Bigger Roads Are Less Safe
by Dom Nozzi
In a great many cases, "improved safety" is a
reason cited as a rationale for adding travel lanes to a road
("widening" a road). Indeed, because "improved safety" is a
"moral high ground" argument (i.e., the argument should be accepted
for ethical reasons), the safety rationale is perhaps the most common reason
given for why a road "must" be widened.
In effect, public policy makers, when confronted with the
"public safety" justification, are forced into an uncomfortable
position when a decision must be made to widen or not widen a road: Either
agree to the widening, or take a position that seems to suggest an uncaring
attitude toward public safety.
It comes as no surprise that a large number of
decision-makers are persuaded solely on the basis of the public safety
argument.
Because road widenings are enormously expensive, speed up
car traffic, and can profoundly worsen quality of life as well as accelerate
strip commercial development and urban sprawl, we must be certain that road
widenings do, in fact, deliver on the promise of dramatically improved safety.
The Forgiving Road
The "Forgiving Road" is a road that
"forgives" a motorist when a driving mistake is made. That is, being
reckless, or driving at high-speeds, or driving inattentively is not followed
by the "punishment" of consequences such as crashing into something
on the side of the road. For several decades, we have designed forgiving roads.
We have been pulling buildings, parked cars, pedestrians, bicyclists, trees and
other "obstructions" away from the sides of roads so that even an
unskilled motorist can travel at high speeds without crashing into something.
The forgiving road was thought to be a way to promote
"safety"(the hidden agenda, for many, was to promote high-speed
travel by large volumes of car traffic).
Of course-human nature being what it is-such a design
encourages reckless, high-speed, inattentive driving because human psychology
compels us to tend to drive at the highest speed that still feels safe. After
all, we are always "running late." We are always in a hurry. And we
are so busy.
The forgiving roadway lulls us into a false sense of
security. Vigilance and concentration wane on the forgiving road. Is it any
wonder that today, we increasingly see motorists driving at high speeds with
one hand, while putting on make-up, drinking coffee, or chatting on the cell
phone with the other?
Since we tend to be busy and in a hurry, forgiving streets
deliver lots of motorists who drive as fast as they can and "multi-task"
while driving. Why? To save time.
The predictable result: An increase in crashes due to
speeding, inattentiveness, and recklessness.
Ironically, motorist safety declines and driving skills
atrophy, because the forgiving street conditions motorists to be less careful
drivers, and lowers the need to maintain or improve driving skills.
Increasingly, American motorists drive dangerously, and more ineptly.
Three Lanes vs Four Lanes
Some safety analysts point out that 3-lane roads are
noticeably safer than 4-lane roads, in part because, when comparing 3 lanes to
4, average vehicle speeds are reduced, there is less variability in vehicle
speeds, and there is less speeding. In addition, there is a significant
reduction in what engineers call "conflict" points, and an increase
in "sight distance" for turning and crossing traffic on a 3-lane
versus 4-lane road (Welch, undated).
This is particularly important for senior citizens who are
motorists, because fewer conflict points and increased sight distances means
fewer decisions and judgements have to be made to enter or cross a 3-lane road.
Similarly, a 3-lane road reduces the street-crossing
distance for pedestrians. Compared to a 4-lane road, a 3-lane can create
"refuge" areas where a pedestrian can safely wait until there is a
safe gap in traffic before crossing the other half of the street. A refuge is
also created for motorists with 3 lanes.
A review of the research on this question raises significant
questions as to whether wider roads are safer roads.
Fewer Travel Lanes
A study published in 2002 (Huang, Stewart, Zegeer, 2002)
reported that in Oakland CA, a street carrying 24,000 trips per day was
converted from four lanes to three. The number of annual crashes went from 81
before to 68 after. On another street in Oakland was narrowed, crashes went
down 52 percent. In Minnesota, a road diet resulted in a 33-percent reduction
in injury crashes. In Billings MT, a road diet resulted in 62 percent fewer
crashes after travel lanes were removed. In Lewistown PA, removal of travel
lanes saw the number of crashes drop to almost zero. Finally, these researchers
found that in Seattle WA, a number of road diets were analyzed, and a
34-percent reduction in total crashes and a 7-percent drop in injury crashes
was noted.
The Surface Transportation Policy Project (1999) released a
study in 1999 that found a strong link between aggressive driving deaths and
increased road capacity. Those living in states with the largest number of lane
miles per capita were 65 percent more likely to die in an aggressive driving
crash than in states with less lane miles per capita. Similarly, those metro
areas that added the most lane miles over a five-year period had higher levels
of aggressive driving deaths. See their 2003 report for additional
information about how big roads are less safe.
The Iowa Department of Transportation (2001) has found that
converting a four-lane undivided road to three lanes can improve safety while
retaining an acceptable level of service. Their review of research found that
when such conversions occurred, there was a reduction in average speeds, a
significant reduction in speeding, and a substantial reduction in the total
number of crashes.
According to Engwicht (1989), straighter, wider roads
encourage greater speed. Accidents that do happen are therefore more severe,
resulting in more injuries or a greater likelihood of death.
There is a large body of research which suggests that
increasing the safety of a car or road simply encourages the driver to take
greater risks. Drivers are willing to take a certain amount of risk in exchange
for the benefit of faster traveling time. This risk is added to the safety
limits of the car or road. The new safety features lull the driver into a new
sense of security. Vigilance, concentration and attentiveness wane.
Welch (Welch, undated) conducted an analysis of converting a
two-lane road to a four-lane road in Ft. Madison IA. This conversion resulted
in a 4 percent increase in traffic volume, a 4 percent increase in corridor
travel delay, a 2.5 mph increase in mid-block 85th percentile speed, a 14
percent increase in accidents and an 88 percent increase in injuries. The
report also found that traffic traveling more than 5 mph over speed limit
increased from 0.5% to 4.2%.
Welch reports that in Billings MT, when a four-lane was
converted to a three-lane road, the number of reported accidents decreased from
37 in the 20 months before to 14 in the 20 months after conversion. No increase
in traffic delay was found.
Despite initial apprehension from the local community and
its engineers, Welch indicates that a conversion from four lanes to three in
Storm Lake IA (US 71) resulted in an observed improvement in safety ("an
immediate large reduction in accidents"). The Iowa DOT Office of
Transportation Safety has begun actively promoting conversion of four-lane
roads to three-lane when a concern about safety is expressed. In Helena MT, an
urban primary highway (US 12) was converted from four lanes to three. (City
staff and other state staff engineers now support the conversion after
observing an improvement in traffic operations and a reduction in accidents.)
In a study conducted for the Minnesota DOT, it was found that the highest urban
corridor accident rates are found on four-lane undivided roads. In fact, the
collision rate was 35 percent higher than on urban three-lane roads. Howard
Preston, who conducted the study, stated that he would convert most four-lane
roads with less than 20,000 car trips per day to three-lane roads "in a
heartbeat."
In Duluth MN, a conversion from four lanes to three (21st
Ave East) was initially opposed by many. After conversion, the Duluth
News-Tribune editorial had this to say: "When Duluth officials announced
they would convert busy 21st Avenue East...from four lanes to two, with a turn
lane in the middle, some armchair analysts predicted it wouldn't work. The
News-Tribune Opinion page was among them. Well, it works. About everyone
agrees-from city traffic officials to neighbors-that the change has eased
congestion and reduced drivers' speed making it safer for pedestrians..."
Frequently, according to Welch, emergency vehicles find it
difficult to travel down four-lane roads. Emergency vehicles typically need to
wait for traffic to move over to the curb lane to get out of the way. But a
center two-way left-turn lane usually has less vehicle conflicts, and often
produces less delay for emergency vehicles traveling down it.
Hoyle (1995) points out that widened roads are alleged to be
safer roads based on data provided by those in favor of many road widenings.
However, data showing a decrease in crashes per vehicle mile don't take into
account the fact that widened roads encourage extra car trips that would not
have happened had the road not been widened. Widened roads also encourage
longer trip lengths. When such factors are taken into account, crash rates per
trip or per hour spent on the road remain nearly the same.
Michael Ronkin (2001) suggests that the most effective way
to reduce vehicle speed is by reducing the number of road lanes. "With two
lanes in each direction, regardless of width, a driver who wants to move faster
than the car in front can get into the adjacent lane and pass. With one lane in
each direction, the slowest car sets the pace for all cars behind it."
While driving in Boston recently, he found that "lanes are narrow, very
narrow, but on multi-lane one-way streets, cars zipped along at incredibly high
speeds for urban streets, around 40 MPH..."
Ronkin notes a great deal of misunderstanding among
pedestrian advocates about the speeds. "Pedestrians are more threatened by
the occasional car going much faster than reasonable, than by cars travelling
at an average speed." On multi-lane roads, "the crossing pedestrian
has several threats and challenges: the possibility of a car going faster than
the rest of traffic could be invisible as it is masked by another car, its
speed may not be apparent to the pedestrian. That makes it very difficult to
judge adequate gaps. With one lane in each direction, a gap is a gap."
One of the most frequent types of fatal crashes "is the
multiple threat-a driver stops to let pedestrian cross on a multi-lane road,
and the pedestrian is struck (and usually killed) by a driver passing in the
adjacent lane." Ronkin points out that this type of crash is not possible
if there is no adjacent lane.
For Ronkin, another important contributor to crashes,
besides speed, is the "complexity" involved in crossing a street.
After analyzing a great many fatal crashes, he concludes that many of those
crashes presented both the pedestrian and the driver with a relatively complex
situation. According to Ronkin, "there just wasn't enough time for both parties
to react to an unforeseen event." He concludes by pointing out the
importance, in designing a road crossing, of creating an environment that that
minimizes the number of decisions that must be made simultaneously..
In sum, Ronkin indicates that there have been
"demonstrated reductions in crashes" when a road had lanes
removed-convincingly so.
Narrow Lanes
Joseph R. Molinaro (1991) reports that wider travel lanes
are more dangerous because they encourage higher-speed driving. Larger
neighborhood collector streets work well with only 26 feet of width, and
smaller neighborhood streets are safe at 20-24 feet. He also points out that
residential streets should use tighter turns in order to force slower motorist
speeds. With a smaller turn radius, motorists are more likely to come to a full
stop than a more dangerous rolling stop.
The ITE Transportation Planning Council Committee (ITE,
undated) cites the American Association of State Highway Officials, which found
that "'[t]he number of accidents increases with an increase in the number
of decisions required by the driver.' A corollary to this truism is that the
actual and potential effects of each driver-decision become more significant as
the speed of the particular motor vehicle increases."
It is quite common for engineers to design a road for the
rare large truck. Such design requires large turning radii and wide travel
lanes. These relatively large dimensions far exceed those of passenger cars
most common on residential streets. The overscaled design of these roads
encourage faster passenger car speeds by the most frequent motor vehicles on
these roads.
"Clearly, reducing the width of a street,"
according to ITE, "has the effect of reducing vehicular speeds."
The Conservation Law Foundation (1995) finds that vehicle
speeds increase when roads are widened because there is an extra "safety
cushion" provided by the increased lateral distances and increased sight
distances. Psychologically, the wider road tells the motorist that it is safer
to speed up, and since motorists tend to drive at the fastest speed they feel
safe at, faster speeds are seen on wider roads with a higher perceived
"safety cushion." In addition, the field of vision of the motorist
shrinks as speed increases, which reduces the ability of the motorist to see
things (such as cars or pedestrians) that are ahead.
The Foundation also points out that designing for faster
driving speeds, while possibly reducing the frequency of crashes, also
increase the severity of car crashes.
Swift, Painter, and Goldstein (1998) conducted a study that
analyzed the safest street widths with regard to accident frequency. Their
study found that "as street width widens, accidents per mile per year
increases exponentially, and that the safest residential street width is 24
feet (curb face)."
Indeed, crash rates were 18 times higher on 48-foot wide
streets than on 24-foot wide streets.
The authors concluded, in part, by calling for a
re-evaluation of public safety. That local governments recognize that the
chance of injury or death due to, say, a neighborhood fire, is quite small
compared to the much higher probability of injury or death in a neighborhood
due to speeding traffic. That the reduced number of injuries or deaths resulting
from wide streets and allegedly faster fire truck response time is tiny in
comparison to the comparatively large number of injuries or deaths that occur
due to speeding cars-a problem that increases in frequency due to widened
streets. The local government should "ask if it is better to reduce dozens
of potential vehicle accidents, injuries and deaths [through the creation of
more modest streets], or provide wide streets for no apparent benefit to
fire-related injuries or deaths."
Even if more modest streets increased fire injury risks
slightly (a problem not found by the study), modest streets would still
be safer than wide streets because the risk of car injuries is so much higher
than fire injuries.
In other words, by focusing public safety on life safety,
rather than fire safety, a much larger number of community injuries and
deaths can be managed and perhaps reduced.
A large number of firefighters are starting to understand
that over-sized streets have resulted in streets that are not safe for
families, while providing few, if any, benefits regarding fire safety and
emergency response times, according to Siegman (2002).
Siegman relates a story from Dan Burden, a colleague who
works in the field of safe street design:
While in Honolulu last week doing two school traffic calming
charrettes our team had two tragic nights. In both cases a squad of firemen
were with us for the evening, learning about and giving good input into traffic
calming their neighborhoods. They had their truck with them in case they received
a call. When asked by a member of the audience what they thought of the traffic
calming plan the Captain said that they rarely, if ever, can expect a fire in
the area....and that their concern is to lessen the speeds on area roads so
that they are protecting rather than rescuing lives. They had good reason to
say this ... during the evening the firemen were called out to respond to a
pedestrian tragedy several blocks from our meeting room, and in our project
site.
The next school traffic calming meeting we again had four
firemen, and their apparatus. We had just settled them down to a design table
to design traffic calming solutions when they leaped up to attend a call. They,
too, came back before the meeting was over. They had provided first assistance
for a head-on crash of two motorists.
The meeting ended at 9:00. At 9:05 a bicyclist was hit (and
presumably attended by these firemen). The cyclist was a star athlete on the
University of Honolulu campus. She was killed one block from our school, in one
of our crosswalks.
"Many firefighters," according to Siegman,
"realize that traffic crashes are a far greater hazard in our communities
than fires, because they so often have to pick up the pieces."
Siegman reminds us that "for every one person killed in
a fire, more than eleven die in traffic crashes. And that for every one person
injured by fire, 148 are injured in traffic crashes."
A great many firefighters also tell us that fire truck
response time does not depend simply on the width of a street.
For example, Siegman tells us that fire departments know
that response time is a product of the speed of travel and the distance from
the firehouse.
When streets are walkable and connected as they were in
traditionally designed neighborhoods, they "usually allow far more direct
routing than disconnected cul-de-sac designs." Even when narrow (or
"skinny"), the connected streets, Siegman points out, "can often
deliver equal or better response times." Connected streets also reduce the
probability of traffic congestion, and congestion slows response times.
"That understanding," notes Siegman, "is apparently not yet
reflected in fire codes, which discuss street width, but...have no
specifications whatsoever on directness of routing, or distance from home to
the arterial, or to the fire station."
Siegman points out that a number of other fire departments
are "no longer ordering U.S.-made fire engines, choosing instead the more
maneuverable European models, which work well with smaller, safer,
pedestrian-friendly street designs."
According to Siegman, "we aren't yet at the stage where
all firefighters have excellent training in street design and traffic
safety." He wonders "how many communities still design their streets
and intersections to accommodate the largest fire truck in the fleet, without
having weighed pedestrian safety effects as part of the truck purchase."
In conclusion, Siegman presents us with the following
eye-opening statistics for fire and traffic fatalities and injuries in 1999 in
the United States. In that year, "3,570 civilian (i.e.
non-firefighter)" fire deaths occurred, and 21,875 civilians were injured.
In addition, 112 fire fighters died while on duty-11 of them in traffic
crashes. He also reports that "41,611 people were killed and 3,236,000 people
were injured in the estimated 6,279,000 police-reported motor vehicle traffic
crashes. 4,188,000 crashes involved property damage only."
As reported by Finch (1994) and Preston (1995), every one
mph reduction in traffic speed, in general, reduces vehicle collisions by five
percent, and reduces fatalities to an even greater extent.
Narrowing travel lanes made things safer unless the
narrowing was done to accommodate more travel lanes, according to a report from
the Transportation Research Board (1994).
References Cited
Conservation Law Foundation. Take Back Your Streets.
Boston MA. May 1995.
Engwicht, D. (ed.) Traffic Calming: The Solution to Urban
Traffic and a New Vision for Neighborhood Livability. 1989.
Finch, D.J., Kompfner, P., Lockwood, C.R., Maycock, G. Speed,
Speed Limits and Accidents. Transport Research Laboratory (Crowthorne, UK),
Report 58, 1994.
Hoyle, C. Traffic Calming. American Planning
Association. Planning Advisory Service Report Number 456. 1995.
Huang, H.F., Stewart, J.R. and Zegeer, C.V. Evaluation of
Lane Reduction "Road Diet" Measures on Crashes and Injuries.
Transportation Research Record 1784: 80-90. 2002.
Iowa Department of Transportation. Guidelines for the
Conversion of Urban 4-lane Undivided Roadways to 3-lane Two-Way Left-turn Lane
Facilities. April 2001.
ITE Transportation Planning Council Committee, Traditional
Neighborhood Development: Street Design Guidelines. 5P-8. Undated.
Molinaro, J.R. Rethinking Residential Streets.
Planning Commissioners Journal. Vol. 1:1. November/December 1991.
Preston, B. "Cost Effective Ways to Make Walking Safer
for Children and Adolescents," Injury Prevention, 1995, pp.
187-190.
Ronkin, M. Pedestrian & Bicycle Program Manager, Oregon
Department of Transportation. March 27, 2001.
Siegman, P. Siegman & Associates, Town &
Transportation Planning, 260 Palo Alto Avenue, Palo Alto, CA 94301. August 4,
2002 email submitted to a Dan Burden/Walkable Communities internet discussion
group.
Surface Transportation Policy Project. Aggressive Driving.
Washington DC. April 1999.
Swift, P., Painter, D. and Goldstein M. Residential
Street Typology and Injury Accident Frequency. Copyright Peter Swift, Swift
and Associates. 1998.
Transportation Research Board. Low-volume rural roads
(Roadway Widths for Low-Traffic-Volume Roads). Transportation Research
Board. NR362, 1994.
Welch, T.M. The Conversion of Four-Lane Undivided Urban
Roadways to Three-Lane Facilities. Transportation Research Board. TRB
Circular E-C019: Urban Street Symposium. Undated.
Back to Dom's Urban Design home
page.