We use Social Survey data for 2014–16 and Google Maps data to study the distribution of employees in Israel by their travel modes—and in particular, their dependence on private vehicles. The analysis was conducted from two perspectives: one allows geographical mapping of the localities in Israel by their accessibility to job localities via public transportation (relative to private vehicle), and the second examines, at the individual data level, the impact of the accessibility and of individual characteristics on the choice of travel mode.

Mapping Israeli localities by an index of relative accessibility via public transportation to workplaces, calculated in this paper, indicates notable gaps. The more distant home localities are from the metropolises’ core, the less relative accessibility there is. In most localities in the periphery—and particularly in Arab localities—the relative accessibility is low due to the limited supply of public transit. In small Jewish localities in the periphery, accessibility is low, but high socioeconomic levels of these localities may indicate that the low accessibility derives from residents’ preference for private vehicles (given the level of public transportation that can be provided to such localities). In ultra-Orthodox cities and localities, the relative accessibility is high. In many localities with a lower socioeconomic background—particularly in the Arab sector—the relative accessibility is low, while in tandem there are organized shuttle services provided by employers. This mode is efficient in the sense of distance covered in a given time. However, a lack of alternatives creates dependence on such shuttle transportation, which reduces the residents’ employment possibilities and creates among them a dependence on a small number of employers.

Analyzing individual effects on the travel mode (using the Discrete Choice Model) given a limited number of alternatives indicates a small (but statistically significant) effect of the trip’s travel time. However, the proximity of bus/train stations and frequency of service markedly increase the probability of choosing those modes of transportation, and reduce the use of private vehicles. In contrast, car maintenance benefits and a company car lead to the choice of private vehicles.

The findings also indicate a correlation between a low socioeconomic level (in terms of wages, schooling, and housing density) and a greater tendency to use a bus, and a low attractiveness of the bus mode for upper income deciles. The probability of choosing a bus is markedly lower among private vehicle owners and among workers who are eligible for car maintenance benefits from their employer. Concerning trains, the findings indicate a higher readiness to choose them, when the train is available, both among vehicle owners and among other commuters.​


For a PDF of the full article in Hebrew​ ​