Arbeitszeiten in Deutschland - Eine Frage der Messung? (Working hours in Germany * a question of survey design?)

"This article deals with the sometimes large differences which exist when collecting information on actual working hours by means of various surveys. In contrast to the IAB’s calculation of annual working hours based on macro-data, more and more often data on actual working hours are being gathered by interviewing people. However, the range of the average weekly hours of work calculated using such surveys is very wide. Using data from the European Labour Force Survey, the Socio-Economic Panel, the Working Time ’99-Survey of the ISO Cologne, the European Working Conditions Survey and the Employment Options of the Future Survey, this study shows what differences exist and where they come from. The key element of argumentation here is the role played by the position of a question within a questionnaire (halo effect) and by the formulation of the question. On the basis of the different average weekly hours of work and the distribution of the weekly hours of work (hour brackets), it was possible to prove that the positioning of a question within the questionnaire as well as the wording of the question asking about the working week have a strong influence on the response to the question, and thus on the average weekly hours of work ascertained. Those surveys which ask about the weekly hours of work normally performed and do not mention overtime in the question or in the context of the question show significantly lower weekly hours of work when compared with those which ask about the average working week and do mention overtime in the question or in the context of the question. Hence, when analysing actual weekly hours of work based on a survey, it is imperative that the aspects of the formulation of the question and the context of the question be taken into account in order to make an adequate interpretation possible." (Author's abstract, IAB-Doku) ((en))