Content uploaded by Hal Hershfield
Author content
All content in this area was uploaded by Hal Hershfield on Feb 21, 2022
Content may be subject to copyright.
Ann. N.Y. Acad. Sci. ISSN 0077-8923
ANNALS OF THE NEW YORK ACADEMY OF SCIENCES
Issue: Decision Making Over the Life Span
Future self-continuity: how conceptions of the future self
transform intertemporal choice
Hal E. Hershfield
New York University, New York, New York
Address for correspondence: Hal E. Hershfield, New York University, Department of Marketing, 44 W. 4th St., New York, NY
10012. h-ersnerhershfield@kellogg.northwestern.edu
With life expectancy dramatically increasing throughout much of the world, people have to make choices with a
longer future in mind than they ever had to before. Yet, many indicators suggest that undersaving for the long term
often occurs: in America, for instance, many individuals will not be able to maintain their preretirement standard of
living in retirement. Previous research has tried to understand problems with intertemporal choice by focusing on
the ways in which people treat present and future rewards. In this paper, the author reviews a burgeoning body of
theoretical and empirical work that takes a different viewpoint, one that focuses on how perceptions of the self over
time can dramatically affect decision making. Specifically, when the future self shares similarities with the present
self, when it is viewed in vivid and realistic terms, and when it is seen in a positive light, people are more willing to
make choices today that may benefit them at some point in the years to come.
Keywords: future self-continuity; behavioral economics; intertemporal choice; temporal discounting; retirement
saving
Introduction
People are living longer than they ever have before.1
In the Western world, for example, more years have
been added to average life expectancy in the last
century than in all of the previous millennia com-
bined.2In the blink of an eye, sixty-five has quite
literally become the new fifty-five: the typical 65-
year-old living in modern America can now expect
to live as long as the average 55-year-old living in
1935 America once did.3Although the addition of
these extra years can be viewed as an unprecedented
example of human innovation and technological
progress, there is still a downside to this new de-
velopment. For most of human history, planning
for the distant, long-term future was not neces-
sarily essential. When reaching one’s 60th birthday
was relatively uncommon—as it was in the early
1900s1—considering one’s preferences at age 70 or
80 would have been considered an unnecessary exer-
cise (and one that most likely would have taken time
away from making important and more consequen-
tial decisions in the present). But with newly added
years to the life span, plans must now be made with
a longer future in mind. Saving behavior, though,
has not kept pace with increasing life expectancy.
In what follows, I will first briefly review previous
research on the psychological determinants of un-
dersaving, most of which examines the way indi-
viduals evaluate present and future rewards. I will
then overview a burgeoning body of theoretical and
empirical work that takes a different viewpoint, one
that focuses on how perceptions of one’s self over
time can dramatically influence decision making.
Saving behavior and temporal discounting
For Americans in the earlier part of the twentieth
century, the average length of time spent in re-
tirement was approximately two years,4and con-
sequently, one did not need to worry too much
about accumulating a huge nest egg to ensure a
comfortable end-of-life period. In today’s world,
however, people are retiring earlier than they have
in years past, living longer, and saving less. In the
United States, for instance, the average male adult
doi: 10.1111/j.1749-6632.2011.06201.x
30 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
spends approximately 17.1 years in retirement, and
for women it is 20.1 years.5Yet, planning behavior
has not kept pace with these pronounced increases
in life expectancy; humans are not suddenly gain-
ing more foresight as more and more years have
been added on to their lives. The National Retire-
ment Risk Index,6which assesses the percentage of
households who will fall short of meeting their re-
tirement goals, for example, has been steadily in-
creasing over time, and significantly so after the 2008
financial crisis.7Two-thirds of Baby Boomers will
not be able to maintain their preretirement stan-
dard of living in retirement,8and perhaps most
shockingly, more than half of working-age Amer-
icans have accumulated less than $25,000 in sav-
ings.9Although the sheer magnitude of this un-
dersaving may be surprising, the fact that it occurs
may not be so. Indeed, according to research in
psychology and economics, people characteristically
care less about future outcomes than they do about
present ones, a phenomenon known as temporal
discounting.10–12
Theorists have suggested that such discounting
may arise for several reasons,13 not the least of which
is that the needs of the present may simply over-
whelm the needs of the future. A given individual
may want to save for the future but not have the
financial means to do so. The 44 million Americans
who fall below the poverty line,14 for instance, can-
not be expected to contribute to tomorrow’s saving
account when today’s groceries need to be bought.
Yet there are many people who can afford to save but
do not. To take one example, Choi, Laibson, and
Madrian15 surveyed employees across seven large
companies and found that between 20% and 60%
failed to even contribute an amount to their retire-
ment accounts that their employers would match.
By not doing so, these employees are essentially re-
jecting a “free lunch” that causes them to lose ap-
proximately 6% of their income annually.
On a more psychological level, present rewards
may feel more arousing and emotional than fu-
ture rewards do. Mischel and colleagues16–18 have
shown experimentally that children are much bet-
ter able to postpone a desirable reward when they
think about the “cool,” informative properties of
that reward (e.g., a marshmallow’s shape and color)
rather than the “hot,” consummatory aspects of it
(e.g., the marshmallow’s taste). Thinking about the
“hot” aspects of the reward can lead to a frustrating
situation in which the future reward becomes even
more desirable, yet still unavailable.17 Focusing on
a reward’s “cool” properties allows an individual to
alleviate this frustrating situation without actually
consuming the reward.19 Thinking about rewards
in this manner has a similar effect on decision mak-
ing as taking on an abstract level of construal does
(i.e., considering the “big picture” of an event rather
than the details), which according to Trope and col-
leagues20–22 is a key aspect of mitigating temporal
discounting.
People also have a fundamental inability to
project their thoughts and feelings into the distant
future.23–25 One might think, for example, that life
will be happier and brighter if this week’s lottery
ticket is a winner; in a related fashion, assistant pro-
fessors might view a future without tenure as bleak
and depressing. But lottery winners and nonlottery
winners are equally happy, and professors who fail
to receive tenure still go on to lead meaningful, pro-
ductive lives.26 In other words, people overestimate
the degree to which they will feel good about a pos-
itive outcome and bad about a negative outcome.
According to Gilbert et al.,27 individuals make poor
predictions regarding the impact of a future emo-
tional event because they tend to focus too much on
the event in question and by so doing, fail to take
into account the influence that other events and
situations will be exerting on their feelings. Such
errors in affective forecasting can frequently have
deleterious results for decisions made in the present
that have consequences in the future, or intertem-
poral choices. A hypothetical consumer might be
inclined to spend his paycheck on a new electronics
purchase, erroneously thinking that this will bring
him more happiness in the future than, say, putting
that same money into his 401(k) would. Conversely,
losses or deficits may contribute less to unhappiness
than anticipated (e.g., people consistently underes-
timate how happy they would be after a traumatic
accident or injury28 ).
The previous accounts explain planning failures
in the context of rewards that can be consumed now
or at a delay. Recent work, however, suggests another
possibility, one that deals not with present and fu-
ture rewards, but wit h present and future s elves. This
review draws on theory from philosophy, psychol-
ogy, and economics, and surveys empirical work
suggesting that the way the future self is perceived is
a strong predictor of decision making over time.
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 31
Future self-continuity and choice Hershfield
Theoretical underpinnings of the future
self-continuity model
Theorists have argued that problems with intertem-
poral decision making occur because of conflicts
between temporally distinct selves.11 For example,
the self has been conceptualized as an organization
of individuals composed of a long-term planner and
a short-sighted doer.29–33 Although the planner-self
may plan to lose weight by this summer’s beach sea-
son, the doer-self may indulge in several Girl Scout
cookies in the office break room. In certain domains,
such as dieting and substance abuse, people are often
sophisticated enough to recognize that tensions will
arise between the wishes of today’s self and the de-
sires of tomorrow’s self.34 It is this recognition that
most likely leads many people to engage in precom-
mitment strategies that attempt to constrain future
behavior.35,36 Ulysses plugged his ears with wax so
that he would not succumb to the fatal songs of the
beautiful Sirens;37 alcoholics have taken Antabuse
to avoid imbibing; and the Save More Tomorrow fi-
nancial planning program allows employees to pre-
commit to increasing their retirement allocations
once they receive a paycheck raise.38
One problem with precommitment strategies is
that it is not always an easy exercise to determine
which self’s interests to satisfy (i.e., the planner-self
who advocates dieting, or the doer-self who sup-
ports indulging).36 Although I may say today that I
want to shed five pounds by June,w hy should the self
who states that desire get privilege over the self who
wants to eat the cookie in two weeks’ time? It is dif-
ficult, according to Schelling,36 “for two selves that
do not simultaneously exist to compare their pains,
joys, and frustrations” (p. 10). Rather than trying
to determine who the authentic self is—that is, the
self whose wishes should be met—one could instead
conceptualize a person as a collection of selves over
time, all of whom have authentic wants and desires.
According to this view, espoused by Parfit39,40 and
Strotz,41 among others, the self is a collection of
distinct identities that overlap with each other over
time. Each one of these overlapping selves may share
a strong degree of psychological connection with the
next one, or they may not. Parfit cites Proust42 to
illustrate this point: “...we are incapable, while we
are in love, of acting as fit predecessors of the person
whom we shall presently have become and who will
be in love no longer...” (p. 631). We may at one
point in time think that our preferences and actions
are reflective of our true, authentic selves, but as the
years advance and situations change, we may find
that we are suddenly completely different from the
person we once thought we were.
Naturally, there will be less overlap between selves
with a greater degree of temporal distance between
them: with more time, psychological connectedness
of oneself in the present with oneself in the future
grows more tentative. People feel more connected
to their potential self of five years than their po-
tential self of forty years. Consequently, one might
logically identify less with a more temporally dis-
tant future self to the point at which an extremely
distant future self may seem like a different per-
son altogether,39,43–45,athough Pronin et al.45 find
that even a not-so-distant self (i.e., one that is just
afewdays,weeks,ormonthsaway)issometimes
treated like another person. Here, it is not impor-
tant whether in various senses, a person actually
does change over time; rather, what matters is how
much a given individual feels he or she will be the
same person over time. Although trait-level person-
ality characteristics47 and general interests48 remain
relatively consistent over the course of a lifetime,
many other aspects of our selves change over time:
people can alter their names, noses, and reputations
beyond recognition. What matters, for long-term
planning, however, is that one person has but one
identity, and it is with this link that the assets of the
present and future selves are tied together. Individ-
uals who feel as though the future self is a different
person fail to acknowledge this connection, that is,
fail to identify with themselves in the future.
aThis idea is not just germane to British philosophy, as
it has found its way into popular culture as well. The
comedian Jerry Seinfeld, for example, has a humorous
take on distinct current and future selves: “I never get
enough sleep. I stay up late at night because I’m ‘night
guy’.‘Nightguy’wantstostayuplate.‘Whataboutgetting
up after five hours of sleep?’ ‘Oh, that’s morning guy’s
problem. That’s not my problem—I’m night guy! I stay
up as late as I want.’ So, you get up in the morning, with
your alarm, and you’re exhausted and groggy...Oh, I hate
that ‘night guy’! Ya see, ‘night guy’ always screws ‘morning
guy’. There’s nothing ‘morning guy’ can do. The onlything
‘morning guy’ can do is try to oversleep often enough so
that ‘day guy’ loses his job and ‘night guy’ has no money
to go out anymore.”46
32 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
Recent research has demonstrated that, in impor-
tant ways, people often treat the future self as if it is
in fact another person. On a general level, individ-
uals make attributions about the future self in the
same manner that they do for others, for example,
by attributing the future self’s behavior to disposi-
tional factors rather than situational ones,44,49 and
to make decisions for the future self using a similar
process that they use to make decisions for other
individuals.45 These findings have important impli-
cations for intertemporal choice. If people tend to
consider the future self as a stranger—that is, if that
distant future self feels on an emotional level like
another person—then they may rationally have no
more reason to save their money for their future
selves than to give that money to a stranger.
It is important to consider that Parfit’s39,40 model
does not posit that the degree to which the future self
is considered a stranger is a stepwise function: the
future self does not necessarily have to be thought
of as either totally connected to the current self or
totally disconnected from it. Rather, just as there can
be degrees of connection between people,50 so too
can there be gradations of connectedness between
today and tomorrow’s self. Critically, then, the de-
gree to which an individual feels disconnected from
his or her future self should correlate with the degree
to which that individual discounts future rewards.
The more continuity a person shares with his future
self—that is, the more that future self feels like a
direct extension of who he is now—the more moti-
vated he will be to act in ways that will benefit him-
self in the future. Conversely, the more the future self
feels like a stranger—that is, the more disconnected
a person is from his future self—the less motivated
he will be to plan for the future. In the domain of
financial decision making, for example, if the future
self is lacking in continuity with the current self,
a given individual may be inclined to spend in the
present rather than save for the future.
The composition of continuity
But what aspects matter when judging connected-
ness to the future self? First, continuity with one’s
future self may be determined by the extent to which
one feels similar to it. Parfit40 explicitly discusses
similarity in terms of one’s likes, beliefs, values, ide-
als, etc. in making judgments about the self in the
future. Indeed, a host of research examining em-
pathy and prosocial behavior has found that people
are more likely to help others whom they perceive as
similar along many of these dimensions (see Eisen-
berg and Miller51 for a review). Recent work, in fact,
has shown that on microfinance websites, people are
more willing to donate when recipients are judged
as being similar to the self.52 Similarly, if the future
self is viewed as being similar to the current self, one
will be more likely to “donate” to that future self by
way of saving money in the present.
Second, Parfit40 observes “when we imagine pains
in the further future, we imagine them less vividly,
or believe confusedly that they will somehow be less
real, or less painful” (p. 161). The reverse also seems
to be true: Tversky and Kahneman53 noted that out-
comes or events that are easier to picture in one’s
mind are more “available” and, hence, seem more
likely to occur. Along these lines, Loewenstein19 the-
orized that a more vivid impression of oneself en-
gaging in some action in the future might intensify
the emotions that are linked to thinking about that
scenario. These intensified emotions might, in turn,
allow an individual to be better informed regarding
the future consequences of a present decision. For
example, pulmonologists tend to smoke less than
other doctors, perhaps because seeing blackened
and withered lungs on a daily basis increases the neg-
ative emotions that are associated with smoking.19 A
victim who is portrayed in vivid terms is more likely
to elicit a sense of connection and sympathy, and
subsequent charity than one who is not.54,55 Sim-
ilarly, if the future self is more vividly imagined—
that is, if it is easier to picture—one should feel a
greater sense of connection to it and consequently
be motivated to save for the future.
Third, previous work has implied that the actual
attitude one holds toward the future self—that is,
whether it is viewed in positive or negative terms—
can be a predictor of long-term decision making.
In the health domain, for example, Levy et al.have
used attitudes toward the elderly as a proxy for atti-
tudes toward the future self, and found that negative
views of the elderly held earlier in life can lead to
worse cardiovascular health later in life (controlling
for one’s own health).56 Moreover, holding positive
views of one’s own aging process is associated with
increased longevity.57 Theoretically, it should be eas-
ier to feel connected to another individual if that
person is viewed in positive terms. Indeed, when
forming a new relationship, people feel a greater
sense of overlap with another person who elicits
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 33
Future self-continuity and choice Hershfield
Figure 1. Future self-continuity scale that has been used to assess similarity between current and future selves.59
positive emotions.58 Along these lines, one should
be more likely to perceive continuity with a distant
self who is viewed in positive terms.
In what follows, I will review recent empirical
work that has been conducted that speaks to each of
these three domains of continuity: similarity, vivid-
ness, and positivity.
Similarity to the future self
Measuring the relationship between future
self-similarity and saving
To examine the degree to which people feel similar
(or dissimilar) to their future selves, my colleagues
and I59 developed a novel psychometric task based
on existing measure of self-other similarity.60 In the
task, participants choose a set of overlapping Eu-
ler circles that range from depicting no overlap to
almost complete overlap (see Fig. 1). Participants
picked the set of circles that best represents how
similar they feel to their selves in ten years time. Im-
portantly, in early research, we found that our mea-
sure of future self-similarity had strong test-retest
reliability and correlated with scores on a “me/not
me task” that examined the match between present
and future self-descriptions.
To understand how similarity to the future self is
related to saving behavior, we gave a group of col-
lege undergraduates the future self-similarity task
as well as an ecologically valid measure of tempo-
ral discounting.61 Each trial included one smaller
immediate reward paired with one larger delayed
reward. We had hypothesized that participants who
felt a greater degree of similarity to their future selves
would choose in a way that would benefit those fu-
ture selves, and indeed, this is exactly what we found:
there was a significant positive correlation between
similarity to the future self and the number of
larger later rewards that were chosen in the temporal
discounting questionnaire. We have found this re-
sult using both hypothetical choices and incentive-
compatible ones (i.e., participants were paid in ac-
cordance with their choices on the temporal dis-
counting task).59
To what extent, however, would the relationship
between future self-similarity and saving behavior
be found outside of the laboratory context, among
people making actual financial decisions? To address
this question, we surveyed a diverse group of work-
ing adults who lived in the San Francisco Bay area
and who ranged in age from 18 to 86. In line with our
prediction, we found a significant positive correla-
tion between perceived similarity to the future self
and the number of assets that had been accumulated
over time.59 Not surprisingly, age was also positively
correlated with future self-similarity: as people grow
older, life circumstances most likely become more
stable, and each passing interval of time feels sub-
jectively shorter (i.e., ten years represents a smaller
fraction of time in the life of a 50-year-old than in the
life of a 20-year-old). Nevertheless, the relationship
between future self-similarity and accumulated as-
sets remained robust even when controlling for age,
as well as other factors that have been previously as-
sociated with saving behavior such as education.62
Tomakethisresearchparadigmassimpleaspos-
sible, we deliberately left the concept of similarity
open to interpretation for our research participants.
Recent research by Bartels and Rips,63 however, has
used a more specific conception of similarity, and
found comparable—if not even stronger—results
to ours by asking participants to judge similarity to
their future selves in terms of likes, dislikes, values,
ideals, and interests. Moreover, although Ersner-
Hershfield, Garton et al.59 found a correlational
relationship between future self-similarity and in-
tertemporal choice, recent work has shown that per-
ceptions of similarity to the future self can be subtly
manipulated. Bartels and Urminsky,64 for example,
had participants read a passage describing research
in psychology that demonstrated stability in many
34 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
core aspects of identity over time. In another con-
dition, the passage described research that came to
the opposite conclusion. The researchers found that
participants in the stability condition rated a higher
degree of overlap with their future selves and also
made more patient intertemporal choices. In addi-
tion, Bartels and Urminsky64 demonstrated that fu-
ture self-similarity accounted for unique variance in
decision making, even when controlling for related
constructs such as uncertainty about the future, af-
fective appraisal of future outcomes, and the extent
to which one is biased toward the present.
These previous findings, however, have relied on
self-report to assess similarity to the future self,
and thus the credibility of the results could ben-
efit from complementary research that uses fewer
explicit techniques. To this end, my colleagues and
I drew on previous research in the field of social
neuroscience to more implicitly gauge future self-
similarity.
Neurobiological basis of future self-similarity
There is a growing body of neuroimaging research
that suggests that the human brain differentially
represents thoughts about oneself and thoughts
about another person.65,66 Previous work suggests
that people show decreased activation in cortical
midline structures when considering information
about others versus the self,65,67 and increased ac-
tivation when engaging in self-reflection or intro-
spection.68–70 In an early examination of this topic,
Kelley et al.66 scanned subjects with event-related
functional magnetic resonance imaging (fMRI) as
they judged whether trait words applied to them-
selves or another person. The investigators found
that judgments of self-relevance selectively main-
tained activation in the medial prefrontal cortex
(MPFC) at a baseline rate, while judgments of
other-relevance decreased MPFC activation below
baseline.
More broadly, Northoff et al.67 found that
processing other- versus self-relevant information
elicited decreased activation in cortical midline
structures, including the MPFC and rostral ante-
rior cingulate (rACC). These neural differences are
not only a function of thinking about the self ver-
sus another person; they also emerge when think-
ing about the self in the past: D’Argembeau et al.71
found that processing past self-relevant information
versus current self-relevant information decreased
activation in these same cortical midline structures.
Furthermore, the neural differences that normally
arise from comparing self to another are attenuated
to the extent that the target other is perceived as be-
ing similar to the self. Mitchell et al.,72 for example,
found that a ventral region of the MPFC was more
strongly active when participants made judgments
about the mental states of others who were perceived
to be similar to oneself compared to those perceived
to be dissimilar from oneself.
We wanted to leverage these previous findings
to further test the relationship between future self-
similarity and intertemporal choice.73 If people
think of the future self as being dissimilar from
the current self, judgments about the future ver-
sus current self should elicit reduced activation in
cortical midline structures (just as judgments about
another person compared to judgments about the
self elicit a similar pattern of activation). Further,
on an individual level, the greater the difference
between activation elicited by the future self and ac-
tivation elicited by the current self, the more future
rewards should be discounted. The more the future
self is represented neurally as another person, in
other words, the less likely a given individual should
be to save for the future. To test these hypotheses,
we scanned research participants with event-related
fMRI while they made judgments about the appli-
cability of trait adjectives to their current self, their
future self, a current other, or a future other. A week
later,par ticipants returned to the lab to complete the
temporal discounting task we used in the laboratory
studies mentioned above.61
Replicating previous research, we found that there
was greater activation in the MPFC and rACC for
self-judgments compared to other-judgments. Im-
portantly, within this same region of the brain,
thoughts about the current self elicited significantly
greater activation than did thoughts about the future
self. On a general level, then, neural activation asso-
ciated with thinking about the future self showed a
similar pattern to activation associated with think-
ing about another person (see Fig. 2A and B).
As expected, there was individual variability in
these neural differences: for some participants,
thinking about the future self elicited neural activa-
tion patterns that were almost exactly like patterns
that were associated with thinking about another
person; for other participants, thinking about the
future self elicited neural activation patterns that
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 35
Future self-continuity and choice Hershfield
Figure 2. Neural activation differences between current self and future self-trials correlate with discounting rates. (A) Greater
activation in the rostral anterior cingulate cortex (rACC) for current self versus future self trials; threshold P<.005, uncorrected.
(B) Activation time courses for each condition in the right rACC volume of interest; CS, current self; FS, future self; CO, current
other; FO, future other. (C) Scatterplot of individual differences in discount rates (log(k) +10) and individual differences between
peakcurrentselfandfutureselfactivationintherACCvolumeofinterest(r=0.47, P<0.05). Note: For display purposes and ease
of interpretation, a constant of 10 was added to the log(k) values. Figure adapted from Ref. 72.
were more or less in line with patterns associated
with thinking about the current self. In line with
our prediction, participants who showed the biggest
difference between activation elicited by the current
self and activation elicited by the future self also
discounted future rewards most steeply. There was
a positive correlation, in other words, between the
extent to which the future self “looked” like an-
other person on a neural level, and the degree to
which participants chose to consume smaller re-
wards in the present rather than delay larger re-
wards for the future (see Fig. 2C). In general, this
line of work identifies the neural systems support-
ing current self–future self overlap and points to
one neurobiological mechanism that contributes to
far-sighted behavior.
Vividness of the future self
In an early examination of the relationship between
vivid perceptions of the future self and intertempo-
ral choice, Klineberg74 measured the extent to which
adolescents were able to imbue future events in their
lives with a sense of reality by asking them to list a
number of typical life events in chronological order.
The more these participants perceived their futures
to follow a realistic and vivid course, the more likely
they were to delay gratification on a subsequent in-
tertemporal choice task (i.e., they opted to wait to
consume a larger candy bar in a week’s time rather
than eat a smaller candy bar immediately).
Such work, however, did not manipulate the ex-
tent to which the future self was seen as vivid, and
the extent to which such manipulations influence
saving decisions. My colleagues and I75 thus con-
ducted several experiments to examine the associ-
ation between a vivid perception of one’s self in
the future and the propensity to save more for the
long term. We used a combination of software and a
novel technology, immersive virtual reality (VR), to
make the perception of the future self more realis-
tic. Namely, we used preset algorithms from a com-
puter software package [Facegen Modeler76], which
(i) locates key points on a research participant’s face
from a photograph, (ii) builds a three-dimensional
model of that face, and (iii) morphs the shape and
texture of the model to simulate the aging process
to create a persuasive visual analog of a 68-year-old
version of a current college student (see Fig. 3 for
an example of this procedure). To create the aged
avatars—or digital representations of people—the
age-progression algorithm of the FaceGen Modeler
software package was applied with identical settings
to each photo. Because the softwareis specifical lyde-
signed to manipulate facial features and not hair, an
artist next digitally retouched each image to change
the original hair color from the photograph of the
participant to gray. We used an identical procedure
(i.e., creating the three-dimensional model) for the
nonaged avatars, except that we did not use the ag-
ing algorithm nor did we change the color of the
hair.
In one study, college undergraduates entered a
VR environment in which they could view images
ofthemselvesinavirtualmirror.
77 A control group
of college undergraduates viewed a digital repre-
sentation of their current selves in a virtual mirror,
while the experimental group of college undergrad-
uates saw an age-progressed version of themselves in
the mirror. While viewing these avatars, all partic-
ipants spent approximately five minutes answering
36 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
Figure 3. Example of age progression procedure. (Left) Photo of the author; (Middle) computerized rendering of the author using
Facegen and Photoshop; and (Right) age-progressed version of the same photo.
questions about themselves that have previously
been shown to enhance identification with a viewed
avatar.78 Upon exiting the VR environment, all
participants were given, among other tasks, a hypo-
thetical monetary allocation task. As hypothesized,
those participants who were exposed to their fu-
ture selves were subsequently more likely to allocate
money toward a hypothetical retirement savings ac-
count than were control participants.
It is possible, however, that participants who were
shown the age-progressed images of themselves were
merely primed with the concept of aging, and this
primepromptedthemtosavemoreforretirement.
79
As such, in a follow-up study, we exposed partici-
pants to either their own aged avatar or another
research participant’s aged avatar, and did so after a
lengthy experimental delay during which time par-
ticipants completed filler tasks (to eliminate the pos-
sibility that results from our first study were simply
due to demand effects). Results indicated that par-
ticipants in the future self condition demonstrated
significantly more patience on a series of intertem-
poral choice tasks than did participants in the future
other condition. These results suggest that relevance
matters: the extent to which the individuals identify
with the aged image is a key component of positively
affecting saving behavior.
Although these findings are encouraging, this ap-
proach to VR is expensive and time-consuming for
participants, and most companies or banks will not
be able to use immersive VR to convince their em-
ployees or customers to adopt a longer-term per-
spective when making decisions about retirement
savings. Moreover, experimental condition partic-
ipants in the studies mentioned above were sim-
ply shown a neutral image of their future selves.
Loewenstein’s19 work suggests that in order to make
a true connection to the future self the future emo-
tional consequences of current decisions must be
made clear. Indeed, previous research has demon-
strated that exposure to virtual cause-and-effect
actions can change actual behavior. For example,
compared to a control group, when participants
were shown a virtual version of themselves gain-
ing weight, they were more likely to go to the gym.80
Accordingly, in two additional studies, my colleague
Dan Goldstein and I used a more accessible format
to address the cause-and-effect nature of retirement
decision making.75
Namely, we used a web-based study design in
which all participants were shown a slider bar that
they could move to make allocations from a hy-
pothetical paycheck to a hypothetical retirement
account. As they moved the slider bar toward fu-
ture consumption, their annual take-home pay de-
creased (indicated in today’s dollars), but their an-
nual retirement income increased (again, indicated
in today’s dollars). In one condition, participants
were shown the monetary amounts as well as an
image of their current self, which changed emo-
tional expression as a function of the allocations
that they chose to make (sadder as more money
was allocated toward future consumption, happier
as more money was allocated toward present con-
sumption). In another condition, participants were
shown the monetary amounts as well as an image
of their future self, which changed emotional ex-
pression as a function of the allocations that they
chosetomake(happierasmoremoneywasallo-
cated toward future consumption, sadder as more
money was allocated toward present consumption).
Again, results indicated that participants who were
exposed to their future selves allocateda sig nificantly
higher percentage of pay toward retirement than
did participants who were exposed to their current
selves.75
It is possible, however, that instead of results of
the previous study being due to exposure to the
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 37
Future self-continuity and choice Hershfield
future self, they were due to participants reacting to
the valence the different faces present. Rather than
being motivated to save more by the presence of
the future self, participants could have merely been
moving the retirement slider bar toward whichever
face was smiling at them. To correct for this poten-
tial confound, we conducted an additional experi-
ment that was identical in form, except that instead
of being presented with emotional images of their
current or future selves, participants were exposed
to neutral versions of these selves (i.e., faces that
neither frowned nor smiled). This design also al-
lowed us to explore whether, in this nonimmersive
environment (i.e., the web-based design) an emo-
tional image of the future self was necessary to en-
hance saving behavior; as in the VR studies, could
saving be boosted after exposure to an image of a
future self with a neutral expression? Furthermore,
to test for generalizability, we used working-aged
adults as participants instead of college undergrad-
uates. Once more, participants elected to contribute
a significantly higher percentage of pay toward a re-
tirement account when they were exposed to their
future selves than when they were not.
Positivity of the future self
When taking into account how positive views of the
future self can influence saving behavior, consider
that in an earlier examination of this relationship, we
directly asked research participants how much they
liked and cared for their future selves.59 Neither con-
struct, however, mapped robustly onto intertempo-
ral choice. This may not be surprising given that
most healthy adults have a bias toward seeing them-
selves improving over time.81 Indeed,par ticipants in
the Ersner-Hershfield et al.59 study seemingly held
positive opinions of their future selves: both liking
and caring variables were positively skewed and nei-
ther exhibited substantial variance. Yet, for any given
topic (including positivity of the future self), peo-
ple may quite obviously maintain different attitudes
than the ones they report holding.82 One method of
eliciting more variance on a response item, then, is to
ask people how they feel about others in the domain
of interest. Accordingly, in recent work, Galinsky
and I have attempted to ascertain feelings about the
future self by assessing and manipulating attitudes
about the elderly in general.83 Namely, we examined
one aspect of positivity, the degree to which people
respected (or disrespected) older people.
Our first study examined the relationship be-
tween respect for the elderly and savings at a macroe-
conomic level.83 L¨
ockenhoff et al.84 surveyed par-
ticipants from 26 countries around the world about
their attitudes toward the elderly, and asked par-
ticipants to indicate in general how positively or
negatively their society viewed the elderly. Using the
same years in which the original L¨
ockenhoff sur-
vey was conducted (2006–2007), we mapped these
country level attitudes onto gross national saving
rates (expressed as a percentage of gross domes-
tic product), as reported by the World Bank.85 We
found a positive relationship between attitudes to-
ward the elderly and national saving rates, and im-
portantly, this association remained robust when
controlling for a host of relevant variables (percent
of population over age 65, amount of contact with
the elderly, the GINI index, literacy rates, and land
size).
Using experimental designs, we then assessed the
causal relationship between respect toward the el-
derly and personal saving behavior. Because previ-
ous research has shown that perspective-taking en-
hances respect and positive attitudes toward social
groups,86 we showed participants a photograph of
an individual (who was either an elderly or young
male) and asked them to write a day-in-the-life
narrative of this photographed individual by either
taking his perspective or remaining detached. The
experiment had three conditions: (1) take the per-
spective of an elderly man, (2) take the perspective
of a young man, or (3) take a detached objective
perspective of an elderly man. After the perspective-
taking manipulation, all participants completed the
Kirby and Marakovic61 measure that my colleagues
and I used in earlier work.59 Participants in the
elderly perspective-taking condition chose signifi-
cantly more larger later rewards than did partici-
pants in the other two conditions; the detached per-
spective of the elderly and the perspective-taking
of the young man did not differ significantly from
each other. If these findings were a function of sim-
ply thinking about the elderly, then we would have
observed elevated saving behavior in the elderly de-
tached perspective condition. Similarly, if these re-
sults were merely due to taking any person’s perspec-
tive, we would have observed increased saving in the
young perspective-taking condition. Rather, a boost
in saving behavior only occurred when participants
took the perspective of an elderly individual.
38 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
Our next experiment directly manipulated re-
spect for the elderly and examined its effects on
saving by randomly assigning participants to one of
four conditions: participants wrote about a time that
they showed respect or disrespect toward a peer or
an elderly individual. Participants then completed
the same temporal discounting task from the previ-
ous studies. Again, participants in the elderly respect
condition opted for more larger later rewards than
did participants in any of the other three condi-
tions; the elderly disrespect, peer-respect, and peer-
disrespect conditions did not differ from each other.
Only when participants thought about a time they
respected the elderly did we observe increased saving
behavior relative to all other conditions.
In a follow-up study, we examined saving over
a longer time frame and also used a manipulation
that was considerably more implicit than the previ-
ously described manipulations. Drawing on seman-
tic priming procedures,87 we presented undergrad-
uates with a word search task (tables that contained
words hidden among rows and columns of letters)
that included either words related to respect (e.g.,
“esteemed”), words related to the elderly (e.g., “el-
derly”), or words related to both the elderly and
respect (e.g., “elderly”, “esteemed”). Indeed, previ-
ous work has found that when words related to two
separate concepts are put into the same word puz-
zle, implicit links between the two concepts are cre-
ated.88 In a control condition, participants did not
complete a word-search task. Participants in all four
conditions then answered a version of the Kirby
and Maracovic61 task in which they made choices
between smaller rewards they could obtain imme-
diately (e.g., $2,694 tonight) and larger rewards that
they could obtain at a significant delay (e.g., $93,050
in 39 years). Demonstrating the robust link between
positive perceptions of the elderly and saving behav-
ior, participants primed with words related to the el-
derly and respect opted for significantly more larger
later rewards than did participants in the other three
conditions.
Remaining questions and directions
for future research
The research reviewed above provides initial evi-
dence for the ways in which different conceptions
of the future self can affect intertemporal decision
making. Yet, several questions still remain. Below, I
review a sampling of them that provide possibilities
for future research.
Relationships among similarity, vividness,
and positivity
To start, the relationship among the concepts of
similarity, vividness, and positivity is largely un-
known. In the VR studies, exposure to images of
the future self enhanced feelings of similarity to that
self.75 We do not yet know, though, whether ma-
nipulations that increase future self-similarity also
produce more vivid perceptions of the future self.
Theoretically, in order to judge whether one is sim-
ilar to the future self, that future self may have to
possess some degree of vividness and realism. Alter-
natively, it is possible that judgments of similarity
do not necessarily rely on vividness and are instead
a function of characteristics of the future self that
are less concrete.
With respect to similarity and positivity,we found
a significant correlation, albeit a relatively weak one,
between similarity to the future self and how much
one liked the future self.59 Although liking and gen-
eral positivity are not necessarily the same concept,
it is not difficult to imagine a positive bidirectional,
causal relationship between similarity and positiv-
ity: people often feel similar to (or at least want to feel
similar to) those whom they hold in high regard,89
and they like and respect people with whom they
feel similar.90 Future work could benefit from or-
thogonal manipulations of both of these constructs.
It is less clear, though, whether exercises that en-
hance perceptions of vividness can also lead to more
positive views of the future self and vice versa. The
two might be associated, but the strength of the
connection may depend, in large part, on the effec-
tiveness of the vividness manipulation. A degraded
or extremely unattractive image of the future self
might, not surprisingly, lead to a less positive con-
ception of that self. Figure 4 depicts one possible
model of the relationships among these three com-
ponents of future self-continuity (similarity, vivid-
ness, positivity) and intertemporal decision making.
Future research should certainly test which of these
links exists and which does not.
Changes in future self-continuity with age
My colleagues and I have found that future self-
similarity becomes positively skewed as people grow
older.59 As noted earlier, such a shift is somewhat
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 39
Future self-continuity and choice Hershfield
Figure 4. A possible model of future self-continuity. Untested relationships are represented by dashed lines.
expected given that life circumstances become more
stable with age and many major life choices have al-
ready been decided in early adulthood (e.g., who to
marry, what career to pursue, where to live, etc.).
Nonetheless, there are still several important in-
tertemporal choices that arise toward the end of
life that may be impacted by future self-continuity.
Upon entering retirement, for example, one must
decide the optimal way to use the money that has
(hopefully) accumulated in one’s retirement ac-
counts over time.91
Certainly, one factor that influences retirement
spending choices is an individual’s perceived sense
of time left in life.92 However, the way the future self
is perceived may also play a role in these spending
decisions, and in line with the research presented
above, similarity, positivity, and vividness may all
influence decision making during this decumula-
tion phase of life. But whether they affect intertem-
poral decisions in the same way that they do for
younger adults is an area of inquiry that should be
undertaken by future research.
Moreover, it is possible that similarity, vividness,
and positivity may not be the only aspects of fu-
ture self-continuity that matter when one is older.
Namely, to the extent that close loved ones are often
part of the self-concept,93 the image of the future self
at the end of life may not be confined to just the self,
but may also include one’s children and grandchil-
dren. In this way, the future self may continue exist-
ing after one has died in the form of one’s offspring.
The extent that they—that is, the future kin—are
seen as similar, vivid, and positive, could influence
saving and spending by influencing decisions re-
garding intergenerational transfers of wealth. If a
given individual perceives continuity between him-
self and his children, then he may be more motivated
to continue saving for their benefit;94 if, on the other
hand, no continuity is perceived, he may be more
inclined to spend in the present.
Connections to the past self
Promoting self-continuity may help people not only
prepare better for the future, but also come to better
terms with the past. Theoretically, as people move
through the life course and are faced with major
life decisions, more opportunities for potential re-
gret accumulate.95,96 Yet, not everyone is paralyzed
by the negative emotions that are associated with
regret.97
One explanatory factor for these individual dif-
ferences may be the way in which people relate to
their past selves (i.e., the selves that made regrettable
decisions). Just as a stronger sense of continuity to
one’s future self could promote saving behavior, so
could a stronger sense of continuity with one’s past
self promote fewer severe regrets? It is possible that
being able to take the perspective of one’s past self
might lead to better understanding and ultimately
acceptance with past regrettable decisions.
Conclusions
Our differing conceptions of our future selves can
dramatically influence the long-term choices that
we make. Future work will hopefully shed light on
the various ways that similarity, vividness, and pos-
itivity interact so that future self-continuity, and its
downstream effects on decision making, can be best
understood. Nonetheless, the empirical work pre-
sented in this review suggests that when the future
self seems similar to the present self, when it is im-
bued with realism and vividness, and when it is seen
in positive terms, people are more willing to make
sacrifices today that may benefit them at some point
in the years to come.
Acknowledgments
The author is grateful to Jennifer Heil, Garriy
Shteynberg, and two anonymous reviewers for ex-
tremely helpful comments on earlier versions of this
article.
40 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
Conflicts of interest
The author declares no conflicts of interest.
References
1. Arias, E. 2007. United States life tables, 2004. Natl. Vital Stat.
Rep.56: 1–40.
2. Carstensen, L.L. 2009. A Long Bright Future: An Action Plan
for a Lifetime of Happiness, Health, and Financial Security.
Random House. New York, NY.
3. Shultz, G.P. & J.B. Shoven. 2008. Putting Our House in Or-
der: A Guide to Social Secruity and Heal Care Reform.W.W.
Norton & Company, Inc. New York, NY.
4. Barrett, V. 2009. New age thinker. In Forbes 184: 22–23.
5. OECD. 2011. Pensions at a Glance 2011: Retirement-Income
Systems in OECD and G20 Countries., Vol. 2011.
6. Munnell, A.H., A. Webb & F. Golub-Sass. 2007. Is There
Really a Retirement Savings Crisis?: An NRRI Analysis.Center
for Retirement Research,Boston College. Chestnut Hill, MA.
7. Munnell, A.H., A. Webb &F. Golub-Sass. 2009. The National
Retirement Risk Index: After the Crash.CenterforRetirement
Research,BostonCollege.ChestnutHill,MA.
8. McKinsey Global Institute. 2008. Tal k i n ’ ‘ B o u t My G e n e r-
ation: The Economic Impact of Aging U.S. Baby Boomers.
McKinsey & Co. San Francisco, CA.
9. Helman, R., C. Copeland & J. VanDerhei. 2011. The 2011
retirement confidence survey: confidence drops to record
lows, reflecting “the new normal”. Employee Benefit Res. Inst.
Issue Brief 355: 1–40.
10. Chapman, G.B. 1996. Temporal discounting and utility for
health and money. J. Exp. Psychol. Learn. Mem. Cogn.22:
771–791.
11. Frederick, S. 2002. Time preference and personal identity.
In Time and Decision: Economic and Psychological Perspec-
tives on Intertemporal Choice. G. Loewenstein, D. Read &
R.F. Baumeister, Eds.: 89–113. Russell Sage Foundation.
New York.
12. Lynch, J.G. & G. Zauberman. 2006. When do you want it?
Time, decisions, and public policy. J. Public Policy Mark. 25:
67–78.
13. Soman, D., et al. 2005. The psychology of intertemporal
choice: why are distant events valued differently from prox-
imal ones? Marketing Lett.16: 347–360.
14. Morello, C. 2010. About 44 million in U.S. Lived Below
Poverty Line in 2009, Census Data Show. Washington Post.
Washington, DC.
15. Choi, J.J., D. Laibson & B.C. Madrian. $100 bills on the
sidewalk: suboptimal investment in 401(k) plans. Rev. Econ.
Stat. 93: 748–763.
16. Mischel, W., E.B. Ebbesen & A. Raskoff Zeiss. 1972. Cog-
nitive and attentional mechanisms in delay of gratification.
J. Pers. Soc. Psychol. 21: 204–218.
17. Mischel, W. 1974. Processes in delay of gratification. In Ad-
vances in Experimental Social Psychology. Vol. 7. L. Berkowitz,
Ed.: 249–292. Academic Press. New York.
18. Mischel, W., Y. Shoda & M.L. Rodriguez. 1989. Delay of
gratification in children. Science 244: 933–938.
19. Loewenstein, G. 1996. Out of control: visceral influences on
behavior. Organ. Behav. Hum. Decis.65: 272–292.
20. Trope, Y. & N. Liberman. 2000. Temporal construal and
time-dependent changes in preference. J. Pers. Soc. Psychol.
79: 876–889.
21. Trope, Y. & N. Liberman. 2003. Temporal construal. Psychol.
Rev. 110: 403–421.
22. Fujita, K., et al. 2006. Construal levels and self-control.
J. Pers. Soc. Psychol. 90: 351–367.
23. Frederick, S., G. Loewenstein & T. O’Donoghue. 2002. Time
discounting and time preference: a critical review. J. Econ.
Lit. 40: 351–401.
24. Laibson, D., A. Repetto & J. Tobacman. 1998. Self-control
and saving for retirement. Brookings Pap. Econ. Act. 1998:
91–196.
25. Schelling, T.C. 1982. Ethics,Law,andtheExerciseofSelf-
Command. Harvard Institute of Economic Research.Boston,
MA.
26. Wilson, T.D. & D.T. Gilbert. 2003. Affective forecasting. In
Advances in Experimental Social Psychology, Vol. 35. M.P.
Zanna, Ed.: 345–411. Academic Press. San Diego.
27. Gilbert, D.T., et al. 1998. Immune neglect: a source of dura-
bility bias in affective forecasting. J. Pers. Soc. Psychol. 75:
617–638.
28. Gilbert, D.T. &T.D.Wilson. 2007. Prospection: experiencing
the future. Science 317: 1351–1354.
29. Benabou, R. & M. Pycia. 2002. Dynamic inconsistency and
self-control: a planner-doer interpretation. Econ. Lett. 77:
419–424.
30. Benhabib, J. & A. Bisin. 2005. Modelling internal commit-
ment mechanisms and self-control: a neuroeconomics ap-
proach to consumption-saving decisions. Games Econ. Be-
hav. 52: 460–492.
31. Brocas, I. & J. Carillo. 2008. The brain as a hierarchical
organization. Am. Econ. Rev.98: 1312–1346.
32. Fudenberg, D. & D.K. Levine. 2006. A dual self model of
impulse control. Am. Econ. Rev.96: 1449–1476.
33. Thaler, R.H. & H.M. Shefrin. 1981. An economic theory of
self-control. J. Polit. Economy 89: 392–406.
34. O’Donoghue, T. & M. Rabin. 2000. The economics of im-
mediate gratification. J. Behav. Decis. Making 13: 233–250.
35. Hoch, S.J. & G. Loewenstein. 1991. Time-inconsistent
preferences and consumer self-control. J. Cons. Res. 17:
492–507.
36. Schelling, T.C. 1984. Self-command in practice, in policy,
and in a theory of rational choice. Am. Econ. Rev. 74: 1–11.
37. Elster, J. 1977. Ulysses and Sirens: a theory of imperfect
rationality. Soc. Sci. Inform.16: 469–526.
38. Thaler, R.H. & S. Benartzi. 2004. Save More Tomorrow
(TM): using behavioral economics to increase employee sav-
ing. J. Polit. Economy 112: S164–S187.
39. Parfit, D. 1971. Personal Identity. Philos. Rev. 80: 3–27.
40. Parfit, D. 1987. Reasons and Persons. Clarendon Press.
Oxford.
41. Strotz, R.H. 1956. Myopia and inconsistency in dynamic
utility maximization. Rev. Econ. Stud.23: 165–180.
42. Proust, M. 1927/1981. A la recherche du temps perdu (A
remembrance of things past).RandomHouse,Inc.NewYork,
NY.
43. Butler, J. 1736. The Analogy ofRelig ion, Naturaland Revealed,
totheConstitutionandCourseofNature:ToWhichareAdded
Two Brief Dissertations: I. Of Personal Identity. II. Of the
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 41
Future self-continuity and choice Hershfield
Nature of Virtue. Printed for James John and Paul Knapton.
London.
44. Pronin, E. & L. Ross. 2006. Temporal differences in trait
self-ascription: when the self is seen as an other. J. Pers. Soc.
Psychol. 90: 197–209.
45. Pronin, E., C.Y. Olivola & K.A. Kennedy. 2008. Doing unto
future selves as you would do unto others: psychological
distance and decision making. Pers. Soc. Psychol. Bull.34:
224–236.
46. Gammill, T. & M. Pross. 1993. The Glasses. Seinfeld. Sony
Pictures, Inc. New York, NY.
47. Roberts, B.W. & W.F. DelVecchio. 2000. The rank-order con-
sistency of personality from childhood to old age: a quantita-
tive review of longitudinal studies. Psychol. Bull.126: 3–125.
48. Low, D.K., et al. 2005. The stability of interests from early
adolescence to middle adulthood: a quantitative review of
longitudinal studies. Psychol. Bull.131: 713–737.
49. Wakslak, C.J., et al. 2008. Representations of the self in the
near and distant future. J. Pers. Soc. Psychol.95: 757–773.
50. Jones, B. & H. Rachlin. 2006. Social discounting. Psychol.
Sci. 17: 283–286.
51. Eisenberg, N. & P.A.Miller. 1987. The relation of empathy to
prosocial and related behaviors. Psychol. Bull.101: 91–119.
52. Galak, J., D.A. Small & A.T. Stephen. 2010. Micro-finance
decision making: a field study of prosocial lending. J. Mar-
keting Res. In press.
53. Tversky, A. & D. Kahneman. 1973. Availability – heuristic
for judging frequency and probability. Cogn. Psyc hol.5: 207–
232.
54. Small, D.A. & G. Loewenstein. 2003. Helping avictim or
helping the victim: altruism and identifiability. J. Risk Un-
certainty 26: 5–16.
55. Kogut, T. & T. Ritov. 2010. The identifiable victim effect:
causes and boundary conditions. In The Science of Giv-
ing: Experimental Approaches to the Study of Charity.D.M.
Oppenheimer, & C.Y. Olivola, Eds.: 133–148. Taylor and
Francis. New York.
56. Levy, B.R., et al. 2009. Age stereotypes held earlier in life
predict cardiovascular events in later life. Psychol. Sc i.20:
296–298.
57. Levy, B.R., et al. 2002. Longevity increased by positive self-
perceptions of aging. J. Pers. Soc. Psychol.83: 261–270.
58. Waugh, C.E. & B.L. Fredrickson. 2006. Nice to know you:
positive emotions, self-other overlap, and complex under-
standing in the formation of a new relationship. J. Positive
Psychol.1: 93–106.
59. Ersner-Hershfield, H., et al. 2009. Don’t stop thinking about
tomorrow: individual differences in future self-continuity
account for saving. Judgment Dec. Making 4: 280–286.
60. Aron, A., E.N. Aron & D. Smollan. 1992. Inclusion of other
in the self scale and the structure of interpersonal closeness.
J. Pers. Soc. Psychol.63: 596–612.
61. Kirby, K.N. & N. Marakovic. 1996. Modeling myopic de-
cisions: evidence for hyperbolic delay-discounting within
subjects and amounts. Organ. Behav. Hum. Decis.64:
22–30.
62. Lea, S.E.G., R.M. Tarpy & P. Webley. 1987. The Individual in
the Economy: A Textbook ofE conomic Psychology. Cambridge
University Press. New York, NY.
63. Bartels, D.M. & L.J. Rips. 2010. Psychological connected-
ness and intertemporal choice. J. Exp. Psychol. Gen.139:
49–69.
64. Bartels, D.M. & O. Urminsky. 2011. On intertemporal self-
ishness: the perceived instability of identity underlies impa-
tient consumption. J. Cons. Res.38: 182–198.
65. Amodio, D.M.& C.D. Frith. 2006. Meeting of minds: the me-
dial frontal cortex and social cognition. Nat. Rev. Neurosci.
7: 268–277.
66. Kelley, W.M., et al. 2002. Finding the self? An event-related
fMRI study. J. Cogn. Neurosci.14: 785–794.
67. Northoff, G., et al. 2006. Self-referential processing in our
brain–a meta-analysis of imaging studies on the self. Neu-
roimage 31: 440–457.
68. Johnson, S.C., et al. 2002. Neural correlates of self-reflection.
Brain 125: 1808–1814.
69. Gusnard, D.A., et al. 2001. Medial prefrontal cortex and
self-referential mental activity: relation to a default mode of
brain function. Proc.Natl.Acad.Sci.USA98: 4259–4264.
70. Raichle, M.E., et al. 2001. A default mode of brain function.
Proc. Natl. Acad. Sci. USA 98: 676–682.
71. D’Argembeau, A., et al. 2008. Self-reflection acros time: cor-
tical midline structures differentiate between present and
past selves. Soc. Cogn. Affect. Neurosci.3: 244–252.
72. Mitchell, J.M., C.N. Macrae & M.R. Banaji. 2006. Dissocia-
ble medial prefrontal contributions to judgments of similar
and dissimilar others. Neuron 50: 655–663.
73. Ersner-Hershfield, H., G.E. Wimmer & B. Knutson. 2009.
Saving for the future self: neural measures of future self-
continuity predict temporal discounting. Soc. Cogn. Affect.
Neurosci.4: 85–92.
74. Klineberg, S.L. 1968. Future time perspective and preference
for delayed reward. J. Pers. Soc. Psychol.8: 253–257.
75. Hershfield, H.E., et al. Increasing saving behavior through
age-progressed renderings of the future self. J. Marketing
Res. In press.
76. Singular Inversions, I. 2004. Facegen Modeller 3.1. Vancou-
ver, B.C.
77. Blascovich, J. & J.N. Bailenson. 2011. Infinite Reality .
William Morrow. New York, NY.
78. Yee, N. & J.N. Bailenson. 2007. The Proteus effect: self trans-
formations in virtual reality. Human Commun. Res.33: 271–
290.
79. Bargh, J.A. & T.L. Chartrand. 1999. The unbearable auto-
maticity of being. Am. Psychol. 54: 462–479.
80. Fox, J. & J.N. Bailenson. 2008. Virtual exercise in the third
person: identifcation, physical similarity, and behavioral
modeling. Presented at Communicaton and Technology
Commission of the International Communication Associ-
ation, Montreal, Canada.
81. Wilson, A. & M. Ross. 2001. From chump to champ: people’s
appraisals of their earlier and present selves. J. Pers. Soc.
Psychol.80: 572–584.
82. Wilson, T.D. 2002. Strangers to Ourselves: Discovering the
Adaptive Unconscious. Harvard University Press.Cambridge,
MA.
83. Hershfield, H.E. & A.D. Galinsky. 2011. Respect for the El-
derly Predicts National Savings and Individual Saving Deci-
sions.NorthwesternUniversity.Evanston,IL.
42 Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences.
Hershfield Future self-continuity and choice
84. Lockenhoff, C.E., et al. 2009. Perceptions of aging across 26
cultures and their culture-level associates. Psychol. Ag ing 24:
941–954.
85. The World Bank. 2010. Gross Savings (% of GDP).
Vol. 2010.
86. Galinsky, A.D. & G.B. Moskowitz. 2000. Perspective-
taking: decreasing stereotype expression, stereotype acces-
sibility, and in-group favortism. J. Pers. Soc. Psychol.78:
708–724.
87. Bargh, J.A., M. Chen & L. Burrows. 1996. Automatic-
ity of social behavior: direct effects of trait construct and
stereotype activation on action. J. Pers. Soc. Psychol.71: 230–
244.
88. Lammers, J., et al. 2008. Illegitimacy moderates the effect of
power on approach. Psychol. Sci. 19: 558–564.
89. Hoffner, C. & M. Buchanan. 2005. Young adults’ wishful
identification with television characters: the role of perceived
similarity and character attributes. Media Psychol.7: 325–
351.
90. Byrne, D. 1961. Interpersonal attraction and attitude simi-
larity. J. Abnorm. Soc. Psychol.62: 713–715.
91. Scott, J.S., W.F. Sharpe & J.G. Watson. 2009. The 4% rule –
At what pr ice? J. Invest. Manage.7: 31–48.
92. Kalemli-Ozcan, S. & D.N. Weil. 2010. Mortality change, the
uncertainty effect, and retirement. J. Econ. Growth 15: 65–91.
93. Aron, A., et al. 1991. Close relationships as including other
in the self. J. Pers. Soc. Psychol.60: 243–253.
94. Gale, W.G. & J.K. Scholz. 1994. Intergenerational transfers
and the accumulation of wealth. J. Econ. Perspect.8: 145–160.
95. Bauer, I., C. Wrosch & I. Jobin. 2008. I’m better off than
most other people: the role of social comparions for coping
with regret in young adulthood and old age. Psychol. Aging
23: 800–811.
96. Wrosch, C. & J. Heckhausen. 2002. Perceived control of life
regrets: good for young and bad for old adults. Psychol. Ag ing
17: 340–350.
97. Gilovich, T. & V.H. Medvec. 1995. The experience of regret:
what, when, and why. Psychol. Rev.102: 379–395.
Ann. N.Y. Acad. Sci. 1235 (2011) 30–43 c
2011 New York Academy of Sciences. 43