Is travel behavior change durable ?

During decades, car travel growth was stimulated through moderate energy prices, a higher average income, suburbanization, decreasing vehicle costs, and the liberal attitude of public authorities towards the car. However from the 1990's, a downturn took place in public policies, from then willing to regulate car use to the benefit of alternative modes. This new paradigm, driven by environmental concerns, was further promoted from the early 2010's under the goal of energy transition, notably through the implementation of a carbon tax, the growing taxation of diesel fuels, and accessibility restrictions to city hearts for the most polluting vehicles. For authorities involved in such policies, assessing their long-term efficiency is vital. In particular, the durability of travel behaviour change might be more or less consistent with their expected outcomes. Besides, rising oil prices and increasing taxation may have generated strategies among households to mitigate energy vulnerability, either through reducing car use, buying more energy-efficient cars or shifting their residential location. Finally, the multiplication of emergency events - for instance the quarantine measures that were implemented during the coronavirus epidemic - has produced great disturbance in travel patterns, and might challenge the ability of individuals to recover their previous habits or reorganize in order to preserve their way of life. Thanks to the existence of long-term data collection, econometric literature is helpful in understanding such mechanisms. It notably confirms the existence of travel behaviour elasticity to price and income effects (e.g., Johansson and Schipper, 1997), assuming symmetrical response to opposite variations in economic conditions. However, mechanisms of imperfect reversibility may also apply, where returning to initial conditions does not result in the status quo, because of new travel patterns adopted meanwhile. Though initially found out in the field of agricultural economics (e.g., Traill et al., 1978), this plastic behavior was also highlighted for travel patterns (e.g., Gately and Huntington, 2002) and within the energy field (e.g., Broodstock et al., 2011). In particular, rising fuel prices generate durable incentives to rationalize energy consumption through travel patterns and the renewal of the vehicle fleet (e.g., Gately, 1992) - notably because of loss aversion (Wadud, 2017) - that would reinforce the long-term efficiency of environmental taxation measures (e.g., Dargay, 2007). Yet, despite strong market volatility, fuel prices manifest a rising trend (Bonnafous et al., 2010), coming out from both oil prices and environmental taxation, which resulted in a plateau in car travel in many industrialized countries (e.g., Millard-Ball and Schipper, 2011). However, the strategies involved might not differ fundamentally from those implemented during the preceding oil shocks, which also resulted in durable reductions in car use (e.g., Gately, 1992). In particular, there is strong evidence that this threshold was mainly the result of fuel prices skyrocketing (e.g., Bastian et al., 2016). As a result, this apparent shift could only be temporary if fuel prices were to decrease again, an interpretation that was further supported by the recovery in traffic growth since 2012 (Maurin, 2017). In this research, we will question the durability of travel behaviour change in response to changing economic conditions among French households, by seeking evidence for mechanisms of imperfect reversibility, accounting for changes in income, fuel prices, vehicle efficiency and vehicle prices. We will address this issue through French Car Fleet Surveys, which display information about car ownership, car mileage and vehicle efficiency, by considering a restricted subsample of households that were kept into the panel over a long time period. The modelling framework will consist of a trivariate model structure, built of one main equation for the total car mileage, with car ownership and fuel efficiency as endogenous regressors, and two auxiliary equations for car ownership and fuel efficiency. To test the existence of asymmetry, exogenous variables will be split into monotonic series, describing the historical sequence of income and price variations. Studying the perennity of behavior change will provide a complementary perspective on the efficiency of economic incentives to reduce car travel and fuel consumption. In particular, as transition in travel behavior resulting from mechanisms of imperfect reversibility tends to reinforce the efficiency of policy measures based on economic incentives, these could be mitigated to account for household dependence on fossil fuels and equity concerns.