June 26, 2006
How Tariff Rebate Passthrough would work
Dave Siegel posted a challenge to my Tariff Rebate Passthrough net neutrality proposal, claiming that technical implementation would be unduly burdensome, and also touching on the fact that consumers generally prefer flat-rate to metered pricing. I think the best response would be to spell out, in a little more detail, how it would work. Along the way, I think I can answer Dave’s (and anybody else’s) concerns.
- The ISP would implement a few levels of QOS, and an assignment scheme that would by default assign each kind of traffic to a certain QOS level. Nothing hard there; they just have to buy the proper equipment, and if they don’t do that this whole discussion is moot.
- Information providers could contract with individual ISPs to subsidize the tariff on behalf of consumers. Naturally, the level the information provider pays for is the QOS level that would be assigned.
- Since the information provider’s only choice is to pay the ISP’s price or not, implementing the payment/contracting scheme would actually be pretty easy. Somebody will make good money creating connectivity and a clearing house for these highly-automated deals, presumably on a SaaS basis. Basically, if you own a website you can hook up to the exchange and pick a price you’re willing to subsidize. Price variability by region, etc. can be built in pretty straightforwardly. This application software is NOT going to be what makes the system fail.
- On computers and other GUI devices, there would be a small pop-up window showing “the meter running.” This should give the opportunity to change the QOS so that the cost is zero (base-level QOS won’t be metered). If the cost is zero, either for subsidy or because only base-level QOS is being delivered, the meter by default won’t appear. Devices that can’t provide that meter will generally be the same kinds of small, mobile devices that are expensive to use on a per-minute or per-message basis already.
- As a practical matter, most web sites would remain unmetered. Most websites would be unmetered to the consumer, either because they don’t need above-minimum QOS, or because the owner has an actual business model (advertising or subscription) and can subsidize the QOS. That would put competitive pressure on the others to follow suit.
- Telephony might well be metered. So what? It is today.
I continue to think that Tariff Rebate Passthrough is a compromise that meets pretty much everybody’s legitimate needs, and a reasonable fraction of everybody else’s other desires to boot.
Comments
6 Responses to “How Tariff Rebate Passthrough would work”
Leave a Reply
[…] One of the best features of Tariff Rebate Passthrough is that, even with pricing flexibility, it can be implemented using simple legislative language. There only have to be three stipulations: […]
[…] If it were not for Edisonet, extreme net neutrality would be fairly harmless. If it were not for the huge public benefits of Jeffersonet, letting the telecom carriers have their way on non-neutral pricing wouldn’t be so bad. But given the presence of both, a middle course is needed. Fortunately, one is available that gives appropriate treatment to Jeffersonet and Edisonet alike, without giving tasks to regulatory agencies that are much different from the kinds they actually do a pretty good job of performing already. Tariff Rebate Passthrough shows the way. […]
How would this tarrif be applied in the case where the connection between the information provider and consumer traverses multiple service providers? (As in the provider’s ISP, a transit network, and the consumer’s ISP.)
This proposal is waaaay too confusing and complicated. Why not price the internet like electricity? Charge everyone a minor (about $5 / month) connection fee, and then charge a flat rate per gigabyte downloaded. Heavy bandwidth users will end up footing the bill for their usage, and market forces will push ISPs to continually lower their price per gigabyte. This will also mitigate problems with uncapped cable modems, as customers who abuse it will end up paying large usage fees. In addition, light internet users might be tempted to connect to broadband as their total bill could be something like $10 a month.
Andy,
So we have to PAY to receive advertising?
That could screw up the advertising-supported business model.
Which in turn could screw up the availability of things that we DO want.
I think publishers should have the option of paying for delivery, as they do now.
CAM
[…] I wrote about network neutrality in 2006-7, the issue was mainly whether broadband providers would be allowed to ship different kinds of data […]