I've been following Seattle Transit Blog lately, and thinking about the constant mass transit crises going on in American cities. It's not just Seattle, with its high-powered NIMBYs preventing dense inner-city development, lack of state income tax (and resulting extremely regressive tax impacts), and tortuous geography... just for a few of the obstacles faced by transit here. Chicago and New York seem to constantly have transit funding problems, too.
The root of the problem, from an economic perspective, might be that mass transit service is offered to the public below-cost. Expanding the system to serve more users means the transit agency has greater losses instead of greater profits like most businesses. But raising fares up to the level of per-rider costs doesn't really look so good in most places either — think $6 bus fares in most of Seattle. Maybe popular commuter routes to employment centers with scarce parking could cover their costs from the farebox, because they'd still be a good deal compared to paying for parking.
The government subsidy for transit service is one of the biggest targets for transit opponents. Transit proponents hearing this will shout back that the road networks are given away to drivers, mostly for free. A tiring argument follows, filled with conjecture, over who is subsidized more, transit users or drivers. I don't think there can be an answer to this question — surely it varies considerably in different circumstances, which doesn't help much if you're trying to create cost-effective transportation strategies.
Maybe the answer is that we shouldn't subsidize transportation very much at all. All motorized transportation requires lots of energy, and subsidizing lots of energy usage is about the last thing we need to do. If all the major highways were tolled to cover road construction and maintenance costs (probably with the toll roughly tied to vehicle weight, as heavy vehicles including city buses put a lot of stress on the roads), and transit fares reflected the cost of offering the service, all the costs of traveling would be included in the price for traveling instead of being externalized onto everyone else. The incentive, then, would be to act in ways that lower everyone's total costs. If mass transit is the most cost-effective way to get around, its fares will be cheaper than the cost of driving.
Another thing that's pretty obvious and affects thinking about transportation generally is climate change. People don't really like talking about this, but greenhouse gas emissions cause climate change, the earth has a limited supply of carbon sinks, and given past and present levels of emission, rich countries like the US have something like a moral imperative to reduce emissions. Because of this, we probably need to set something like a price on pollution itself. The goal would be to set the price of polluting at a level so that we weren't taking more than our fair share of the earth's ability to take carbon emissions. The incentive then becomes to travel when necessary, but to avoid unnecessary travel that adds cost and environmental damage.
In a place like Washington state, where taxes are extremely regressive, subsidies on transportation are probably, in terms of benefit distribution by income, one of the more progressive things going. King County's proposed $20 car tab fee puts this in relief: a regressive tax to provide a progressive benefit. This might be mitigated by replacing general sales taxes with general income taxes. Another thought is that after a truly significant carbon tax is in place (which would have to be at the federal level, not the state), the government would not have to raise as much money through income tax. Per-head refunds from this extra money would be a good mitigation for regressiveness of carbon taxes. The overall problem is not one to overlook; largely removing subsidies on an area as broad as transportation without fixing wealth distribution is basically a non-starter.
Ultimately, all this stuff would be a heavy disincentive towards economic activity, compared to the status quo. We don't just use energy because it's there, we use it to do stuff. But if we're really serious about fixing our emissions problems we might have to set markets to work in slowing down the economy in the wisest possible ways...
I'm currently reading Amartya Sen's Development as Freedom, which, among other things, makes the case that it isn't always economic growth that brings development (he, as the title suggests, measures development by people's freedom to live in a way they have reason to prefer). What we'd have to do in a world with less travel and less economic activity is to thoughtfully choose the things that matter most to us. We might choose to take this action ourselves to stem climate change; we might have it thrust upon us by environmental collapse or peak oil. Either way, I really believe we'll have to consider it within my lifetime.