The state's push to end car-first street planning could ripple across the country.
SAN FRANCISCO—The report that Michael Schwartz sets down on the table is truly enormous. It looks like it has eaten several smaller reports and laughed as they tried to run away screaming. The document is some 700-pages long and several inches thick; that's not counting the second volume or the thousands of pages of technical supplements. Schwartz has posted a photo series showing his newborn son alongside the report on the door of his office at the San Francisco County Transportation Authority. "I finished this two weeks before he was born," says Schwartz, with a look of tired pride. The baby doesn't appear markedly bigger than the document until the photo labeled "five months."
The photo series is titled "BRT Baby." It's not immediately clear whether the name refers to the child or the report. It's safe to say both required significant labor.
The tome in question is the environmental impact report for a bus-rapid transit line that will run two miles down Van Ness Avenue in downtown San Francisco. When finished, the Van Ness BRT will be a transit marvel. Passengers will board from a central median and travel in an exclusive lane with traffic-signal priority. Even now, sharing lanes with cars, buses carry a third of all trips in the corridor (roughly 16,000 riders a day), and with BRT that figure could soar. "It's sort of the first of its kind to say, we're doing full-feature BRT in a dense urban right-of-way with cross-streets," says Schwartz. All told, the line will anchor a $126 million street upgrade.
San Francisco has been waiting for Van Ness BRT a long time. The line was a signature project in the half-cent sales tax referendum, Proposition K, that city voters approved in 2003. The original plan called for Van Ness to be up and running by late 2009. The latest timeline has BRT beginning operations in 2018—a full decade and a half after the Prop K vote (which itself came years after the route concept emerged). Big city infrastructure projects get pushed back for countless reasons, but in the case of Van Ness BRT, a major source of the delay was the need to produce this massive report. It didn't receive final approval until late 2013, and was part of a preparatory phase that, all told, took 6 years and cost $7.6 million.
Understanding the delay requires a quick primer on the California Environmental Quality Act, the state's environmental law. Under CEQA, major planning projects like Van Ness BRT are analyzed for their potential impact on 18 areas of life—air quality, water quality, noise, land use, traffic, and so on. If an initial analysis shows that a project will have no significant negative impact on any of these areas, its leaders can prepare a short report. If a negative impact is unavoidable, project leaders must prepare a slightly longer report explaining their plans to mitigate or offset the damage. And if a negative impact is unavoidable but can't be offset, they must prepare a full environmental report like the one on the table.
Here's the sad thing about the Van Ness BRT report: The only area where it had an unavoidable negative impact that couldn't be offset under CEQA was traffic. "So this whole document was prepared because of the traffic impact," says Schwartz, nodding at the enormous report. And here's the really sad thing about CEQA traffic impacts: They're determined using a car-friendly metric known as "level of service" that bases a project's transportation performance on driver delay. In other words, Van Ness BRT required all the trouble of preparing this massive report because, in the twisted eyes of California law, public transit is considered a greater enemy to the environment than car travel.
That's the bad news. The encouraging news is that this law is about to change. California will soon reform traffic analysis under CEQA by replacing "level of service" with another metric more in line with its environmental and urban mobility goals. So transit projects and transit-friendly development are about to get much, much easier in California cities—and some think the shift in mindset will spread across the country.
"For a project like this," says Schwartz, tapping the report, "it's huge, CEQA reform." Five-month-old baby huge.
• • • • •
Level of service was a child of the Interstate Highway era. The LOS concept was introduced in the 1965 Highway Capacity Manual, at the very moment in American history when concrete ribbons were being tied across the country, and quickly accepted as the standard measure of roadway performance. LOS is expressed as a letter grade, A through F, based on how much delay vehicles experience; a slow intersection scores worse on LOS than one where traffic zips through. Planners and traffic engineers use the metric as a barometer of congestion all over the United States.
In California, LOS has an especially high-profile. As the primary arbiter of traffic impacts under CEQA—adopted in 1970 by Governor Ronald Reagan—the metric not only determines the fate of many transportation and development projects, but has the awkward role of promoting car use within a law designed to protect the environment. "We have one section of CEQA saying we've got to reduce greenhouse gas emissions," says transportation consultant Jeffrey Tumlin of Nelson\Nygaard, "and another section of CEQA saying we need to accommodate unlimited driving."
It was the state's love for basketball, oddly enough, that compelled a change. In September 2013, the legislature passed Senate Bill 743, a bill securing a new downtown basketball arena for the Sacramento Kings. The law also instructed the Governor's Office of Planning and Research to recommend a new CEQA metric for traffic analysis to replace level of service. As an extra caution, the law expressly forbade CEQA from relying on LOS in California's "transit priority areas"—in other words, its cities.
In late May, just a short walk from where that Kings arena will one day reside, I met Chris Ganson of the Governor's Office of Planning and Research to discuss which way the state was leaning. Ganson is just the person to be involved in mobility reform; in addition to being a Sacramento native and lifelong Kings fan, he has degrees in environmental science as well as planning and transportation engineering. He showed up in grad school right when An Inconvenient Truth came out. "My interests lined up nicely with the timing in the world," he says.
Ganson told me OPR was planning to recommend "vehicle-miles traveled" as the new "central metric" under CEQA. He says VMT meets all the state's major criteria for a traffic evaluator: fewer greenhouse gases, more multimodal networks and urban infill developments, a general boost to both the environment and public health. Where LOS encouraged public projects to reduce or eliminate driver delay at city intersections, VMT would encourage them to reduce or eliminate driving at all.
"If we're using delay metrics to rate our progress, we're going to look like we're doing bad, even as we're doing exactly what we're trying to do," says Ganson. "Even as we're meeting not just our environmental goals, but our goals for the fundamental purpose of transportation—providing access to destinations. Getting people places."
The most obvious advantage that VMT provides over LOS as a CEQA metric pertains to public transportation projects. In the eyes of LOS, street elements like crosswalks, bike lanes, and transit lines downgrade a project by increasing driver delay. In that sense, says Ganson, LOS mischaracterized multimodal projects as impediments to transportation rather than legitimate modes in their own right. In the eyes of VMT, projects that give street space to pedestrians, cyclists, or transit riders will score well even if—especially if—that means less room for cars.
Transit-oriented development should benefit right alongside transit in California. LOS favors sprawl to smart growth, because the traffic generated by remote development creates little delay at any single intersection when dispersed over a full road network, especially compared to compact infill placed right at a city corner. VMT favors the reverse pattern: while a single-family development in the exurbs generates a great deal of driving mileage, a new mixed-use building near major transit lines and walkable cores should generate very little.
"People want to live in what are essentially low-VMT neighborhoods," says Ganson. "People don't want to have to get in their car and go far for everything. In a lot of ways, we're ripe for it."
Specifics of the shift are still being determined. (Legislation required a draft recommendation by July 1, but OPR has delayed its formal announcement; until then, some details are subject to change.) There will be room for technical derivatives—say, VMT per capita for a residential building, or VMT per employee for an office—and ultimately local governments set the precise parameters for CEQA metrics. In other words, cities themselves decide how many vehicle miles constitute a project failure. But if the current direction holds, a new CEQA metric with VMT at its core will be adopted sometime in 2015.
"We think it's going to really facilitate where many cities want to go," says Ganson.
• • • • •
Michael Schwartz is explaining how a BRT Baby is made. To be fair, I asked. I wanted to know how a project like Van Ness BRT might have developed in a California where VMT, and not LOS, determined traffic impacts under environmental law. I wanted to know if the same process, done sometime in the future, would have given birth to the same enormous report.
Schwartz opens the report to the traffic section. Back in 2007, as part of its preliminary CEQA analysis, the San Francisco County Transportation Authority estimated level of service grades at 139 intersections on and around Van Ness Avenue. Even without a BRT project, some of these intersections experienced LOS failures by the year 2015 simply based on city growth patterns, says Schwartz. The failures are shown on the map as partially black (for LOS E) or full black (LOS F) dots.
Schwartz then turns the page to the LOS estimates at the same intersections assuming the Van Ness BRT project got built by 2015. Even though the BRT will claim one existing travel lane for its buses, congestion didn't get worse at many intersections, according to the analysis. In fact, car traffic found a way to disperse itself such that there were only two spots where LOS failed in ways that couldn't be avoided. The biggest failure was at Franklin and Market (below, toward bottom), where cars routing off Van Ness clogged the intersection.
The visuals demonstrate the limits of LOS as a traffic measure, let alone as an environmental measure, with striking clarity. Van Ness BRT would produce barely any significant car delay by 2015 in the entire 139-intersection area—to say nothing of the mobility benefits it would provide for transit riders. And yet the entire project fails under CEQA using an LOS traffic metric. (In a 2035 estimate, also required under CEQA, the project failed at several other intersections.)
That left SFCTA with two choices. One was mitigating or offsetting these failures. That would have meant widening the roads or adding turn lanes at the critical intersections, something that, if it were even physically possible, would go against both the city's transit-first policy and environmental logic. With that option off the table, there was only one way forward: a full environmental impact report.
"Not a huge difference," says Schwartz, flipping between the maps. "But this is the CEQA thing: If there's potential for any significant and unavoidable impact, where you can't mitigate it to less than significant … then you have to prepare an EIR."
Under a VMT-based metric, it's pretty safe to assume that Van Ness BRT would have performed better in its traffic analysis. (No one can tell for certain until the new metric is finalized and cities, including San Francisco, set their exact VMT failure thresholds.) The BRT project might add a bit of mileage to some car trips, in the form of drivers routing around Van Ness more than before, but the increased transit use among riders who once drove would more than make up for that. On aggregate, BRT would almost certainly pull vehicle mileage off the transportation network.
More simply put, it's hard to see how the very same project, proposed once LOS reform is in place, would have triggered the need for a full environmental report. ("We all sit here and say, I can't believe we just did all this," says Schwartz.) That doesn't mean the entire six-plus years and seven-plus million dollars spent on the Van Ness BRT report would have been saved; that process included some of the careful preliminary vetting, unrelated to CEQA, that goes into every major project. But Schwartz ballparks the savings, in this hypothetical case, as somewhere in the range of "multiple years and likely millions of dollars."
Later on, we met with Viktoriya Wise of the San Francisco Planning Department, which handles most of the CEQA preparation for city projects. Wise agrees that the planning process should speed up noticeably in a post-LOS era. "We hope—and that's the intention—that the new metric will be easier to calculate, more transparent, and faster," she says. It's rare for a project to need a full environmental report for just one CEQA failure, as was the case with traffic for Van Ness BRT, but Wise thinks planners could save several months on the initial transportation impact study—about a third of the average time. And if a project avoids a full environmental report, the time savings will be even greater.
"To that end, perhaps the transportation system will look like what we plan a little bit faster than otherwise," says Wise. "At the end of the day, when I go home, that's kind of my sincere hope."
"That it wouldn't take seven years for BRT, basically."
• • • • •
The way it works right now, you might say public projects born into California cities grow up in a broken home. The planning profession, which looks favorably on dense mixed-use environments and multimodal networks, has long since been separated from the environmental law under LOS, which looks favorably on remote development and more road capacity. With VMT as the CEQA traffic metric, the marriage of urban planning and environmental policy should be a more harmonious one.
In addition to speedier completion, that union should have numerous indirect benefits. Take, for one, the so-called "last-in" problem that exists with LOS under CEQA. Let's say a developer wants to plan a building at an intersection that's already on the verge of an LOS failure. If the project tips the scale to failure, its leaders are on the hook for offsetting a problem largely caused by developers who came before them. The potential to trigger a last-in failure might lead a developer to water down a project, or to move it somewhere else, or to scrap it all—an invisible negative impact on a growing city.
Questionable lawsuits should decline, too, say advocates of LOS reform. The most infamous example is the case of San Francisco's ambitious bike plan, which was entangled in years of litigation on the grounds that its traffic impact under CEQA should require a full environmental impact report. "All transit and bike projects get greatly delayed because of the 'dire effect' they might impose to the environment," says Tumlin, the transport consultant. Once CEQA's traffic metric and modern planning ideas fall into step, such counter-intuitive legal attacks will lack a solid standing.
It's not just that a new metric should encourage transit and transit-oriented development, it's that it should discourage car-first planning, says Amanda Eaken, deputy director of urban solutions for the Natural Resources Defense Council, who pushed strongly for S.B. 743. Projects that previously scored well under the LOS framework of CEQA—a highway expansion, say, or an exurban development—may now grade out as detriments to the environment.
And even though CEQA is unique to California, Eaken believes the spirit of LOS reform will have an effect on planners across the United States. "I think probably, if you looked, you'd see LOS as a barrier to sustainable communities in every state in the country," she says. "It's a conversation that's long overdue and I'm very pleased we're leading the way toward coming up with a replacement metric. I hope and do expect our efforts will have some traction elsewhere and help others to make similar change."
Some of the country's more progressive cities have already started to leave LOS behind. The New York City Department of Transportation has used "reliability," rather than delay per se, to evaluate some of its street performance. What matters is not so much that traffic might average 10 miles an hour, but that sometimes it travels at 20 m.p.h. and sometimes it travels at 2 m.p.h. Jamie Henson of the District Department of Transportation in Washington, D.C., says LOS can come in handy even for transit projects—intersection delay hurts buses as much as cars, after all—but that it's best seen as part of a broader toolbox of planning measures.
"I think it's a measure, as opposed to the measure," says Henson. "We're like every other place: It ends up being a central element of what we do. But we do at least try to keep it in its appropriate context."
Richard Dowling, a senior principal engineer at Kittelson and Associates, has done extensive research on alternative metrics to LOS—primarily a suite of measures called "multimodal LOS" that aims to leverage the benefits of LOS with the needs of other transportation modes. He says LOS should be a data point at the start of the planning process, not a basis for decisions at the end of it. At the same time, he says, there's no single "magic performance measure" out there capable of resolving every question about an urban planning project by itself.
"LOS is not a perfect measure; it doesn't cover everything," he says. "Let me know when somebody creates one that does it all."
• • • • •
As a kid, says Michael Schwartz, "I was a transit geek." He rode the elevated train around Chicago and grew obsessed with maps; later on, he quit a job in advertising to bicycle across the country. The journey took 10 weeks, traveling east to west, and when he and the group finished in San Francisco, he knew he wanted to move there. He joined SFCTA in the fall of 2008, excited to enact the big transportation ideas he'd dreamed of in his younger days. "We're actually looking at these things," says Schwartz. "We want to build BRT. We're studying congestion pricing."
Changing the CEQA metric from LOS to VMT will help put those big ideas in motion, he says. That's especially true in San Francisco, whose longstanding "transit-first" policy clashes by nature with a car-first metric like LOS. And contrary to some fears that urban planning agencies will now impose their ideology on city residents, the public feedback process means projects will always need to address local concerns, says Schwartz. If an environmental impact report is appropriate, of course planners will do one. But it will be planning for the sake of planning, not the sake of litigation.
"This is why I'm in this field," says Schwartz. "To be able to explain important concepts to people who care and get input and make this the best project it can be."
One morning, I rode the No. 49 Muni bus from its origin, at North Point Street, south along Van Ness Avenue. The No. 49 was not a bad ride, as far as city bus rides go. Travelers could board from the rear door. Stop names were displayed on an overhead LED sign. The bus smelled like a bus, as one rider loudly declared upon boarding, but the seats themselves were clean.
Still, there was clear room for improvement. Sharing the right lane with regular traffic meant the bus often waited for cars to turn or maneuvered around those that were double-parked. By O'Farrell Street the bus had standing room only, despite the trip occurring well outside of rush hour. After a 12:02 start, we reached Market Street at 12:26, two minutes behind schedule. That's roughly 2 miles in 24 minutes—or 5 miles an hour. And that felt like a good performance.
If the Van Ness BRT performs as expected, it will do noticeably better. The route will absorb the No. 49 (as well as the No. 47) between Lombard and Market streets when it's finished. The bus stops, which are now often nothing more than a posted sign, will be high-quality shelters with rapid boarding procedures; the bus itself will get an exclusive lane and make only 9 stops. The No. 49 makes 17. Travel time in the corridor is supposed to fall by a third.
I got off the bus at Market and Van Ness, which is set to become the final southbound stop on the BRT line. There's already a light rail station here, and a street-level trolley. Twitter's headquarters is a block away on Market, and the city's main civic area—with City Hall adjacent to a performing arts center—is a few blocks north on Van Ness. On one of the corners, there's a Honda dealership, of all things; the site was recently sold to a residential developer, and could one day house as many as 700 units.
This was Michael Schwartz's point when he talked about planning for the sake of planning: San Francisco is already moving in the direction of transit and transit-oriented development, despite all the challenges posed by an environmental law with LOS as its traffic metric. So post-LOS planning isn't going to alter the course of the city; it's going to help the city travel a course it's already on. It occurred to me that the difference between transportation planning before and after LOS reform might be a little like the difference between the No. 49 bus and BRT along Van Ness. They're both working for the same purpose, both traveling the same route, both heading toward the same place, but one will get there quicker than the other.