24
due to the way people adjust their time-use. What makes the unemployed better off is that
they use much more of their available time for activities that are more satisfying than working
and work-related activities. By contrast, aspirations do not seem to adjust to the new
circumstances. Employment sets the benchmark to which one compares one’s own
achievements in life: being in employment is better than being unemployed - despite the fact
that being at work makes one unhappier than not working. What determines aspiration,
whether it is the pursuit of valuable activities (Raez 1994), the search for a meaningful life, or
a question of controlling one’s own life, is an open question for further research.12 The results
indicate, however, that these factors do not affect hedonic adaptation as strongly. Our findings
thus provide additional support for the claim that a “shift in attention is not the only possible
explanation for adaptation, however. Substitution of activities, for example, may also play a
role. ... Measures of well-being that are connected to time use have the potential to uncover
such shifts.” (Kahneman and Krueger 2006, p. 18)
The problem with the Day Reconstruction Method is that it is only a snapshot. To validate
our hypothesis that long-term unemployment causes hedonic adaptation but not a lowering of
aspirations, it would be ideal to collect panel data that follows individuals through the entire
adaptation process - from still being in employment, via their short-term unemployment
experience, up to their long-term well-being. Alternatively, it would also be useful to extend
existing cross-section time-use surveys by adding well-being questions and to apply the Day
Reconstruction Method to people who have just received their notice of dismissal, to people
just being laid off, and to people with an unemployment spell of up to six month. For this
purpose, the newly defined measure of episode satisfaction may turn out to be more pragmatic
than asking respondents about a large number of emotions as it allows us to learn about
experienced utility within time-use surveys by only asking one question, instead of a
multitude of questions, per episode without losing too much information.
12 For a discussion, see Loewenstein (2009).