Men At Work

How much does road construction really snarl up traffic?

Cars stuck in a traffic jam.
A traffic jam 

The highway sign that cheerfully advises “Your Tax Dollars at Work!” typically comes as small comfort to the driver caught in a simmering mile-long queue. This summer, there’ll be a new twist: Drivers will also see signs with slogans like “Putting America to Work,” signs that tout the “American Recovery and Reinvestment Act,” the federal stimulus package that allotted $27 billion to road projects and promises to send a ferocious torrent of asphalt and orange cone-shaped hail raining down on highways across the land. “People will see more construction. There’s no question about it,” Florida’s Department of Transportation chief told USA Today.  

The good news, of course, is that there hasn’t been a better time in half a century to undertake a massive infrastructural overhaul: Thanks to the recession, there are fewer cars on the road to be inconvenienced by repairs. Total vehicle miles traveled dropped 1.2 percent from March of 2008 to March of this year (including the largest month-to-month drop in car travel ever recorded), with congestion plunging 30 percent on major urban roads during peak hours. But still, given that the vast majority of the stimulus is going toward repairing existing “mature” roads, which tend to have the most traffic, drivers can expect retro-reflective vests to be one of the most-spotted fashion trends of the summer. Which makes it an apt moment to examine whether road construction is as bad for traffic as we suspect—and whether there’s anything we can do about it.

The bane of drivers is the “work zone,” that place where a bustling four-lane highway whittles down to three (or worse). The old Northern joke says there are only two seasons, winter and road construction, and it’s been estimated that in a typical summer (which this is not) there are more than 6,400 work zones in the United States, covering some 20 percent of the National Highway System. These work sites are said to cause 482 million hours of vehicle delay each year, accounting for one-quarter of what’s called “nonrecurring congestion”—all those random events and “incidents,” like crashes, that by some accounts cause more urban congestion than physical bottlenecks or simple lack of capacity.      

The negative traffic impact of a work zone isn’t hard to fathom: Losing one lane of four represents a 25 percent drop in capacity. But that’s theoretical capacity. Inefficient (or hostile) merging may make things worse, which is why engineers have struggled to come up with better ways to move motorists through work zones.(Perhaps the best approach, often used by European road agencies on their comparatively denser road networks, is to simply not lose a lane to begin with, instead restriping drivers into the usually vacant hard shoulder.) Drivers may also be distracted by the events going on in the work zone; the quick and often unexpected slowing that happens approaching a construction zone is in fact a significant source of crashes. It is a vicious cycle: The congestion caused by the work zones leads to crashes, which leads to further congestion.

Although the creeping motorist may not appreciate it, there have been a number of innovations in how road projects are executed that has helped dampen their negative effects. Much of this was borne by necessity: From the early 1980s to the early 2000s, as “vehicle miles traveled” increased nearly 80 percent, the number of new “route miles” added to the nation’s road network grew by only 3 percent. “Rush hours” essentially doubled in length during that period. At the same time, the “design life” of many segments of the U.S. Interstate System was reaching its end. There was more work to be done in a shorter window.

The easiest strategy—in principle, if not in actuality—is simply to build roads that require less maintenance. The U.S. Interstate Highway System was built on a rapid 19-year schedule (Congress mandated it be finished by 1975; it was not), with a greater emphasis on quantity than quality (giving way to such things as the “rutting epidemic” of the 1980s—we’re talking asphalt, not antlers). That approach has largely shifted, in places like California, to emphasize so-called “long-life pavements,” with an intended design life of 40 years. Long-life roads typically use fast-setting hydraulic-cement treatments rather than asphalt, though there is a vigorous debate on the virtues of concrete vs. asphalt, “white” vs. “black” roads, that itself could consume an entire column—is the world ready?—but that can be roughly summarized as: Concrete is more expensive and more time-consuming to pave, whereas asphalt is cheaper, easier to work with—and often smoother—but doesn’t last as long.

One common response has been to shift almost entirely to nocturnal construction. (Rush-hour roadworks are a very rare sight these days.) On highways like the Connecticut Turnpike, which as a RAND report notes operates during peak hours at “close to 185 percent of its target capacity,” most road construction happens between 8 p.m. and 6 a.m. While this means paying higher wages to workers (adding, according to one study, about 6 percent to the base cost of a lane improvement) and often extends the time it takes to complete projects, the other benefits—less congestion, fewer work-zone crashes, reduced emissions, etc.—on average seem to far outweigh the added costs. Another strategy has been to make the road crews themselves more efficient. So-called “cost-plus-time” bidding (or “A+B bidding” as it’s known in the business) makes completion time an essential variable in the bid, while the even more effective “lane rental” strategy actually charges the contractor for the right to “occupy” a travel lane (a form of “congestion pricing,” if you will) in increments as small as 15 minutes.

Perhaps the most radical but effective solution is known as the “full-road closure.” This no doubt sounds apocalyptic (and a touch counterintuitive) to the commuter, but road agencies have gradually learned that shutting down one direction of travel or an entire highway—typically round-the-clock during several successive weekends but sometimes for longer—is faster than trying to complete road projects while active traffic whizzes by. No traffic means no work-zone crashes and less interference from drivers, and it also ensures that trucks hauling gravel and other material and equipment no longer have to sit in the same congestion as everyone else en route to the work site. A study by the Federal Highway Administration looking at six full-closure projects found time savings ranging from 63 percent to 95 percent versus projects that tried to maintain traditional traffic.

Closing much or all of a highway requires extensive upfront planning—what’s called the “Traffic Management Plan.” Drivers have to be warned and given viable and signed alternate routes, and attention must be paid to the overall “network effects.” For example, when Caltrans (California’s DoT) repaired the I-710 (the Long Beach Freeway) over a set of eight 55-hour weekend partialclosures, it monitored suggested detours and other nearby routes for traffic impacts. * The agency found it was able to reduce traffic through the construction zone by some 37 percent, while traffic on nearby arterials, not surprisingly, increased; a certain number of drivers, meanwhile, were simply “no-shows,” deciding the added hassle wasn’t worth it. But Caltrans noticed an interesting pattern: Traffic through the congested work area was lowest the very first weekend but gradually grew over successive weekends, as a kind of “learning curve” set in and drivers tested varying strategies for getting around the detours.

So, a lesson for this summer may be: Drive through the construction you’ve been warned about when you first hear the warnings. After that, it’s only going to get worse.

Correction, June 11, 2009: This piece initially referred to the Long Beach Freeway as the 701. It’s the 710. (Return  to the corrected sentence.)