Greater Greater Washington

Posts by Ben Ross

Ben Ross was president of the Action Committee for Transit for 15 years. His book about the politics of urbanism and transit, Dead End: Suburban Sprawl and the Rebirth of American Urbanism, is now available in paperback. 


Careful jaywalking saves lives

To make streets walkable, we need to re-think the basic principles of how people on foot and people in cars share the roadway. This is the first of a multi-part opinion series.

Pedestrians put themselves in danger if they wait for a walk signal instead of crossing the street whenever and wherever it looks safest. There are no definitive studies, but that is what available evidence strongly suggests.

Photo by nydiscovery7 on Flickr.

Most research on traffic safety focuses on narrow questions posed by the highway agencies that fund it. Basic premises, like the idea that "jaywalking" is intrinsically unsafe, are rarely investigated.

In the absence of systematic studies, one must turn to indirect statistical evidence.

One useful data set was collected for New York's Vision Zero program. That city, where residents routinely ignore signals when they cross streets, can be thought of as a natural experiment. The majority of pedestrian deaths, and a far larger majority of non-fatal crashes, occur while crossing the street legally in a crosswalk.

Why might that be? Drivers hit pedestrians when turning more often than when they are driving straight ahead. At a red light, drivers who are about to turn wait alongside pedestrians. The changing signal sends both into the intersection at the same time—maximizing the opportunities for collisions.

Other researchers, working in places with less foot traffic and fewer striped crosswalks than New York, got results that point in a similar direction. They found that pedestrians crossing big highways are more likely to be struck at marked crosswalks than at unmarked ones. On smaller roads, they found little advantage either way.

The Federal Highway Administration took these findings to mean that putting stripes on highway pavement makes it more dangerous to cross there. It used them to justify a ban on new crosswalk markings, except at traffic lights, on wide high-speed roads. A far more likely explanation is that pedestrians are better judges of their own safety than are traffic engineers, whose first concern is usually to move cars fast.

The concept of jaywalking was invented in the 1920s by motoring lobbies to empty streets of other users. Drivers wanted to go faster and automakers sought to sell more cars. Safety, as Peter Norton has shown in his book Fighting Traffic, was no more than an afterthought.

Almost a century has now passed, and our traffic laws are still not geared to safety.


You don't have to push this button to cross the street

If you walk to a corner and see a button to activate the walk signal, you might need to push it. Or you might not. It might only be there to activate a chirping noise for people with vision impairments. Unfortunately, there's no way to tell.

Connecticut Avenue and N Street in DC. Photo by David Alpert.

Some intersections keep "don't walk" signals lit during both red and green phases of a traffic light unless someone pushes a "beg button"—technically an "actuated pedestrian push button"—before the light turns green.

The sign on the picture above clearly implies that that's what will happen when people wanting to cross the street push the button.

But the button actually has nothing to do with the walk signal. The walk signal comes on whether you press the button or not.

What the button does is turn on a loud chirping noise that speeds up when the walk signal begins. The misleading signs have appeared in large numbers in DC, Montgomery County, and elsewhere over the past year, on local roads and state highways.

Unless you can't see the sign, pushing this button won't help you cross Bethesda Avenue. Photo by the author.

Why is this?

Federal guidelines, known as the Manual on Uniform Traffic Control Devices (MUTCD), authorize only certain standard signs for pedestrians. Among them are several variants for buttons that control the walk signal, but no sign for buttons that merely activate the audible signal for people with visual impairments.

Image from the Federal Highway Administration.

In downtown Bethesda, chirper buttons have appeared in large numbers over the last half year, all accompanied by the standard sign. Frequent passers-by soon recognized that the sign conveyed a falsehood, and now, few people push the button.

From my observation, it has become more common for people to simply cross streets wherever and whenever they feel safe. The streets seem no less safe.

Highway agencies take great care to ensure that signs meant for drivers are accurate and unambiguous—and doing so helps keep all who use the roads safe. Pedestrians, as these pushbutton signs illustrate, get very different treatment.

By taking such a nonchalant attitude toward those on foot, traffic engineers implicitly recognize something their profession refuses to officially admit: Drivers in the wrong place endanger others, but pedestrians do not.


There's a new group opposing a more urban Bethesda

Montgomery County is working on a new master plan for downtown Bethesda that would promote continued development and might allow the population to increase from 7,210 today to over 18,000. But a group of people who oppose urbanization are gearing up to fight it.

Downtown Bethesda today. Image from Google Earth.

In an email to Bethesda civic leaders on Saturday, former Town of Chevy Chase Mayor Pat Burda revealed plans to form the Coalition of Bethesda Area Residents. Organizers say that they are "angry" at plans for increased density, with special ire toward the quantity of new housing.

While CBAR hopes to draw support from single-family neighborhoods on all sides of downtown Bethesda, the initiative comes from the Town of Chevy Chase, a 1,200-home enclave located southeast of the downtown commercial area. (The town is just one of the many neighborhoods that make up what is generally considered Chevy Chase.)

A report submitted to the town council earlier this month presents what CBAR is trying to accomplish. The central premise of the 13-page document, which is framed as a history of downtown Bethesda development, is the primacy of the single-family house and its owner. Apartments, stores, and offices are welcome only to the extent they serve the residents of nearby homes.

The document is clear in its opposition to development, skipping the usual pieties about planning and community participation. Land use decision-making, it explains, is an "inherently political process." If one county council can upzone, the winners of the next election can equally well downzone.

The document's author is Scott Fosler, who was among the early architects of the anti-development movement that has flourished in the Chevy Chase area since the 1970s. This movement, acting under a variety of organizational umbrellas, has had a strong influence on land use policy throughout Montgomery County. It was most recently in the news in 2012, when it tangled with county planning director Rollin Stanley after he referred to some of its leaders as "rich white women."

Fosler himself served two terms on the Montgomery County Council in the 1980s. He earlier chaired the town's zoning committee and then served on the town council.

While Fosler insists that homeowners have every right to change the zoning of adjoining properties, there is no reciprocity. He would find it unthinkable to rezone the Town of Chevy Chase to better serve the remainder of the county.

Fosler sees Bethesda's urban center, even when properly subservient, as an alien intrusion that is best kept at a distance. He outlines the strategy by which the Town of Chevy Chase and its allies have obtained the separation they desire. The downtown is encircled with a "comprehensive cordon" of land occupied by parking lots, parks, and house-sized structures. The function of this territory is simply to be as empty and little-used as possible. Property on Elm Street that was made a park would serve the purpose just as well if it were a parking lot.

In the past, the anti-development movement had great influence over land use in Bethesda and Chevy Chase. The 1976 Bethesda master plan imposed a five-fold reduction in the allowable square footage of the downtown. The 1998 plan for Friendship Heights limits buildings on some stretches of Wisconsin Avenue to three stories.

But public opinion is shifting as the demand for urban living grows. Divisions have emerged even within the Town of Chevy Chase. Whether CBAR can achieve the same political power as its predecessors remains to be seen.


In some places, living near Metro has become more affordable

Washington-area neighborhoods in walking distance of Metro are wealthier and whiter than their surroundings, according to a new Census Bureau study. But for many places outside the District, living near Metro has become more affordable.

How earnings of workers who live near Metro and elsewhere have shifted. Click the image for definitions. Tables from the Census Bureau.

Working at the Census Bureau, Brian McKenzie can see data that privacy rules keep from other researchers. His new research paper is chock-full of interesting data.

McKenzie was able to compare surveys taken in 2006-08 and 2011-13, and compare DC residents to those who live in the five-county area of Alexandria, Arlington, Fairfax, Montgomery, and Prince George's. Using the individual addresses of people who have jobs and answered Census surveys, he separated those who live on street blocks within a half-mile of Metro stations from those who live elsewhere.

This isn't a perfect way of identifying who can walk to the train, but it's far superior to what other researchers have been able to do.

Among McKenzie's key findings is that more people are living near Metro, and more of them are riding the transit system. In five years the number of workers in walking distance of a station rose 23%. The working population living farther from Metro increased just 5% region-wide, and dropped slightly in the District. Over 15% of the region's 1.8 million workers, including a majority of the 321,000 who live in DC, now have a short walk from home to Metro.

The data also show the increasing use of non-auto transportation, cycling even more than transit. This trend is strongest among DC residents who live near Metro. Although this population is growing rapidly, the number of drivers among them hardly changed, so that the percentage who drive to work alone plummeted from 30% to 25%.

Transit use in this group increased from 57,000 to 72,000, and the number of bicycle commuters soared from 3300 to 7900. Trends among other demographics are similar in direction, but slower.

Trends in travel mode to work.

The most widely noted finding of this study is the increasing affluence and whiteness near Metro. This, however, is essentially a DC phenomenon. In the surrounding counties, the income spread between walk-to-Metro housing and elsewhere is, if anything, shrinking.

All ethnic groups grew in absolute numbers near non-DC Metro stations; the share of both whites and blacks declined as Asians and Hispanics moved in faster. And where the percentage of these counties' residents with income over $100,000 was 2.1% higher near Metro than elsewhere in 2006-08, the difference fell to 1.7% five years later. Most other income groups show a consistent pattern of shifts.

Because these changes fall within the margin of error, it's not clear whether the difference between incomes near Metro and farther from it is really closing, but the gap is not growing wider as it is in the District.

New apartments and inclusionary zoning have helped with affordability

Why hasn't Metro-accessible housing in the outer counties become less affordable? Most of the credit undoubtedly goes to the smart growth zoning that has opened up stations in Montgomery and Arlington Counties to new apartment construction. Builders there are required to include a percentage of more affordable units in new construction. And while newly built market-rate apartments in Bethesda or Clarendon aren't cheap, they are (with the occasional exception) less expensive than the single-family houses nearby.

The District, meanwhile, lagged behind in enacting inclusionary zoning and then stalled on implementing it. And it has allowed little new construction near Metro in the upscale neighborhoods west of Rock Creek.

Demographic and cultural change may be the motive force behind shifting living patterns, but public policy makes a big difference in how things play out.


Congress gives Metro riders an early Christmas present

If you're a federal government worker, you'll soon get up to $255 a month to pay for transit under a tax bill Congress agreed on last night. Or, if your employer allows setting aside pre-tax earnings for transit, you will also be able to reserve more. This will translate into badly needed fare revenue for Metro—perhaps as much as $15 million a year.

Photo by Elvert Barnes on Flickr.

The federal transit benefit lets employees at many organizations put pre-tax earnings aside to cover transit or parking (and sometimes also bicycling). In addition, some employers, in our region most notably the federal government, pays for employee transit up to the allowed transit benefit amount.

For the last two years, the transit benefit has been capped at $130, but $250 for parking The new law—which is expected to pass both houses and be signed within the week—will set both limits at $255, with increases to match future inflation.

Under an executive order signed by President Bill Clinton, the benefit to federal employees automatically increases to match the tax-free maximum. So federal employees whose commutes cost more than $130 a month (the benefit cannot exceed your actual fare) will soon see more money on their Metro farecards.

Lowering the cost of transit for long-distance commuters will attract more riders and bring in more fare revenue. A study by Metro estimated that the system lost 6,000 daily round trips as a result of the January 2014 cut in the transit benefit from $230 to $130. If these riders return at an average fare of $5 each way (only high-fare commuters are affected), the agency would gain $15 million a year in revenues.

MARC and VRE commuter trains can also expect more riders, as can express buses from outer suburbs.

Congress made the new limit retroactive, but few riders will benefit from this provision. Employers will not be able to make payments for past months (although it is still possible to make a payment for December). However, anyone who has been getting a transit benefit of over $130 a month and paying taxes on it will be able to deduct the entire amount (up to $255 a month) for the whole year.

Details of how the new limit will be implemented vary among federal agencies. It's probably best, however, to apply quickly for the full amount of your commute expenses so that the added benefit will start as soon as possible.


Montgomery's most walkable streets are also its safest

Downtown Bethesda, Silver Spring, and Rockville are where it's easiest to walk around in Montgomery County. They are also where drivers are least likely to kill someone on foot.

A decade of traffic fatalities in the Wheaton-Glenmont area. A red person represents a crash where the driver died. Orange is passenger, yellow is pedestrian, and purple means multiple people died. Image by Max Galka.

An extraordinary new interactive map by Metrocosm shows the location of all traffic fatalities in the United States between 2004 and 2013.

Zeroing in on Montgomery County, the map shows that pedestrian death comes in clusters that center on high-speed suburban arteries. Drivers killed eight people on foot in Aspen Hill, for example, four in downtown Kensington, and five on a 3500-foot stretch of Route 118 in Germantown Town Center.

The places where people walk the most are far safer. One pedestrian died in Bethesda's downtown, one in Rockville's, and three in Silver Spring's—and all five of these killings occurred on the fringes of the urbanized centers.

Drivers killed 4 people walking in downtown Kensington (top) and one on downtown Bethesda's more urban streets (bottom), even though Bethesda's downtown is bigger, has more people on foot, and is hardly as easy to walk in as it might be. Photos by the author.

These downtowns are hardly walking paradises—they contain many of the county's identified hotspots for frequent pedestrian crashes—but they share some characteristics that seem to prevent fatalities. Streetcorners are close together, stores front directly on the sidewalk, and speed limits are reduced.

Trends elsewhere in the region are similar. In the District, roads engineered for incoming commuters, New York Avenue in particular, are deadlier than downtown streets where far more people are on foot. Old Town Alexandria and, to a lesser degree, the Rosslyn-Ballston corridor stand out as islands of safety in the Virginia suburbs.

Indeed, researchers who looked at data from the entire country found that there are fewer traffic deaths of all kinds, and especially fewer killings of pedestrians, in counties with denser population and smaller blocks.

Unfortunately, Montgomery County continues to build the kind of roads where drivers kill. County and state transportation officials have made concessions in the long-running battle over street widths and speed limits in the rebuilt White Flint, but elsewhere they continue to resist life-saving urban street designs.

The highway engineers have been especially obstinate in insisting on dangerously large street blocks. At Glenmont Metro, for example, the State Highway Administration rejected a street connection for being too close to another corner. With evidence accumulating that smaller blocks are safer, the agency will be on very shaky legal ground if it tries to issue such vetoes in the future. Under Maryland law it may only deny a builder access to a state highway "to promote safety."

The new data show a way forward to make pedestrian killings the rare events they should be. The urban places that the market now demands are not only more pleasant, but safer too. Rebuilding suburban highways as city streets saves lives.


Suburban-style streets don't fit a busy Bethesda corner

Cyclists leaving the Capital Crescent Trail and entering Bethesda at the corner of Woodmont and Bethesda Avenues are in serious danger. Pedestrians and drivers, too, suffer unneeded delay at this busy crossroads. There's a common cause for these problems: planners using suburban highway engineering practices that are unsuited to an odd-shaped urban street crossing.

There's a curb where the Capital Crescent Trail meets Bethesda Avenue. Cyclists must dismount or ride on the sidewalk. Photos by the author.

Newly built stores and apartments and a big underground garage have replaced a pair of parking lots on the south side of Bethesda Avenue. Responding to long-standing public complaints about safety, the designers of the new buildings moved Woodmont Avenue slightly and removed a high-speed turn lane. But county transportation officials vetoed suggestions from bike and pedestrian advocates for more drastic changes in the layout of the streets.

The pavement here, at the point where the trail comes into downtown Bethesda, carries a continual flow of bicycles. Theaters, restaurants, and shops draw heavy foot traffic. Plenty of cars come through too, although removal of the surface parking lot, a magnet for drivers, has made traffic on Woodmont Avenue lighter.

Image from Google Maps.

Immediate danger to cyclists

As construction now winds down, county traffic engineers are restriping traffic lanes and moving crosswalks. Their new layout compounds a glaring weakness of earlier designs: lack of attention to cyclists' turns between roadways and the trail.

The county has built a curb at the intersection of the trail and Bethesda Avenue, removing a ramp put there soon after the trail opened. Northbound cyclists arriving on the trail now must either dismount, walk into the street, and climb back onto their bikes while standing in a traffic lane, or ride on the sidewalk.

From this sidewalk, cyclists continue north across Bethesda Avenue at a crosswalk. Maryland law requires cyclists to wait at the curb ramp for the walk signal.

View from the Barnes and Noble at the northwest corner of where the streets cross. Conflicting movements of cars (blue) and bicycles (green) when Woodmont Avenue has green light. Official bike path alignment shown in orange.

The walk signal to cross Bethesda coincides, however, with a green light on Woodmont Avenue. The signal sends law-abiding, on-road cyclists diagonally across four lanes of moving automobile traffic to reach the northbound Woodmont Avenue bike lane or the Georgetown Branch Trail. Drivers in these lanes have the green light and do not expect cyclists to be crossing in front of them. (Few adept cyclists follow the bike trail mapped by the county, which directs them to dodge pedestrians and street furniture on three crowded sidewalks, make repeated 90-degree turns, and, if headed for the bike lane, wait at slow traffic lights three times.)

The county could eliminate this obvious danger by giving cyclists and pedestrians their own green signal, with cars stopped in all directions. Or the trail could get a curb cut, with or without its own traffic light. But such fixes would do little for those on foot and make things worse for drivers.

The walk signals at the corner are too short

As things stand, pedestrians crossing Woodmont Avenue get a seven-second walk signal (the shortest allowed under federal guidelines). A 113-second interval follows when it is illegal to start walking.

The seven seconds when pedestrians may enter the crosswalk is just the time that cars are most likely to be in their way. The walk signal begins at the moment the light changes. That's when drivers stopped by a red light on Bethesda Avenue make their turns. During the long don't-walk interval, on the other hand, few cars come by.

People who walk in this area have reacted as one would expect to its nonsensical signal sequences: they now pay little heed to signs or signals of any kind.

As with the bike safety problem, it's easy to suggest a modest improvement in the Woodmont walk signal: It could be lengthened without changing the timing of the traffic light. But this would leave a crossroads still unfriendly to people on bikes, people on foot, and drivers alike.

The real problem is the traffic engineering doctrine

One fix after another has failed here because Montgomery County remains wedded to old-style traffic engineering. Two of the profession's basic tenets are to blame. First, engineers design streets for cars and see other users as obstacles. Second, they design the roads to minimize rush-hour delays rather than to work best at all hours. When you measure success by counting cars in rush hour, wide streets and slow lights are solutions rather than problems.

At this corner, with its oblique angle and many diverse travelers passing through, the weaknesses of this doctrine are manifest. The more different groups that must wait in turn for others, and the more time each group needs to get across the wide-angled pavement, the longer all are delayed. The long waits make pedestrians and cyclists feel unwelcome, and drivers suffer too. A redesign based on fundamental rethinking would be better for everyone.

A single-lane roundabout could easily handle all the cars that pass through this intersection. Traffic might back up at the busiest hour on Friday and Saturday evenings, but the time lost in those delays would be much less than the time drivers would save the rest of the week—and pedestrians and cyclists would benefit at all hours.

Concept for intersection redesign. Blue: curb extensions. Orange: crosswalks. Dark green: cycletracks. Light green: bike lanes. Red: angle parking. Pink: possible traffic lights. Base from Google Earth.

If all cars feed into a single roundabout lane, turn lanes serve no purpose and the streets can be narrowed. Narrow streets make drivers slow down—a needed remedy for Woodmont Avenue's late-night speeding problem. There would be room for angle parking and protected cycle tracks, and pedestrians would need less time to cross the street.

The two main bicycle and pedestrian crossings could be moved away from the roundabout. Traffic lights might be needed at these crossings, but if so the narrow roadway would allow them to have short cycles and impose little delay.

My sketch surely needs refinement, and some entirely different concept may turn out to be better. But whatever the ultimate solution is, it will require new thinking. This corner has vexed Montgomery County officials, and it has vexed them for a reason. Car-centered engineering practices don't make good places for people.


Maryland officials won't talk about more MARC trains

Maryland transportation officials have nothing to say about running more MARC trains between Union Station and Frederick County. That's what state legislators and commuters learned at a recent public meeting in Germantown.

A MARC train at the Dickerson station. Image by Samual Ruaat on Wikipedia Commons.

Frederick County Delegate Carol Krimm led a group of state legislators in organizing a meeting to hear from the state officials about plans to help residents of northern Montgomery and Frederick Counties travel to jobs located closer to DC.

Delegate Krimm asked the Maryland Department of Transportation to address two topics: Governor Hogan's recently announced plans to spend $100 million on I-270, and potential upgrades to MARC's Brunswick Line, which serves the same corridor.

Greg Slater, chief planner for the State Highway Administration, talked for some time about the Hogan plan for I-270, but he was short on specifics. He said the state will spend $100 million on "innovative traffic congestion relief strategies," but he couldn't say what those will be. SHA plans to issue a Request for Proposals and see what ideas the bidders come up with.

Kevin Quinn of the Maryland Transit Administration followed, speaking at equal length on transit other than MARC. On the Corridor Cities busway and Montgomery County's BRT planning, he simply recapped previous public presentations.

The one transit topic where new information emerged was investigations into running buses on the shoulders of interstates. In many places, Quinn said, SHA had found obstacles like bridge pillars and roadbeds too weak to bear the loads, but it expected to soon locate a stretch of interstate for a trial of the concept.

Finally, MARC's Erich Kolig gave a very short talk focused mostly on new equipment recently purchased by MARC. He said nothing about what the audience had come to hear about: what it would take to expand the Brunswick Line beyond the one-way rush-hour service now provided. Most of the half-dozen legislators and 50-odd members of the public in attendance sported "More & Better MARC" stickers distributed by the Action Committee for Transit, which had leafleted train stations about the meeting.

MARC's Brunswick Line. Map from MTA Maryland.

Legislators and the public peppered the visitors with skeptical questions, complaining of insufficient action on MARC and other fronts. State officials answered inquiries about I-270 and Corridor Cities with detailed information about travel patterns and previously released studies. But they said almost nothing in response to repeated calls for more trains. The most Kolig could offer was that cars with space for bicycles might soon run on the Brunswick Line on Mondays and Fridays.

Another meeting on this topic will take place early next year.


Complex traffic signals make streets less safe

Streets across the United States are often difficult and dangerous to walk on because wide lanes invite drivers to speed. That isn't all that makes them dangerous, though: many also have signals that distract drivers and draw their eyes away from the road.

Arlington's Intersection of Doom. Drivers who want to turn right in order to travel north from the eastern side of the intersection have to account for oncoming north-bound cars, people crossing from all directions, and confusing signs. Photo by the author.

A case in point is the "intersection of doom" where the Mount Vernon Trail turns into the Custis Trail at the foot of the Key Bridge in Arlington.

Drivers exiting I-66 are allowed to make right turns on red, except during a brief "leading pedestrian interval" when a walk signal gives pedestrians and cyclists a head start across North Lynn Street before the traffic light turns green. But it's hard to see how drivers can make this turn safely without extra eyes on both sides of their head.

Base image from Google Earth.

Drivers must simultaneously watch for cars coming from the left, cyclists and pedestrians entering the crosswalk from the right, and an overhead signal that went in in January that flashes a no-right-turn graphic for a few seconds during the leading pedestrian interval.

To make things worse, the no-right-turn graphic is hard to see in bright light, and it is flanked by highly visible signs that seem to say turns are allowed.

Diligently watching the short-lived no-turn signal while looking for a gap in oncoming traffic from the left would make it nearly impossible to look to the right. A driver trying to legally turn right on red has no time to look toward the sidewalk on their right and can't see whether someone is about to enter the crosswalk and pass in front of their car.

Overhead signs at intersection of doom, with the no-right-turn sign illuminated.
Photo by the author.

Dangers like these are widespread on American streets. What makes this intersection stand out is the heavy bike and pedestrian traffic, not the arrangement of the signals.

Tell drivers what they need to know, and repeat it

From an engineering point of view, the information traffic signals send to drivers is part of a control system that must operate reliably to keep roadways safe. So is the drivers' reaction to that information.

The traffic engineering establishment certainly recognizes that human behavior affects road safety. But two concepts are conspicuously missing from its guidelines on human factors in signal design: redundancy and parsimony.

Redundancy means backups for missed signals and improper actions. Parsimony means signals aren't excessively complex.

It's easy to see redundancy's value. Intersections with simple red-yellow-green traffic signals are full of redundant information: The movement of vehicles and pedestrians is a cue to when the light changes, so drivers don't need to stare at the signal and can keep watch on the street.

More complexity—turn arrows, walk signs, rules that allow right turns on red—means less redundancy. Demands on the driver's eye and brain increase, and the inevitable moments of inattention do more harm.

Parsimony is a less intuitive idea, but an equally important one. This principle, which originated from the statistical analysis of time series, warns against using too many input variables to control decisions. Adding complexity, when there aren't enough data to do it right, makes outcomes worse.

Consider the countdown clocks attached to walk signals. The Federal Highway Administration mandated them when research indicated that when pedestrians know how much time is left before cars start to move, they get hit by drivers less often. But this information changes the behavior of drivers too. Whether the drivers speed up to beat the light or simply get distracted is not clear, but the effect is real. A recent study in Toronto found that countdown timers cause more collisions than they prevent.

Complex signals have other costs. Turn arrows make signal cycles longer. This gets more cars through the intersection, but everyone waits longer for a green light and the street becomes a barrier to walking. The delays to pedestrians and non-rush-hour drivers may exceed the time saved from reduced congestion—but the traffic engineers won't know without data on off-peak travel, something they rarely measure.

And slow lights create safety hazards of their own. The longer pedestrians are asked to stand and watch a don't walk signal, the more likely they are to ignore it.

It will never be possible to govern traffic with the mathematical precision of electric circuits. But they are both control systems afflicted with random noise, and similar principles apply.

Solving problems by adding more gadgets to an already complex system can do more harm than good. Traffic control operates most reliably when everyone on the road knows that green means go and red means stop.

Turn arrows and separate walk signals should be used sparingly. They squeeze more cars through an intersection in rush hour, but they exact a price in safety, dollars, and travel delay.

Traffic engineers need to balance vehicle throughput against the benefits of redundancy and parsimony. In any control system, and especially in one that relies on the actions of human beings, simplicity has to be a priority.


A Maryland road widening will be more costly than the transit it replaces

Maryland governor Larry Hogan wants to build roads with money saved from cancelling the Baltimore Red Line and cutting back the Purple Line. The governor says the two light rail lines cost too much. But his marquee highway project, a wider Route 404 on the Eastern Shore, looks to be far less cost-effective than either.

Route 404. Photo by Doug Kerr on Flickr.

The Route 404 widening will turn 12 miles of two-lane road between Route 50 and Denton into a four-lane divided highway. The work will cost $204 million: $160 million in new money plus $44 million budgeted earlier.

The governor presents his road plan as a way to speed traffic. But the travel time savings from widening Route 404 will be far more expensive than the time saved by the two rail lines.

The two-lane road only backs up on summer weekends when people drive to the beach. According to Google maps, the average traffic delay on summer Friday and Sunday afternoons varies from zero to six minutes. By a generous estimate, this adds up to 60,000 hours lost each year in traffic backups, making the construction cost $3,400 per annual hour saved.

Building the Purple Line will cost $288 per annual hour of rider benefits, and the number for the Red Line is $456. The amount of money the state is spending to save a minute of travel time on Route 404 is seven and a half times greater than the amount it refused to spend to save a minute of travel time in Baltimore. That means a Baltimore bus rider will wait an hour so that an auto passenger can get to the beach eight minutes faster.

Highway safety is another goal, but widening 404 may not help much

Eastern Shore officials offer another rationale for widening Route 404. There are many fatal crashes on the road, and they suggest that the planned widening will fix that. But it's unlikely that the death toll will go down significantly.

Using web searches and a memorial website, I found descriptions of 11 fatal crashes on Route 404 since 2010. Seven of them were on the 12-mile section of two-lane highway; four on the 12-mile stretch that is already four lanes.

Of the seven collisions on the two-lane road, only two involved vehicles crossing the center line. Three vehicles were hit from the side as they turned onto 404 from side roads. There were two rear-end collisions. On a four-lane divided highway, center-line crossing would be impossible, but turns would be more difficult. These numbers suggest that widening 404 would only modestly improve safety, if at all.

Also of note: five of the 11 crashes involved tractor-trailers. Requiring through trucks to use US 50, which has far fewer intersections without signals, might have prevented most or all of these.

Moreover, former Maryland highways chief Parker Williams has said that Route 404 isn't an especially dangerous road, which implies that highway safety money could be better spent in other places. For the cost of widening 404, the state could install some 2,000 of the flashing crosswalk lights known as hawk beacons. They would undoubtedly have saved a good number of the 630 pedestrian lives lost on Maryland highways between 2009 and 2014.

The highway projects in Governor Hogan's package have never gotten the sort of detailed assessment of costs and benefits that the Red and Purple Line projects were subject to. The numbers for Route 404 suggest that cancelling the Red Line was not at all the cost-conscious decision the governor presented it as.

Support Us