Open-source simulation of the long-term diffusion of alternative passenger cars on the basis of investment decisions of private persons

The role of norms and collective efficacy for the importance of techno‐economic vehicle attributes in Germany

“…The preference shares were calculated using logit choice probabilities (Train, 2016). The calculations are based on the following assumptions displayed in Table 7 regarding average vehicle characteristics in a moderate scenario for the year 2020 (Senkpiel et al, n.d.).…”

“…Next, we identified lower and upper bounds of attribute developments and conducted own calculations for the years 2020, 2030, 2040 and 2050. For reasons of brevity, these assumptions on the future development of the attribute levels are described in more detail in (Senkpiel et al, n.d.). The participants were distributed to the different categories depending on which class they stated to be interested in for a future car purchase.…”

“…We did not offer a choice card specifically referring to the status quo as an alternative, since the subjective perceptions probably differ tremendously. However, the current conditions are indirectly represented in the levels that were chosen and also used for simulating the bottom‐up diffusion of electric vehicles in Senkpiel et al (n.d.).…”

“…HB has been shown to be more effective than multiple regression or multinomial logit models (Wellman & Vidican, 2008) as utilities are calculated on an individual level and, therefore, take the heterogeneity of the population into consideration (Gelman et al, 2014). For this study, it was particularly important to obtain values on the individual level since, in a second step, we developed a diffusion model to simulate electric vehicle diffusion based on individual utilities (see Senkpiel et al, n.d.). A total of 5000 draws were used for each respondent and 20,000 iterations were performed to estimate the parameters.…”