small-cell configuration to penetrate deeper. But currently, a very low number of cell towers are connected to fiber-optic network. In Thailand, 90 percent of cell towers are connected with fiber optic; in Malaysia, the ratio is 60 percent, in India it’s 27 percent and in Pakistan it’s less than 10 percent. So we need to have much more investment in fiber optic connectivity. Incidentally, there too the custodians of ‘Right of Way’ ask for unreasonably high fees.
And lastly, there is insufficient use-case for 5G at present. As for now, the 5G connectivity is more attractive for use cases like driverless cars, online gaming, online medical care, and to carry out a variety of remote operations where latency is critically important. Even in the West, they are still struggling to identify viable commercial use cases for 5G.
My assessment is that it would be at least another three years before Pakistan can think about entering the 5G space. Part of that assessment is influenced by the above-mentioned limiting factors, and part of it is explained by the nascent 5G ecosystem that is developing globally, like 5G consumer devices. The ITU hasn’t even agreed on 5G specifications yet. Later this year in October-November, they hopefully will. Be that as it may, in the next couple of years, we won’t see 5G deployments going beyond the main urban centers, even in the developed countries.
BRR: You mentioned fiber-optic connectivity as one of the limiting factors. What is the optimum level of fiberisation for Pakistan to adequately serve the urban and rural areas?
PI: I don’t have exact numbers, but Pakistan has more than 30,000 km of optic fiber cables. Most of the fiber network is running along the Indus River. But more important is the spread, as to how dispersed that network is. Is fiber going to each and every union council at least? That is not the case. However, thanks to USF, every Tehsil headquarter has fiber-optic connectivity.
That’s about long-distance fiber. Also important is how much metro fiber there is. Even cities like Islamabad or Rawalpindi do not have fair amount of metro fiber. But in smaller cities, there needs to be even more concentrated effort for it. The government doesn’t have to do this. Operators will do it. The government, however, needs to make it easier for operators to lay the fiber. Presently that is not the case; difficulties with ‘Right of Way’ being one example.
BRR: There is also the issue of inefficient spectrum management. What can be done to free up the unused spectrum and have regular spectrum auctions for private-sector use?
PI: There are three kinds of spectrum. One which has been allocated to wireless operators, and is in their use. One which has been allocated to other users, and is also in use. And one which has neither been allocated to anyone, nor is it being used, so it’s lying idle. This latter kind of unallocated, unused spectrum gives rise to inefficiencies in spectrum management. Spectrum doesn’t have any intrinsic value. It has value only if it is used.
Spectrum that is lying idle is losing money for the country, and hurting the country’s development. Such unallocated spectrum is in hundreds of Megahertz, about the same as spectrum allocated to the operators. More spectrums can also be freed up if the unused spectrum scattered randomly is re-farmed. But that’s secondary. First, the authorities need to auction the unallocated idle spectrum, starting with reasonable floor prices. Even that would stimulate some economic activity.