Imagine that I proposed to you the following argument. I have (let's pretend) added up all the fuel consumed in the process of getting WidgetCo's widgets from the factory to consumers. There's all the gas burned by the trucks that bring the widgets from the factory to the retail store. And then there's the gas each consumer burns driving to and from the store for widgets. And having added up the costs of all this carbon, I discover that WidgetCo is paying only a tiny fraction of the total cost of the fuel consumed in the process of getting widgets from the factory to the homes of customers. Now suppose I claim that this is evidence of some form of outrageous unfairness—WidgetCo is somehow forcing you to subsidize their shipping costs! (The widgets, by the way, are free.)
If, in fact, I were to make such an argument, you would rapidly conclude that there are really only two possibilities: (1) I am a moron, or (2) I must think that you are if I expect you to find this persuasive.
I will let you decide which applies to the author of a "research study" of Google's bandwidth use being pushed by the anti–net neutrality site NetCompetition.org. Using some rather dubious proxy measures—which would be worth further scrutiny as well, if the fundamental premise weren't so manifestly bogus as to render such quibbling moot—telecom shill Scott Cleland estimates that Google and its subsidiaries "used" 16.5% of consumer broadband traffic in 2008, but only paid 0.8% of consumer broadband costs. This, the author brazenly claims, amounts to an implicit subsidy of some $6.9 billion to Google, and proves that Google "uses" 21 times as much bandwidth as it pays for.
