> If, after fixing all the intersections, flow is improved, people who were making do with something other than driving at the peak rush, will show up to consume the new capacity.
I fail to see the problem with "we built something and people used it because they preferred it to their other options"
His premise there was "why simply adding more lanes won't fix the problem of congestion". He didn't claim that it won't lead to more cars using the road.
The problem with this is that if your goal is to build a liveable city where people can reach their destinations in predictable time without pulling their hair out, simply building more lanes does not usually help, for the reasons outlined very clearly in the article.
And that isn't exactly a wild guess, we have decades of data on the subject.
From the article, the problem is "removing a road segment, would block the competition, and actually improve overall flow".
Ie. we built something, people used it because they preferred it, and now everyone gets a worse experience.
The problem is cost efficiency.
It's not cost effective to have a bus line that stops at everybody's house.
However, when you start to have 3 lanes of traffic then a lot of that traffic can be handled by a bus at a cheaper cost to both the drivers and the relevant department of transportation than expanding the road to handle the current traffic.
In addition to the sibling comments, Braess's paradox actually goes beyond this. You can add capacity and every actor acts rationally as you describe and the net result is an objectively worse situation for everyone, in which the decisions made are still the rational ones.
Assuming nobody actually wants to move slower.