Thanks for the link. A lot depends on where a flat price ends up of course. There's benefit to predictability but not if I pay 2x what I would under a usage-based scheme.
That flat price is going to depend on things like marginal costs, to what degree flat pricing drives up average costs, and how much the heavy users drive up costs for others. As the link suggests, flat pricing seems to work pretty well for broadband but it's unclear how practical it is (today) for wireless.
> Thanks for the link. A lot depends on where a flat price ends up of course. There's benefit to predictability but not if I pay 2x what I would under a usage-based scheme.
Not necessarily. The combination of the value of the services you decided to forgo to not rack up metering charges and the mental cost of doing those calculations can be arbitrarily large, because the mental cost is a sunk cost[1] by the time you know whether paying for the additional metered bandwidth is worth it. You could theoretically spend all day worrying about using too much bandwidth without ultimately deciding to use any at all, so your bill would be much lower but at the cost of spending all day worrying about it.
There are also knock-on effects of not having unmetered service. It causes people to use less bandwidth (obviously) but that means they're deriving less value from the network, which means they're not willing to pay as much for service, which means slower upgrades, which means potentially higher customer costs in the long term as the price per GB stays higher longer.
> That flat price is going to depend on things like marginal costs, to what degree flat pricing drives up average costs, and how much the heavy users drive up costs for others. As the link suggests, flat pricing seems to work pretty well for broadband but it's unclear how practical it is (today) for wireless.
The strange thing about wireless is that it seems likely to eat itself. Spectrum is scarce and towers are expensive, but charging a high monthly fee for unlimited usage or a high metered rate is likely to drive customers to use WiFi rather than cellular whenever possible, which reduces demand for cellular and pushes the prices back down.
I suspect the equilibrium is going to be "unlimited" wireless service where only the first X amount of MB per month is "fast" and then the service works but is throttled. The result then is that everybody uses WiFi whenever they can to preserve their "fast" cellular bandwidth (or because they've already used it and then WiFi is much faster) but nobody actually has to worry about using too much because there is never any bill for the overage. And then you start demanding that WiFi access points be installed at whatever locations you were in where you used all your cellular bandwidth.
I pretty much agree. Quite a while back, Clay Shirky wrote about how one of the problems with micropayments (for online content) was the mental transaction costs that they entailed. There's a lot to that argument.
OTOH, as you suggest, there need to be proper incentives around scarce resources--such as using WiFi where available rather than cellular. The equilibrium you describe seems sensible--perhaps with throttling less extreme than AT&T's current version seems to be and with greater transparency.
That flat price is going to depend on things like marginal costs, to what degree flat pricing drives up average costs, and how much the heavy users drive up costs for others. As the link suggests, flat pricing seems to work pretty well for broadband but it's unclear how practical it is (today) for wireless.