Subjective Well-Being and Hedonic Editing: How Happy People Maximize Joint Outcomes of Loss and Gain

Subjective well-being (SWB) is a trait-like, stable state of happiness, which is consistent across time and situations (Diener et al. 1999; Sandvik et al. 1993). Findings from happiness research have shown that individuals with greater SWB are more likely to employ strategies that enhance hedonic outcomes than individuals with lower SWB. Recently, Robinson and Kirkeby (2005) demonstrated that one’s level of life satisfaction shapes how one organizes self-knowledge pertaining to emotional experiences and suggested that trait-like happiness should be regarded as a belief system. Thus, the present authors treated SWB as an independent variable that influences one’s behavior including decisions in the hedonic editing paradigm. In addition, it should be noted that in this article “happy people” refers to “those with high SWB” and the term “happiness” indicates “subjective well-being.”

In order to understand the sequence by SWB interaction, we should first address the general preference for longer intervals in the gain-first sequences than in the loss-first sequences. There are several ideas that could explain why longer intervals were preferred for the gain-first sequence than for the loss-first sequence. For one, Loewenstein and Prelec (1993)’s “sequence frame effect” suggests that the strong preference for happy endings diminishes when the two events are presented with a longer interval and consequently cannot be perceived as a sequence. If there is no temporal relation between two events, people tend to follow the temporal discounting rules: take gains as soon as possible. Participants who chose the gain-first sequence might have aimed to minimize the discounting effect and tried to segregate two events. Another possible explanation could be found in the “contrast effect” (Elster and Loewenstein 1992; Tversky and Griffin 1991). The impact of a loss would be greater if it comes after a gain because people are sensitive to relative changes. Participants who chose the gain-first sequence might have tried to separate the loss from the gain in order to avoid the negative contrast. Conversely, shorter intervals for the loss-first sequence would provide positive contrasts that facilitate more favorable temporal comparisons. However, these explanations for general preferences cannot account for the finding that participants with higher SWB were more likely to place shorter intervals between the loss and gain events, even for the gain-first sequence. Happier individuals preferred to arrange the events so that the two events were temporally closer, regardless of the loss-gain sequences. It seems that for happier individuals, the loss-buffering effect outweighed the other effects. Although an interesting finding, we will not further discuss this in the current article since the sequence by SWB interaction was significant only for the same-domain pairs of the mixed-loss condition and because the cross-domain pairs with a social gain are our main focus. From Study 2 on, we discuss only the loss-first sequence that the majority of the participants preferred.

We adopted the Freeman–Halton extension of the Fisher test (Freeman and Halton 1951) to compensate for a small sample size. The Chi-square test produced the same results for financial losses, χ ² (2) = 6.39, p < 0.05; for academic losses, χ ² (2) = 12.57, p < 0.01; and for social losses, χ ² (2) = 2.48, ns.