The number and share of miles that trucks drive empty is a critically important metric for the freight industry. However, as we have documented, there is a wide range of estimates of empty miles that are not necessarily comparable.
There are no industry-wide standards or guidelines for calculating and reporting empty miles. Some carriers do so voluntarily or to inform investors, but it is unclear the degree to which these metrics — even when similarly labeled — are truly comparable across carriers.
There are two main data sources for empty miles estimates: surveys and tracking devices, each with its own advantages and limitations.
Surveys: Breadth but with Possible Bias
In general, surveys provide broader coverage but are vulnerable to a number of known biases. Decisions made in designing the survey — for example, who is contacted to take the survey, how they are identified, how they are contacted and what happens if respondents fail to respond — as well as decisions about question wording, order and the length of the survey can all shape the outcome.
There are also potential biases in how people respond. Questions that ask about the past are vulnerable to biases or lapses in people’s memory, and people can be inadvertently misleading in their responses to questions that ask about controversial or competitive topics. In the case of empty miles, there is a clear competitive incentive to understate empty miles.
Tracking Devices: The Still Untested Frontier
Devices such as ELDs and smartphones provide more detailed location data based on actual truck movements — and therefore have the potential to produce more precise estimates of empty miles. But this approach also requires complementary data or some degree of inference to identify when a truck is loaded and when it is empty.
Until recently, the use of location data from ELDs and smartphones was only common on company or asset-based carrier fleets, effectively overlooking the large pool of owner operators. When they have been disaggregated in survey data, owner operators have historically reported higher numbers of empty miles making this omission particularly important.
Beyond the Data
Beyond differences in the underlying data, there is scope for subjectivity in deciding what exactly to count as an “empty” mile. Most industry participants would probably agree that empty truckloads between consecutive loads should count as empty, but there are a number of less obvious grey areas including:
- Miles driven between a home, dispatch or resting spot and the start or end of a load;
- Less-than-truckload (LTL) loads or partial loads;
- Miles driven to a pickup that is canceled on late notice;
- Tractor and trailer repositioning for drop loads. (Drop loads are when the trailer is filled and emptied without a driver and tractor present, which means that only the tractor is driven between loads lowering fuel consumption.)
All of these ambiguities are important to consider to make empty miles estimates truly comparable across carriers. The chart below shows some suggested question to ask in order to understand what exactly is going into a particular empty miles calculation.
View our economic commentary disclaimer here.