Traditional Neighborhood Development: Will the Traffic Work?

Presentation by Walter Kulash at the 11th Annual Pedestrian Conference in Bellevue WA, October 1990.

 

Definition of Traditional Neighborhood Development (TND)

As participants in the Eleventh Annual Pedestrian Conference, this audience probably has a better feel for what TND means than any other group that could be assembled. Traditional neighborhood development, variously called “neotraditional” development or “urban villages” refers to a style of urban or suburban development, evolving since the 1970’s, that revisits many of the features of urban neighborhoods of 50 to 100 years ago.

If we had to give a single most distinguishing feature of TND, I would suggest that it is its continuous fabric of intimately blended land uses, arranged so that travel between them can be made by a variety of methods (walk, bicycle, transit, taxi) in addition to the usual privately-operated auto.

The land use in TND is mixed in an intimate blend, not, as in typical suburban development, in globules of single use parcels arranged in isolation from other uses.

The street design in TND is arranged to support this intimate blend of land uses. TND streets are small, and connected into dense networks. On these streets, there is an emphasis on non-motorized travel, and on the overall quality of travel for the automobile traveler. There is, at the same time, a de-emphasis of the narrowly defined performance standards (mainly travel capacity and speed) that are dictating what our streets and suburbs look like today.

The traffic engineering features of TND are attracting considerable attention and debate, because they seem to fly in the face of long-held principles of traffic engineering and subdivision planning:

Network of streets — the TND concept calls for a dense network of highly connected streets. In traffic terms, dense network means multiple available routes for a given trip. If the primary route for a trip is unacceptable because of traffic conditions, alternates are available. The dense network is in stark contrast to the sparse branched pattern of most suburban growth.

Street cross-section — the TND concept calls for street cross-sections that are typically no greater than two travel lanes plus on-street parking, which translates into a maximum pavement width of 40 feet. TND calls for a street right-of-way sufficient to contain this street cross-section, but not intended to

accommodate a wider pavement at later stages. Typically, a right-of-way width of 70 feet can accommodate the TND street.

Reduced or non-existent hierarchy of streets — the TND concept either eliminates or greatly reduces the ‘hierarchy’ of conventional functional classifications that are assigned to streets. In the conventional system, the base of the hierarchy is local streets, intended for immediate property access.

The next level is the collector, intended to gather traffic from local streets and feed it to the arterial system. The final level is the arterial street, intended for longer distance mobility and not intended to serve as immediate access to properties (although this function is almost inevitable).

Lateral clearance — TND guidelines permit and even encourages the reduction in lateral clearance between street and the fixed objects (trees, street furniture) on the side of the street.

On-street parking — on-street parallel parking is basic to TND. This parking is designed to buffer pedestrians on the sidewalk from moving vehicles in the traveled lanes to provide street activity (drivers entering and leaving their vehicles), for the supply of parking itself, although this source of supply serves only a small part of the overall parking need in a business district, and to “enclose” the sidewalk space.

Short traffic signal cycles — traffic signal lengths of no greater than 60 seconds are compatible with TND. Short traffic signals are pedestrian-friendly. They also create more frequent gaps in traffic for midblock pedestrian crossings.

Two-phase signals — these are signals that simply turn green for the entire approach, with no turn arrows. These are possible where there is a dense street network, because there is a much greater choice of locations for left-turn movements. The Conventional Suburban Development concept concentrates left-turn movements at a few major intersections, creating the need for multiphase signals. Two-phase signals convey a sense of small scale, to both drivers and pedestrians, that contrasts strongly to heavy-duty multiphase sequences. Two-phase operation permits a greatly reduced cycle time.

Curb radii — TND calls for a greatly reduced curb radii, typically 10 feet or less, at intersections to reduce the speed of turning automobiles and to greatly reduce the in-street walking distance required for pedestrians crossing the street.

Alleys – the TND concept frequently includes alleys serving the rear of all properties. These alleys eliminate the need for curb cuts for driveways in the streets, and permit continuity of buildings along a block front, Curb continuity further increases the amount of on-street parking that can be obtained in the design. Alleys are also intended to provide a utility corridor, thereby removing utilities, particularly power lines, from the streetscape.

Some other features, while not directly related to traffic, are highly characteristics of TNDs:

Architectural Themes — Most TND’s have vigorously pursued an architectural theme, generally following historical styles, further recreating the fueling of communities of 50- to 1 00-years old.

Many of these TND street features are old ideas revisited. TND proponents hold that these are ideas whose time has come around again, in a new and innovative way, analogous to the return of the four-cylinder engine or all-natural fabrics. Critics hold that these are just old worn-out ideas. The debate is getting interesting.

 

Will the Traffic Work?

Much of the criticism and suspicion that has been directed at the TND concept stems from a belief that the vehicular traffic will not work. it is frequently pointed out that many of the TND elements – narrow streets, small blocks, closely abutting buildings — are urban design features that were found to be incompatible with good vehicular traffic flow many years ago. The last 40 or 50 years of traffic engineering and land development regulation have indeed been directed at securing diametrically opposed features in our new development.

The debate about the traffic pros and cons of TND has been producing some interesting smoke and heat, but so far little light. Andres Duany, who with his partner, Elizabeth Plater-Zyberk are easily the most widely known proponents of TND, has, by way of traffic analysis, asserted that traffic engineers are in the lower academic half of their graduating engineering classes. At first, I thought that this was a compliment, implying that half of us in fact graduated. Presumably this contribution to the understanding of the traffic issues of TND was provoked by some comparably relevant criticism of TND on the part of traffic engineers.- The field of technical analysis of TND traffic is ripe for detailed traffic engineering analysis, and the results of precisely this type of analysis are what I will be addressing this afternoon.

In order to answer this question of whether or not the traffic works, let’s examine the three criteria that currently drive our traffic planning: vehicular capacity, travel speed (and therefore travel times) and safety.

How about other considerations, such as Pedestrians, or how the streets look? Remember, we said we are looking at the criteria used in traffic planning. Therefore, non-motorized travel and aesthetics can not, by any stretch of the imagination, be considered as serious driving forces in street design. To settle the question of whether the traffic works, we will focus on only those three ‘hard-nosed’ engineering measures.

 

TND has superior traffic capacity

A network of small interconnected streets has more traffic capacity than the same street area arranged in a sparse hierarchy of large streets.

This superior capacity is unrelated to the REDUCTION in travel demand or SHORTENING of travel distances that also results form the TND pattern. These decreases in total vehicular travel are also important advantage to TND, but we need to carefully isolate them in our analysis of TND traffic. The feature that we are focusing on is that for a given amount of traffic demand (i.e., same number of vehicular trips) the TND will simply out-perform the conventional street design.

 

Large Streets Have Deficiency of Scale

The fundamental reason why a dense network of small streets out-performs a sparse hierarchy of streets is that streets become less (not more) efficient as their size increases. So instead of an EFFICIENCY of scale as the street gets larger, we experience instead a DEFICIENCY of scale.

The reason is in the intersections. Intersections control the capacity of any network, dense or sparse. Think about it — if it weren’t for the intersections, every street would have essentially the capacity of a freeway lane, ideally 2,000 vehicles per hour (vph) and eroded, on surface streets, by various friction elements to 1,500 to 1,600 vph.

But unfortunately, surface streets have to share the intersections with other surface streets. So their capacity is immediately cut in half. Then the streets have to further share with left turns. If the streets are big enough, left turns are in all directions, left turns need their own piece of the intersection signal time, still only 60 minutes per hour, regardless of how big the intersection is.

Let’s take a somewhat abstract example of two different street patterns that illustrate this feature of traffic flow. In the first example, typical of the Conventional Suburban Development, we have a single large intersection of a four-lane divided and six-lane divided arterial street: each having left-turn lanes and protected left-turn signals. In other words, a well-engineered major intersection.

Let’s assume that this intersection is operating at close to peak-hour conditions, and that traffic service is beginning to be noticeably affected by the congestion. This corresponds to a level-of-service (LOS) ‘D’ at the intersection. This corresponds to traffic volumes of 3,000 vph on the six-lane street and 2,000 vph on the four-lane street, and turning movements of 300 and 200 vph, respectively, for the major left turn movements. The minor left turn movements are overshadowed, with respect to the signal time that they need by the major movements and are not significant in this analysis. The intersection described above is operating the limit of LOS ‘E.’ No further traffic can be added, in any of the major directions, without the LOS deteriorating to ‘F,’ or unacceptable.

Now let’s put the same amount of traffic on the same amount of pavement, but on a differently configured road system — a pair of two-lane streets intersecting three parallel two-lane streets.

In a corridor sense, that is, in the east-west and in the north-south direction, the total number of lanes remains the same as in the Conventional Suburban Development example. Also notice that the total amount of pavement stays the same. The radical difference between the two plans is in the number of intersections in each system — the TND has six times as many as the Conventional Suburban Development.

This large number of intersections reduces the turning movement load at any given intersection to a fraction (one-sixth in this example) of the turning movement load that exists in the Conventional Suburban Development pattern. Consequently, the entire system can carry greater traffic volumes at the same level of traffic service.

 

Turning Movements are More Efficient on Small Streets

The highly connected grid of streets that is built into the TND provides numerous, redundant opportunities to make a left turn. This contrasts to the pattern under Conventional Suburban Development, in which left turns are gathered up from multiple locations and focused at a single location.

Many, perhaps most of these left turns in a TND network can be accomplished in gaps, and do not represent the loss of any green time to opposing through traffic. Obviously, left turns can proceed simultaneously at different intersections, in effect multiplying the turn movement capacity of the entire network. On the other hand, when left turns are all gathered and focused at intersections, EVERY left turn represents a loss in green time to the opposing through traffic.

Not only does TND offer many more places to make turns, but it also decreases the difficulty of a given turning movement. It is far easier to make a left turn across a given volume of traffic in a single lane than across twice that same volume of traffic in two lanes, and so forth. So with multilane streets, you lose almost all of the capacity to make network left turns (street-to-street) and are left with only the ability to make turns into driveways.

This finding is another one of those unexpected conclusions, and is worth illustrating with an example.

When we make a left turn at an unsignalized location, we wait for an acceptable “gap” (around six seconds) to appear in the opposing traffic stream.

Even with a fairly normal heavy traffic flow, reflective of a peak-hour, with 900 vph, you will get a large number of acceptable gaps. The statistics say you will wait an average of 14 seconds until you receive an acceptable gap.

Now what happens as we simply double the traffic situation: twice the traffic and also twice the size of the road. Do we still have the same gaps? Do we still wait an average of 14 seconds?

We would if the cars came paired perfectly.

But, of course, they don’t come paired. And in fact, three things start to go wrong with the left turn movement: (1) the cars rearrange themselves, so that two-thirds of them are in the outside lane. We don’t need to know why, we just know they do it. So, some of our gaps are gone. Then, the additional traffic, of course, does not pair perfectly with the existing traffic. The second lane is, in fact, also a separate distribution of gaps, totally independent from the first lane. Only when we get occasionally lucky do the gaps coincide. Now the speeds are different, and cars start to change lanes and present the left turner with a more uncertain picture of what’s happening. Finally, the distance that we, as leftturners, have to clear has doubled, from one lane to two, and so we need a gap of not six seconds, but twice that or 12 seconds.

So all of these things working together mean not just a little more difficulty in making the turns, but a drastic drop in the ability to make them, to the point of impossibility. What is our solution?

In the Conventional Suburban Development, there is only one solution: move every last one of those turning movements to a signalized location, where, as we now know, they rob capacity

from all the other traffic movements.

 

Real-Time Route Decisions Occur on TND Networks

“Real-Time Route Decisions” is jargon for simply playing it by ear. In a TND environment, the driver can choose from the many routes available on the basis of what they see out on the street. In a TND, drivers make turns in advance or after their primary choice of turn location.

With the well-connected network in a TND, the driver can take alternative routes in the full confidence that the network is complete, and that they can find a reasonable route to their ultimate destination. Those of you that have lived in Phoenix with its highly connected network of streets know how this works. The importance of real-time decisions is that left turn movements stay out of the major intersections, and that minor streets pick up an important part of the turn movement load.

Real-time route selection is most likely to happen in peak-hours, when congestion is present.

This further explains the tendency of traffic volumes on local TND streets, to be more ‘peaked’ than traffic in general. Local streets are more likely to be pressed into service in peak-hours by drivers making work trips than in off-peak periods by drivers making mid-day trips.

 

Uninterrupted Traffic Flow is More Likely on TND

Uninterrupted flow conditions — meaning absence of traffic signals — are more likely to be obtained on TND systems than on sparse network. Freed from the sharing of right-of-way with other vehicles, the capacity of the street becomes essentially its free-flow capacity, ideally 2,000 vph and in reality, after considering side friction and other imperfections, around 1,500 to 1,600 vph.

Interrupted flow — as soon as a signal is installed — cuts the flow of the street to 800 to 900 vph, depending on the amount of time that is shared with the other movements at the intersection. So clearly, if we can stay beneath the threshold for signalization, we can preserve uninterrupted flow on streets.

At cross-streets, the need for signalization is reduced by greatly reducing the load on any given intersection, owing to a large number of intersections sharing the traffic load.

The fine-grained land use along the TND street also works toward the containment of the need for traffic signals. Small commercial properties, each with their own access, do not individually warrant signals, particularly in light of the relatively low main street volumes and the two-lane cross-section of the TND streets on which they front. Also, the probable commercial site on TND streets is small scale, and may in fact turn out to be junior versions of our currently familiar suburban community and regional shopping destinations.

 

What the Book Says About Capacity

The definitive method for measuring traffic performance, the 1985 Highway Capacity Manual developed by the Transportation Research Board and evolving from earlier methodologies over a 30-year period, is the definitive statement of traffic performance in the US.

The procedures within the 1985 HCM clearly confirm the basic premise of TND traffic flow; namely that there is no advantage of scale in large unconnected streets, and that a well-connected network of small streets will out-perform the sparse unconnected network of larger streets in terms of capacity.

For analyzing urban streets, the heart of the 1985 HCM methodology is the intersection analysis, since intersections, as we have seen, control the capacity of urban streets. in its planning method, (a short-cut approximation of the full HCM procedures) the 1985 HCM clearly states the diminishing returns on adding more traffic lanes.

Under the full detailed methodology, computationally complex, the deficiency of scale is not stated in simple rules, but is clearly visible when comparing capacity analysis results for identical traffic loadings applied to single, large intersections as opposed to a grid of smaller ones. The only way to really get at this, because of the inaccessibility of the method, is to run identical traffic volumes through different types of intersections and compare them. A series of several such comparisons shows the thrust; the grid of small intersections consistently out-performs a sparse hierarchy, for the same amount of traffic and under the strict application of 1985 HCM methodology.

In the view of many traffic engineers, the 1985 HCM method for unsignalized intersections is of marginal interest because unsignalized intersections are virtually designed out of conventional suburban design. However, the unsignalized intersection looms much larger in the analysis of TND, because of its ability to operate a large percentage of its intersections as unsignalized intersections. The unsignalized intersection methodology, more penetrable by manual analysis, very clearly shows the deficiency of scale of larger intersections.

These procedures, to put it simply, say that it is significantly more difficult for intersecting or turning traffic to cross a given volume of traffic in two lanes than to cross the same volume in a single lane of traffic. Consequently, the unsignalized intersection becomes less efficient as it gets larger.

 

What the Models Say About TND Street Capacity

We get the same conclusion about TNDs performance through the standard transportation modeling process. In one analysis, we took a development program that was proposed by one of our clients for a 700-acre site — a square mile. The development program was a typical planned unit development with single-family and multifamily homes, retail, and office park, schools and churches. We laid out this project in two different ways.

We then tested the traffic capacity of both the examples, using the standard transportation planning methodology. Some of you will know that this involves generating the trips, distributing them to their probable destinations and then assigning them to the street network that is in place. The computerized algorithm that does this — the model — simulates the decisions made by drivers in actual practice. Drivers take the best available route. When a route gets congested to the point that its speed degrades, drivers begin to divert to other routes, if available.

What we found in our prototype was that the TND was perfectly capable of carrying the traffic. The level of traffic service on the arterial streets actually improved in our prototype, because of the diversion to local streets. Collector street traffic virtually disappeared. Local street service, despite the shift of traffic to them, was virtually unchanged. The explanation lies in the ability of the large mileage of connected local streets in TND to absorb large amounts of traffic.

 

TND Has Lower Travel Speed But Comparable Travel Time to Conventional Suburban Development

TND has a shorter trip length than Conventional Suburban Development. This is due simply to the geometry of a dense network of streets, which minimizes the travel distance for any given pair of origins and destinations.

The shorter trip distances in TND are due to its layout. The TND pattern does not have any enclaves of development, and so therefore there is no element of enclave time — the time needed to drive into the enclave and drive out of it. In TND, a more direct routing is possible.

Another reason why TND trips are shorter is the absence of street hierarchy, with its need for routing all traffic onto a sparse network. Because the Conventional Suburban Development network is by definition sparse, it is therefore not direct, and forcing all traffic onto it creates the need for additional travel distances.

So far, we have been assuming an identical density of development, with both the TND and the Conventional Suburban Development developed to the same density. The argument for shorter trip distances under TND becomes even stronger when we project the increase in intensity that is intended as a part of the TND experience.

Under the TND, the interwoven mixture of trip destinations (shopping, personal business, employment, school) mixed in with the residential trip origins shortens the average trip length. Even without increasing the overall gross density, the TND land use pattern reduces the total trip distance for any given origin/destination pair.

This reduction in travel time illustrated by the analysis of the prototype TND and Conventional Suburban Development projects that we discussed earlier.

The travel speed profile for a TND generally shows a lower peak speed and shorter, more frequent intersection delays than on a Conventional Suburban Development.

The Conventional Suburban Development trip, made mainly on major arterial streets, is typified by a pattern of high speeds for short segments of road, interspersed with long traffic signal delays at individual traffic signals. The TND trip, on the other hand, with its greater use of minor arterial, collector and local streets, is characterized by low maximum speed, more frequent short delays at intersections and a greater number of turning movements.

For shorter trips, the TND pattern tends to produce lower trips times. This is largely due to the elimination of the large “threshold” of delay in getting out of the origin enclave and into the destination enclave for a typical Conventional Suburban Development trip. This threshold, usually involving entering and leaving a major arterial street at a long (120 to 180 seconds) traffic signal cycle, is eliminated in the TND pattern with its small accessible fronting streets and its interwoven land use pattern.

The travel time pattern illustrates a major and interesting features of the TND pattern of city design. The TND pattern gives good traffic service to the short trips (60 to 70 percent of household trips) that are made for daily needs. The TND is not as friendly as the Conventional Suburban Development to the long trips (30 percent of household trips) that are primarily for employment.

 

Quality of the Automobile Trip

The criteria of traffic capacity and speed measure only a limited aspect of the total travel experience. We submit that there are important factors, other than capacity and speed, that affect the quality of travel. At the present time, we don’t consider these measures when designing a road system, nor do we consider them in gauging the performance of the road system once in operation. From observation, however, we feel that these measures have a large bearing on the level of service that is actually experienced by the traveler.

 

The Traveler’s View is Awful

The typical arterial, designed for a 50-mile-per-hour (mph) design speed and carrying 40,000 to 50,000 vehicles daily (the “50/50 Arterial”) is a miserable driving experience. We should simply stipulate this as a basic premise. You don’t see these arterials in chamber of commerce brochures. You don’t see visitors coming here to Bellevue and asking to see strip commercial. Personally, I don’t like them, can’t find anyone who does, and think we can all agree that these things are simply a rotten environment for the drivers.

 

Why are they such a poor atmosphere?

*Pavement, everywhere. This is not exciting pavement. Not graceful, no beauty of form and function that we do see in some transportation works. This is ugly pavement.

*No people.

*No trees.

*Boring.

*Hot blazing in the sun in the sunbelt. Bleak anywhere else.

*Monumentally ugly traffic control de vices.

*Forest of overhead wires.

*Long, exasperating delays at traffic signals, with nothing in the blighted view to offset the waiting.

 

The 50/50 arterial, by its very nature, has an inherent drive toward ugliness; or perhaps certain ugly things are looking for 50/50 streets to settle on.

 

The Inevitable Sellscape

The packaging of 50,000 daily vehicles (and therefore, a total daily population of 60,000 to 70,000 drivers plus passengers) into a single arterial street leads inevitably to the irresistible urge to sell things to this population, and creates a sellscape along the street.

Containing this type of use of 50/50 streets is far beyond the will and ability of the typical local government. The 50/50 arterial is a gift-wrapped, gold-plated, gift to strip development. Once in place, almost no power on earth will stop its march toward strip commercial. Time spent berating local governments (counties and cities) for not doing better with these monstrosities (and I’ve done my share of this) is satisfying to the critic, but is unproductive. Once in place, it is too late to do much about the 50/50 arterial.

It is bad enough that we inevitably get a sellscape on our 50/50 roads. What is even worse is that get a miserable looking sellscape. Because of the inherent design feature of the road — the 50- to 60-mph design speed — we evolve a sellscape that is geared largely to the motorist passing by at 45 to 50 mph.

Visually, the sellscape is focused on a narrow cone, ranging several hundred feet from the driver. This full sellscape is enjoyed only by those drivers in the outer lane; things get worse as you get into the interior lanes. This 50-mph sellscape can be interesting and can entertain, particularly at first viewing. Some superior design and behavioral science talents have gone into making sure that this scene attracts your attention. The 50-mph sellscape is the highway equivalent of newspaper headlines: they catch your first attention, but don’t ‘impart anything more with closer inspection or repeated exposure.

At slower speeds, the driver’s field of vision broadens out to a wider angle. And what does the sellscape offer as you look at it more closely? Does the 50-mph feature now reveal a 20-mph texture, a finer grain, a deeper level of stimulation? No, it doesn’t. lt’s hollow, empty. It has nothing further to offer.

And at the stopped condition (where we spend a good portion of our travel time) how does the sellscape look? Even worse than at 20 mph, of course.

To follow the newspaper analogy, there are no bylines and no news story. No richness of information. No inverted narrative. If you’re not moving, the sellscape doesn’t want you, and can’t use you.

The street sellscape, like its television counterpart in the 30-second and 15-second commercials, has to go for the quick hard-sell. You do this by being strident, out-shouting the next seller, demanding attention or raising the decibels.

The individual land uses on the 50/50 strip may be attractive in other settings. The blight comes from how these are assembled into a 50/50 environment. For example, offices can have great driver eye appeal or zero appeal. Fast food restaurants that we like to criticize as auto-dominated can be part of the ordinary sellscape or contribute substantially to driver eye appeal. Commercial tourist attractions, a mainstay of our Florida growth, can also range from terrific driver eye appeal to absolute zero.

 

Boredom

The 50/50 strip may very well yield an interesting experience the first time or first few times you drive it. The message, however, wears thin and cannot stand up to repeated trips through it.

A typical work trip by a suburban resident, having only a single route available, will be forced through the same section of arterial 500 times annually. Many suburban residents are also hostage to the same strip of highway for daily shopping needs, thus raising their total exposure to given road segments to 1,000 or 1,500 times annually. The strip, with no depth of information or stimulation, is giving little more on the five-hundredth or one-thousandth trip than it did on the third or fourth trip.

 

Why is Auto Sellscape Any Worse Than Pedestrian Sellscape?

An interesting devil’s advocate question can be raised: Why are we being critical of the auto-oriented sellscape along the 50/50 arterial, and at the same time enthusiastic over the street-level pedestrian sellscape.

We value this greatly when we have it, we envy other countries for it, and as Andres Duany points out, we make tourist shrines out of the few places that achieve it. So why is it OK for pedestrians, but awful for motorists?

The answer, I believe, is depth of texture. A pedestrian-level sellscape, because of the low speeds of the passersby (and the ability to stop altogether) can offer a tremendous depth of possible interest. The pedestrian sellscape is interesting at the full pedestrian speed (four feet per second and 15-foot distance). This same sellscape, because of its richness of detail, can absorb your attention for large amounts of time at zero speed. There is a near-zero threshold of access to the greater detail: as a pedestrian, I can stop within seconds and delve into the deep layers of detail.

On the other hand, the sellscape on 50/50 arterial streets has no depth. It plays well (at least at first) at 45 mph; it doesn’t play at all at low or zero speed. Unlike the pedestrian sellscape, the threshold of access is huge. To get more detail, I not only have to leave the traffic stream, enter a parking lot and park, but also get out of my car and walk through a hostile parking lot environment.

We DO have interesting vehicle-oriented sellscapes — we just don’t seem to get them on the 50/50 arterials. Here, for example, is something from a vehicle oriented seascape that plays at 45 mph, at 25 mph and a distance of 50 feet, and even at the pedestrian scales of four feet per second and 15-foot distances.

The more we look at why we don’t have appealing vehicle-oriented sellscapes, the more we are led to the conclusion that we almost CAN’T have them at the 50/50 level; and that if we want more of them, we simply need to have less of our pavement in the form of 50/50 highways.

 

TND Yields a Superior Overall Trip Profile

One of the primary reasons for our current notion of arterial streets instead of a more livable pattern is that we have no criteria, in the planning and design process, that even measures livability in any way. Our dominating design criteria is basically one-dimensional: speed, or its related variable, capacity. Safety sometimes enters as a distant second.

Using this criteria alone, we will always select the solution that yields the greatest capacity, and therefore, the greatest travel speed.

What we have long felt intuitively, but are only starting to appreciate, is that our perception of travel is not one-dimensional at all, but rather considers a host of all kinds of other factors along with the ‘hard’ measures of time and speed. We could aggregate all of these qualitative factors into a single measure called “trip quality” an add this dimension to the strictly time measure and get a resulting profile of overall travel quality.

Now we are portraying not only the single dimension of time, but also a measure of the ‘goodness’ or ‘badness’ or the time as felt by the driver and passenger.

Using this time profiling notion, let’s examine two trips made for identical trip purposes: from home to the grocery store, located in the typical community shopping center.

In the conventional subdivision layout example, the trip begins at home, and first travels on a local street.

The trip then proceeds to a collector street, and from there to a typical arterial street.

At the destination, the traveler then moves onto what is essentially an internal collector within the shopping center.

The traveler parks and walks to the destination.

The first part of the trip, on the local street, is a pleasant, high-quality experience, as indicated graphically by the color (green) and the height of the color. However, after this initial segment of high quality, things begin to deteriorate. The next part of the trip, on the collector, is a boring, low-interest experience, due to the deliberate removal of all interaction from the roadside and replacing with a continuous wall. Things go from mediocre to bad as we reach the arterial street. Our introduction to the arterial street is a 70- to 90-second wait at a traffic signal, with nothing else to occupy our attention. Once on the arterial, we are exposed to the full dose of the 50/50 sellscape — the pavement, parking, lack of people. On our profile of travel quality, the level descends down into the deep red.

Things stay at that level as we enter our destination site — the shopping center parking lot. Our introduction to the site is along the ‘throat,’ essentially a piece of collector roadway whose purpose is to absorb clots of traffic from the arterial street, and let them make the transition from 45 mph travel to parking in a safe, off-street environment. It serves that purpose well. In overall quality terms, it rates more deep red.

At last, we come to the walk from parking space to store. Typically, this is a big negative. There are exceptions, certainly, but remember — we are talking about your typical community shopping center.

 

Typical TND Trip Profile

Now let’s make the same trip, between the same origin at home and the same destination at a grocery store, but this time on a TND network. As in the Conventional Suburban Development case, the trip begins with run on a local street. Because neighborhoods are linked with local streets (and not segregated into enclaved pods) the travel on local streets is likely to continue for several more blocks.

From the local street we move onto a collector, in this case a three-lane street with scattered commercial.

Following the design concept of TND, the collector street directly serves the shopping destination. Because of the smaller scale of the use, and the absence of the 50/50 arterial, the entire scale of the destination is reduced. As in the case of Conventional Suburban Development, we park in a lot and walk to the store.

The quality profile of this trip made on a TND network is radically different from the same trip made under Conventional Suburban Development conditions.

The TND trip sustains the high-quality throughout the entire vehicular travel, because it is made entirely on local and small collector streets. So instead of a continuously increasing mass of red, the TND trip maintains a consistent green, or high-quality profile. At the end of the trip, when we again find ourselves in the store parking lot, the quality of travel is, as in the case of the Conventional Suburban Development, negative. Standing on a big piece of asphalt is never going to come out green on this profile. However, even this negative aspect of the TND trip is less severe in the TND experience, compared to the Conventional Suburban Development. Because of the TND scale of development, the severity of the negative impact of the parking lot is less. Further, because it is smaller, we spend less time in it. So we reduce both the severity and the duration of our negative parking experience.

 

Comparison of TND and Conventional Suburban Development Profiles

A graphical comparison of the two profiles for our hypothetical trip show that they are radically different. Recall that when we applied only the one-dimensional measure of speed, the Conventional Suburban Development pattern yielded the shorter trip, and therefore was the higher-performing system.

However, when we add the dimension of travel quality we find that the quality aspect starts to outweigh and overwhelm the original time dimension, and the ” weighted average” experience of our entire trip quality Is very large under the TND model.

The net result of this type of quality difference is that the perceived time becomes significantly less for the TND trip.

For example, we could consider the time on local streets, because of their environment, to be perceived at 80 percent of clock time. Arterials, because of their offensive environment, have the opposite impact — a perceived time that is 30 percent greater than clock time. The time spent in the parking lot, the most offensive of all environments along our trip, is weighted at 120 to 150 percent of true clock time.

Working through the arithmetic points out what happens to the perceived (as opposed to the clock) trip times. Because it is largely on local streets, the TND trip computes out at only about three quarters of the time of the Conventional Suburban Development trip.

And even though this is just a contrived example, you can see that the results are what we engineers call robust. In other words, you will reach the same bottom line conclusion even if you vary the stipulated conditions considerably. For example, we could penalize the TND trip by making local street travel only slightly better than neutral, and by reducing the arterial penalty, and still retain the same conclusion of a lower overall perceived time for TND. What the example is telling us is that if we keep our local trips on local streets and downscaled collectors, we will overcome substantial differences in travel speed through the perceived quality of travel. The land use features of TND — smaller commercial sites interwoven into the urban fabric — will further reduce the terminal time for TND trips.

 

The Broken Profile

Another thought that these profiles suggests is that the continuity of a high quality travel experience may figure largely in our impression of our entire community. We may well have high-quality origins and destinations, but if every trip connecting them involves our being dragged out onto low-quality arterial streets, our overall impression of our community is marred. Even if you don’t see the blighted arterial streets from either your origin or destination, the realization that any travel is captive to them will spoil the concept of urban village that we are trying to foster.

The value of a continuously positive profile, or conversely, the damage done by a broken profile is one of those areas that we can only guess at now. We need more research here.

 

Overall Quality – A Long-Time Feature of Other Products

The notion of overall profile of quality determining its real value, while a novel, untested, unknown experience to us transportation planners, is old stuff to anyone who sells almost any kind of product. Take, for example, the product that runs in our networks — the cars.

As an engineer, I’ll tell you what that hard, measurable performance measures are: speed and capacity. So, do we try to sell cars by improving their measurable performance qualities? Absolutely not. Thirty years ago, the standard mainstream sedan had a capacity of six passengers, a top speed of 100 mph or so and a legally usable top speed of 60 to 65 mph.

Thirty years later, the standard mainstream vehicle has the same performance -six-passenger capacity and 100 mph, of which we can legally use 65 mph. In terms of the hard measurable performance, then, we haven’t improved t e cars at all. But have we improved the overall profile of the product? Immensely, through comfort, convenience and luxury.

With respect to automobiles, then, we have long ago “maxed out” the hard measurable performance measures of capacity and speed. Over the last 30 or 40 years, we have continued to improve the product, by adding to the nonmeasurable aspects.

You could repeat this pattern for almost any common product you can name. Yet, in the case of road travel, we are still trying, against increasingly difficult obstacles, to improve a single dimension of performance without paying attention to what really constitutes the quality of the experience in the user’s eye.

 

The Opportunity for Driver Interaction

In creating the sparse system of major arterials, we make driving an increasingly passive process for drivers. We create a situation in which drivers cannot, through their own efforts, improve their own driving experience — either the hard measurable features (speed) or their perceived quality — for a particular trip.

 

Limited Choice of Routes

The sparse hierarchy patterns of roads in conventional systems takes away any element of route choice for the driver. In the sparse network, there is only one route that can be taken. As a driver, I can’t match my route to my trip purpose — I can’t take a leisurely scenic drive home to unwind after work, I can’t impress an out-of-town visitor with a route that emphasizes the unique quality of my city.

Driving a sparse network will never develop skill from repetition. This is frustrating, and contrasts to our experience in other activities. In many of our other day-to-day activities, for example, grocery shopping or for many of us who still encounter a parking “problem,” there is considerable satisfaction with ‘learning the ropes’ and making satisfying decisions from the myriad of options available.

The ability to choose routes is important not only in improving the quality of the travel aspects, but is also a large factor in converting travel from single-purpose/ single-destination travel to a far more productive pattern of ‘linked’ trips, in which multiple trip purposes are served in a single series of connected trips.

When we consider the multitude of routes opened up by a dense network, and further consider the type of commercial response to dense networks, we can appreciate the potential for greatly changing the pattern of travel.

Nor, in the sparse hierarchy, do I get to exercise my skill, experience, judgment, powers of observation, or intuition to improve my travel time on my single available route. On the sparse hierarchy, in fact, I am limited to a meager choice of options for improving my performance: I can start my trip earlier or later. Further, I can cocoon myself off from the entire miserable experience by listening to an elaborate sound system or talking on my cellular phone. Or, through strenuous efforts, I can try to improve my travel speed and get the whole thing over with earlier. This approach, largely the province of young male drivers, involves competitiveness, hostility, aggression and a generally anti-social behavior. This type of driving is a large contributor to the “red” in the quality profile we saw earlier. This type of driving, despite its high visibility and the sheer quantity of unpleasantness dealt out to other drivers, produces interestingly small benefits to the practitioner, and none at all to traffic in the aggregate. For example, an extensive travel time survey in a southwest city found that for a three-mile trip, the difference in overall travel time between orderly driving in the traffic mainstream and highly aggressive driving was 17 seconds, or two percent of the total travel time.

Improving the individual driving time, furthermore, is not a systems benefit. Because traffic is a stream of incompressible vehicles, an aggressively driven vehicle imposes its advance in the flow only at the expense of the travel time of other vehicles — a win/lose situation and not a win/win situation.

Once the traffic volumes have neared saturation (70 percent or greater of capacity) aggressive driving takes the form of darting into any available gap (space) between vehicles ire adjacent lanes. The size of these gaps is set by the following drivers, and is their idea of safe following distance, based on speed, roadway condition and their own abilities. An aggressive driver darting into this space simply causes the following driver to slow down in order to restore the initial gap.

Even worse for traffic flow, this slowing, particularly if the driver brakes, may be followed by a “wave” of slowdowns extending backward (“upstream”) from-the point of the lane-change action.

Aggressive driving by a minority of the drivers does not improve the overall traffic flow and therefore the capacity of the road. Aggressive driving simply works the non-aggressive majority of drivers toward the back of the line.

 

TND and Non-Motorized Travel

TND is Friendly to Non-Motorized Travel

Since one of the driving motivations behind TND is to create pedestrian environments, it should come as no surprise to find out that in fact they perform well as pedestrian environments. It is worthwhile to consider the actual mechanics of why TND works so well, and to be able to defend the features that produce this friendliness to non-motorized travel.

 

Interwoven Urban Fabric Creates More Walkable Origin/Destination Pairs

Perhaps the single biggest underlying factor in the pedestrian-friendliness of the TND approach is the concept that land uses are interwoven in an intimate mix. This is something that, try as we might, we are simply not achieving in all of our so-called mixed-use developments. From a traffic point of view, even in the best of our mixed-use developments, we are afraid of putting origins and destinations together. It’s almost as though we have systematically worked to keep origins and destinations apart. It raises the interesting question of what market force is driving this separation.

It’s not within the scope of this traffic discussion to explain why our planned mixed uses don’t ever really mix, and our unplanned mixed use becomes delightful and transportation-efficient fabric. I feel that it has a lot to do with the size of ownership. If there are only large projects, there will never be a fabric. If there is a certain critical threshold of small, individually owned projects, then there will be the fabric.

Christopher Alexander, in his interesting book A New Theory of Urban Design, holds that the key to livable urban development is a mix of all sizes of projects.

 

More Origin/Destination Pairs Due to Intimate Mix

The intimate mix of origins and destinations in the TND concept places a large number of origins and destinations within walking or bicycling distance. It does this simply by breaking down the size of the origin only (residential) pods and the destination only (commercial) pods. This type of intimate mixing, even without any change in density, but simply by its geometry, brings large numbers of origins and destinations together.

 

TND Improves the Routes of Pedestrian and Bicycle Travel

More Direct Routing

TND, because of its intimate mix of activities and the density of network, provides a shorter travel distance for pedestrian trips. This is in most cases due simply to geometry. A dense network simply provides a shorter travel distance between any two points. The TND concept eliminates enclave pods of development, which typically have a 600- to 800-foot walking distance threshold, and a particularly ugly one at that.

The TND concept replaces this walking distance threshold with a greatly reduced threshold. The TND concept also removes the enclaves, and provides access in all directions to the site. So even in the absence of any land use changes, just the network in the TND brings many more origin/destination pairs into walking and bicycle distance.

 

A Better Bicycle and Pedestrian Environment

A series of small streets yields a better bicycle and pedestrian environment than a hierarchy of a few larger streets.

A given lane volume of traffic is more hostile to walkers/bikers on a multilane street than on a two-lane street that is the mainstay of the TND network. The reason for this deterioration in walk/bike environment is the enlarged intersection size on multilane intersections.

Specific problems with larger arterials are large-radius high-type traffic engineering features, shallow-angle crossings at ramps and turn-lanes, monstrous pavement expanses to be crossed, hostility of dual left turn-lanes to any type of human habitation, and general feeling, by walkers and cyclists, of being in an alien moonscape.

Just the sheer size of a single intersection exceeds the walking distance radius for even good environments, not to mention poor environments. Tortuously long traffic signals add further to these already improbable walk times.

Some other factors, not easy to measure, figure in the hostility of multilane roads to walk/bike travel. Competitive and aggressive driving is more prevalent on multilane arterials, because of the opportunity for drivers to improve their position at the expense of other drivers, with a corresponding loss of attention to walk/bike traffic. The sense of auto-domination at intersections of arterials has a profoundly discouraging impact on pedestrians. We all know the ingredients of this hostility. They are blighted streetscape, endless parkingscape, absence of any tree cover whatsoever, presence of overhead power.

 

Better Environment: Subjective Factors

A 20-mph environment along a TND street is vastly preferable for walk/bike travel to a 5O-mph environment along an arterial street. The 5O-mph corridor, even with high-type walk/bike facilities, is a poor environment because of the lack of interest.

What little visual interest there is in a 50-mph environment is lost when the speed drops to walking speed. What the walker/cyclist then notices at that the speed is the coarseness of roadside experience. In a TND environment, the visual texture of the dense fronting properties sustains a high interest, to the passerby, at speeds down to a slow walk. There are densities of one structure per 50 feet, and individualistic touches on all of the separate properties.

 

TND Multiplies the Number of Alternative Routes Available for Non-Motorized Travel

With the TND concept, there are an almost endless combination of alternative routes available for a given origin/destination pair. The walker/cyclist can select routes in response to real-time conditions (for example, a separate route for peak traffic periods versus weekend or midday travel). The multiplicity of routes available also lets the walker/cyclist match the route to their particular skills. For example, expert cyclists can choose to take their place in traffic as a fully-vested vehicle, while low-skill cyclists can travel on small, two volume, possibly more circuitous routes. All in all, TND presents a vast difference in opportunity over that which exists in the sparse hierarchical system, in which all walk/bike trips are forced up and down the street hierarchy.

The alternative routes can give large number of bicycle and walk trips a profile that totally excludes the use of arterial streets.

 

Roadblocks to TND

If TND is as attractive a development pattern as we have been arguing over the past few chapters, than why isn’t it the dominant form of new development?

We can offer some reasons, and we suggest that understanding these reasons can help advance the adoption of TND concepts.

 

Suburban Sociology – Throwing the Baby Out with the Bath

Conventional suburban street design first appeared in the wave of suburbanization that began in the late 1940’s after the close of World War 11. This first wave of suburbanization brought numerous benefits — first-time ownership of an affordable house, renewed availability of household appliances, resumption of automobile production. At the same time, the traditional neighborhoods, which the new suburbanites were leaving behind, became closely associated with undesirable features of the old life-style: burdensome family commitments, intrusive church presence, machine politics, single-bath homes, transit dependency, difficult parking for the new cars, and so on.

The association of the suburban community with everything “new” and desirable, and, conversely, the association of the traditional neighborhoods with everything “old” and undesirable has persisted for 40 years. The hierarchical street system and enclave project pattern became an inseparable part of the suburban ‘package.’ Along with the nice suburb you automatically got, as standard equipment, a nice unconnected street system.

The momentum for Conventional Suburban Development has persisted long after its original motivations have disappeared. The first cul-de-sac streets, for example, were intended for children to play in them safely, a reasonable idea 40 years ago with larger family sizes and no television. That design motivation for cul-de-sac streets, however, is long gone — children are fewer in numbers, and are all indoors watching TV instead 6-f playing in the street. Another example of an old design idea persisting: the employment pattern of 40 years ago, with strong downtowns and industrial concentrations, favored a hierarchical street system to carry the “many-to-one” traffic pattern that strong downtowns and industrial areas generated. That pattern has faded.

Similar to the people in the fable who accidentally discovered roast pig when their barn burned down, and then continued to make roast pig by burning barns down, we persist in trying to recreate the advantages of suburban living by reproducing a 40 year-old street design.

 

The Functional Class Myth

Traffic engineers, technicians by temperament and training, believe in things that can be measured. They like to sort, to find a structure, a harmony. This search for structure and order has evolved one of the fundamental ideas making our cities look the way they do. This concept is the functional classification hierarchy for streets.

This concept, now at the very foundation of all of traffic engineering thinking, has been around a long time and became standardized in the 1950s. Some versions of its origin say that it was meant to foster Civil Defense evacuation of cities during nuclear attack.

The functional class idea has an immediate plausibility — that at one end of the spectrum we should have streets that are meant for high speed long distance travel, carrying large volumes, and not hindered by local access. And at the other end of the spectrum, a local street, meant to feed other streets, carrying small volumes at low speeds. And in between, a collector street that does just what its name says — collects from local streets, maybe has fronting commercial uses like commercial, and feeds the arterial system.

Unfortunately, this hierarchy exists only in the minds of traffic engineers and planners. In reality, something entirely different happens. The idea that you can keep local access off the arterial streets is simply preposterous. I can tell you from our day-to-day work in development approval, that the 50,000 ADT arterial street is a gift wrapped, gold plated irresistible invitation to develop strip commercial. Think about it — we bundle together 50,000 vehicles with 60,000-70,000 occupants into a captive market. We make sure we don’t give them any other route. We ruin the roadscape, by the size of it, for anything else. And then we, in theory, expect strip commercial to stay off? Get serious.

It’s interesting to see what happens as the realization sinks in that that dense commercial activity access is not going to be kept off the arterials, but is in fact going to be attracted to it. The typical reaction, incredibly enough (but understandable in light of the emphasis on capacity) is to make the street bigger, now through frontage roads!

These are actually POLICY in some parts of the US such as California and as Texas. Fortunately, these are so excessive, in terms of cost, that they are not frequent and are not spreading fast.

In reality, a more believable model of access and mobility reads like an inverted curve, that says at very high volumes and at very low volumes access is the main feature of streets.

The concept of functional classification has parallels in all sorts of transportation systems. For example, we have 4 inch ‘local’ gravity feet sanitary sewer feeding to 8 inch collectors, then 12 inch force mains to central plants. Or power, where we have interurban 24,000 KV transmission, the ‘Interstates’ of the power grid, stepping down to 2,400 V transmission lines, to substations and to the 220 V local line into your house. These utility hierarchies are functionally classified in the strictest sense. For example, you will never see an individual home with “access’ directly onto a 24,000 KV transmission line.

There are also numerous examples of hierarchical structure in nature. For example, drainage basins start with creeks (locals) working their way up to rivers (arterials).

The body’s circulation system, which even furnishes the term ‘arterial’, its largest pipeline, to the largest size element of a traffic network, is a hierarchy.

Everywhere we look, in both nature and in man-made systems, we see hierarchical systems of flow. Why do these work so well everywhere else and work so poorly for traffic?

For one thing, all the other flows we gave as examples are focused at a single origin or a single destination — they are “one-to-many’ flows (e.g. electric power) or “many-to-one” flows (e.g., sanitary sewer). Traffic on the other hand, is a “many-to-many” flow, and is becoming ever more so.

More importantly, for all the examples we have mentioned, the measure of service is very simple — capacity and nothing else. There is no aspect of quality of experience for whatever is being transported in the system. To continue with the sanitary sewer example, the design of sewer systems properly does not consider any aesthetics as seen by the material being transported. In the case of highways, we unfortunately continue the same idea of service, and are paying no attention to the quality of the travel as experienced by the traveler.

Perhaps the most important distinction between traffic networks and all of the other flow networks both natural and man-made, that seem to be analogous is that of all of these systems, traffic alone involves human behavior. As we have noted, the presence of 70,000 persons daily passing along a road sets off strong, almost unstoppable series of development actions (“sell them something”) that greatly change the intended operation of the arterial street.

As traffic engineers working with private development, we see these actions on a daily basis. On the other hand, we don’t see the same drive to locate along 24,000 KV lines, or along a 12 inch force main.

 

No Constituency within Traffic Engineering Industry

We need to understand that the industry that we look to for moving people — the traffic engineering industry — has no constituency for doing so.

Let’s consider the range of possible goals of traffic engineering, as practiced on any level, local up to statewide, and see which one jumps out as the most likely.

The traffic engineering profession, which has perhaps the greatest influence of any group on how our new urban growth looks, is interested in one thing only — moving the maximum number of cars.

The profession is fixated on this goal. It does not see its charge as moving people in an attractive setting; it does not even see its goal as moving people.

Before we criticize this vision any further, we should understand that it is totally predictable, and is in fact simply a manifestation of a mature bureaucracy. We don’t even think about or question the fixation of other transportation modes on their own industries. For example, do we expect the airlines or their governmental regulators to recommend the development of rail corridors in air markets that are getting too dense for efficient air travel? Of course not. We expect the air industry to expand, build more airports, build more traffic control capacity — in other words, to just keep on expanding their own capacity.

What is useful to us is to understand that the only possible direction from change is outside the mainstream of traffic engineering.

 

Belief in Efficiency of Scale

The traffic engineering industry is based on an implicit assumption of efficiency of scale; that bigger roads are, like a bigger power plant or wastewater plant, more efficient. This is true for power and sewer. It is not the case for roads.

In fact, the OPPOSITE is true for roads. Rather than and EFFICIENCY of scale, roads have a DEFICIENCY of scale. We saw this earlier in the detailed analysis of capacity.

There is a counterpart feeling that administratively there is also an efficiency of scale, and that a large regional road agency will be much more beneficial than a myriad of smaller one. We are getting into something more subjective here, and there is no equivalent to 1985 Highway Capacity Manual to support the notion that agencies (like road capacity) are less effective with scale.

It certainly appears that the capacity for doing damage to communities through road design goes up with the size of agency. Further, road planning and building, a mature science, is fully within the capability of small sophisticated suburban jurisdictions with their professional staffs. Our host, City of Bellevue, Washington, with its population of 189,000, can plan, design and put to bid a given section of road for the same ultimate bid price as the State of Washington.

 

Holdovers from Strong Core Days

Much of our hierarchical street system concept is probably a holdover from the strong center days, when this kind of street system was indeed a valid response. I am going to spend zero time preaching to this choir that we are in a post-strong central city situation, and that we should think about retiring the hierarchical concept.

It is interesting that design of products and systems is LESS hierarchical. For example, the cellular phone breakthrough substituted an all-local grid for the sparse, strongly centralized hierarchy (two-way radio) that it replaced. Twenty years ago, all US automobiles were of the frame-and-body design, a hierarchical concept. Now 96 percent of all cars sold in the US are unitized body, in which the frame, and therefore, the hierarchy, disappears.

 

Conventional Suburban Development is a Great Cost Exporter

With conventional suburban design, much of the transportation cost of new development is shifted away from the private participants (the builder and the initial buyer) and shifted to the public at large. This ability to show a lower initial cost of new development is, needless to say, highly appreciated by builders and buyers, and they are quite happy to see the situation continued.

A typical enclave project in Conventional Suburban Development is, by definition, placing ALL of its external travel demands on the public street system outside the project.

The limitation of access to a single point, furthermore, increases the amount of travel distance that is needed to make any given trip. The pod itself is performing NO role whatsoever in carrying the systems traffic — it is all exported to the external, publicly built system.

The individual TND project, on the other hand, with its interconnected streets, carries a share of the regional traffic.

How successful are we at capturing the external cost from new development through impact fees and other assessments that are meant to exact the full cost of new development from that development itself? If we account for only the raw cost of building the new roads to serve Conventional Suburban Development, we can, as has been shown in Florida, quite readily pay for half or so of the cost of our current road patterns. However, in terms of degradation of environment, excess travel time and distance, and loss of opportunity for non-motorized travel, we can conclude that we are paying for enclave pods at the expense of degraded public realm.

 

Lack of Input from Other Relevant Fields

Despite its large influence on our daily life, the traffic engineering industry has a limited amount of input from other fields. As a result, we get street designs that very nicely meet the narrow engineering criteria that we adopt, only to find that, because of behavioral realities, they produce an unexpected and sometime totally contrary result.

We can best illustrate this kind of backfiring action by examples. Take protected left turns, for example.

These are supposed to make it easier to turn for the individual, and increase capacity. The actual impact, over time? Drivers are collectively forgetting how to make left turns at ordinary intersections, are clamoring for turn arrows everywhere, and system capacity is going down. A prime example on a citywide scale is Phoenix, which under public pressure had to finally install a large number of protected left turns, subsequently degrading the remarkable performance of its highly-connected street network.

Driver’s education is a classic example of a decision made without enough behavioral content. Everybody knows that driver’s education is a good thing, helps make more responsible drivers, and reduces accidents, right? Actually, just the opposite. The Insurance Institute for Highway Safety has concluded that drivers education INCREASES the number of accidents, and is on record in opposition to the inclusion of drivers education in schools.

One more recent example of what happens when only a single viewpoint tries to deal with a problem is just too good to pass up. A traffic engineering study, just recently reported in the ASCE Journal, was conducted in Huntsville, Alabama, to analyze the problem of vehicular collisions with roadside trees.

This analysis, done by traffic engineers, did a good job of data assembly and analysis, and identified an interesting profile of the typical driver involved in a collision with roadside trees. These accidents predominantly involve young male drivers, alcohol- or drug-impaired, in small group size, in early hours of the morning, in single-car accidents.

The profile was very conclusive, and suggests the following range of recommended remedies:

*Trees are doing a great job. Plant more trees.

*Get drugged drivers off road.

*Cut down trees.

Which of these three options do you think the traffic engineers recommended? You guessed it, “Cut the Trees Down”.

The list of similar apparently contradictory measures goes on and on. The common theme that emerges is that apparently sound traffic policies, when examined in light of all of their human behavior consequences, may in fact be counter productive. This concept is similar to the notion of counter-productive social programs that Edward Banfield presented so persuasively in his book “The Unheavenly City” several years ago.

 

Where Do We Go Now?

We have demonstrated, in the course of this discussion, that the traffic “works” in the TND pattern of land use. We feel strongly that further investigation into the technical traffic engineering features of TND will further deepen the support for its claimed circulation advantages.

We already knew intuitively that TND yields strong advantages for non-. motorized travel. We have shown, in this discussion, that the features of TND that promote non-motorized travel are complementary, and not necessarily hostile, to vehicular traffic circulation.

We all appreciate that there are large obstacles to the application of TND. This raises the question, then, of what are the feasible ‘do-able” measures that we can take to further enable the actual use of TND and related approaches to travel that seek to an alternative to the automobile-dominated design that we see in all our new growth?

The answer is that there is plenty of things that we can all do, starting right now.

Most of us here today are in the grouping referred to in the graphic as “Builder.” It goes without saying that developers and their designers and consultants fall in this category. Less obvious is the inclusion of many of you present at this conference — public works staff, city engineering and city and county administration — in the “Builder” category. This grouping is consistent, however, when we consider the point of contact with the development process – the entire group deals, on a daily basis, with decisions that immediately determine the built environment.

These daily decisions — for example, on street network configuration in new projects, on allowing enclave project, etc., will have an immediate impact on the shape of development. In the longer run, the “Builder” group is the logical focus of various ‘tactical’ actions, such as establishing the market feasibility for TND projects, and countering the misinformation surrounding the safety and liability aspects of unconventional street design. In the even longer run, we can raise the standing of TND and related design concepts to fully acceptable designs in the view of sanctioning bodies such as the institute of Transportation Engineering (ITE), and so forth.

The “Policy Maker” group, which applies to many of you present at this conference, are the logical candidates for the enormously important tactical task of getting TND language into local plans. This tactic, already meeting with success in numerous locations, is one of the most promising ways to advance the entire concept quickly.

…

“Personal Advocates” — individuals interested in the TND concept but not professionally involved in the planning/design/development fields, will most likely constitute a growing source of enthusiasm for TND as the public awareness of the concept continues to grow. We are certainly seeing this happen at the present time. The target activity for the “Personal Advocate” group is intervening in the local development review process (hearings, citizen review committees) and in seeking changes in local comprehensive development plans and land development codes.