Utility Perception in System Dynamics Models
Saeed P. Langarudi! and Isa Bar-On?
‘New Mexico State University
?Worcester Polytechnic Institute
July 14, 2017
Abstract
Utility that individuals perceive is believed to be different from the
utility that they actually experience. Sy implicitly cate-
gorize this phenomenon as a form of bounded rationality and traditionally
ystem dynamici
employ a simple smoothing function to capture it. We challenge this gen-
eralization by testing it against an alternative formulation of utility per-
ception that
particular, traditional smoothing formulation is compared with peak-end
uuggested by modern theories of behavioral economics. In
rule in a simple theoretical model as well as in a medium-size model of
electronic health record ii ion problem. Experi ion with
the models reveals that the way utility perception is formulated is impor-
tant and might affect behavior and policy implications of system dynamics
models.
Decisions regarding preferences involve the utility derived from the preferred
outcome. Utility in this context can be either experienced utility or decision
n be notable differen en the two (Kahneman et al.
Experienced utility is the pleasure (or its opposite) that we gain from an
action at any moment. Decision utility is revealed by the choices we make,
and is also called revealed preference. Experienced utility can be summarized
over time to obtain the total utility as determined by some valid measure. For
decision utility it has been shown that it isproportionally affected
by the peak and end values of an experience, while duration of experie
to be less significant (Fredrickson, Kahneman|, [Kahneman et al]
stem dynamics models represent utility-based decisions using the con-
struct of perceived utility. Perceived utility (or perc preference) is for-
mulated as a first order smooth. One example is the reaction of customers to
long delays in delivery. Customer satisfaction from delivery time (x) will be
perceived (2) with a delay (T):
scems
Spa = E+ SO dt (1)
T
This perception of utility, is usually symmetrical and uniformly:smoothed
over'a particular time period., Ag euch it is akin to the sum of experienced
utilities, ‘The representation of utilities used in system dynamics models is
developed from the use of amoothis ini the formulation of peresived information.
Perceived inforination includes: ai information delay (amooth) of thé actual
information to répresént the diserapancy between actual information and what
agents perceive as information. Decision utility is qualitatively different as it
includes psychological factors, These may lead to a disregard of time duration
and emphasis on extreme values and more recent experiences thus leading to
the Peak-End rule.
‘The Peak-End tule had ‘originally been documented for pain experienced
during medical procedures where’ patients werd
asked during the procedute to rate the pain that they expérienced during the
procédure in fixed intervals. After'the procedure ‘the patients ware asked to
rate thé total amount of pain that they had experienced during the procedure:
The results indicated that this later measure was not reflective of the sum of
s during the procedure, but-was better represented by the average of
the peak and last values—thus the term Peak-End value. This Peak-End rule
has 4i n documented for a range of situations (Stone et al
the valu
U_ - Perceived (remembered) utility
P - Utility from most intense experience (PEAK)
E - Utility from most recent experience (END)
This formulation is substantially different from that of an averaged expres-
sion of utility. In this paper we pose the question that whether or not alternative
formulations of utility functions for preference decisions will substantially alter
the results of the modeling exere
In this paper we apply Kahneman’s decision utility to two system dynamics
models and compare the results to those obtained from traditional perceived
utility formulations. Specifically, we use the peak-end rule for comparison with
the traditional formulation. This rule, a result from behavioral economies, has
been used successfully to describe preferences in a range of applications.
Experiments with a simple theoretical model that focuses on the different
utility formulations indicate that there may be
outcomes. The peak-end rule is then applied in a medium size model of elec-
health record (EHR) implementation. The model has many (positive and
negative) feedback loops and includes 7 utility perception equations and four
different payment schemes. One thousand Monte Carlo simulation runs are per-
formed for each formulation and for each of the four payment policies, yielding
8000 total runs. on between simulation runs across different models
and different seen in outputs of different models is
considerable. The extent of dis cy, however, depends on the initial setup
of the models. Results also show that it i likely that the two models lead
to diametrically opposed policy recommendations.
Based on these results we conclude that different utility formulations for
preferen tually matter. The point of this investigation was not to prove
that the peak-end rule is necessarily the better formulation for the EHR model.
Rather, we were trying to show that the formulation of the decision utility for
preferences may affect recommended policies. This situation is complicated as
studies in behavioral economics seem to indicate that decision utility can, in
addition, be influenced by time d, repeated experiences
chological factors for which there are no 2
formulations to date.
Despite the lack of consei
sibility of alternative formulatioi
nificant differences in model
tronic
ve
and other psy-
pted utility
in utility formulation, we argue that the pos-
such as the peak-end rule (or others) should
ases where there is a reason to assume that they might be
ion utility description or as a matter of good practice. These
different formulations should be investigated together with a wide parameter
space in order to identify potential policy recommendations that differ from
those obtained with the traditional information.
Much more work needs to be done to understand the effect of alternative
utility formulations on the outcomes of system dynamics models, and a good
starting point might be the investigation of generi s and their
tivity to such formulations.
be investigated in
the dominant de:
Finally, we recommend that alternative formulations for utility perception be
included in system dynamics software packages so that applications of theories
such as peak-end value become readily available to modelers.
References
Clark Andrew E., Georgellis Yannis. Kahneman meets the quitters: peak-end
behaviour in the labour market. 2004.
Fredrickson Barbara L., Kahneman Daniel. Duration neglect in retrospective
evaluations of affective episodes // Journal of Personality and Social Psy-
chology. VII 1993. 65, 1. 45. bibtex: fredrickson_duration_1993.
Kahneman Daniel. Experienced utility and objectiv
based approach // The ology of economic de
bibtex: kahneman_experienced_2003.
¢ happiness: A moment-
ons. 2003. 1. 187-208.
Kahneman Daniel, Fredrickson Barbara L., Schreiber Charles A., Redelmeier
Donald A. When More Pain Is Preferred to Less: Adding a Better End //
Psychological Science. XI 1993. 4, 6. 401-405.
Kahneman Daniel, Thaler Richard H. Anomalies: Utility Maximization and
Experienced Utility // Journal of Economic Perspectives. II 2006. 20, 1.
221-234. bibtex: kahneman_anomalies:_2006.
Kahneman Daniel, Wakker Peter P., Sarin Rakesh. Back to Bentham? Ex-
plorations of Experienced Utility // The Quarterly Journal of Economics. V
1997. 112, 2. 375-405. bibtex: kahneman.back_1997.
Langer Thomas, Sarin Rakesh, Weber Martin. The retrospective evaluation of
payment sequences: duration neglect and peak-and-end effects // Journal of
Economic Behavior & Organization. IX 2005. 58, 1. 157-175.
Miron-Shatz Talya. Evaluating multiepisode events: Boundary conditions for
the peak-end rule // Emotion. 2009. 9, 2. 206-213.
Redelmeier Donald A, Kahneman Daniel. Patients’ memories of painful medical
treatments: real-time and retrospective evaluations of two minimally invasive
procedures // Pain. VII 1996. 66, 1. 3-8.
Stone Arthur A., Broderick Joan B., Kaell Alan T., De
E. G., Porter Laura E. Does the peak-end phenomenon ob:
pain studies apply to real-world pain in rheumatoid arthritics?
of Pain. I 2000. 1, 3. 212-217.
aul Philippe A.
ved in laboratory
// The Journal
