1 Introduction
People discount future events both by preferring to obtain an immediate gain
(even if it is smaller than the one that may be obtained in the future), and
by trying to avoid an immediate loss, even if it is smaller than the one that
may incur in the future (Frederick, Loewenstein, & O’ Donoghue, 2002). In this
paper we propose that this phenomenon is likely to take different forms across
cultures because of cultural biases in attention allocation.
Consistent with recent work on cultural psychology (Markus & Kitayama,
1991; Masuda & Nisbett, 2001), we assume that people engaging in Western cul-
tures (Westerners) tend to focus their attention on the magnitude of a reward in
lieu of its context, including both a delay until receipt and a distant ob ject. Ac-
cordingly, these individuals may be much more strongly affected by a proximal
object than by the distant one with respect to the near future. Moreover, they
may be relatively impervious to the exact length of the time interval between
the two objects, because they may tend to focus on each temporal fragment
(a divided portion of the time interval between more delayed rewards) sepa-
rately, rather than focusing on the undivided time-interval between rewards. In
contrast, individuals engaging in Eastern cultures (Easterners) tend to allocate
their attention more holistically to both an immediate and a distant object,
as well as to the time interval between them. When compared to Westerners,
Easterners may then be expected to discount the future less, and moreover, even
when they do it, they may do so as a steady function of the length of the time
interval.
In the present paper, we first formulate these conceptual predictions in terms
of a mathematical model of future discounting based on Tsallis’ statistics, and
we report a cross-cultural experiment designed to test the specific derivations
of the model. Finally, we propose future directions in cultural neuroeconomics
employing neurocomputational models based on Tsallis’ statistics.
2 Intertemporal choice model based on Tsallis’
statistics and psychophysics of time
When given a choice between two possible times at which an outcome can oc-
cur, people usually prefer to receive it earlier if it is a good outcome, and
later if it is a bad one (Frederick et al., 2002). This phenomenon of temporal
discounting has been the subject of much research in neuroeconomics and neu-
ropsychopharmacology, which has revealed how the strength of the preference
for earlier outcomes over later ones is influenced by factors including the mag-
nitude and sign of the outcomes (Frederick et al., 2002) and temporal cognition
(Takahashi, 2005, 2006; Wittmann & Paulus, 2008). Specifically, (i) people are
patient with respect to the distant future but impulsive with respect to the
near future, when they choose between smaller sooner rewards and larger later
ones (preference reversal due to ”hyperbolic discounting”, also referred to as
”present bias”; for details see Soman et al., 2005), and (ii) people make more
impulsive choice when the length of delay is perceived as a sum of shorter time-
intervals (”subadditive discounting”; see Read & Roelofsma, 2003). We provide
examples of time-inconsistency and impulsivity in temporal discounting in Ap-
pendix A. These inconsistencies in intertemporal choice cannot be accounted