(Bloomberg Opinion) — If President Donald Trump follows through on his recent threats to cut off federal funding for California’s long-troubled high-speed rail project, it would be better for all concerned: For all intents and purposes, this thing went off the rails (sorry) a long time ago.

Escalating costs have made it clear that no money was or ever would be available to realize the vision of a modern bullet train between Los Angeles and San Francisco. What’s under construction is a segment through California’s Central Valley, where costs are cheap compared to other parts of the system but which offers almost no economic value. The whole thing has become a zombie project that nobody with clout in state politics can either rescue or kill. A hated outsider officially ending it would let the state’s Democrats complain while also allowing them to acknowledge the reality that it’s not going to happen.

The tragedy is that the basic concept of high-speed rail for California makes a lot of sense.

Los Angeles and San Francisco are two large metropolitan areas that are about as far apart as Rome and Milan (about 380 miles). Trains between those two Italian cities have a 68% market share relative to airplanes, and the competition puts downward pressure on airfares. At this kind of distance, many passengers prefer the comfort of a train to the speed of a plane, and the convenience of train stations to airports. A train could also provide frequent service to intermediary locations such as Bakersfield, Modesto and Fresno — cities that in the aggregate have a large population, but by themselves aren’t large enough to support a lot of flights to LAX or SFO. And finally, once the core HSR line was built, spurs to San Jose and Sacramento, and an extension to San Diego, would be relatively straightforward.

These are all real benefits. But they depend on connecting Los Angeles and San Francisco with a train that is both fast and cost-effective to build.

The failure to achieve this has become a legendary case study in progressive excess, but the original sin was committed by a Republican — Michael Antonovich, then a member of the LA County Board of Supervisors — in 1999. Planners wanted the train to head north from Los Angeles along the route of Interstate 5, but Antonovich successfully pushed to detour the train through his district. That made the project more expensive and increased travel time.

Unfortunately, this set the template for almost every subsequent decision around the project. To build a fast train between Los Angeles and San Francisco in a cost-effective way, it is important to prioritize making the train go quickly between Los Angeles and San Francisco. There may be tradeoffs between expense and speed. But it should never cost more to make the train slower. Yet it happened again with another major decision to get from the Central Valley to San Francisco via the Pacheco Pass rather than the more northerly Altamont Pass.

There are many more details, complexities and decisions that went into this fiasco, but the basic story is pretty simple: They couldn’t build a cost-effective fast train between Los Angeles and San Franciso because they kept making choices that deprioritized that goal. It is of course understandable that elected officials who represent places other than LA or San Francisco would have other priorities. But regularly deferring to the wishes of those who weren’t aligned with the core goal of the project undermined it.

The way to do these things is to avoid precommitments. California should have invested a modest amount of money for a cost-effective proposal, and then asked the legislature to support it. If it said yes, great. If it said no, fine. Either way, you wouldn’t end up with a bottomless money pit — and no train.

A new high-speed rail proposal for the East Coast, from the Transit Costs Project at New York University, shows what sound planning looks like. Rather than copying Amtrak’s official proposal — which starts by asking every stakeholder what they want, then rolls it into an impossible $117 billion plan — the NYU study looks for the cheapest way to send trains from Washington to Boston in just under four hours. Its plan involves modest amounts of new construction and significant changes to commuter rail operations. But the whole thing comes in at about $17 billion, which is a very modest cost for a program with large benefits given New York’s constrained airspace, and leaves most train commuters better off.

Yes, some existing riders would lose out, as would some Amtrak customers in less populated cities. The politics of making this plan a reality aren’t simple. But the upside — especially to “in between” cities such as Baltimore, Providence and Philadelphia — would be huge. It’s an idea creative politicians should take up.

More important, politicians throughout the country should pay attention to the enormous price gap between the “do it as cheaply as possible” plan and the “accommodate as many as possible” plan, because the basic point is applicable to all kinds of infrastructure projects in all kinds of places: If something is worth doing, it needs to be made a priority. If it’s not important enough to be prioritized over other considerations, better to give up and do something else instead. Otherwise, like California’s politicians, they may be left with not much more than a lot of wasted time and money.

Elsewhere in Bloomberg Opinion:

For more, subscribe to our newsletter.

This column reflects the personal views of the author and does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

Matthew Yglesias is a columnist for Bloomberg Opinion. A co-founder of and former columnist for Vox, he writes the Slow Boring blog and newsletter. He is author of “One Billion Americans.”

More stories like this are available on bloomberg.com/opinion