Online TDM Encyclopedia - Evaluating Nonmotorized Transportation

Nonmotorized Modes (Walking and Bicycling, and their variants such as Wheelchairs and Small Wheeled Modes, also called Active Modes and Human Powered Transport) are important components of the transportation system, are often critical to the success of TDM programs.

· They are resource-efficient travel modes (i.e., they consume minimal road and parking space, impose minimal costs on consumers and the environment) that support TDM objective.

· They provide Basic Access. Nonmotorized modes are often critical for trips that society considers particularly valuable, such as access to essential services, education, employment, and social activities by people who are transportation disadvantaged.

· They are a primary component of Universal Design (transportation systems that accommodate people with disabilities and other special needs).

· They provide Transportation Choice and consumer savings.

· They provide Healthy Exercise and enjoyment.

· They help create more Livable Communities.

· They provide access to Public Transit and so are critical to efforts to make transit more practical and popular.

· They support efficient land use, such as New Urbanism, Location Efficient Development and Transit Oriented Development.

Homo sapiens are walking animals. Environments that are conducive to walking are conducive to people. Walking is a fundamental and critical activity for physical and mental health. It provides physical exercise and relaxation. It is a social and recreational activity. Walking is also a critical component of the transportation system, providing connections between homes and transit, parking lots and destinations, and within airports. Often, the best way to improve another form of transportation is to facilitate walking.

However, nonmotorized travel is often overlooked and undervalued. Conventional travel surveys find that only about 2% of total travel is by nonmotorized modes, which implies that it is unimportant, and improving nonmotorized conditions can do little to solve transport problems. But conventional surveys undercount nonmotorized travel because they ignore short trips, non-work travel, travel by children, recreational travel, and nonmotorized links. Actual nonmotorized travel is usually three to six times greater than these surveys indicate (Litman, 2003).

Conventional transport planning assumes that society is better off if somebody spends 5 minutes driving for an errand than 10 minutes walking or cycling, since it applies an equal or greater cost value to nonmotorized trips than motorized trips, only considers vehicle operating costs (vehicle ownership costs, and external impacts such as congestion and parking costs are ignored), and no value is assigned to the health and enjoyment benefits of nonmotorized travel. Such assumptions tend to skew countless planning decisions toward motorized travel at the expense of non-motorized travel. For example, it justifies expanding roadways to increase vehicle traffic capacity and speeds, requiring generous amounts of parking at destinations, and locating public facilities along busy suburban roadways, in order to facilitate automobile transportation although each of these tends to reduce walking accessibility.

Nonmotorized travel tends to be stigmatized. Some people consider walking and cycling outdated, unsophisticated and unexciting compared with motorized modes, or even as symbols of poverty and failure.

Basic Data and Performance Indicators

Certain types of data are useful for evaluating nonmotorized transportation trends and activities. Performance indicators are data collected specifically to measure progress toward objectives. It is useful to establish standard nonmotorized data collection procedures to allow comparisons between different locations and times. The table below lists some types of data that are useful for nonmotorized transport evaluation. Some of this data may already be collected, others will require new data collection activities. Note, however, that conventional travel surveys often undercount nonmotorized travel, particularly walking, because they ignore short trips, and walking links of motorized trips, as described later in this chapter, so improved travel survey methods may be needed. “Disaggregation” describes how this data should be classified.

Table 1 Nonmotorized Transport Data (based on ABC, 2000)

Data Type	Disaggregation
Activity
Percentage of total trips by walking and cycling.	User demographics, trip purpose and geographic area.
Average length of walking and cycling trips.	User demographics, trip purpose and geographic area.
Portion of population that walks or cycles on an average day.	User type, trip purpose and geographic area.
Facilities and Conditions
Length (miles or kilometers) of walking facilities (sidewalks and paths).	Type of facility, quality, geographic area.
Portion of streets and roads with walking facilities.	Type of facility, quality, geographic area.
Length (miles or kilometers) of cycling facilities (bike lanes and paths).	Type of facility, quality, geographic area.
Portion of streets and roads with cycling facilities.	Type of facility, quality, geographic area.
Percentage of bicycle network that is continuous.	Type of facility, quality, geographic area.
Quality of cycling conditions on road network.	Type of facility, quality, geographic area.
Bicycle parking and changing facilities at major destinations.	Type of facility, quality, geographic area.
Equipment
Number of bicycles owned per capita.	Type of bicycles, demographics of owners.
Number of bicycles sold annually.	Type of bicycles, demographics of owners.
Number of bicycles stolen per capita.	Type of bicycles, type of theft.
Safety
Number of police-reported walking and cycling crashes, number of hospital-treated walking and cycling injuries, and number of walking and cycling crash fatalities per capita per year.	Victim demographics, type and geographic location of crash.
Portion of cyclists wearing helmets.	Cyclist demographics, type of facility, riding conditions.
Number of bicycle training program graduates.	Type of training program, demographics of participants.
Portion of children who have participated in a bicycle training program when graduating school.	Type of training program, demographics of participants.
Planning and Promotion
Number of nonmotorized planning programs.	Type of program.
Funding of nonmotorized planning programs.	Type of program.
Number of specialized nonmotorized planning staff.	Type of program.
Number and quality of nonmotorized encouragement programs.	Participant demographics, activity type and geographic location.
Portion of recreation, mobility management, health, safety and sport programs that include nonmotorized encouragement activities.	Participant demographics, activity type and geographic location.

This tables lists various types of data that can be collected, as much as possible, for nonmotorized planning and evaluation. Some of this data (such as quality of cycling conditions) are discussed later in this chapter.

Measuring Nonmotorized Travel In Conventional Travel Surveys

How transport is measured can have a significant effect on transportation planning decisions (Measuring Transportation). Current practices tend to undercount shorter trips, non-work trips, off-peak trips, nonmotorized links of motorized trips, travel by children, and recreational travel. As a result, there are usually far more nonmotorized trips than what conventional travel Models recognize.

Conventional transport surveys and models often only count the “primary” mode used between Transportation Analysis Zones (TAZs). Some only count peak-period travel or commute trips. Nonmotorized trips to access motorized modes are often ignored in transportation surveys, even if they involve travel on public paths and roads. For example, if a traveler takes 10 minutes to walk to a bus stop, rides on the bus for five minutes, and takes another five minute walk to their destination, this walk-transit-walk trip is usually coded simply as a transit trip for analysis, even though the nonmotorized links take more time than the motorized link. Walking trips from a parking space to a destination, or between nearby buildings, are often ignored.

Similarly, many types of pedestrian activities are ignored in conventional field surveys. For example, conventional traffic surveys often ignore people who are sitting or waiting on sidewalks, skaters and skateboarders, and people walking from cars or buses to buildings (Haze, 2000).

Weinstein and Schimek (2005) discuss problems obtaining reliable nonmotorized information in conventional travel surveys, and summarize walking data in the U.S. 2001 National Household Travel Survey (NHTS). They find that about 10% of total measured trips involved nonmotorized travel. Respondents average 3.8 walking trips per week, but some people walk much more than others. About 15% of respondents report walking on a particular day, and about 65% of respondents reported walking during the previous week. The median walk trip took 10 minutes and was about ¼ mile in length, much less than the mean walking trip (i.e., a small number of walking trips are much longer in time and distance). The table below summarizes walking trip data.

Table 2 NHTS Walking Trip Attributes (Weinstein and Schimek, 2005)

Purpose	Frequency	Mean Distance	Median Distance	Mean Duration
	Percent	Mile	Mile	Minutes
Personal business/shopping/errands	48%	0.44	0.22	11.9
Recreation/exercise	20%	1.16	0.56	25.3
To transit	16%	N/A	N/A	19.6
To or from school	7%	0.62	0.33	13.3
To or from work	4%	0.78	0.25	14.1
Walk dog	3%	0.71	0.25	19.0
Other	2%	0.57	0.22	14.8
Totals	100%	0.68	0.25	16.4

This table summarizes the results of NPTS walking trip data. N/A = not available.

They reach the following overall conclusions about U.S. walking activity:

· Most Americans walk very little. The vast majority, 84%, reported no walk trips in their daily diaries. More than one-third reported no walk trips within the previous week.

· Exercise and recreational trips account for more than one quarter of walk trips, a significant share. Because these average twice the distance of other walk trips, they account for about half the share of total distance walked. However, the determinants of exercise trips are completely different than those of utility walk trips.

· Transit access trips are a significant component of total walking, comprising 16% of all walk trips. This finding suggests that improving the pedestrian environment might be an important component of making transit more attractive, and that increased transit use could significantly increase daily physical activity.

· The mean trip distance for utility walk trips, 0.5 miles, was remarkably constant across many demographic groups and in different neighborhood densities. This may suggest that it is reasonable for planning purposes to take this figure of half a mile as a maximum that many Americans would be willing to walk in ordinary circumstances.

· For those who do walk, walking can make a significant contribution to the Surgeon General’s suggestion of at least 30 minutes of daily exercise. Respondents who walked to transit averaged 26 minutes of walking per day (2.0 trips x 12.8 minutes/trip), those who walked or jogged for recreation averaged 41 minutes per day (1.6 trips x 25.7 minutes/trip), and those who walked for other purposes averaged 28 minutes per day (2.2 x 12.5 minutes/trip).

· Increased land use density by itself has only a modest impact on walking activity compared with other factors.

The UK 2003 National Travel Survey (NTS) (www.transtat.dft.gov.uk) includes short walking, although they are probably underreported due to the methodology used (walking trips under a mile were only recorded on the last day of a seven-day survey period, when participants were probably tired of the process, and scaled up to estimate total walking). According to this survey, an average person walks 192 miles per year, out of 6,840 total miles of travel (3% of the total), and spends 64 hours a year walking, out of 361 total hours spent on travel (18% of the total).

Table 3 Average Annual Travel By Mode (DfT, 2003)

	Travel		Travel Time		Trips
	Miles	Percent	Hours	Percent	Trips	Percent
Walk	192	2.8%	64	18%	245	25%
Bicycle	34	0.5%	5	1.3%	14	1.5%
Motorcycle/Moped	36	0.5%	1	0.4%	3	0.3%
Car or Truck Driver	3,466	51%	140	39%	401	41%
Car or Truck Passenger	2,047	30%	82	23%	226	23%
Other private vehicles	162	2.4%	7	1.9%	8	0.8%
Public Transit	897	13%	62	17%	92	9.3%
Totals	6,833	100%	361	100%	990	100%

Walking represents just 2.8% of personal mileage, but a much larger portion of travel time and trips.

Rietveld (2000) finds that the actual number of nonmotorized trips is six times greater than what conventional surveys indicate. Similarly, in Germany only 22% of trips are completely by walking, but 70% include some walking (Brog, Erhard Erl and Bruce James, 2003). The Southern California Metropolitan Transportation Authority recently increased the portion of nonmotorized travel in their models from about 2% of regional trips (based on conventional travel surveys) up to about 10% (based on more comprehensive travel data from the 1995 National Personal Transportation Survey). The 2001 National Household Travel Survey (www.bts.gov/nhts) obtained more detailed information on walking trips than most travel surveys. It found that walking represents 8.6% of total personal trips, about 50% more than reported in the 1995 National Personal Travel Survey, which used more conventional survey methods. Chu (2003) used NPTS data to calculate average minutes walked by various demographic groups.

Besser and Dannenberg (2005) used the 2001 National Household Travel Survey to analyzes the amount of walking associated with public transit trips, and factors that affect this activity. They found that Americans who use public transit on a particular day spend a median of 19 daily minutes walking to and from transit, and that 29% achieve the recommended 30 minutes of physical activity a day solely by walking to and from transit. In multivariate analysis, rail transit, lower-income, age, minority status, being female, being a nondrivers or zero-vehicle household, and population density were all positively associated with the amount of time spent walking to transit.

Winters, et al (2007) used data from the 2003 Canadian Community Health Survey (CCHS) and various other statistics to evaluate factors affecting utilitarian cycling rates. They found that the proportion of the urban population reporting bicycling in a typical week was 7.9%, with students cycling more than nonstudents (17.2% vs 6.0%). In the general population, older age, female gender, lower education, and higher income were associated with lower likelihood of cycling. More days of precipitation per year and more days of freezing temperatures per year were both associated with lower levels of utilitarian cycling, although there was less variation in the proportion of students who cycled by age and income.

Some surveys include general information on nonmotorized travel. For example, a survey by the Bureau of Transportation Statistics found that nearly half (46%) of driving-age adults (16 years or older) have access to a bicycle, and 54% of those with a bicycle used it during the last month (BTS, 2004).

To their credit, many transportation professionals give nonmotorized transportation more consideration than what is implied by the available travel survey data. They realize that nonmotorized travel has many critical functions in an efficient and balanced transportation system, some of which are difficult to measure. However, this occurs in spite of, rather than supported by, conventional transportation data analysis. There is much that can be done to improve transportation planning and Modeling to better evaluate nonmotorized transportation.

Information on current walking and cycling travel can be gathered in the following ways (IHT, 2000; Price, 2001):

1. Travel surveys can be designed to elicit sufficient responses concerning nonmotorized travel. For example, “travel” should be clearly defined to include walking and bicycling trips. Short, non-work and recreational trips, and trips by children should be counted.

2. A special survey targeting cyclists and pedestrians (such as survey forms distributed through bicycle shops, sport clubs, recreation centers, colleges, and schools). Surveys can be handed out to cyclists and pedestrians as they travel along a street or path. Surveys should include special user groups, such as people in wheelchairs and elderly pedestrians, particularly in areas they frequent.

3. Traffic counts that gather information on pedestrian and bicycle travel. These can include photoelectric counters installed on trails, electronic counters installed on cycle paths and bike lanes, and manual counts. Volunteers from pedestrian and cycling organizations may also be mobilized to perform manual counts for nonmotorized travel.

Surveys should gather the following information on nonmotorized travel:

· Who – Demographic information such as age, gender, residence location, employment status, and income.

· Where – Origin and destination of trips, including links by other modes (such as transit).

· When – Time, day of the week, day of the year, and conditions, such as weather, road conditions, and traffic conditions.

· Why – Purpose of trip. What factors affected travel choice (for example, would a cyclist have chosen another route or mode if road conditions or facilities were different).

Measuring Nonmotorized Travel Demand

Nonmotorized travel demand refers to how much people would use nonmotorized modes under various circumstances. A number of specific factors can affect demand for nonmotorized transport in a particular situation (Schwartz, et al, 1999; Porter, Shurbier, and Schwartz 1999; Moudon, 2001; Katz, 2001; Dill and Carr, 2003; Schneider, Patten and Toole, 2005; Raford and Ragland, 2006; McDonald, et al., 2007). These include:

· Attractions. Certain activity centers tend to be major attractors for walking and cycling, including commercial districts, school-college-university campuses, employment centers, recreation centers and parks.

· Trip distance. Most walking trips are less than a mile, and most bicycling trips less than 5 miles in length, although recreational trips are often much longer (Iacono, Krizek and El-Geneidy, 2008).

· Demographics. Young (10-20 years), elderly, and low-income people tend to rely more on walking for transport. Young and low-income people tend to rely on cycling for transport. Households with lower vehicle ownership rates tend to rely more on nonmotorized modes than those with one vehicle per driver.

· Land use patterns (density and mix). Walking and bicycling for transportation tend to increase with density (i.e., number of residents and businesses in a given area) because higher density makes these modes more efficient.

· Travel conditions. Wide roads with heavy, high-speed vehicle traffic can form significant barriers to nonmotorized travel. Special facilities for nonmotorized travel (sidewalks, wide curb lanes, and paths), their condition and connectivity can have a significant impact on the amount of walking and bicycling that occurs.

· Topography and climate. These factors can affect walking and bicycling, but not as much as might be expected. For example, the cities of Seattle, Portland and Missoula report significantly higher levels of cycle transportation than many “Sunbelt” cities that are flat and have mild climates.

· Community attitudes. Local attitudes can have a major impact on the level of cycling in a community. For example, it may be unremarkable that cycling tends to be high among college students and staff, but many college towns find that cycling is also relatively common among people who have not formal affiliation with the college simply because it has become an acceptable form of transportation. This indicates that some people hesitate to cycle, but will if they perceive it to be more socially acceptable.

· Time and geographic scope. It may take several years for a community to fully achieve its full nonmotorized travel potential. First year impacts are frequently modest, but tend to increase as individuals become more accustomed to nonmotorized travel and as additional support facilities (pedestrian and bicycle network, bicycle parking, etc.) develop.

Using available travel surveys Barnes and Krizek (2005a) estimate that on average roughly 1% of adults in the United States ride a bicycle during a particular day. Over large geographic areas such as metropolitan areas or states, this number ranges roughly between about 0.3% and 2.5%. Over smaller areas such as specific parts of metropolitan areas, the range could go as high as 15%. They concluding that total cycling can be estimated in a particular area as 0.3% plus 1.5 times the commute share.

Phillips, Karachepone and Landis (2001) and Katz (2001) describe various ways to estimate demand for motorized transportation. Transportation surveys such as the National Personal Transportation Survey (Transportation Statistics) provide information on nonmotorized travel. Rossi (2000) describes technical information on nonmotorized demand models developed in the Boston, Portland and Philadelphia regions. University of North Carolina (1994) and Clarke and Tracy (1995) summarize data from various studies on nonmotorized travel demand, including survey data from various communities, and discussion of factors that affect walking and cycling activity. Desyllas, et al (2003) used Multiple Regression Analysis to model pedestrian travel demand in the city of London, taking into account take walkway conditions, nearby land uses (trip generators), street network connectivity and transport accessibility (proximity to Tube stations and other transport terminals). McDonald, et al. (2007) develop a model for predicting nonmotorized travel demand and the impacts that on- and off-road walking and cycling facilities will have on the use of these modes.

Schneider, Patten and Toole (2005) describe nonmotorized travel surveys used in various communities, including manual counts, automated counts, surveys targeting non-motorized users, surveys sampling a general population, inventories, and spatial analyses. Cao, Handy and Mokhtarian (2005) used a travel survey performed in Austin, Texas to evaluate the effects of land use patterns on strolling trips (walking for pleasure or exercise) and utilitarian walking trips. The found that the pedestrian environment at the origin (home) has the greatest impact on strolling trips, while the pedestrian environment at the destination appears to be at least as important for utilitarian trips. They also found that people are more likely to stroll around or walk to the store when fewer vehicles travel residential and commercial streets. They found that strolling accounts for the majority of walking trips, and tends to be undercounted.

Petritsch, et al. (2008a and 2008b) develop models for predicting the increases in cycling activity, reductions in motorized travel, and resulting health benefits and energy conservation that is likely to result from cycling facility improvements. They find that a cycling network’s overall quality has a greater influence on the volume of cyclists in an area than any specific facility, indicating significant network effects.

Pucher and Buehler (2006) find that despite a colder climate, Canadians cycle about three times more than Americans. Reasons for this difference include Canada’s higher urban densities and mixed-use development, shorter trip distances, lower incomes, higher costs of owning, driving and parking a car, safer cycling conditions, and more extensive cycling infrastructure and training programs. The researchers point out that most of these factors result from differences in transport and land-use policies, and not from intrinsic differences in history, culture or resource availability. They suggest that it is possible to significantly increase cycling levels in the United States by adopting Canadian policies that have promoted cycling and enhanced its safety.

Latent Demand Score

The Latent Demand Score (LDS) is a relatively easy-to-use technique for estimating potential demand for bicycle travel. Landis (1996) describes the model and examples of its application. It is a probabilistic gravity model, similar to conventional four-step models, but uses a number of simplifying assumptions to reduce data requirements. It estimates the probability of bicycle travel on individual road or street segments based on area demographics and the proximity, frequency and magnitude of adjacent bicycle trip generators. Bicycle trip attractors/generators (employment, shopping, parks and schools) are geocoded and stratified according to proximity. Bicycle trips are predicted using the LDS equation, which has different bicycle trip probabilities versus distance for each trip purpose. This information allows planners to estimate how much bicycle use would increase if a bicycle facility were developed on a particular corridor.

Pedestrian Location Identifier

Moudon (2001) developed Pedestrian Location Identifier methodologies (PLIs), which use GIS and remote sensing tools to identify suburban, postwar areas that have land use characteristics likely to create substantial latent demand for pedestrian travel. Two tools were developed that use different sets of databases. The Pedestrian Location Identifier One, relies on readily available census data, GIS software, and aerial photographs. It is a manual method that requires an individual analyst’s judgment when comparing the data from the census to aerial photographs to delineate clusters. The Pedestrian Location Identifier Two uses parcel-level data with GIS software, which requires information that only MPOs, large cities, and some counties and states may have. It is an automated method, which uses the GIS software to define and analyze pedestrian locations.

National Survey of Pedestrians and Bicyclists Attitudes and Behaviors (www.walkinginfo.org/pdf/bikesurvey.pdf)

The 2002 National Survey of Pedestrians and Bicyclists Attitudes and Behaviors, which involved phone interviews with more than 9,600 adults age 16 and older throughout the U.S., found the following:

· Nearly 80% of adult Americans take at least one walk of five minutes or longer during the summer months, while fewer than 30% ride a bike, according to a major new survey released today by the U.S. Department of Transportation.

· Bicycling is most common among younger residents. Nearly 40% aged 16 to 24 ride a bicycle during the summer, 26% aged 45 to 54 cycle, but only about 9% of those age 65 and older report they cycle.

· Walking declines slightly as people age. Eighty-two percent of those aged 16 to 24 take walks, 80% aged 45 to 54 walk, and 65% aged 65 and older report taking walks.

· Only half of all adults are satisfied with their communities’ designs for bicycling safety, whereas three out of four adults are satisfied with their communities’ designs for pedestrian safety.

· Survey respondents were also asked to recommend changes to their communities for either bicycling or walking. Most persons suggested changes in bicycle and pedestrian facilities. For those recommending changes, 73% wanted new bicycle facilities, such as trails, bicycle lanes and traffic signal, and 74% wanted pedestrian facilities including sidewalks, lighting and crosswalks.

· People who do not take walks cite these reasons: disability or other health problems (25%); unfavorable weather (22%); and too busy or no opportunity (19%). Those who do not bike cite lack of access to a bicycle (26%); too busy or no opportunity (17%); disability or other health problems (10%).

· Males are more likely to take a bike ride during the summer than are females. However, both groups are about equally likely to take walks during the summer.

Mode Shifts

The benefits of a nonmotorized program are affected by their travel impacts, including increases in nonmotorized travel and reductions in motorized vehicle travel. Shifts from automobile to nonmotorized modes are measured by mode substitution rates, that is, the ratio between increased nonmotorized person-miles and reduced motor vehicle-miles.

When automobile travel is reduced in response to disincentives such as increased vehicle fees or vehicle restrictions, a significant portion (typically 10% to 50%) of reduced trips shift to nonmotorized modes (Transport Elasticities). Shorter trips (less than three miles) shift to nonmotorized modes, and longer trips shift to combined transit and nonmotorized trips. For example, when UK residents were asked how they could reduce short (less than 8 kms) vehicle trips, respondents indicated they could shift 31% of these trips to bus, 31% to walking, and 7% to bicycle (Mackett, 2001). After Canadian fuel prices increased about 15% in 2001, a federal Competition Bureau survey found that about a quarter of motorists shifted some automobile travel to other modes, of which 46% took transit, 36% walked, 24% cycled, and 20% shared car rides. Parking Cash Out (allowing commuters to exchange a free parking space for cash) caused a 13-point reduction in automobile trips, a 9-point increase in carpooling, a 9-point increase in transit use, and a 1-point increase in nonmotorized commute trips.

When nonmotorized travel increases due to improved conditions, not all new walking and cycling trips substitute for automobile trips, some reflect increased total travel (including recreational trips) or shifts from transit or ridesharing. Typically, 20% to 50% of increased nonmotorized travel substitutes for motorized travel, depending on conditions.

In addition to person-miles shifted from motorized to nonmotorized travel, increased nonmotorized transportation tends to leverage additional vehicle travel reductions. A short walking or cycling trip often replaces a longer automobile trip, for example, people may choose between walking to a nearby store or driving to a more distant shopping center. Pedestrians and cyclists often use shortcuts unavailable to drivers. When people shift to nonmotorized travel for a particular trip, or when households reduce their vehicle ownership due to improved nonmotorized conditions, they tend to reduce their total vehicle mileage by avoiding discretionary trips. Nonmotorized transport supports Smart Growth land use patterns (more compact, mixed, multi-modal development) that reduce travel distances and total motorized travel.

Figure 1 shows average per capita annual vehicle mileage in U.S. cities categorized by nonmotorized commute mode split. As nonmotorized travel increases, vehicle mileage declines. Although nonmotorized mode split is small (representing less than 5% of trips and probably less than 1% of person-miles, since nonmotorized trips average less than a fifth the distance of motorized trips), the mileage differences are large. Each percentage point increase in nonmotorized transport is associated with about 700 fewer annual vehicle-miles. Assuming commute mode split is representative of total personal travel, urban residents average 10,000 annual person-miles, and nonmotorized trips average one mile in length, each nonmotorized mile is associated with seven reduced vehicle-miles.

Figure 1 U.S. Nonmotorized Vs. Motorized Transport (Census and FHWA Data, 2000)

Per capita vehicle mileage tends to decline as nonmotorized travel mode split increases.

International data also indicate that increased nonmotorized transport is associated with reduced driving, as indicated in Figure 2. Of course, association does not prove causation. Not every walking or cycling trip causes seven miles of reduced driving. The lower vehicle mileage in cities with relatively high nonmotorized mode split reflects various land use and transport system factors, such as density, mix, street design, parking supply, and pricing which affect the relative attractiveness of motorized and nonmotorized travel. But programs that increase nonmotorized travel tend to create such communities, which is to say that smart growth supports nonmotorized travel and nonmotorized travel supports smart growth. As a result, mobility management programs that increase nonmotorized transport usually leverage reduced motorized travel, causing proportionately larger reduction in vehicle-miles, although exactly how much depends on the situation.

Figure 2 Nonmotorized Vs. Motorized Transport (Kenworthy and Laube, 2000)

International data show that vehicle travel tends to decline as nonmotorized travel increases.

These leverage effects probably apply only to nonmotorized travel used for transportation purposes, not to recreation walking and cycling. For mobility management evaluation an important question is the degree that factors that can be changed through public policies can increase nonmotorized travel and leverage reductions in motorized travel in the short or medium term. If higher nonmotorized transport and lower motor vehicle mileage in different geographic areas completely reflect the legacy of patterns established decades earlier, it may be futile to try to change them. However, at least some of factors can be changed by public policies in medium-term, including nonmotorized facility quality, traffic management practices, financial incentives (such as road and parking pricing) and public information and attitudes can be changed in the short term, and other factors, such as the location of public facilities, the design of new buildings, and community redevelopment practices. Many communities have experienced significant nonmotorized travel growth and reductions in nonmotorized travel over a few years due to policy changes and mobility management programs (Success Stories).

Some experts conclude that walking and cycling can do little to solve transportation problems because they only consider current commute trips that can shift completely to nonmotorized modes (Comsis, 1993). But other studies give more positive assessments of potential travel impacts. According to some studies, 5-10% of urban automobile trips can reasonably be shifted to nonmotorized transport (ADONIS, 1999; Mackett, 2000; Socialdata Australia, 2000; Cairns et al, 2004).

Figure 3 Urban Mode Split (Pucher and Lefevre, 1996)

This figure shows the portion of urban travel by different modes in various countries. Nonmotorized travel varies significantly from one country to another.

For example, the Australian TravelSmart program uses various incentives to encourage residents to use alternative travel modes. Before-and-after surveys find that automobile trips decline by 5% to 14%, and that about half of these reductions result from shifts to nonmotorized travel. Rates of nonmotorized travel vary significantly from one community to another, depending on land use patterns, transportation system design factors, and community attitudes, as indicated in Figure 3. Even relatively cold and hilly countries, such as Sweden, Switzerland and Germany achieve high levels of nonmotorized travel. Similarly, some North American communities have much higher rates of nonmotorized travel due to supportive Land Use policies.

Most communities appear to have significant latent demand for pedestrian travel, that is, people would walk more frequently if they had suitable facilities and resources. One US survey found that 38% of respondents would like to walk to work, and 80% would like to walk more for exercise (STPP, 2003). The table below summarizes a Canadian public survey indicating high levels of interest in cycling and walking.

Table 4 Active Transportation Survey Findings (Environics, 1998)

	Cycle	Walk
Currently use this mode for leisure and recreation.	48%	85%
Currently use this mode for transportation.	24%	58%
Would like to use this mode more frequently.	66%	80%
Would cycle to work if there “were a dedicated bike lane which would take me to my workplace in less than 30 minutes at a comfortable pace.”	70%	NA
Support for additional government spending on bicycling facilities.	82%	NA

This survey indicates a high level of interest in cycling and walking.

Evaluating Existing Conditions – General Techniques

Field Surveys

Various methods are used to evaluate existing walking and cycling conditions (Moudon and Lee, 2003; Roberts-James, 2003). Planners, community members and public officials can walk around an area to survey of walking and cycling conditions. A Walkable Places Survey, described below, is an organized framework for involving community members in field surveys. Some transportation agencies use volunteers or hired college students to perform field surveys. Below are some of the features that should be evaluated in site surveys. Also see the Level of Quality Guidelines (Burden, 2003), which illustrate how specific roadway conditions affect walking, bicycling, traffic calming, transit access and street crossings.

Field Survey Data to Collect

· Nonmotorized traffic volumes and speeds.

· Sidewalk, path, and trail conditions (effective width, surface condition, sight distances, etc.).

· Security, cleanliness, vandalism, litter, and aesthetic conditions.

· Public washrooms and other services along trails and bike routes.

· Curb cuts, ramps and other universal access facilities.

· Pedestrian road crossing facilities.

· Vehicle traffic volumes and speeds.

· Lighting along streets and paths.

· Special hazards to walking and cycling.

· Roadway and road shoulder widths and pavement conditions (for cycling).

· Presence of parked cars adjacent to the traffic lane.

· Presence of potholes and dangerous drain grates.

· Bicycle Parking.

When evaluating facilities it is important to maintain a distinction between nominal (“in name”) and functional (“actual condition”) dimensions. For example, many sidewalks and paths are nominally 1.8 to 2 meters wide, but functionally they may be much narrower, due to objects such as telephone poles and signposts, and surface failures such as cracks and potholes. As a result, a walkway that meets technical specifications may be inadequate for some potential users (particularly wheelchair users and people with strollers). Similarly, a bike lane may be useless if it has poor surface conditions or is frequently used for vehicle parking.

Evaluating Current Policies and Practices

The efficiency of walking and cycling transportation is highly affected by land use factors such as land use mix, street Connectivity, and site design (see Land Use Impacts on Transportation), local street design standards and development ordinances should be evaluated. Current zoning codes, such as minimum parking and lot size requirements, tend to discourage pedestrian-oriented design (New Urbanism), and current planning practices often undervalue nonmotorized safety and mobility impacts (Elvik, 2000; Hillman, 2001; Goodman and Tolley, 2003).

Transportation plans, planning practices and municipal budgets can be evaluated to determine whether they give nonmotorized travel sufficient consideration:

· Do transportation plans include walking and cycling improvement components?

· Do transportation surveys and Models incorporate data on nonmotorized travel, including short trips, off-peak trips, non-motorized components of linked trips and travel by children?

· Is there a pedestrian/bicycle planner or program coordinator within the transportation agency?

· Do nonmotorized improvements receive adequate funding, taking into account the role of nonmotorized as a form of basic mobility, as a complement to transit travel, and as a form of recreation?

Safety

Transportation Risk and Safety can be evaluated in several different ways, which tend to give different conclusions about the relative safety of walking and cycling. When measured per unit of travel (per mile or kilometer), nonmotorized travel has about tend times the fatality rate as driving. But the health risk from nonmotorized travel is less than these estimates indicate because:

Non-motorized travel imposes minimal risk to other road users.
Non-motorized transport encourages shorter trips and land use patterns that reduce travel distances over the long term. Put another way, when people shift from nonmotorized to motorized travel they often increase their total annual travel, and therefore their exposure to crash risk.
Bicycling offers significant Health Benefits that offset accident risk.
As nonmotorized travel becomes more common in a community, crash rates tend to decline, apparently because drivers become more cautious and safer facilities are provided.

Empirical evidence indicates that shifts from driving to nonmotorized modes can reduce total per capita crash risk. Jacobsen (2003); Wittink (2003); Robinson (2003); and Turner, Roozenburg and Francis (2006) found that per capita collisions between nonmotorized travelers (pedestrians or cyclists) and motor vehicles decline in areas with higher rates of nonmotorized travel suggesting that drivers become more cautious when they see more walkers and cyclists. Jacobsen calculates that the number of motorists colliding with pedestrians and cyclists increases at roughly 0.4 power of the number of people walking or cycling (e.g., doubling NMT travel in a community will increase pedestrian/cycling injuries by 32%), and the probability that a motorist will strike a nonmotorized traveler declines with the roughly -0.6 power of the number of people walking and cycling in a community (e.g., as a pedestrian, my risk of being hit by a motor vehicle declines 34% if walking and cycling double in my community). Robinson (2005) found similar results using Australian data: doubling bicycle travel reduces cyclist risk per kilometer by about 34%; and conversely, halving bicycle travel increases risk per kilometer about 52%.

The San Francisco Department of Public Health developed an Vehicle-Pedestrian Injury Collision Model which predicts how demographic, geographic and land use planning factors affect the number of collisions resulting in pedestrian injury or death in an area (SFDPH, 2008a). The model indicates that pedestrian injuries and deaths increase with motor vehicle traffic volume, vehicle traffic speeds, pedestrian volume, and various intersection and street design factors.

Level of Service Ratings

Transportation facilities have traditionally been evaluated using Level of Service (LOS) ratings that range from A (best) to F (worst, or failure). Multi-Modal Level-of-Service rating systems indicate the convenience and comfort of other modes. Several LOS ratings have been developed for pedestrian and cycling facilities (Henson, 2000; Phillips, Karachepone and Landis, 2001; FDOT, 2002; Moudon and Lee, 2003; Hiatt, 2006; FHWA, 2006). Below are some examples of these ratings.

Walkability Checklist

The Walkability Checklist: How Walkable Is Your Community, by Partnership for a Walkable America and the Pedestrian and Bicycle Information Center (www.walkableamerica.org/checklist-walkability.pdf), provides an easy-to-use form for evaluating neighborhood walkability.

Bikeability Checklist

The Pedestrian and Bicycle Information Center (www.bicyclinginfo.org) produced a community bikeability checklist (www.walkinginfo.org/cps/checklist.htm). It includes ratings for road and off-road facilities, driver behavior, cyclist behavior, and barriers, and identifies ways to improve bicycling conditions.

Shared-Use Path Level-Of-Service Ratings

The U.S. Federal Highway Administration (FHWA, 2006) developed a Shared-Use Path LOS (SUPLOS) model, which is a mathematical formula that uses select inputs describing conditions along a trail to calculate an LOS score. This is based on detailed research that included the creation of path traffic flow theory, an extensive effort to collect data on path operations, and a survey during which path users expressed their degree of satisfaction with the paths shown on a series of videos.

The resulting method requires minimal input and produces a simple and useful result. The method requires only four inputs from the user: One-way user volume in the design hour, mode split percentages, trail width, and presence or absence of a centerline. Factors involved in the estimation of an LOS for a path include the number of times a typical bicyclist meets or passes another path user and the number of those passes that are delayed. The method considers five types of path users when calculating adult bicyclists' LOS, including other adult bicyclists, child bicyclists, pedestrians, runners, and in-line skaters. The FHWA provides step-by-step instructions on how to use the LOS procedure and spreadsheet calculation tool.

The basic SUPLOS model equation is (see the guidebook for more detailed information and cautions on using this method):

SUPLOS = 5.446 – 0.00809(E) – 15.86(RW) – 0.287(CL) – (DPF)

Where:

E = Events = Meetings per minute + 10 (active passes per minute)

RW = Reciprocal of path width (i.e., 1/path width, in feet)

CL = 1 if trail has a centerline, 0 if trail has no centerline

DPF = Delayed pass factor

The SUPLOS model generates a LOS score between zero and five.

The resulting SUPLOS scale can be converted to letter grades. An A is the highest score, excellent, and an F is the lowest score.

LOS Score

X ≥ 4.0 = A

3.5 ≤ X < 4.0 = B

3.0 ≤ X < 3.5 = C

2.5 ≤ X < 3.0 = D

2.0 ≤ X < 2.5 = E

X < 2.0 = F

Interpreting LOS grades.

A: Excellent. Trail has optimum conditions for individual bicyclists and retains ample space to

absorb more users of all modes, while providing a high-quality user experience. Some newly

built trails will provide grade-A service until they have been discovered or until their

ridership builds up to projected levels.

B: Good. Trail has good bicycling conditions, and retains significant room to absorb more users,

while maintaining an ability to provide a high-quality user experience.

C: Fair. Trail has at least minimum width to meet current demand and to provide basic service

to bicyclists. A modest level of additional capacity is available for bicyclists and skaters;

however more pedestrians, runners, or other slow-moving users will begin to diminish LOS

for bicyclists.

D: Poor. Trail is nearing its functional capacity given its width, volume, and mode split. Peakperiod

travel speeds are likely to be reduced by levels of crowding. The addition of more

users of any mode will result in significant service degradation. Some bicyclists and skaters

are likely to adjust their experience expectations or to avoid peak-period use.

E: Very Poor. Given trail width, volume, and user mix, the trail has reached its functional

capacity. Peak-period travel speeds are likely to be reduced by levels of crowding. The trail

may enjoy strong community support because of its high usage rate; however, many

bicyclists and skaters are likely to adjust their experience expectations, or to avoid peakperiod

use.

F: Failing. Trail significantly diminishes the experience for at least one, and most likely for all

user groups. It does not effectively serve most bicyclists; significant user conflicts should be

expected.

Highway Capacity Manual

The Highway Capacity Manual published by the Transportation Research Board is a basic reference widely used by transportation planners and engineers for evaluating roadway conditions. It provides Level-of-Service ratings for roads and intersections based on traffic density (an indicator of congestion). The same approach was applied to pedestrian Level-of-Service ratings, as illustrated in Figure 4, but this has been criticized as being inadequate, since it is based on just one variable.

Figure 4 Pedestrian Level of Service (TRB 1997)

This figure illustrates an early pedestrian LOS ratings based only on density. This has been criticized as being too simplistic.

The 2000 Highway Capacity Manual incorporates a number of factors into pedestrian Level-of-Service ratings for roadway crossings, reflecting pedestrian delay, as indicated in Table 5. Kim, et al. (2008) evaluate the impacts that various types of street furniture (benches, bicycle racks, planter boxes, trees, mail boxes, brochure bins, trash cans, vending and coffee carts, and tables and chairs) have on pedestrian level-of-service, depending on their type, size, shape and use. They recommend specific design and management practices based on type of furniture, sidewalk width, pedestrian volumes, and the potential number of users or customers.

Table 5 Pedestrian Road Crossing Level of Service (Milazzo, et al, 1999)

Level of Service	Signalized Intersection	Unsignalized Intersection	Likelihood of Pedestrian Noncompliance
A	<10	< 5	Low
B	10-20	5-10
C	20-30	10-20	Moderate
D	30-40	20-30
E	40-60	30-45	High
F	60+	45+	Very High

Average Delay Per Pedestrian in Seconds. Crosswalk walking speeds are estimated at 1.2 meters per second for most areas, and 1.0 m/s for crosswalks serving large numbers of older pedestrians.

Dixon

Dixon (1996) describes LOS ratings for walking and cycling conditions. The ratings take into account the existence of separated facilities, conflicts, speed differential, congestion, maintenance, amenities, and TDM. These are relatively easy to use methods for evaluating non-motorized roadway conditions that may be more practical than other methods that are more data intensive.

Tables 6 and 7 summarize a simplified method for evaluating walking and cycling level-of-service. The results are then scored to determine the LOS rating in Table 8.

Table 6 Pedestrian Level-of-Service (Dixon, 1996)

Pedestrian

Points

Facility

(Max. value = 10)

Not continuous or non-existent

Continuous on one side

Continuous on both sides

Min. 1.53 m (5’) wide & barrier free

Sidewalk width >1.53 (5’)

Off-street/parallel alternative facility

Conflicts

(Max. value = 10)

Driveways & sidestreets

Ped. Signal delay 40 sec. or less

Reduced turn conflict implementation

Crossing width 18.3 m (60’) or less

Posted speed

Medians present

0.5

Amenities

(Max. value = 2)

Buffer not less than 1m (3’5”)

Benches or pedestrian scale lighting

Shade trees

0.5

Motor Vehicle LOS

(Max. value = 2)

LOS = E, F, or 6+ travel lanes

LOS = D, & < 6 travel lanes

LOS = A, B, C, & < 6 travel lanes

Maintenance

(Max. value = 2)

Major or frequent problems

Minor or infrequent problems

No problems

-1

TDM/Multi Modal

(Max. value = 1)

No support

Support exists

Table 7 Bicycle Level-of-Service (Dixon, 1996)

Bicycle

Points

Facility

(Max. value = 10)

Outside lane 3.66 m (12’)

Outside lane 3.66-4.27m (12-14’)

Outside lane >4.27m (14’)

Off-street/parallel alternative facility

Conflicts

(Max. value = 10)

Driveways & sidestreets

Barrier free

No on-street parking

Medians present

Unrestricted sight distance

Intersection Implementation

0.5

Speed Differential

(Max. value = 4)

>48 KPH (>30 MPH)

40-48 KPH (25-30 MPH)

24-30 KPH (15-20 MPH)

Motor Vehicle LOS

(Max. value = 2)

LOS = E, F, or 6+ travel lanes

LOS = D, & < 6 travel lanes

LOS = A, B, C, & < 6 travel lanes

Maintenance

(Max. value = 2)

Major or frequent problems

Minor or infrequent problems

No problems

-1

TDM/Multi Modal

(Max. value = 1)

No support

Support exists

Table 8 Level of Service Ratings

LOS Rating	Points
A	>17
B	>14-17
C	>11-14
D	>7-11
E	>3-7
F	3 or less.

Bicycle Compatibility Index

Harkey, et al (1998) describes the Bicycle Compatibility Index, a practical tool for evaluating the suitability of urban and suburban roadways for cycling. It incorporates curb lane width, traffic volumes and traffic speeds. The report describes how to gather the necessary data and apply the method, discusses case study examples, and includes a spreadsheet model to facilitate the analysis. Tables 9 and 10 show how this information is evaluated to indicate cyclist stress level and suitability ratings for a specific roadway.

Table 9 Roadway Cyclist Stress Level Values

Stress Rating	Speed	Volume	Trucks	Curb Lane	Hindrances
	Posted speed limit (km/hr)	Vehicles/hr per traffic lane	Percentage of truck traffic	Curb lane width (m)	Commercial driveways and intersections per km
1	<40	<50	<2%	>4.6	<6
2	50	51-150	4%	4.3	13
3	60	151-250	6%	4.0	19
4	65	251-350	8%	3.7	25
5	>75	351-450	>10%	<3.3	>31

These values are used to calculate Cycling Suitability Rating in Table 8.

Table 10 Cycling Suitability Rating

Summed Values	Average Stress Level	Road Suitability for Cycling
< 7	1	Road is reasonably safe for all types of cyclists.
7-12	2	Road accommodates casual and experienced cyclists, but needs improvement to accommodate child cyclist.
13-17	3	Road accommodates experienced cyclists, but needs improvement to accommodate casual and child cyclists.
18-22	4	Needs improvements to accommodate experienced cyclists, not recommended for casual and child cyclists.
>22	5	May be unsuitable for all cycling.

This table indicates a roadways Cycling Suitability Ratings, and the type of cycling that it can accommodate.

Additional factors considered in the Bicycle Compatibility Index are listed below:

· Presence of bicycle lane or paved shoulder.

· Bicycle lane or paved shoulder width.

· Curb lane width.

· Curb lane volume.

· Other lane volume.

· Average traffic speed.

· Presence of parking lane with more than 30% occupancy.

· Type of roadside development.

· Truck volumes.

· Parking turnover.

· Right turn lanes.

Cycle Demand and Hazard Scoring

Landis (1996) describes relatively easy-to-use techniques for estimating potential bicycle travel demand (the Latent Demand Score), and evaluating roadway conditions for cycling in a particular area (the Interaction Hazard Score). These are similar to other models used by traffic engineers that require various demographic, geographic and road condition information. This information allows planners to estimate how much bicycle use would increase if a bicycle facility were developed on a particular corridor.

Community Cycling Conditions

Most cycling Level-of-Service rating systems are designed to evaluate conditions on a specific road. Other systems evaluate overall cycling conditions in a community. IHT (1998) describes how to perform a cycle audit and cycle review, which are standard frameworks for evaluating roadway conditions and transportation plans in terms of their suitability for cycling.

Below are some additional factors to consider when evaluating the quality of cycling in a community.

Cycling rates (per capita annual cycling trips or cycling mode split).
Cycling fatality and injury rates.
Availability and quality of Bicycle Parking and changing facilities.
Bike/Transit Integration, including the ability to carry bikes on transit vehicles, and the availability of bike parking and rental services at train and bus stations.
Quality of Bicycle Education and Encouragement Programs.
Degree to which common destinations (schools, commercial centers, transit stations, etc.) are located and designed for cycling access.

Acceptable Walking Distance

The distance that people are willing to walk is often an important factor in transportation and land use planning. It determines the optimal size of a commercial district or urban village, the area served by a particular public transit service, and the acceptable distance between parking facilities and destinations. This can be called a walkable area or a ped-shed (based on watershed).

The table below indicates Level of Service ratings for pedestrian access. For typical urban conditions, LOS A is less than one block, LOS B is 1-4 blocks, LOS C is 4-8 blocks, and LOS D is more than 8 blocks between a destination and its parking facilities.

Table 11 Level of Service By Walking Trip Distance (in Feet) (Smith and Butcher, 1997)

Walking Environment	LOS A	LOS B	LOS C	LOS D
Climate Controlled	1,000	2,400	3,800	5,200
Outdoor/Covered	500	1,000	1,500	2,000
Outdoor/Uncovered	400	800	1,200	1,600
Through Surface Lot	350	700	1,050	1,400
Inside Parking Facility	300	600	900	1,200

Acceptable walking distances are affected by degree of weather protection, climate, line of site (whether pedestrians can see their destination), and “friction” (interruptions and constraints along the way, such as cross traffic). Of course, people’s abilities and preferences also vary; some may be able to walk much father than others, so it is important to accommodate people with mobility constraints in pedestrian planning. For example, it may be appropriate to reserve some parking spaces close to destinations for people with disabilities and delivery vehicles, so they have shorter walking distances.

Walking Speeds

Walking speeds are an important factor in the design and management of walking facilities, particularly traffic signals. Pedestrian signals that provide inadequate crossing time can create danger and discomfort for slower pedestrians. Walking speeds vary depending on various demographic and geographic factors. Healthy adults typically walk 4.0 feet (1.2 meters) per second or faster, but lower walking speeds are common for older and younger people, people with disabilities, people carrying baggage or pushing handcarts, or when walking on rough surfaces or up a hill. The Institute of Transportation Engineers recommends that traffic signal timing be based on maximum walking speeds of 3.0 feet (0.9 meters) per second to safely accommodate slower pedestrians (LePlante and Kaeser, 2004).

Krizek, et al. (2007) developed methods for calculating travel times by walking, cycling and public transit modes. The researchers used information on networks and speeds to construct a series of maps that graphically depict various non-auto travel networks at different points in time between 1995 and 2005. The maps break down origins and destinations into several zones (similar to watersheds). This technique makes it possible to see changes in travel time between different “travel-sheds” over time.

Walking Security Index

Wellar (1998) uses a Walking Security Index to evaluate pedestrian crossing conditions at roadway intersections, taking into account a wide range of variables that affect pedestrian safety, comfort, and convenience, as summarized in Table 12. This indicates that increased road width, traffic volumes, traffic speeds, vehicle mix, and various other factors affect the mobility, safety and comfort of pedestrian travel.

Table 12 Walking Security Index Variables

Infrastructure

Vehicle Traffic

Pedestrian

Performance

Behavior

1. Number of lanes.

2. Speed

3. Grade (incline).

4. Turning lanes.

5. Curb cut at intersections.

6. Stop bar distance from crosswalk.

7. Sight lines

8. Peak vehicle volumes.

9. Vehicle types.

10. Trip purpose.

11. Turning movements.

12. Pedestrian volumes.

13. Pedestrian age.

14. Right-turn-on-red.

15. Signage.

16. Ice/snow/slush removal.

17. Pedestrian-vehicle collisions.

18. Pedestrian-vehicle conflicts.

19. Vehicle moving violations.

Other researchers emphasize that street crossings must be properly located, that is, they must follow pedestrians’ natural walking patterns and sight lines, without requiring extra walking distance (Stonor, Beatriz de Arruda Campos and Smith, 2001). “Faced with a badly located and badly designed crossing, pedestrians often do one of two things: either they do not cross, and remain instead on one side of the road (with economic consequences for two-sided retailing) or they cross, but do not use the crossing (with road safety consequences).”

Barrier Effect (Severance)

The Barrier Effect (also called Severance), refers to the tendency of roads and traffic to create a barrier to nonmotorized travel (Litman, 2002). Severance usually refers only to the impacts of a highway facility itself, while the barrier effect refers to the combined impacts of the roadway and vehicle traffic, and so increases with traffic volumes. It represents a degradation of the pedestrian and bicyclist environment that reduces the viability of these modes, often leading to increased driving. This is not to imply that drivers intentionally cause harm, but rather that such impacts are unavoidable when fast, heavy and hard vehicles share space with vulnerable road users. Although it could be argued that impacts are symmetrical, because nonmotorized modes cause traffic delays to motorists, pedestrians and cyclists impose minimal accident risk, noise and dust on motorists so the costs they bear are inherently greater then the costs they impose.

Scandinavian Barrier Effect Models

Both the Swedish National Road Administration (Reyier, 1986) and the Danish Road Directorate (1992) roadway investment evaluation models incorporate methods for quantifying barrier effects on specific lengths of roadway. Both involve two steps. First, a barrier factor is calculated based on traffic volumes, average speed, share of trucks, number of pedestrian crossings, and length of road way under study. Second, the demand for crossing is calculated (assuming no barrier existed) based on residential, commercial, recreation, and municipal destinations within walking and bicycling distance of the road. The Swedish model also adjusts the number of anticipated trips based on whether the road is in a city, suburb, or rural area, and the ages of local residents.

Other Barrier Effect Modeling

Rintoul (1995) describes the development and application of a model for evaluating the barrier effect. He calculates that a 5.3 kilometer stretch of major highway crossing through a medium size city imposes barrier effect costs of $2.4 million Canadian annually, or about 83¢ per capita each day. This highway carries 13,600 average annual daily trips, so this cost averages about 8.7¢ Canadian per vehicle kilometer. Kjartan Sælensminde (1992) estimates that the total cost of the barrier effect in Norway equals $112 per capita annually (averaging about 1¢ per vehicle mile), which is greater than the estimated cost of noise, and almost equal to the cost of air pollution.

F.N. Tate (1997) evaluates various ways to quantify the barrier effect, and proposes that this can be measured by asking parents whether they would be willing to allow a child to cross a street unaccompanied, under various road and traffic conditions. Clark and Hutton (1991) describe methods for assessing highway traffic impacts on pedestrian and bicyclist mobility.

Crossing Ratios

John Russell and Julian Hine suggest that the barrier effect can be evaluated using “crossing ratios,” which is the number of pedestrians who cross a road as a portion of total pedestrian flow along that segment. They argue that a low crossing ratio typically indicates the barrier effect, although other factors may also influence such behavior.

Walkability

Walkability reflects the overall support for pedestrian travel in an area. Walkability takes into account the quality of pedestrian facilities, roadway conditions, land use patterns, community support, security and comfort for walking. Walkability can be evaluated in various ways and at various scales (Nabors, et al., 2007). At a site scale, walkability is affected by the quality of pathways, building accessways and related facilities. At a street or neighborhood level, it is affected by the existence of sidewalks and crosswalks, and roadway conditions (road widths, traffic volumes and speeds). At the community level it is also affected by land use Accessibility, roadway Connectivity, such as the relative location of common destinations and the quality of connections between them. Walkability takes into account:

· Pedestrian network quality (quality of paths, sidewalks, street crossings).

· Pedestrian network connectivity (how well sidewalks and paths are connected, and how directly pedestrians can travel to destinations).

· Security (how safe people feel while walking).

· Density and accessibility (distance between common destinations, such as homes, shops, schools, parks).

For example, a busy suburban arterial can have a high pedestrian LOS rating, provided it has sidewalks and pedestrian crossings at intersections, although walking is actually quite difficult and impractical as a form of transportation due to the wide road widths and dispersed land use patterns. Walkability can be enhanced by increasing Clustering and land use mix, by creating pedestrian shortcuts and mid-block pedestrian connections, and by locating commercial buildings close to the sidewalk, rather than being set back behind large parking lots. Walkability is also concerned with the ability to stop in the public right-of-way, for example, to rest, enjoy a viewpoint or shop window, have a conversation or play. Pedestrian Level of Service standards do not encompass these factors, although they are critical to the overall utility of walking as a form of transport.

WalkScore (www.WalkScore.com) automatically calculates a neighborhood’s walkability rating by identifying the distance to public services such as grocery stores and schools. www.WalkScore.com uses Google maps and business listings. It works for any street address in the United States of America and Canada, assigning points based on the distance to local amenities, then averages the score.

The Walkability Tools Research Website (www.levelofservice.com) provides detailed information on methods for evaluating walking conditions and identifing areas of poor walkability and walking limitations, proposed improvements and funding priorities (Abley, 2005). It applies these techniques:

· A Community Street Audit which qualitatively evaluates the quality of public spaces – streets, housing estates, parks and squares from the viewpoint of the people who use it.

· A Walkability Rating System that quantitatively evaluates walking conditions in order to identify problem areas.

Loukopoulos and Gärling (2005) find that on average people will drive rather than walk for a distance over 1,236 meters, with higher walking thresholds for women, and people who frequently walk, and lower values for more difficult walking conditions and people who frequently drive. The authors conclude that improving walking conditions and marketing campaigns can decrease the frequency of short automobile trips.

Defining “Walkable Community”

By Dan Burden of Walkable Communities (www.walkable.org).

A “walkable community” is designed for people, to human scale, emphasizing people over cars, promoting safe, secure, balanced, mixed, vibrant, successful, healthful, enjoyable and comfortable walking, bicycling and human association. It is a community that returns rights to people, looks out especially for children, seniors and people with disabilities and takes aggressive action to reduce the negative impacts of sixty-plus years of auto-centric design and uncivil driving practices. It is also a community that emphasizes economic recovery of central neighborhoods, promotes the concepts of recovering and transforming suburban sprawl into meaningful villages, and especially takes ownership and action to protect and preserving open space.

A walkable community, like a livable community, smart growth community, or sustainable community, makes a neighborhood, hamlet, village, town, city or metropolis into a place where many people walk, ride bicycles and use transit, and where anyone who drives a car moderates their behavior in a way where they take nothing from the rights of those who wish to stay healthy and active by taking part in activities outside the car.

A walkable community is one that is old, historic, well worn, restored sensibly and worthy of protection. A walkable community is one that is compact, new, fresh, invigorating and teaming with people enjoying their streets, parks, plazas, buildings and other physical space.

Below are ten indicators of prosperous, walkable, healthy and livable communities:

Compact, lively town center.
Many linkages to neighborhoods.
Low speed streets.
Neighborhood schools and parks.
Public places packed with children, teenagers, adults and people with disabilities.
Convenient, safe and easy street crossing.
Inspiring and well-maintained public streets.
Land use and transportation mutually beneficial.
Celebrated public spaces and public life.
Many people walking.

Also see “Key Principles of Building Healthy Communities,” Building Communities With Transportation: Distinguished Lecture Presentation, Transportation Research Board, Walkable Communities (www.walkable.org/download/TRBpaper.doc), January 10, 2001.

Nabors, et al. (2007) review various methods for evaluating walkability. Some of these methods are described below.

Pedestrian Environmental Factor

PBQD (1993) describes how to evaluate urban roadway pedestrian conditions based on Pedestrian Environmental Factors (PEF). The four criteria below are each rated on a scale from 1-3, the total of which represents the Pedestrian Environmental Factor (PEF). The results were found to correlate well with the use of non-automobile travel in an urban area. Urban neighborhoods with a high PEF tend to have twice the walk/bicycle mode share as the overall average, as much as five times greater than areas with the lowest PEF.

Pedestrian Environmental Factors

· Ease of street crossings. This is based on street width, traffic volumes, and speeds.

· Sidewalk continuity. Sidewalks that do not connect create barriers to pedestrian travel. A pedestrian network is only as good as its weakest link, particularly for people with physical disabilities. Even problems that appear minor to able-bodied pedestrians may be a major barrier to people with significant mobility constraints.

· Local street characteristics (grid vs. cul de sac). A grid street system provides continuity, allowing more direct access to destinations.

· Topography. Steep slopes create barriers to pedestrians.

Walkability Audit Tool

CDC (2004) provides a walkability audit tool which consists of an evaluation form, shown below, for rating a particular travel segment or area in terms of eight factors, with higher weights for factors considered more important. A total rating of 70-100 is considered good, ratings of 40-69 are considered medium, and a rating under 40 is considered poor.

Walkability Audit Tool

A. Pedestrian Facilities (High Importance): Presence of a suitable facility, such as a walking path or sidewalk.

No facility – pedestrians walk on road or dirt path.

Paved walkway on one side of road, minor discontinuities that present modest barrier to walking.

Continuous paved walkway on both sides of road or completely separated from roadway.

B. Pedestrian Conflicts (High Importance): potential for conflict with motor vehicle traffic due to driveways, high speed and volume traffic, large intersections, poor pedestrian visibility, etc.

High conflict potential

Low conflict potential.

C. Crosswalks (High Importance) presence and visibility of crosswalks at intersecting roads. Traffic signals have functional ‘walk’ lights that provide sufficient crossing time.

Crosswalks not present despite large intersections.

No intersections, or crosswalks clearly marked

D. Maintenance (Medium Importance): buckling pavement, overgrown vegetation, standing water, etc.

Major or frequent problems.