Is Long-Distance Travel Influenced by Regional Geography?: An Analysis of Tripmaking Between Sociogeographic Clusters of US Counties

As long-distance travel volumes continue to grow, decision makers seek data to help manage the travel-related impacts. Household travel surveys offer the detail necessary for developing policy-responsive forecasting models and other analyses, but more efficient sampling plans are needed to capture long-distance travel in a more cost-effective manner. To support the development of sampling approaches for long-distance surveys, the objectives of this study were to (a) develop a sociogeographic cluster classification system for stratifying U.S. counties and (b) measure differences in long-distance travel behavior across these county classifications. Using fuzzy clustering techniques, U.S. counties were successfully classified into sociogeographic clusters for the purpose of analyzing tripmaking behavior. We measured mean household total trip rates, trip rates by mode (air, vehicle, other), and trip rates by purpose (leisure, work, other) using the weighted 2001 National Household Travel Survey. The results indicated that households within each county cluster pursue (a) statistically similar long-distance trip rates with other households in their cluster and (b) statistically different trip rates from households in other clusters. The differences in household long-distance trip rates also extended to the census region level, meaning census regions as well as sociogeographic cluster assignment were significant predictors of long-distance trip rates. Therefore, rather than collect a long-distance travel survey from a sampling framework focused on being nationally representative, these results suggest we might target data collection from regional sociogeographic clusters.