1. Give transit its own right-of-way and good terminals
2. Make it hard to use cars
3. Make it expensive to use cars
In response, Alon Levy writes:
You don't need steps 2 and 3. Transit already has a high modal share on the market it serves well, namely suburb-to-CBD commutes. Even Metro-North, which competes with numerous free bridges and parkways, achieves an 80% market share for commutes into Manhattan.
Alon is more or less right, in that Metro-North has a tremendous advantage over cars in terms of rush-hour trips to Manhattan. However, I would argue that Steps 2 and 3 are partially being followed. It's not like it's smooth sailing and completely free to drive from Scarsdale to Midtown. Either you pay the toll on the Henry Hudson or the Triboro, or you sit in traffic at the "free" Fifth Avenue Bridge. Either you pay a tremendous amount of money for parking every month, or you've got a "free" parking placard. Any of these scenarios, you probably spend at least half an hour stuck in traffic somewhere. It's not ideal, the way it would be if we had true cordon pricing or market-rate parking with no placards. But it is a deterrent to driving, and probably accounts for that high mode share.
So Metro-North has an 80% mode share, but it only has 59% farebox recovery. This is actually the best ratio of any commuter railroad in the country; the next highest are the MBTA with 54.1, New Jersey Transit tied with the South Shore Line in Chicago with 51.5 each, SEPTA with 50.9, LA's Metrolink with 50.3, and the LIRR with 46.3. Why are they all so low? Well, Metro-North's ridership (passenger miles per revenue mile) is only 37.4. This may seem high compared to the buses, but remember that Metro-North's M7 train cars have 101 or 110 seats, depending on whether they have the bathroom or not.
I know what you're thinking: it's that fucking middle seat! Nobody wants to sit in it. But you'd be wrong. If the trains had 64 people per car, they would make a profit and no one would ever have to sit in the middle seat.
One of the shortcomings of the Magic Formula is that it only applies to a particular origin/destination pair, and it succeeds for trips to Manhattan. In the rush hour, the trains are pretty much full. If there are some middle seats free, there are probably more people standing. In other words, those trips are making a profit. It's the system as a whole that's not making a profit, and that must mean that the off-peak trips aren't making a profit. Certainly if you ride Metro-North at night or on weekends, you'll see plenty of cars that are less than 64% full. Although there are plenty of carfree households in Westchester, there are plenty of people who use the train for their commute and a car for shopping and socializing.
On one level, of course, it's a good thing that the trains run all day, seven days a week. Metro-North could probably make a profit by only running trains during rush hour, like VRE. But then Westchester would look even more like Northern Virginia than it already does. As Jarrett wrote, "A profit-oriented operator will tend to skimp on late-evening service, for example, but you need really good late-evening service, even if unprofitable, to make lower car ownership viable on a large scale."
So if we wanted Metro-North to be profitable, we'd have to make evening and late-night service profitable. One way of doing that is to save money on fare collection. The conductors are quaint and helpful, and I don't want to see anyone get laid off, but we seriously don't need four or five people on every train. A proof-of-payment system would probably save a ton of money and make it possible to break even with forty or fifty people per train. Maybe Jay Walder will do that.
We could also apply the Magic Formula to the most popular night and weekend routes. What if we restored the parkways (Bronx River, Saw Mill, Hutch, Taconic and Cross County) to parkway design standards - four lanes with green space on the sides and a multi-use trail? What if we ran trolleys down the middle of Routes 1, 9, 22 and 100, in dedicated rights-of-way appropriated from cars where possible? What if we tolled the major crossings of the Croton Reservoir and River system?
Of course, if we made the whole Metro-North system profitable, those profits would probably be used to cross-subsidize the other MTA services. So any profitability boosting strategies should be applied across the board.