Key Transportation Indicators: Summary of a Workshop (2002)

A basic point of departure for the development of system performance indicators is to understand the phenomenon being measured. The concept of mobility has been interpreted in a variety of ways, some inappropriately. In addition, the scale of application could have an important influence on the definition of mobility. For example, mobility at the metropolitan level might be defined differently than a mobility measure at the corridor or subarea level, which itself might be very different from mobility as perceived by an individual traveler. For purposes of this subgroup, the following working definition of mobility was used to guide the discussion:

Mobility: Mobility refers to the time and costs required for travel. Mobility is higher when average travel times, variations in travel times, and travel costs are low. Indicators of mobility are indicators of travel times and costs and variability in travel times and costs.

An important distinction to keep in mind (as noted by consultant Alan Pisarski) is that there are people indicators and there are system indicators. System speeds and individual travel times can in fact measure the same phenomenon but have different meanings.

Mobility indicators can serve many purposes. They should provide trend information from which implications for transportation can be drawn or from which transportation policy and investment decisions are made. They can provide a basis for comparisons among metropolitan areas. They can provide the public with a sense of whether system performance is improving or getting worse. And in conjunction with intelligent transportation system strategies, they could be used on a real-time basis to inform travelers of current “mobility conditions” of the transportation system so that travel decisions can be made with full knowledge of what to expect. This subgroup concluded that for purposes of the BTS initiative, mobility indicators are best used for trend analysis and comparisons.

It seems likely that with the right selection of indicators, the public and the media would be quite interested in this information. The Texas Transportation Institute congestion index receives yearly national attention when the latest data are released. A “travel temperature” index developed at Georgia Tech, which uses travel time contours to give up-to-date information on expected travel times, received international attention because the traveling public easily understood it. Information such as this is likely to be well received by transportation system users.

There was a strong sentiment for having at least a single mobility index that could be used by local officials to assess how their metropolitan area is doing over time. We could easily provide a set of measures that provide indications on a variety of system characteristics, as well as one overall mobility index.

It was also noted that an unreliable system places greater emphasis on safety, security, and, of course, reliability. Such system performance characteristics definitely need to be part of a performance measures set, although not necessarily in a mobility index.

The above definition of mobility leads to the identification of several performance characteristics that should be included in any set of mobility indicators that reflects what is of interest to the traveling public. These characteristics include:

• “Reasonable” travel time,

• Lost time (or perception thereof),

• Reliability or degree of unexpected delay,

• Physical condition of the transportation system, and

• User travel cost.

The definition of “reasonable” deserves a great deal of discussion. One way of getting at it is to monitor the share of household expenditures that goes to transportation obtained from the Consumer Expenditure Survey of the Bureau of Labor Statistics. This measure shows that not only do people spend more on transportation with rising incomes, but they also spend an increasing share of income, thus indicating the importance of transportation to the household. This phenomenon should be captured in any indicator of mobility.

Including physical condition of the transportation system in a set of system performance indicators was considered very important for a national, state, and metropolitan-level perspective on system condition and performance. Some members of the subgroup felt that condition measures should be included in the mobility category, while others thought that they should be a separate, albeit an important, category.

Transportation system indicators would seem more compelling to users if they were actionable—that is, if a person could actually react or respond to the information being provided. Indicators that directly affect the lives of every person in a metropolitan area are most likely to be received with interest. Transportation fits into this category. It was also noted that one of the biggest complaints from the public is construction-related delay. If a measure could be developed that includes this performance characteristic, there was a sense that it could be used very effectively in developing strategies to minimize construction-related delay time.

We note that these mobility performance characteristics are important for both passenger and freight travel. For example, a recent study of travel time variability in California by the Transportation Research Board estimated a value of $12.60 per hour of standard deviation in travel time, compared with a value of travel time of $5.30 per hour for normal travel time. This indicates that travelers place a much greater value on travel time reliability.¹ A similar result was found in a study of the value of time

associated with travel time variability due to freeway incidents.² Whereas the reliability of travel time is important for passenger travel, it is probably more important for freight movement, for which delivery schedules are often related directly to financial costs. It seems important therefore to incorporate the certainty (or the uncertainty) of travel time into a performance index.

Many indicators have broad impact because they have economic implications. To the extent that such economic implications can be incorporated into mobility measures (and to some extent they are through travel time measures), they will be more relevant to citizens and local officials. The most frequently cited mobility measures fall into six major areas: congestion related (e.g., level of service, volume/capacity, and delay); trip time; amount of travel (vehicle miles traveled, vehicle hours traveled); mode share; transfer time; and transit performance. In considering these candidates and given the discussion above, the following mobility indicators are proposed as initial candidates for consideration as part of a set of national transportation indicators.

• Average daily hours of travel per person,

• Average minutes per mile,

• Average vehicle minutes of delay,

• Total passenger- and ton-miles traveled,

• Reliability factor (for example, percentage of a person’s travel time that is no more than 10 percent higher than average; the particular percentage chosen would depend on the distribution of the data),

• Personal or household consumption expenditures on transportation, and

• Travel rate index, which shows how much time is added to a trip during rush hour conditions compared with free-flow conditions.

It was suggested that perhaps a better approach to measuring mobility

would be to develop different measures for different modes and levels of analysis, and that weights be used (perhaps based on corresponding mode-related dollars or time budget) to aggregate to a metropolitan scale. This approach would reflect typical travel market or geographic segments. With segment-specific indices, officials would be able to gauge the level of mobility by important trip purposes, population groups, and geographic locations.

One member of this subgroup, David Greene, Oak Ridge National Laboratory, raised three issues. First, a good measure of performance reliability would be the variability of travel times for a specific trip. Second, no one knows the relationship between objective measures of transportation systems and subjective measures of these systems. Valuable research could be done to understand how the performance of the system affects the satisfaction of the customers of the system. Third, Greene noted that the mobility indicators subgroup focused mainly on travelers, not on freight. An equivalent set of measures should be developed for freight.

Randall Halvorson, Minnesota Department of Transportation, suggested that a measure of mobility must include elements of accessibility and elements of the infrastructure. Alan Pisarski also mentioned that elements of accessibility are important in a measure of mobility. An example of this type of measure is to record the percentage of travel opportunities within 20 minutes as a measure of what the transportation system offers customers. Pisarski also mentioned that it is important to have some basic system indicators. For example, how many miles of highway per square mile of land area are there? How many passenger-miles of transit service are being provided per square mile, per person? How many airline miles? To how many cities can you fly from this city nonstop?

Anthony Smith, University of Pennsylvania, suggested that a new BTS web site could be developed that would include maps of various transportation measures for different areas of the country, updated on a regular basis. Since it is difficult to identify the potential users of transportation data, this would be a useful way to make a variety of data available. Timothy Lomax, Texas Transportation Institute, noted that a “one number” approach to transportation data may also be appropriate so that more casual data users could understand the impact of transportation indicators.

A variety of perspectives were expressed on the general usefulness of

indicators and most certainly on their definition. It was clear from the subgroup’s interaction that such indices were considered important and likely to be of great interest to national and local leaders, as well as the public. The subgroup felt that travel time, reliability, and economic costs should be incorporated into measures of mobility.