I just received a rather interesting email about a new study by NETCompetition.org, one of the groups that fights for net neutrality and other largely worthwhile causes. The study, authored by NETCompetition chairman Scott Cleland, claims that Google sucked up almost 17 percent of the total consumer Internet bandwidth in 2008, while effectively paying for only a fraction of the cost of supporting that bandwidth. Cleland then comes to a rather incendiary conclusion: Google gets an “implicit bandwidth subsidy” of $6.9 billion from American consumers.
That’s a pretty big number, so it’s important to note that this is merely an estimate gleaned by taking Internet usage data from Cisco and other sources, and extrapolating Google’s share based on its relative market share. Google doesn’t reveal anything regarding its Internet usage or payment for that usage, not that any company I know does.
The report goes on to lambaste Google’s public support for affordable, unlimited bandwidth access through a group called Internet for Everybody, particularly in light of this alleged $6.9 billion free ride Google is taking for itself. Assuming Cleland’s data is remotely good, there is something to be said for asking such a successful and profitable company to shoulder more of the burden of the infrastructure that it depends on.
More importantly, and somewhat regardless of the magnitude of any potential individual Google “subsidy” is the question of what this issue of broadband use has to do with the advent of cloud computing. If the various cloud platform proponents have their way, an enormous amount of potential or actual internal IT infrastructure will effectively be shifted to the cloud. With that shift, and I’m pretty sure it will amount to a significant amount of resources moving over to the cloud, there is the real prospect that the usage of broadband resources by some of these cloud platform companies will exceed their net contribution to the infrastructure’s costs.
It’s an interesting public policy question to add to mix as we consider a clouded future. Internal IT gets to pay for all its resources, or the very vast majority, including the high-bandwidth connections it uses to communicate to the outside world. There is definitely a net savings coming to internal IT costs from the shifting of resources to the cloud — those savings being the main raison-d’etre for cloud computing. But the question of what happens to cloud dependent resources like Internet bandwidth as this shift occurs — mostly in the form of who pays for them — isn’t something that’s been explored very much, if at all.
Cleland’s report shows overall Internet traffic doubling between 2007 and 2009, and while it doesn’t mention anything specific to cloud computing, it’s easy to assume that, considering how much Microsoft is spending on building out its Azure platform — one Microsoft insider told me he measures their platform growth in megawatts, not CPUs — that there’s a lot of bandwidth about to be soaked up by cloud computing.
So, the question is, who should pay? If the freight forwarders have a good year, they pay more in taxes for using the roads and rails than they would in a slower year. Same with the airlines: though our tax dollars built the airports and finance the air traffic control system, the airlines pay fees and taxes relative to their use of these resources. Shouldn’t cloud computing platforms, and everyone else using major chunks of subsidized Internet bandwidth do the same?
Seems to me that, when budgets are strained all over the economy, giving a free ride to wealthy companies might not be the smartest move. I certainly pay for my broadband usage — and so do you. Why should Google, Amazon, Microsoft, and anyone else not shoulder their fair share?