Individuals attribute social status and personal image to owning and using a car ('car pride'), which may interact with their travel behavior in complex ways. This study explores the multi-directional relations among car pride, car ownership, and car use for a sample of 1236 adult commuters in New York City, NY and Houston, TX. Applying multivariate structural equation modeling and incorporating instrumental variables, we find evidence of a feedback loop among car pride, car ownership, and car use. Our results suggest that an individual with higher car pride is more likely to own a vehicle, and, enabled with this ownership, use it more frequently. And individuals who use their car more frequently are likely to feel more pride in owning and using their vehicle.
This exploration of causal multi-directionality in transportation attitude-behavior relations has important implications for behavioral research, model development, and policy interventions. For researchers, potential bidirectionality must be anticipated from the outset of research design and accounted for appropriately in modeling to address underlying endogeneity. For policymakers, our results suggest that there are multiple intervention points within the reinforcing cycle of attitudes and car consumption. Policies could directly target car ownership and use or could consider influencing behavior through attitude change.

The car fulfills not only instrumental transportation functions, but also holds important symbolic and affective meaning for its owners and users. In particular, owning and using a car can be a symbol of an individual's social status or personal image ('car pride'). This paper introduces and validates a standard measure of car pride estimated from 12 survey statements using a cross-sectional sample of 1,236 commuters in New York City and Houston metropolitan statistical areas. We find that car pride is higher in Houston than in New York City. We then empirically examine the bidirectional relation between car pride (attitude) and household car ownership (behavior) using structural equation modeling. To identify the bidirectional relationship we use an individual’s general pride as the instrumental variable (IV) for that same individual’s car pride; in the opposite direction, we use the average household vehicle ownership in the respondent’s census block group as the IV for the respondent’s household car ownership. We find that positive and statistically significant relations exist between car pride and car ownership in both directions. However, on average and in both city subsamples, the relation from car pride to household car ownership (attitude-to-behavior) is much stronger than the reverse (behavior-to-attitude). In fact, in our models car pride is more predictive of car ownership than most individual and household socio-demographics included in traditional ownership forecasting models, including income. Empowered with a well-validated, standard measure for car pride and a robust approach for exploring reciprocal attitude-behavior relations in cross-sectional data, future research can extend the current understanding presented in this paper to explore car pride’s relation with other travel behaviors, the dynamics of these attitude-behavior relations over time, and their implications for policies to promote sustainable travel behavior.