A GOOD WALK, UNSPOILED: A Few Ways to Improve Foot Traffic

The cliché is true: every trip begins and ends on foot.  The basics of what makes for a good walking experience are pretty straightforward:  smooth, wide, unobstructed and well-lit surfaces; stress-free street crossings; pleasing aesthetics; busy but not overcrowded; hassle-free and crime-free areas; opportunities for social or commercial activity; and a way to get to where you want to go.  In fact, policies about pedestrian facilities are generally decent – it’s the implementation that gets controversial.

Although the increasing numbers of people who step into a street with their eyes lost in a screen and their ears blocked with music are asking for trouble, the starting point for road safety must be that as the most vulnerable of road users, pedestrians (on foot or wheelchair) deserve the most protection and should be yielded to by everyone.  However, because every situation has its own particularities, most policies and standards appropriately provide enormous lee-way for on-the-spot judgement by the individual designer or traffic signal controller.   And the devil is always in the details of what gets decided, particularly in the (deliberate or implicit) weighting given to car traffic versus pedestrian (and bicycle) movement.  Here are some ideas about how things should go.


The greatest challenge pedestrians face today is not having enough time to cross during a Walk signal.  Currently, pedestrians are generally assumed to move at about 3.5 to 4 mph which, along with the width of the road, is supposed to be the basis for determining the minimum time that the walk signal stays lit.   Newer guidelines, taking into account the needs of disabled people and the increasing age of our population, suggest using 3 mph.  The occasional intersection regularly stuffed with large crowds of walkers need even more Walk time in order to give people at the back of the pile time to get out of danger.   But, in order to reserve as much time as possible for car movement, many intersections don’t follow even the old standard.   For many traffic engineers, pedestrian signals are seen as “stealing time from cars,” whereas in reality we should be designing our traffic signals around pedestrians first. Instead of asking what the minimum time to cross the street should be, engineers should start their signal timing exercise by asking how to reduce the wait time for pedestrians and maximize the amount of time pedestrians have to cross.

Even in dense urban settings, when an intersection is big or busy enough to need signals, people should not be forced to wait for more than 45 seconds before being allowed to step off the curb.  In many situations, there are more people walking than driving through an intersection. This is especially the case when vehicles are nearly empty, carrying a single passenger.  But the traditional imperative to keep cars moving often forces even longer pedestrian delays, prompting people to take matters into their own hands and step into the road when vehicles still have a green light; a situation which is bad for everyone.

Most corners and crossings are unsignalized, many don’t even have painted crosswalks.  We count on each other’s basic intelligence and curtesy.  At high traffic volume intersections, the major method used to control crossing vehicles and facilitate pedestrian crossing are traffic signals – traffic lights and walk signs.   However, not every corner should be signalized.   In general, two-lane roads (one lane each direction) with low or moderate amounts of traffic are usually safer and more efficient with 4-way stop signs rather than traffic lights:  cars don't speed up to beat the red, no one has to wait very long, everyone must pay attention – and it’s a lot cheaper to maintain.


Push-buttons are inconvenient and often confusing, since they sometimes don’t work, or take too long to have an effect, or are only actually required to be pressed at certain times of day – all of which are very confusing. In general, the Walk signal should occur automatically as part of every cycle rather than require a button-press. (Push buttons can be useful in mid-block locations or areas of low pedestrian activity, particularly when car traffic levels are high.)

Many intersections are programmed with an exclusive walk phase or “all-way pedestrian scramble,” where all car traffic stops (with no Right On Red allowed*) and people can walk in any direction. While in some situations this really is the safest approach, allowing too long for the traffic flow phases in both directions (and for turns) often results in acceptably long waits for pedestrians and minimal time to cross the street, leading many people to jaywalk instead.  In most situations, Walk signals should be “concurrent” with parallel traffic movement.  This allows walkers to “go with the green” and reduces their wait time. (Note that in high pedestrian activity locations, it can be useful to include an all-way pedestrian scramble in addition to concurrent walk phases, particularly if crossing diagonally is appealing in that location.) With concurrent phasing the Walk signal should stay on as long as the parallel traffic light allows cars to move.  Walk signal count-down displays, which should be included by default in upgrades or new fixtures, should not show zero until the next parallel red light appears.  (Anyone who walks in Boston regularly will notice that concurrent signals there often do NOT follow these guidelines, sometimes not being automatic, and often times reaching zero long before the parallel green turns to red. This is something LivableStreets, WalkBoston, and others are trying to address through Vision Zero.)  

*Allowing Right On Red is a safety hazard.  At the least, high numbers of crossing pedestrians, or the presence of people who move more slowly (elderly, the disable) or erratically (e.g. children) should be an automatic trigger for installation of No Right On Red signs. 


After the crossing road’s traffic light turns red, but before the parallel traffic goes green, pedestrians should be given a 5 to 7 second head start – called a Leading Pedestrian Interval (LPI).  Current practice, when there is an LPI at all, is to provide a 3 second head start; this often isn’t enough time for people who move slowly.  Letting pedestrians go first gives them time to move far enough into the intersection to be visible to, and give them priority over, turning cars.  It also provides part of the protection of an exclusive Walk phase but without the long delays.   (Any road with protected or separated bicycle lanes should also have distinct bicycle signals.  However, at all intersections without separate bicycle signals, cyclists should be allowed to also start when the LPI Walk signal shows – although if they’re turning they must still yield to pedestrians in the crosswalk.)

Pushing the car Stop Line further away from the corner, placing a “bike box” between the Stop Line and the crosswalk, or making the crosswalk particularly wide all complement the LPI and give pedestrians additional time to start across.  (Although not standard practice, and without negating the responsibility of cyclists to yield to crossing pedestrians, I think that where the intersection geometry allows and if the number of pedestrians is not too large, it is better to place the bike box closer to the intersection than both the Stop Line and the crosswalk to increase cyclists’ chance of getting across the intersection before cars come through.)


There was a time when anyone could use the street for nearly any purpose.  There was a lot of contention, and horse manure. The rise of automobiles – heavy, fast, deadly – allowed its proponents to force everyone else to get out of their way and led to the creation of “jaywalking” as a civic crime.          Now we have designated crossing lanes.  The old way of identifying crosswalks -- two white edge-lines – has been increasingly replaced by “zebra stripes”, the wider and bolder the better.   Crosswalks should be 5 feet wide or more, up to the width of the sidewalk, increasing as the number of walkers increases.  And the curb cut/ramp should be as wide as the crosswalk.

It is tempting for road designers to place crosswalks and curb cuts at the “apex” or center of a corner.  This allows them to make one cut only, reducing costs.  But it places the crosswalks inside the intersection, closer to cars and therefore a much more dangerous location.  Apex curb cuts are particularly bad for wheelchair users who have to roll out into the intersection.  With a few exceptions, each corner should have two crosswalks with a separate curb cut for each one, perpendicular to the line of travel.  Crossings should match the desire line of someone walking along the sidewalk.   Mindlessly pulling the curb cut a standard number of feet away from the corner may move it too far away from the straight-across desire line, forcing wheelchair users and baby-carriage pushers to make a difficult and aggravating zig-zag.  It’s another reason why all road plans benefit from public discussion.

Crosswalks should not just be at corners.  When blocks are long and traffic is heavy, mid-block crossings should also be installed along with curb cuts and user-activated (push button) stop lights such as a HAWK Beacon.   Similarly, when transit stops (train, trolley, or bus) are in mid-block locations, crosswalks should allow a direct crossing.  No one should have to walk very far to cross the street for a transit stop.   (See the NACTO Transit Guide for details on preferable bus stop locations.)

Crosswalks sometimes need help.   Whenever possible, Walk signals should last long enough to allow everyone to get across, including people in wheelchairs or with an elderly stride.  But when roads are wide and traffic is heavy, not everyone may always make it all the way across in one phase.  In those cases, a mid-crossing refuge/safety area is needed.   These protective block-end islands are short stretches of raised (curb-protected) pavement.  They are not medians – a continuous centerpiece that often has the perverse effect of increasing car speeds and therefore pedestrian danger by reducing driver’s fears of head-on collision.  Despite their potential aesthetic value, medians also often take up space that could otherwise have been used for wider sidewalks or protected bike lanes.   And, ironically, medians created for primarily aesthetic purposes often lack end area refuges that pedestrians can use.

Pedestrian refuges are also useful when providing unsignalized midblock crossings. In fact, most guidelines say that pedestrians should not be expected to cross more than two lanes at a time without a signal unless a refuge is provided.


Americans with Disability Act (ADA) guidelines, as well as state and municipal regulations, require sidewalks to be wide enough, level enough, and in good enough conditions to allow wheelchair access.  But, in reality, the minimal requirements are seldom enough, especially in downtown or other heavily used areas.  Avoiding broken, uneven, and missing sidewalk pavement is a difficult job, requiring regular inspection, preventive maintenance, and repairs – all of which requires staff and money.  Just as complicated and difficult is keeping sidewalks clear of snow and ice in the winter – WalkBoston’s “snow clearing” recommendations are a good starting point.

Sidewalks are more pedestrian friendly when they have interesting or beautiful surroundings – green landscaping, desirable retail and community destinations – and are structured to facilitate social interactions with places to sit or eat and quiet spots to rest.  Separation from traffic by trees, street furniture, parking, or a bike lane increases the feeling of safety, as does adequate lighting and the presence of other peoples’ “eyes on the street”. 

On the other hand, rather than increasing safety through separation it is sometimes better to do it by increasing togetherness.  Neighborways, also called slow streets or safety zones, have surfaces and other road objects that physically inhibit cars from going above 15 (or even 5) mph.  At the same time, these spaces welcome children and other non-motorized users to walk, play, and socialize.  These limited access streets sometimes evolve out of residents’ efforts to discourage cut-through car traffic. 

Walking can be a meditative exercise in itself.  But from a transportation perspective, the utility of a pedestrian route is that it actually takes you from one place to another.  Advocates and planners proposing to upgrade or create a new pedestrian (or bicycling) route need to justify the expense by showing that significantly more people will use the improved facility.   Prediction is a complicated process, but a new software program from the Metropolitan Area Planning Commission (MAPC) called Local Access provides a useful starting point.


A road may be technically in compliance with all official standards but still remain neither inviting nor safe for children, the infirm, and the elderly.  When non-motorized activity is treated as an afterthought, or as a lower priority, or even as an equal priority to motorized traffic, the odds of injury increase – not just for pedestrians and cyclists but also for car occupants.  Unless the safety of the most vulnerable are given top priority, a road and its intersections turn out to be “unsafe by design”. For everyone.  Slowing traffic, improving sight lines, designing “tighter” intersections requiring sharper turns, eliminating “free right” turns, making everyone pay more attention to what’s going on around them – all reduce car accidents and injuries as well.  

For all our frustration with the way intersections currently make walking difficult, it’s important to give municipal Traffic Departments at least a little slack.  Most cities are stuck with old signal systems that have to be individually and manually adjusted.  It takes a very long time to revise out-of-compliance signals, a problem made much worse by the tendency of old signals to malfunction.  But there is no excuse not to set good policies, invite the public to let officials know which signals are causing the worst problems, set a schedule for fixes and upgrades, and to allow people to double-check that the desired changes are working as intended.   We also need a state-level incentive funding program to help convince municipalities to invest in the large-scale modernization of their traffic systems. This will allow for less-difficult adjustments – as well as reduce future operating costs.


Thanks to Sophie Schmitt, Charlie Denison, Mark Tedrow, and Jacob Meunier who drafted LivableStreets’ response to a draft “Pedestrian Level of Service” evaluation rubric created by the Central Transportation Planning Staff – their letter was the starting point for this blog and they provided enormously helpful feedback on early drafts.  Thanks also to Wendy Landman for her feedback on an earlier draft.  


Related previous postings:

> PEOPLE PRIORITY STREETS:  Neighborways, Slow Streets, and Safety Zones

> QUESTIONING COMPLETE STREETS:  Having the Courage of our Vision and Values

> STABILIZING EQUITABLE COMMUNITIES:  Gentrification, Displacement, and Markets



Take our website survey!