Analyses of commuting distances and times in the household context: The case of Berlin

Kurzfassung

Commuting is the consequence of a spatial discrepancy between the residential and occupational locations. Commuting distance and time are changed either by residential or occupational mobility. Residential locations are by their very nature chosen at the household level. These joint decisions determine commuting distances and times of each employed household member, leading to a sort of commuting collaboration between the spouses. Multiple-earner households have to adapt their residential location to several occupational locations instead of usually only one. In this context, the commuting distances and times of all concerned household members play a role. Some empirical evidence suggests that there is a trend towards longer average commuting distances and times. In order to better understand commuting behaviour, it is therefore necessary to consider not only the individual but the entire household, taking the situation of all employed household members into account. The study at hand is based on data of the household mobility survey for cities, which was established in 1972 and aims to create a database of representative travel behaviour (SrV). The SrV survey, which is carried out by the Chair of Transport and Infrastructure Planning at the TU Dresden, collects behavioural data for weekday travel in various cities. Here, the data for Berlin, Germany, from the year 2008 is analysed in detail. The geo-graphical reference for Berlin is its 12 boroughs which are further divided into 195 statistical areas. The so-called Berlin hinterland is composed of 63 municipalities located in the federal state Brandenburg surrounding Berlin. The sample on commuting includes data of about 11’772 individuals in nearly 8’294 households. The workplaces of these individuals are with 93.5% mainly located in Berlin, whereas merely 4.9% commute to the Berlin hinterland and only 0.6% to other municipalities in Brandenburg. The shares of individuals living in households with one employee, two employees or more than two employees amount to 60%, 36% and 4%, respectively. The empirical analyses start out from statistical explorations of the commuting behaviour in the City of Berlin. Furthermore, models for the commuting distances and times are estimated using regression techniques. The various indicators considered in the analyses include, on the one hand, socio-demographic and socio-economic characteristics of the individual and its household, such as gender, age, education, employment, the ownership of mobility tools (driving licenses, cars and public transport tickets) as well as the household composition and income. The household situation with respect to commuting is taken into account in terms of the sum of the commuting distances and times of all employed household members as well as the share of this sum relating to the respective individual. On the other hand, socio-demographic and socio-economic indicators characterising the place of residence as well as the place of employment are incorporated into the analyses at the level of the statistical areas of Berlin and the municipalities of Brandenburg. These factors cover information about the population with respect to gender and age, the foreign population, the employment by economic sector, the number of enterprises and their turnover, unemployment rates as well as information on spatial features and land use. Based on this data, densities, various shares and ratios, such as, for instance, the employees-inhabitants-balance, are calculated. The locations of the home and work place play an important role with respect to the commuting behaviour. On the one hand, a differentiation between East and West Berlin becomes visible. With respect to the distances and times covered on the work trip, a diametrically opposed commuting pattern emerges. Inhabitants of the Eastern part have longer commutes than persons living in West Berlin, whereas employees in East Berlin tend to travel shorter for work purposes than individuals employed in the Western part. Only in the case, that both, home and work place are located within either East or West Berlin, the trips to and from work are likely to be shorter. On the other hand, a place of residence in the inner ring of Berlin is related to shorter commuting trips, while persons living in the outer boroughs cover more distance and time. However, individuals working in the inner ring commute longer. Comparing the commuting trip in single-, double- and multiple-earner households, certain differences are observable. Both, the average distances and times are considerably longer, as the number of employees within a household increases. The household context has a significant impact on the commute of an individual. Both, the summarised distances and times covered by other employees in the household show a positive effect, indicating that the commuting burden is more or less equally distributed among all household members. The paper will provide a brief review of the literature on commuting, followed by a description of the data used for the empirical analyses. The main part will then concentrate on statistical explorations describing commuting behaviour in the City of Berlin, accompanied by the results of a number of regression models for the commuting distances and times. Finally, conclusions will be presented.