Most consumers don’t know how their Spotify streams get to their phones or wireless speakers, how their cell phone calls are sent from their city to another or how their credit cards are accepted around the world. Instead, they rely without much thought on the internet and electronics communications systems that span the globe. Despite a lack of consumer interest, these pipes comprise more and more of our economy, tand we’re going to need more of them (and more data centers) in the near future to continue our digital transition.
Or so says a report from the OECD released Tuesday. The report, “International cables, gateways, backhaul and IXPs,” outlines the economic impact of interconnections, abundant fiber optic long haul and metro cables and trumpets the need for more internet exchange points where those networks can meet. It also makes the case that developing nations need more cable and that the world as a whole needs ten times its current number of internet exchange points to meet future demand. That’s a significant increase:
“Some in the internet technical community have expressed the view that the number of exchange points around the world has not yet reached a sufficient scale and that expansion should be proceeding at a faster pace. The issue here is not just the traditional challenge of establishing IXPs in countries where they do not currently exist, though that remains a priority. These commentators expect a need for a significant increase in the number of IXPs, in the next decade, from the current 20 major locations to a future with 200 such locations. The basis for this assessment is the increased use of fixed and wireless broadband access throughout the world. A significant proportion of the users of these connections are in countries and regions that are under served. Much of their traffic will be sourced or routed from outside their region if IXPs are not available for content and services providers to further localise this traffic. This results in higher costs for transit.”
How the internet economy works
The OECD did a nice job explaining in 2012 the importance of internet exchange points to the cost of transit, showing how places with more IXPs saw lower prices for transit and in effect became broadband exporters. Europe has moved to this model, while the U.S. has a slightly different for-profit IXP model that may be transitioning over to the more cooperative European model, thanks to the Open Internet Exchange organization. But IXPs are only part of the problem.
First, a bit about backhaul. The internet is a series of networks that connect to each other at interconnection points located in data centers. Your ISP provides what’s known as last-mile service, which goes from a business or home back to a central office or head end owned by the ISP. At that point, your traffic heads out to a transit provider that connects to networks owned by content companies, retailers and clouds. If the traffic needs to cross the ocean, it will travel on submarine cables. These transit providers’ long-haul and middle-mile networks, as well as the submarine cables, are all examples of backhaul — how last-mile traffic gets back to the internet itself. The report concerns itself mainly with submarine cables, though.
Many developing areas need more pipes connecting them to other parts of the world. This is both an economic issue (more competition between pipe owners lowers prices) as well as a resiliency issue (more pipes ensure that cable cuts will not shut down the communications network). Not only do these areas need more submarine cables, but the ownership structure is changing, with big internet companies like Google and Facebook investing in backhaul at the submarine cable level as well as long haul fiber across countries.
Building an economic argument for openness
The OECD report lays out a detailed history of how liberalization and increased competition on the submarine cable side can lower prices and boost demand for internet-based services. In fact, the biggest takeaway from the OECD report — which should resonate with all internet stakeholders, from the ISPs to the content guys — is that the more open the system is in terms of access and peering, the more demand there is for these networks. That means ISPs that are trying to block content from entering their networks or force transit providers to pay for peering are doing their part to take more of the overall internet pie, but doing nothing to enlarge it. Meanwhile, efforts to put more IXPs in place, as Google has done in Kenya, help promote cheaper broadband and demand for more broadband. That’s also the rationale behind the open internet exchange effort in the U.S.
While broadband providers try to appeal to an economic argument when they attempt to put in data caps or force transit providers to pay up, they are relying on an old economic model that doesn’t reflect the way the internet works or how its costs are structured. It’s like trying to apply the business models from the horse and carriage days to those of the railroad era. When ISPs and others ignore this economic model, they ignore it to the detriment of the whole ecosystem — including themselves and end users. It’s time everyone realized this. I hope the OECD reports help.