Indeed, everything's always easier when there's a greater wealth of transportation resources available and people have more options. Ideally there should be enough road and/or train capacity to get everyone where they want to go.
If that's not the case, then congestion pricing works by putting pressure on people to stay off the roads at peak times if they can. This is, in fact, a good thing to do; if not everyone who wants to use the roads can do so at once, because there aren't enough roads, then it's most efficient to give the space to those who most need it by willingness to pay. (Regressive distributional impacts of this can be addressed with the revenue that congestion pricing raises, e.g., by cutting consumption taxes.) Making people sit in traffic does no good at all and is just wasteful.
But if you want not just to eliminate that waste, but to actually increase people's mobility, then you need to build enough roads or trains. Whether that's worth the cost depends, as always, on local conditions.
(To be clear, the goal from this perspective is not to "shift transit modalities"; it's to get people where they're going without making them waste time in traffic. Sometimes that means getting more of them onto trains, and sometimes it doesn't.)
> everything's always easier when there's a greater wealth of transportation resources
It’s not just easier. If there are no other options, it doesn’t matter what congestion toll you put. The only option is car.
> putting pressure on people to stay off the roads at peak times if they can
Sure, but for most people the demand for transportation before and after normal working hours is inelastic. A smart policy takes that into account and doesn’t inconvenience people for going to/from work.
The real effect of congestion charges when there isn't good other options is whoever enacts them is voted out of office and anything they might have done to give people other options is scrapped as well. NYC can get away with them in the dense parts because there are enough options that the people who vote there mostly use transit and so they don't pay the price anyway and thus don't care.
If that's not the case, then congestion pricing works by putting pressure on people to stay off the roads at peak times if they can. This is, in fact, a good thing to do; if not everyone who wants to use the roads can do so at once, because there aren't enough roads, then it's most efficient to give the space to those who most need it by willingness to pay. (Regressive distributional impacts of this can be addressed with the revenue that congestion pricing raises, e.g., by cutting consumption taxes.) Making people sit in traffic does no good at all and is just wasteful.
But if you want not just to eliminate that waste, but to actually increase people's mobility, then you need to build enough roads or trains. Whether that's worth the cost depends, as always, on local conditions.
(To be clear, the goal from this perspective is not to "shift transit modalities"; it's to get people where they're going without making them waste time in traffic. Sometimes that means getting more of them onto trains, and sometimes it doesn't.)