Motor Vehicle Use and Travel Behaviour in Germany: Determinants of Car Mileage

Abstract

In contrast to other countries where official odometer readings are collected when cars are being inspected or whenever there is a change in the registration data, no such information is available in Germany. The published annual figures on mileage of German vehicles result from model calculations, based on different data sources. The last two large surveys on car use were carried out in 1993 and 2002. These data are analyzed to find determinants of car mileage traveled and to check if there was a significant change of average mileage within 10 years. The method used to find determinants of car mileage is a log-linear analysis of variance. In general, average annual mileage for a German passenger car was 13,500 km in 2002, about 5 % less than in 1993 (14,200 km per car). When both privately and business owned cars are included in the models, only car?specific characteristics can be used as explanatory variables. In these models there is a high effect of the survey year on the car mileage even if other variables – as car size, car age, and type of engine – are controlled for. However, if we consider private cars only, additional variables of individual users can be included in the models. In addition to engine type, age of car, horsepower, the age and gender of the driver are central variables explaining car mileage. The dummy variable for the year is significant as well, but its effect on average mileage is lower, although, e.g., fuel prices did rise by 50 % between 1993 and 2002. Obviously the demographic changes in motorization are dominant, while an effect of fuel price increase is not evident – apart from the trend towards diesel cars. These observations confirm research results, stating that individual preference for car use is a high-level inelastic demand.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 7 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.