[lofatdairy]

I didn't see anyone saying induced demand in and of itself is the thing being avoided, just that induced demand for driving caused by highway construction is being avoided, as this implies that traffic won't be alleviated by increased capacity.

Induced demand can be positive, yeah, but I think induced demand is more complicated. It concerns a positive feedback loop phenomena that leads to the saturation of a system beyond its intended capacity. Like more people taking the metro is good, but more people taking it than it can handle can degrade their impression if it.

[s1artibartfast]

Induced demand isn't necessarily complicated. The simple answer is transaction costs.

If people are willing to spend an hour in traffic to go to the beach, you are basically stuck with that travel time. You can only increase the number of people that end up going.

[lofatdairy]

I meant complicated as in not necessarily good/bad. I agree it's not complicated in the general sense that it's an intrinsic part of supply and demand.

That said it's definitely more complicated in how to article is discussing it. The article is clearly talking about a Braess's Paradox-like situation where an individual's optimal decision incentivized by changes in the system is worse for more participants than the previous system, due to anticipatory and dynamic effects. This is inherently more complicated to measure and predict.