2.2. Survey design
In order to design our DRM study in a way that yields results comparable to previous studies,
but also to allow us to specifically analyze how experienced utility depends on a person’s
employment status, we used a questionnaire and an interview setup similar to that presented
in Kahneman et al. (2004a). However, we modified the questionnaire to obtain information on
the respondents’ current employment status, their employment history, and their job search
behavior as well.3
We conducted pre-tests of the questionnaire with 24 randomly chosen employed and
unemployed respondents in January 2008. Between March and July 2008, we interviewed a
total of 1,054 persons, of whom 737 respondents were either employed full-time or
unemployed without being engaged in any type of welfare measure. From these 737 persons,
we had to drop 23 interviews due to lack of understanding and missing answers. The total
number of usable interviews was 714, 366 of which were with full-time employees and 348
with unemployed persons. 366 persons (195 employed and 171 unemployed) were
interviewed in the Magdeburg region, and 348 persons (171 employed and 177 unemployed)
were interviewed in Berlin. Interviews lasted between 30 and 60 minutes.
Both employed and unemployed respondents were selected randomly. The unemployed
were approached directly by the interviewers in the local employment offices and asked
whether they would like to participate in a survey. They could then choose whether the
interview would take place directly on site, at their home, or at the local university. We only
interviewed long-term unemployed persons eligible for the means-tested “Unemployment
benefit II”. Unemployed interviewees received a compensation of 10 euro. About 15 percent
of the unemployed we approached participated in the interview. To recruit employed
respondents, we randomly selected, from the telephone directory, addresses within the district
of the employment offices and sent a letter in which we briefly explained the purpose of our
study (without yet mentioning that we would ask respondents to provide information about
their time-use and feelings) to these households and told them that we had selected them to
participate in the study. Within three days, we gave all these households a telephone call to
make an appointment for the interview, which then took place either at the university or at the
interviewee’s home. Of all the persons contacted, 8 percent were employed full-time and
3 A translated version of the questionnaire is presented in Appendix B.