Immobility and activity levels: a comparison of USA and France

In parallel with "travel" surveys, "time use" surveys are another statistical source that can be used as a reference for individual trips. The relative redundancy of these two types of survey with regard to the observation of mobility and activities raises the question of whether these two sources give very similar results on daily mobility, but also, on the immobility of people. Immobility, i.e. the non-trip making in transport is a challenge for transport authorities, urban planners but also for statisticians. Reported immobility has to be examined in order to determine if the amount of measured immobility is a survey artefact or a real behavior - by whatever reasons. Research conducted by statisticians and surveys methodologists suggest, immobility may be analyzed as a 'soft refusal' e.g. non-response (Axhausen et al. 2002a). Policy makers, transport and urban planners may also interpret immobility as a real immobility: Therefore past research also highlighted, that some population groups are not or less mobile as barriers may restrain their mobility (Bacqué & Fol 2007; Fol 2010; Motte-Baumvol et al. 2015). The comparison of immobility indicators from these two surveys is therefore an interesting opportunity to try to better understand this phenomenon and also to question their possible complementarities. We will first analyze national travel surveys and national time use surveys in the USA and in France to give a general picture of immobility. Then we will analyze national travel surveys in USA and in France to produce a dozen of individual profiles (according to their immobility behavior). Indeed, in both countries we have similar behavior in term of immobility (about 17% of the population over 15 year old) and the same factors cause this behavior: Working status, Health, Age and Zone of residence. These factors define 12 different profiles. Then, we will investigate the Time Use Surveys and analyse the different activities performed by each profiles. This analysis is performed with Optimal Matching Analysis (OMA) (see Abbott and Tsay, 2000). The Optimal Matching Analysis is based on sequence analysis. This method is generally preferred when the event studied can be reproduced often and takes several states during study period (for example, household' motorization during life cycle, see Roux, 2012). Its use was mainly developed in biology (with the study of DNA sequences), and then spread to the social sciences in the late 1980s by Andrew Abbott (Abbott et al. 1986). The sequence analysis makes it possible to take into account the trajectory as a whole and thus to study the succession of states in the trajectory. For each profile's we will analyze the activities sequences during the day. This type of analysis could help us to detect the soft refusal (according to its distance with its profile).