Kennth Cheung at the University of Florida has released an interesting paper about whether carriers have an incentive to build more bandwidth in a non-net neutral situation.
His paper, which he sent me a few months ago, uses game theory to suggest that outcome. Unfortunately, Cheung has not done such a good job of making the paper available — I only heard it was released through the press release, and the SSRN link on the press release is dead. So Prof. Cheung, if you’re reading this, put the paper up somewhere.
In a different draft paper I’ve suggested another, fairly obvious possibility (my point is not original).
Consider this. If a network operator makes its income from its direct and immediate customer (Bill & Keep), it has an incentive to maximize bandwidth, so as to be able to charge more to that customer.
However, if an ISP earns money from charging termination fees to content providers (or, gatekeeper fees), its incentives become more complex. It may have an incentive to keep the gate relatively narrow — keep a control on the supply.
These are among the reasons that termination-fee based systems strike me as likely to lead in suspect directions