All submarine cables connecting the Far East with Europe and Africa transit at India. It has made 12 submarine cables (six owned by consortiums and six privately-owned) hopping into 10 cable landing stations (CLS) at the Indian seashore. Voice and data traffic of 27 international long distance operators (ILDO) are processed through the 10 CLS. Four (Tata, Airtel, Reliance and BSNL) out of the 27 ILD providers own respective CLS in India.

The ILDOs who don’t own CLS told TRAI that Tata Communication and Bharti Airtel together enjoy a 93% market share. They alleged although average cost of submarine cable bandwidth has dropped significantly, the average Access Facilitation charges at the Indian CLS remained unchanged during the last four years (until 2012). They blamed the Access Facilitation charge at CLS being the significant portion of the total bandwidth charges paid by the Indian consumers.

TRAI has invited a public consultation in March 2012 with disturbing revelations. Tata annually charges US$628,100 Access Facilitation charges for an STM64 at its Mumbai CLS while Bharti Airtel charges US$450,600 at its Chennai CLS. The same bandwidth, however, costs less than US$700 per annum at Tuas CLS in Singapore. Therefore, Tata’s and Airtel’s Access Facilitation charge in India is respectively 897-times and 644-times expensive than Singapore’s.

On December 21, 2012 the TRAI has asked India’s submarine cable behemoths to charge US$11, 444 for each STM64 circuit at any CLS. It was made effective from January 1, 2013 “to reduce communication cost for BPOs and small enterprises.”  But Tata Communication has challenged this decree and Madras High Court has stayed TRAI’s order. The Hindu remarked:

The reduced charge would mean lower costs for Internet service providers and long-distance telephony players; this, in turn, could result in lower broadband tariffs….. Foreign long-distance telephony companies such as AT&T, Cable & Wireless and BT had complained to TRAI that cable landing charges were pegged artificially high by Bharti Airtel and Tata Communications. The two Indian companies own most of the landing stations in the country.

Sunil Tagare, an Indian-American submarine cable veteran, however, refuses to blame Tata. He criticizes the Indian government’s policy instead. Citing the example of Hong Kong’s open landing rights and open licensing, Sunil has urged India to follow suit.

Don’t get me wrong.  I agree the RIO charges were artificial and ridiculous and something has to be done about them and reducing them by 90% is a good start.  But the government should ask the complainers like Vodafone why they have not yet built a cable station when they are planning to build BBG.  Why did Vodafone ask BSNL to build a cable station and having failed to get BSNL on board, they now have roped in Infotel, RIL’s subsidiary to build the cable station in Mumbai.

Tata has a record of helping submarine cable entrepreneurs starting with the Flag cable where it was the first carrier to sign the MOU.  Later, Tata also helped the Seacom cable land in India.  So the carrier that has been accused of monopolistic practices, has in fact, helped foster competition compared to the complainers who are looking out for their own selfish interests.

I have a solution for the Indian government.  All they need to do is copy what the Hong Kong regulatory authority, OFCA has done.  First of all, OFCA is the single-point-of-clearance for anybody wanting to land a cable in Hong Kong.  Secondly, OFCA set aside land on an oceanfront in Hong Kong specifically so that cable owners can build and share their cable stations with other cable builders.  If a new cable wants to land in Hong Kong, they can either land in one of the existing cable stations or get the land set aside by OFCA to build their own cable station.  This keeps all the existing cable station owners on their toes and nobody dares to cheat each other.

Hong Kong has an open licensing regime so there is no limit to the number of new cable licenses that can be issued.  In addition, OFCA also provides help in buying backhaul at reasonable costs from existing providers or will provide licenses for cables to build their own backhaul.

Following Hong Kong, however, will not solve the problem unless different government departments work in synch. Currently it takes minimum three to five weeks to repair a submarine cable in Indian waters, which is comparatively higher than in other countries. In last December India’s Telecommunications Ministry has proposed to shrink the duration of repair to three to five days for faster restoration of Internet services. The Hindu said:

To bring in an undersea cable repairing ship, an international long distance (ILD) operator needs to get clearances from as many as six agencies, which could be the reason for the delays. These agencies include the home ministry, defence ministry, Indian National Shipowners’ Association (INSA) and the Director General of Shipping among others.

Reducing time for processing of application at the Defence Ministry to a maximum of three days and obtaining INSA’s permission and special licence from DG Shipping were among other suggestions proposed by the Ministry. Faster Naval and custom clearances, where in a repair ship should be treated as an “ambulance on the road”, were also suggested.

There is no quick-fix to make Internet, and consequently broadband, affordable in India and elsewhere. It should be addressed holistically. India’s The Economic Times, has interviewed me along with Punit Garg of Reliance, Doug Madory of Renesys and Michael Ruddy of Terabit Consulting to identify the problems and solutions of international bandwidth bottlenecks of Indian internet. Its article – Bandwidth prices: Why we pay more for internet services – takes a comprehensive look at home and abroad:

According to International Telecommunication Union’s annual Measuring the Information Society report, the cost of an entry-level broadband plan in India accounts for about 5.5% of Indian per capita income. In comparison, a similar plan accounts for 0.5-0.8% of per capita income in countries such as Singapore, the US or the UK. India lags behind countries such as Sri Lanka (2.9%) and Malaysia (3.2%).

There are a lot of domestic restrictions on hosting data in centres located in India,” says Garg, pointing to conditions imposed by the terms of the ISP licence on types of content hosted as well as issues related to taxation of such services.

Data centres in India face other problems: they need uninterrupted and reliable power supply, a rarity in the country. Most ISPs would also like to be present close to the international gateways in a place like Mumbai; but hosting servers there is expensive because of real estate costs.

Bandwidth costs are also influenced by another factor that ties with our original question about the risks of disconnection. As Michael Ruddy, director of international research at Terabit Consulting, says: “Network planners in the IT industry increasingly try to avoid geographical ‘choke-points’, where multiple cables can be taken out of service due to a single event.

That is why, adds Ruddy, the IT industry’s smart money secures redundant paths that are geographically far removed from the choke points. “In the aforementioned cases the goal would be to procure capacity via South Africa or terrestrially through Asia for example,” he explains.

Wholesale internet prices in Asia, not just India, are higher than in Europe. A relative lack of competition and openness in the market is the main factor, according to Madory. “The second factor is probably the geography of the internet in Asia in which connectivity between Asian countries is dependent on a handful of submarine cables to get from one place to another.”

Abu Saeed Khan, a senior policy fellow with LIRNE Asia, an information and communication technology think tank, has worked on researching ways to reduce bandwidth costs in Asia. “Merely increasing the number of submarine cables does not increase India’s competitiveness in the region,” he says, concluding that “the cost of bandwidth is high, if a country is only dependent on submarine cables.

According to TeleGeography, only 13.8 million Indian broadband users (6% penetration) were connected with 888,859 Mbps international Internet bandwidth in 2012. It also said that 156 million Chinese broadband users (39% penetration) were plugged with 4,210,155 Mbps bandwidth at the same time.

China has balanced the flow of its Eurasian traffic through submarine and terrestrial networks. It may reduce India’s historical dominance as an unavoidable transit in Eurasian telecoms connectivity. Our ongoing research is intended to mitigate the risk of rising monopoly or forming a cartel in Asia’s bandwidth marketplace. Open-access remains central to our advocacy. Indian policymakers should read between the lines of writing on the wall.