Last month, Streetsblog LA reported that Los Angeles Mayor Eric Garcetti was planning to announce a “28 by 28” program for transportation investment. This would entail completing 28 different projects listed under Measure M in advance of the 2028 Olympics.  Thus far, there are scant details available about the program, but it is possible to make recommendations about which projects make the most sense to accelerate.  By picking the right roster of Measure M items to build in the next ten years, Los Angeles stands to substantially reduce its construction costs. And since Measure M has a fixed budget, lower costs translate to more infrastructure.

There are several dozen Measure M projects, as reported by The Source last year. Some, accounting for nearly a quarter of Measure M's $40 billion capital pricetag, are highway projects. A few more are bus projects, such as an Orange Line extension to Burbank and Pasadena, and bus rapid transit on Vermont Avenue feeding the Red and Purple Lines. But the bulk of spending is urban rail, including such projects as future phases of the Wilshire subway to UCLA, an extension of the under-construction Crenshaw light rail line north toward Hollywood, extensions of the Gold Line, and a rail tunnel under the Sepulveda Pass.

It remains to be seen whether anything will come of the 28 by 28 program. In 2010, in the wake of Measure R, Mayor Antonio Villaraigosa pitched a similar concept called 30/10. Using federal loans, backed by future Measure R revenues, Villaraigosa proposed constructing 30 years worth of transportation projects in just ten years. Nothing came of the plan, even with Democratic control of the White House and Congress. Today, with federal loans unthinkable and state loans unlikely, the situation is even more tenuous. This alone should give anyone grounds for skepticism that Garcetti's plan will succeed.

Nonetheless, executing public infrastructure projects speedily is worth the effort. The best argument for this comes from Madrid, which has perhaps the lowest subway construction costs in the world. Los Angeles is building subways for about $900 million per mile. The typical European range is $200-500 million per mile, but Madrid's metro expansion in the late 1990s and early 2000s cost $100 million per mile in today's money, and even more recent tunnels appear to be relatively inexpensive. Madrid Metro's former CEO, Manuel Melis Maynar, has some choice words for how other cities should replicate his city's success in building infrastructure inexpensively:

First, it is important to reduce construction time. One cannot build a dam in separate parts because one has to build the foundation and then build up until the dam is finished. It is not linear but with metal systems it can be divided properly. One starts simultaneously in several places. That is the main idea and every part of the project can be finished within 36 to 42 months. There is no doubt about this. I invite the committee to come to Madrid any time, get into our test cabin and see what can be done in 42 months. We can organise this through our embassy.

It is easy to build a metro. The problem is building the first line because people worry that there will be loud noise under their houses. This is not true. To reduce the construction time, divide the project or the whole design simultaneously and award and sign the big contracts before August. This is possible and the design will be finished by Christmas. Most of the design companies and consultants will ask for many years to do the design. This is a waste of time. They need only six to eight months. Any more than this increases costs and is not useful.

Elsewhere, Melis Maynar has also explained how Madrid Metro would choose which bidder to award each contract to: 50 percent of each proposal's score would come from technical merit, 30 percent from cost, 20 percent from speed. Against the belief that projects can be good, cheap, or fast, but not all three combined, in Madrid the prevailing view is that accelerating construction actually reduces costs.

Los Angeles believes the same, for a different reason: construction cost inflation. The impetus for Villaraigosa's 30/10 plan was partly that finishing Measure R construction faster would reduce costs by avoiding future inflation. In the United States, construction costs across all building types, not just infrastructure, have soared. The most recent figures point to a 5 percent annual increase in recent years, whereas annual inflation has averaged 1.25 percent since 2013. Independent of the savings Melis Maynar identifies, accelerating construction allows cities to avoid this future inflation.

The question is whether Los Angeles is capable of accelerating construction, as Melis Maynar would recommend.

It is usually better to build subway lines to their ends at once, rather than in phases. It may reduce construction cost, if those lines are deep (as the Wilshire subway is), since the city would only need to stage the tunnel-boring machine once. But there are bigger operational benefits, in two directions.

First, urban rail needs rail yards, just like mainline freight rail. These rail yards should ideally be located at the end of the line. In an earlier piece in Urbanize detailing Los Angeles's high subway and light rail operating costs, I noted that Chicago has very low operating costs, and high train operator labor efficiency; national transit activists have suggested that the reason Chicago is efficient is that its 'L' system has its rail yards at the end of each line, reducing deadheading. Los Angeles should seek to emulate this feature of Chicago's by building yards at the end of rail lines; this in turn requires the lines to have a consistent end, rather than incrementally extending farther out every few years.

Building a subway line to its natural terminus also makes it easier to reconfigure the buses. Were there a subway from Downtown to Santa Monica under Wilshire, most likely Metro would not need to run many buses on Wilshire, and would divert the resources to connecting routes, including Whittier going east and several north-south lines intersecting Wilshire. Building only part of the way under Wilshire, missing key north-south bus routes, makes for a less efficient system than building end-to-end. The same is true of rail under Sepulveda Pass continuing on the surface on Van Nuys Boulevard: the east-west buses would want to feed such a line, so it had better stretch all the way to the northern end of the built-up part of the Valley.

Building entire lines is common around the world. Boston-based railfan Alexander Rapp has made historical timelines of almost every global rapid transit system. In the largest cities, some metro lines have never been extended since shortly after their opening; others went decades without any extension. Those cities instead built new lines, leaving the old ones alone. This is especially common in East Asia; Tokyo has built most of its subway lines in just one or two phases over a small number of years each.

This suggests Los Angeles should be focusing any acceleration efforts on finishing individual lines. This includes the Westside subway, to UCLA or even Santa Monica. This also includes Sepulveda Pass, with connecting light rail on both sides, going all the way from the northern end of the Valley to Torrance or LAX.

In contrast, outer extensions of rail lines are less valuable, such as the various Gold Line extensions. The West Santa Ana Transit Corridor is of lower priority as well: its first phase, already scheduled to open by 2028, is the outer one, connecting to the rest of the system at the Green Line, whereas it's the second phase that would connect to Union Station; since the yard is likely to be at the outer end, where land is cheaper, or even somewhere on the Green Line, it is not so important to build the line in one piece.

Los Angeles needs to grow its public transportation system. The Olympics provide a convenient deadline for expansion, which means it needs to prioritize: which projects is it most important to accelerate to ensure they are done by 2028? The answer is that Metro should be building complete lines in the urban core, and plan not to need to extend them in the foreseeable future. Its plans for after 2028 should be to build more independent lines, serving additional corridors in the county. This would optimize the system for regional connectivity as well as low operating expenses.

Alon grew up in Tel Aviv and Singapore. He has blogged at Pedestrian Observations since 2011, covering public transit, urbanism, and development. Now based in Paris, he writes for a variety of publications, including New York YIMBY, Streetsblog, Voice of San Diego, Railway Gazette, the Bay City Beacon, the DC Policy Center, and Urbanize LA. You can find him on Twitter @alon_levy.