Vehicle-miles Traveled (VMT) is the number of miles traveled by motor vehicles in the area of study. This indicator divides VMT by population to account for population growth or loss.
How it Relates to Sustainable Communities
Increases in VMT contribute to traffic congestion and air pollution, causing carbon dioxide and particulate matter emissions. Because of population growth and economic development, many regions cannot feasibly reduce absolute VMT. Reducing per capita VMT can help a region achieve air quality, climate change, and congestion reduction goals without penalizing it for population growth (see Aggregate vs. Per Capita Measurement). Coordinated land use and transportation policies can lower VMT per capita by providing more people with additional transportation choices and shortening trip lengths.
Many regions interested in reducing transportation GHG emissions find that VMT is more straightforward to calculate since it does not account for vehicle fleet characteristics and fuel carbon content. Additionally, transportation planning agencies do not directly influence vehicle technologies and fuels, but their decisions can influence VMT. Measuring VMT also avoids the possibility that unexpected changes in vehicle and fuel characteristics would significantly affect a region’s ability to meet its goals.
However, transportation GHG emissions are affected by factors other than VMT–vehicle fuel economy, fuel carbon content, and the efficiency of system operations– and the usefulness of VMT as a proxy for GHGs diminishes as vehicles and fuels become more efficient. In addition, a VMT metric will not capture the potential GHG benefits of transportation system management and operations strategies, such as lower speed limits, traffic signal improvements, and incident management programs that reduce traffic delay.
Vehicle miles traveled
FHWA's Highway Statistics is an acceptable source of vehicle miles traveled (VMT) estimates for most metropolitan areas. Highway Statistics estimates VMT according to the geographic boundary approach (see Calculation Notes below). Care should be taken when states adjust the boundaries of federal-aid urbanized areas (typically once per decade, when decennial Census results are available), because prior year VMT estimates in Highway Statistics may no longer be comparable.
However, Highway Statistics is not acceptable in three cases:
- Select metropolitan areas that contain multiple federal-aid urbanized areas, and thus have VMT reported in multiple entries in Highway Statistics (such as the San Francisco Bay Area)
- A city within a larger metropolitan area
- Non-urban communities (less than 50,000 in population), which are not reported in Highway Statistics.
In the first and second cases, an MPO travel demand model provides a better source of community VMT. Using trip tables and the travel distance between each zone-to-zone pair, it is possible to calculate VMT for internal trips (both trip ends within the city) and external trips (one trip end outside the city). For external trips, it is recommended that half of that VMT is assigned to the city. Note that estimates produced using the generated trips approach are not comparable to estimates produced using the geographic boundary approach.
In the third case (non-urban communities), there are limited options for estimating VMT. In some states, a statewide travel model could be use to estimate VMT in a rural community. The following states appear to have comprehensive statewide travel models that could produce accurate local VMT estimates: Florida, Indiana, Kentucky, Massachusetts, Michigan, New Hampshire, Oregon, Rhode Island (MPO model), Wisconsin, Vermont, Delaware
In some states, data from the department of motor vehicles (DMV) may be available to estimate VMT and track trends at the county level (based on the county of vehicle registration). Aggregate VMT by county can sometimes be obtained, particularly for government use. If available, this would be an acceptable source of VMT data for a rural community, provided that the community makes up a large portion of total vehicles registered in the county. However, it appears that DMV data is not readily available in most states.
If VMT data is unavailable for a rural community, one option would be to use traffic counts on selected roadways that are determined to be indicative of locally-generated travel. This data would be relatively easy to collect on a consistent annual basis. Multiple days of counts would likely be necessary to average out daily fluctuations. However, it would be difficult to apply this approach in a uniform manner that would allow comparison across different communities.
US Census Population data
FHWA Highway Statistics reports population by federal-aid urbanized area and also reports VMT per capita, so separate population figures are not necessary for this metric in most cases. Otherwise, population data is available from the American Community Survey (ACS) and the US Decennial Census. The decennial census is a full count of population, and is therefore the most accurate option. But it is only available every 10 years. The ACS provides new population estimates for most counties and regions on a yearly basis. See the ACS Tip Sheet for general guidance regarding limitations and best practices for using ACS data.
VMT can be calculated per capita, on an average daily basis (preferred), or per household, per vehicle or in total, and/or on an annual basis. Some communities follow VMT both per capita and aggregate (total VMT).
There are two general approaches to calculating vehicle miles traveled (VMT):
- Geographic boundary approach: Using traffic counts to estimate the amount of vehicle travel that occurs within a given geographic boundary;
- Generated trips approach: Using household survey data, a travel demand model, or vehicle odometer readings to estimate the vehicle travel of residents living within a given geographic boundary.
FHWA VMT data is developed using the geographic boundary approach. This approach is easier to calculate, particularly at a small geographic scale, and it captures travel by vehicles that are not garaged within the area (e.g., visiting workers, shoppers, students, commercial vehicles). It can appear artificially high if the area has a lot of pass-through traffic or can appear artificially low if many of the area’s residents drive long distances to jobs outside the area.
The second approach depends on survey, model, or odometer data, which can be difficult and/or costly to obtain and update regularly. It can properly capture all travel by area residents, regardless of their destinations. Variations of this approach use travel model data and can include trips produced by and/or attracted to a given area, for all trip purposes.
Communities that Use this Metric
- City of Portland, OR
- City of Berkeley, CA
- City of Oakland, CA
- Missoula County, MT
- San Mateo County, CA [see Transportation section]
- Travis County, TX (Austin region)
- Greater Boston Region, MA
- Delware Vally Regional Planning Commission (Philadelphia, PA)
- Long Island, NY
- Puget Sound Regional Council (Seattle, WA)