The implications of long-distance tour attributes for national travel data collection in the United States

Abstract

Despite sparse data on long-distance travel in the United States, there is increasing need for studies to inform policy. Most existing datasets have defined long-distance based on a distance threshold. This paper used a recent longitudinal panel of 628 individuals surveyed monthly online for 1 year about overnight travel to consider multiple distance-based classification schemes for long-distance travel and to characterize tours per year, as well as miles and days away. Negative Binomial Regression Models of tour generation per person per year using numerous typical distance threshold definitions (50–3000 miles) did not produce convincing models. Results suggest that distance thresholds do not bound or define a particularly unique type of travel and as such are arbitrary and potentially vary by region. Distance thresholds should not be the defining method for long-distance data collection programs. Expansion of the existing national program to collect all travel together using passive data collection would eliminate the need for distance thresholds and may best represent the diverse spatial and temporal characteristics of long-distance tours. This would allow subsets of tours to be extracted as needed. Results in this paper illustrate long-distance tour complexity: (1) mixed purposes between stops as well as at individual stops occurred in 14% of tours; (2) spatial complexity including multiple chained stops as well as out-and-back from a hub other than home occurred in 20% of tours accounting for 46% of the miles; and (3) different primary modes on different legs of the long-distance tours were used in 11% of cases.