3 ms·

Is there a framework that helps to reliably determine which side of the two extremes - too cheap to meter (internet access) or induced demand (roads) - the prob

Is there a framework that helps to reliably determine which side of the two extremes - too cheap to meter (internet access) or induced demand (roads) - the problem will fall into and how to nudge it in the desired direction? Is this even the right framing of the question? Seems like an important management and policy problem for the humanity to get a grip on, so we can prevent negative outcomes.

AndrewDucker 3y ago
The problem with inducing demand for roads is that they take up significant amounts of space and disrupt everything around them. We can provide everyone with terabit internet without taking up more than a tiny amount of surface space, so inducing demand on that front isn't a problem.
icegreentea2 3y ago
Induced demand for roads is a problem because road network improvements (especially in the form of road widenings) are typically marketed to the populace as reducing travel time at peak, but in practice all they can do is improve overall throughput (with the temporary decrease in travel time coming as a bonus). And I think road network improvements are actually quite good at achieving their throughput related goals. From that perspective, there isn't actually a problem at all. Internet traffic absolutely "suffers" from induced demand. It just happens that we've been able to keep ahead of the worst of the congestion problems. a) ISPs traffic shape. ISPs have more control over their traffic than cities/highways typically do. b) ISPs don't guarantee minimal speeds, or even give you a sense of their distribution. The inability to provide a minimal tells you that probably at peak traffic, there is significant contention (induced demand has kicked in). It just happens that... internet usage demand is probably less coupled than traffic on a highway. c) The largest traffic loads on the internet (streaming video) are majority provided by centralized corporations that can provide a 'parallel' service (through CDNs/edge caches... oh and I guess really degrading video quality) that minimize the load onto the 'main internet'. This is like taking the train or public transit or something.

AndrewDucker 3y ago

The problem with inducing demand for roads is that they take up significant amounts of space and disrupt everything around them. We can provide everyone with terabit internet without taking up more than a tiny amount of surface space, so inducing demand on that front isn't a problem.

icegreentea2 3y ago

Induced demand for roads is a problem because road network improvements (especially in the form of road widenings) are typically marketed to the populace as reducing travel time at peak, but in practice all they can do is improve overall throughput (with the temporary decrease in travel time coming as a bonus). And I think road network improvements are actually quite good at achieving their throughput related goals. From that perspective, there isn't actually a problem at all. Internet traffic absolutely "suffers" from induced demand. It just happens that we've been able to keep ahead of the worst of the congestion problems. a) ISPs traffic shape. ISPs have more control over their traffic than cities/highways typically do. b) ISPs don't guarantee minimal speeds, or even give you a sense of their distribution. The inability to provide a minimal tells you that probably at peak traffic, there is significant contention (induced demand has kicked in). It just happens that... internet usage demand is probably less coupled than traffic on a highway. c) The largest traffic loads on the internet (streaming video) are majority provided by centralized corporations that can provide a 'parallel' service (through CDNs/edge caches... oh and I guess really degrading video quality) that minimize the load onto the 'main internet'. This is like taking the train or public transit or something.