Neuroeconomics
Media
Part of Neuroeconomics
- Title
- Neuroeconomics
- extracted text
-
Neuroeconomics
IFAT LEVY
Abstract
In recent years, researchers in economics, psychology, and neuroscience have joined
forces in the study of decision-making processes to form the new discipline of
neuroeconomics. Neuroscientists turned to theories in economics and psychology
to make sense of the increasing amounts of neurobiological data. At the same time,
economists and psychologists turned to neuroscience for mechanistic constraints on
their theories. Neuroeconomics studies tackle a host of topics, from financial choices
through reinforcement learning to social decision making. Combining behavioral
techniques with brain imaging in humans and electrophysiological recordings in
animals, as well as complementary techniques, this interdisciplinary research has
already generated new insights about the neural architecture of decision making.
The neural mechanisms of some of the behavioral decision processes are increasingly
understood, but many challenges remain. Extending neuroeconomics research to
psychiatric disorders and incorporating new research tools are promising avenues
for future studies.
THE HISTORY OF NEUROECONOMICS
INTRODUCTION
How does one choose between an apple and an orange? We can ask this
question at several different levels. At the behavioral level, we may be interested in accurately predicting individual choices. At the algorithmic level,
we can explore the neural mechanisms that underlie the observed choice
behavior. Yet another level of inquiry is examining the mental states that give
rise to the observed choice behavior. For centuries, researchers in economics,
neuroscience, and psychology have studied decision making as largely separate disciplines. Researchers in each discipline have employed very different
strategies to study one of these aspects of decision making.
Since the 1930s, neoclassical economists have essentially strove to predict
human choice behavior based on rigorous mathematical models. These models were typically “as if” models, in the sense that they were not required
to accurately describe the algorithm implementing the choice process, but
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.
1
2
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
rather to correctly predict behavior. As Paul Samuelson and others have come
to realize, making a small number of simple and reasonable assumptions,
or “axioms” about choice behavior, such as “if a person prefers apples to
oranges she will not also prefer oranges to apples,” allows us to describe the
decision maker’s behavior “as if” they are trying to maximize some “utility” function. On the basis of these axioms, Von Neumann and Morgenstern
(1944) developed the expected utility (EU) theory, a model for choice between
uncertain outcomes, which was later extended by Savage (1954) to take into
account the decision maker’s subjective estimation of outcome probability
(subjective EU).
At the same time that these economists were seeking to predict behavior,
neuroscientists were aiming to reveal the mechanisms underlying the same
behavior. For decades neurological research essentially depended on brain
lesions in humans and animals. In this approach, behavioral deficits are correlated with the particular brain injury in an attempt to infer the function of
the damaged neural system. Perhaps the most famous example is the 1848
case of Phineas Gage, a young railroad worker whose brain was penetrated
by a steel rod (Neylan, 1999). Despite his remarkable physical recovery following the accident, Gage has exhibited substantial changes in personality
and decision making, providing the first evidence in humans for the involvement of the prefrontal cortex, the part of the brain most affected by the accident, in decision-making processes. While brain lesions were very helpful
in identifying general associations between brain structures and particular
sensory, motor, or cognitive functions, they could provide little information
about the neural algorithms that are implemented in each of these structures.
Advances in both economic theory and neurobiological techniques during
the second half of the twentieth century led researchers in both of these disciplines to look to psychology, and eventually toward each other, in what
turned out to be the birth of neuroeconomics.
On the economics side, economists began to note examples of choice behavior which was not compatible with neoclassical economic theory. In several
cases, human choices violated one or more of the core axioms of the theory.
Allais was the first to describe such violation in what is known as the “Allais
paradox” (Allais, 1953), followed by Ellsberg, who described the “Ellsberg
paradox” (Ellsberg, 1961). In a series of seminal studies published in the 1970s
and 1980s Kahneman and Tversky have documented numerous substantial
behavioral deviations from EU theory, demonstrating that these deviations
were the rule rather than the exception (Tversky & Kahneman, 1974, 1981).
The work of Kahneman and Tversky suggested to many economists and psychologists that economic models could benefit from psychological data and
insights. These economists and psychologists formed the discipline of behavioral economics, a union of economics and psychology.
Neuroeconomics
3
While these processes were taking place in the economics world, neuroscientists were also experiencing considerable transformations in their
research. The introduction of novel data collection techniques markedly
increased the ability of neuroscientists to investigate the neural circuits
underlying higher cognitive functions. The first advance was made in the
1960s, when recording of neural activity in the brains of awake behaving animals became available. The next development occurred in the 1980s–1990s,
when noninvasive imaging methods in humans, in particular positron
emission tomography (PET) and functional magnetic resonance maging
(fMRI) became widely used. These tools enabled neuroscientists to examine
dynamic changes in neural activity while humans and other animals were
engaged in complex behavior. The ability to record the activity of single neurons in awake behaving animals, and to image the activity of populations of
neurons in humans, allowed much more than a simple correlation between
neural activation and observed behavior. The analysis tools for making use
of these new techniques and the large quantities of data that they generated,
however, were not well developed. Similar to economists, neuroscientists
have also begun to look to psychology, in order to use models of cognitive
function in designing their experiments and analyzing their data. Instead of
correlating a damaged brain area with impaired cognitive function, neuroscientists could now look for correlations between neural activity and hidden
variables of the mental models. These studies gave rise to the new discipline
of cognitive neuroscience, a union of neuroscience and psychology.
The formation of behavioral economics and cognitive neuroscience prepared the ground for what followed. Starting in the mid-1990s researchers in
each of these disciplines, who were studying decision making, began to look
to the other discipline. Several cognitive neuroscientists considered the use
of economic theory as a normative theory against which they could examine
their neurobiological data. At the same time, a few economists contemplated
employing the mechanistic and algorithmic properties of the human nervous
system as constraints on their models. Both of these processes set the stage
for the emergence of the new discipline of neuroeconomics.
EMERGING FIELD
Shizgal and Conover were probably the first to explicitly apply economic
theory to neurobiological data (Shizgal & Conover, 1996). In a review paper,
these authors used economic theory to describe the neurobiology of choice
in rats pressing a lever to directly stimulate reward-related neurons in their
brains. Shortly afterwards, Platt and Glimcher (1999) published a highly
influential study in which they showed that neurons in monkey parietal
cortex encoded both the probability and the magnitude of expected juice
4
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
rewards, compatible with a role for these neurons in encoding the EU of
the rewards. A similar approach was adopted in a human study by a group
of experts in economics, neuroscience, and psychology. Breiter, Aharon,
Kahneman, Dale, & Shizgal (2001) combined neuroimaging with economic
theory to study the neural responses to expectation and experience of monetary gains and losses. The experimental design was based on Kahneman
and Tversky’s prospect theory (Kahneman was also one of the authors),
testing two principles of the theory: that the evaluation of an uncertain
option depends on whether it is presented as a gain or a loss, and that losses
loom larger than gains of the same magnitude.
An explicit call for neuroscientists and economists to join forces in the
study of decision making was made shortly afterwards by Glimcher in a
2003 book. Glimcher asserted that economics could provide a normative
theory for the study of the neurobiology of higher cognitive function. In
the next few years, a growing number of neurobiological papers in humans
and other animals relied on economic theory in the design and analysis
of their experiments. These papers examined the neural mechanism of a
variety of behaviors, including reinforcement learning (Lee, Seo, & Jung,
2012), intertemporal choice (Kable & Glimcher, 2007), decision under risk
and uncertainty (Platt & Huettel, 2008), valuation of goods of different types
(Levy & Glimcher, 2012), social decision making (Lee, 2008) and a variety of
decision biases. Some of these studies are described in the next section.
While neuroscientists widely appreciated the potential contribution of
economics to neuroscience and quickly embraced the emerging field, the
economic community was slower in its acceptance of neuroeconomics,
and is still divided regarding the usefulness of neurobiological insights to
economics theory. Perhaps the most famous opposition to neuroeconomics
in the economics community was made by Gul and Pesendorfer (2008).
These scholars argued that the goal of economic theories was to make
predictions about behavior and that the actual machinery by which choice
is accomplished must remain irrelevant to economists. In recent years, however, a growing number of economists have been advocating for considering
neurobiological data in the development of economic theories. Camerer and
colleagues first made the case for neuroeconomics from the economics side
(Camerer, Loewenstein, & Prelec, 2005), arguing that the neural mechanisms
of decision making should provide constraints on possible theories of
decision making and may direct future studies in economics. In other words,
these scholars proposed a shift from the “as if” models to models that
use neural data in order to describe the actual decision mechanisms. As
described later, substantial empirical support for this approach now exists.
In parallel to the first instances of interdisciplinary research combining
economics, psychology, and neuroscience, several meetings and conferences
Neuroeconomics
5
were held to bring together scientists from these different disciplines. These
meetings eventually led to the formation of the Society for Neuroeconomics,
which has been holding annual meetings since 2005, featuring the most
recent studies in neuroeconomics. In 2008, the Society published a volume
entitled “Neuroeconomics: Decision-Making and the Brain,” which was
edited by Glimcher, Camerer, Poldrack, and Fehr, and authored by many
central scholars in the field. A new and largely modified second edition
of this volume was published in 2014 (Glimcher & Fehr, 2014), summarizing nearly all of the most recent advances in neuroeconomics. This book
serves as an excellent introduction to the discipline, as well as a handbook
for researchers in the field and a textbook for students. Many of these
researchers and students belong to specialized centers for neuroeconomics,
which were open in many universities around the world. These centers
provide both training to emerging neuroeconomists and support for established scientists, who, together with scholars in traditional departments
for neuroscience, psychology, and economics, continue to investigate the
behavior and neurobiology of decision making. Although neuroeconomics
has only been around for a little over a decade, researchers in this discipline
have made substantial progress in understanding various aspects of the
neural architecture of decision making. Many questions, however, remain
open, and it will be interesting to witness how the field evolves within the
next few years. Some of these accomplishments and open questions are
described here.
CUTTING-EDGE RESEARCH AND CENTRAL FINDINGS
Decision theories generally agree that the decision-making process consists
of first assigning a “subjective value” to each available option and then
choosing the most valuable option (Kable & Glimcher, 2009). These values
need to be learned, stored, and modified when circumstances change.
Typically, intentions and preferences of other agents also need to be taken
into account. Numerous recent studies have employed a neuroeconomics
approach to study each of these aspects of decision making. Owing to space
limitations I briefly survey some of the most notable studies—the reader is
encouraged to turn to the further readings for additional information.
THE NEURAL REPRESENTATION OF VALUE
To be able to choose between an apple and an orange, the decision maker
needs to estimate the value of each piece of fruit using some “common
currency.” One of the most solid findings that came out of neuroeconomics
research is the identification of a valuation system in the brain, which
6
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
encodes values precisely in that manner. Accumulating evidence from a
large number of studies strongly supports the existence of such a system. A
common feature of most of these studies is that they used the choices that
each participant made to infer her unique preferences, and then searched
for “psychometric-neurometric matches” or correspondence between the
behavioral measurement and the neural measurement. This approach, which
was first employed in studies of perceptual decisions (Newsome, Britten,
& Movshon, 1989), has proved to be very useful in studying economic
or “value-based” decisions. Recording neural activity in the orbitofrontal
cortex in monkeys, Padoa-Schioppa and Assad (2006) identified neurons
whose activity was compatible with a common-currency representation
of value. The monkeys made choices between pairs of different juices at
varying quantities. On the basis of these choices, Padoa-Schioppa and
Assad estimated the subjective value of each juice type at each quantity,
and showed that a substantial number of their recorded neurons encoded
that “offer value” on a common scale. fMRI studies soon yielded similar
results. In one study, Kable and Glimcher examined how the value of
immediate and delayed rewards is encoded in the brain by having subjects
make choices between possible gains of different monetary amounts at
different times of receipt (Kable & Glimcher, 2007). Critically, Kable and
Glimcher used the participants’ behavior to infer the subjective values that
options of different delays and amounts held for each individual participant. They were then able to show that activity in three brain areas—the
medial prefrontal cortex, the striatum (part of the basal ganglia) and the
posterior cingulate cortex—is correlated with this measure of subjective
value. This finding was of importance not just for neuroscientists but also
for economists, because it was not compatible with a prominent economic
theory of intertemporal choice. Thus, this study is a good example for the
potential of neurobiological data to falsify existing economic theories and
to generate constraints for new theories. Two of the same brain areas, the
medial prefrontal cortex and the striatum, were also indicated as a common
currency valuation system in another study published the same year. In that
study, Tom, Fox, Trepel, & Poldrack (2007) compared the effect of anticipated
rewards (monetary gains) and anticipated punishments (monetary losses)
on neural activation in the whole brain. Activity in the medial prefrontal
cortex and the striatum was modulated by the magnitude of both rewards
and punishments. When participants anticipated a larger reward, activity in
these two brain areas increased. When participants anticipated a larger loss,
activity in the same brain areas decreased. Moreover, activation patterns
reflected the individual’s idiosyncratic aversion to losses, as estimated from
their behavior. These two initial studies in humans were followed by a surge
of studies examining various aspects of the neural representation of value.
Neuroeconomics
7
Subsequent studies have shown that the same brain areas encode the value
of a host of different rewards and punishments, including appetitive (Levy
& Glimcher, 2011) and aversive (Plassmann, O’Doherty, & Rangel, 2010) food
items, as well as other consumables (Chib, Rangel, Shimojo, & O’Doherty,
2009). A recent meta-analysis (Bartra, McGuire, & Kable, 2013) of over 200 of
these studies confirmed this notion of a single, common-currency, valuation
system.
THE LEARNING OF SUBJECTIVE VALUES
Where do these values encoded in the medial prefrontal cortex and the
striatum come from? One of the main sources of information about value
is learning by experience, another area where research has made great
strides. Substantial research has pointed to a role of the neurotransmitter
dopamine in this type of learning (Dayan & Niv, 2008). In a seminal work,
Schultz, Dayan, & Montague (1997) have shown that a theoretical algorithm
from computer science (Sutton & Barto, 1998) provides a surprisingly
good description of the activity of dopaminergic neurons in the monkey
brain stem. The firing rate of these neurons seemed to signal the difference
between the obtained and expected juice rewards, or the “reward prediction
error,” a signal that can drive learning. At the start of the experiment, juice
was generally unexpected, and so whenever the monkey received a drop
of juice, the dopaminergic neurons increased their firing rate. With time,
however, the monkey learned that rewards were associated with visual
or auditory cues that preceded them. He learned, for example, that after
hearing a certain tone, he was likely to receive a reward. The intriguing
observation of Schultz and his colleagues was that in the course of learning
the dopaminergic neurons shifted their response from the reward to the
reward-predictive cue. After several presentations of the cue followed
by the reward, the appearance of the cue fully predicted the subsequent
reward, which therefore did not generate a “reward prediction error” any
longer. The cue itself was now the unpredicted rewarding event, and it
was in response to these predictive cues that the dopaminergic neurons
now increased their firing rate. Similar reinforcement learning signals have
now been observed by many groups in several brain areas in both animals
and humans. For example, fMRI studies have identified neural signals that
reflect prediction errors in the striatum, a major target area of dopaminergic
projections (O’Doherty, Dayan, Friston, Critchley, & Dolan, 2003). This type
of learning is known as “model-free”—discrepancies between expected
and actual outcomes directly affect future expectations. More recent studies
have also turned to examine “model-based” types of learning, in which
the animal or the human constructs a cognitive model of the decision
8
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
situation, taking into account internal and external factors beyond the
simple stimulus–outcome associations. These studies have identified neural
correlates of both model-free and model-based learning (Daw, Gershman,
Seymour, Dayan, & Dolan, 2011), and suggest a promising avenue for
probing individual differences in value learning.
SOCIAL DECISION MAKING
Decisions are seldom made in isolation. An increasing number of neuroeconomics studies incorporate social aspects in their experimental designs.
Behavioral economics provides a convenient framework for many of these
studies in the form of game theory. Von Neumann and Morgenstern, the
fathers of the EU theory, also laid out the basis for game theory, in which
decisions are affected by choices made by many players with competing
interests. These decisions are complicated, because they require the decision
maker to infer the beliefs and intentions of other behaving agents, and
these beliefs and intentions in turn also depend on the decision maker’s
actions. Although humans and animals typically deviate in their behavior
from the precise predictions of the theory, game theory, just like EU theory,
provides a useful benchmark for quantifying these behavioral deviations
and searching for their neural correlates. In monkeys, these studies usually
employ simple paradigms, such as the well-known rock-paper-scissors game.
In one example, using the slightly simpler game of matching pennies, Lee and
colleagues (Lee, 2008) have examined how monkeys’ choices are affected
by changes in the strategies of their opponent. In matching pennies, each of
two players chooses one of two available options. One player wins if both
players make the same choice, the other wins if the choices are different.
The monkey played the game against a computer opponent, whose game
strategy was systematically manipulated by the experimenters. What they
changed, essentially, was the degree to which the computer exploited the
monkey’s choice history in making its own choices. The researchers found
that monkeys were, in fact, able to adapt their behavior, making their
choice patterns more random, and thus more difficult to exploit, with each
increase in the computer’s level of sophistication. While the monkeys were
playing that game, the experimenters recorded the activity of neurons in
the dorsolateral prefrontal cortex and found that those neurons encoded
the monkey’s past choices and rewards (Barraclough, Conroy, & Lee, 2004),
providing information that could be used to update the monkey’s estimate
of future rewards.
In humans, more complex game-theory-based paradigms are frequently
used, examining levels of trust, cooperation, and preferences for the
well-being of others. McCabe, Houser, Ryan, Smith, & Trouard (2001) were
Neuroeconomics
9
the first to use game theory in an fMRI experiment, in which participants
had to decide whether to trust another player. Results showed that those
participants who tended to trust their opponents had higher neuronal
activation in regions of the medial prefrontal cortex while playing against
humans compared to playing against computers. A similar trust game was
used a few years later by Kosfeld, Heinrichs, Zak, Fischbacher, & Fehr (2005),
but, critically, in their study the brain levels of oxytocin, a neuropeptide that
is thought to play a role in social attachment, were increased before they
made their decision. Kosfeld and colleagues found that those participants
treated with oxytocin were subsequently more trusting, compared to a
control group. This study is probably the earliest example demonstrating
the potential role for neuroscientific data in shaping economic theory.
Game-theory paradigms are also used to probe mentalizing processes,
namely, the ability of humans to understand the mental states of others and
to predict their behavior based on this understanding (Hampton, Bossaerts,
& O’Doherty, 2008). Finally, a few studies have directly examined charitable giving, revealing neural activity in reward-related brain structures
in response to charitable donations, even when those were mandatory
(Harbaugh, Mayr, & Burghart, 2007). While one should be careful of making
inferences from neural activation about behavior, this result is consistent
with an account of “pure altruism,” suggesting that people are capable of
experiencing rewarding sensations in response to the good fortune of others.
OPEN QUESTIONS AND FUTURE DIRECTIONS
The field of neuroeconomics continues to expand. Many questions have been
at least partially answered, but many more are still waiting to be solved. In
the next few years, neuroeconomics will continue to look for the missing
pieces for a comprehensive understanding of the neural architecture of decision making. At the same time, it is likely to broaden its scope to additional
questions. In particular, a promising avenue for neuroeconomics research is
to apply its rigorous analytical tools to the study of impaired decision processes, which are common in mental disorders.
A COMPLETE MODEL OF CHOICE
One of the major goals of neuroeconomics is to obtain a complete neural
model of the choice process. While we know a great deal about how values
are encoded in the brain, it is less clear how those values are used to produce
choice. How are values compared to each other to select the option of the
highest value? How does this selection in turn guide action? Recent findings
10
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
suggest that at least part of the choice process takes place in the medial prefrontal cortex, the same brain area whose activity also represents the values
of available options. Single-unit recordings in monkeys (Strait, Blanchard, &
Hayden, 2014) and magnetoencephalography (MEG) recordings in humans
(Hunt et al., 2012) identified signals in this brain area whose magnitude was
correlated with the difference between the values of available options, an
essential computation for choice. Choice-related signals were also identified,
however, in the activity of neurons in the posterior parietal cortex (Louie &
Glimcher, 2010). The posterior parietal cortex is ideally suitable to accommodate a choice process. First, it is spatially organized such that activity of
specific neurons is linked to specific spatial locations, and can therefore conveniently encode the subjective values of items in these spatial locations.
Second, a choice generated in this brain region can be directly communicated to the appropriate motor brain areas, to generate the required motor
action that will implement the choice. More research is needed to paint a full
picture incorporating the different components of the choice process. This
picture will likely also involve additional brain areas that have been shown
to have a role in valuation and choice, such as the posterior and anterior cingulate cortices, the amygdala and the insula.
THE NEUROECONOMICS OF PSYCHIATRY
The bulk of neuroeconomics research focuses on healthy individuals, in an
attempt to illuminate the neural mechanisms underlying decision-making
processes in the intact brain. Recently, however, a few researchers have begun
to apply neuroeconomics techniques to the study of mental illness (Sharp,
Monterosso, & Montague, 2012). This extension of the scope of neuroeconomics research from the normal to the abnormal seems obvious. Current
psychiatric research is largely based on self-report questionnaires. While certainly useful, such questionnaires may be biased, as they require the individual to reflect upon behavior. Instead of asking participants what they would
do, neuroeconomics techniques allow researchers to observe what participants actually do, and to study the neural mechanisms of this actual behavior.
Neuroeconomics is likely to offer novel insights on mental disorders, which
could direct more personally tailored interventions.
REFERENCES
Allais, M. (1953). Le comportment de l’homme rationnel devant le risqué: Critique
des postulates et axioms de l’ecole americaine. Econometrica, 21, 503–546.
Barraclough, D. J., Conroy, M. L., & Lee, D. (2004). Prefrontal cortex and decision
making in a mixed-strategy game. Nature Neuroscience, 7(4), 404–410. doi:10.1038/
nn1209 nn1209 [pii]
Neuroeconomics
11
Bartra, O., McGuire, J. T., & Kable, J. W. (2013). The valuation system: A coordinatebased meta-analysis of BOLD fMRI experiments examining neural correlates of
subjective value. NeuroImage, 76, 412–427. doi:10.1016/j.neuroimage.2013.02.063
S1053-8119(13)00218-8 [pii]
Breiter, H. C., Aharon, I., Kahneman, D., Dale, A., & Shizgal, P. (2001). Functional
imaging of neural responses to expectancy and experience of monetary gains and
losses. Neuron, 30(2), 619–639.
Camerer, C., Loewenstein, G., & Prelec, D. (2005). Neuroeconomics: How neuroscience can inform economics. Journal of Economic Literature, 43(1), 9–64.
Chib, V. S., Rangel, A., Shimojo, S., & O’Doherty, J. P. (2009). Evidence for a common
representation of decision values for dissimilar goods in human ventromedial prefrontal cortex. Journal of Neuroscience, 29(39), 12315–12320. doi:29/39/12315 [pii]
10.1523/JNEUROSCI.2575-09.2009
Daw, N. D., Gershman, S. J., Seymour, B., Dayan, P., & Dolan, R. J. (2011). Modelbased influences on humans’ choices and striatal prediction errors. Neuron, 69(6),
1204–1215. doi:S0896-6273(11)00125-5 [pii] 10.1016/j.neuron.2011.02.027
Dayan, P., & Niv, Y. (2008). Reinforcement learning: The good, the bad and the ugly.
Current Opinion in Neurobiology, 18(2), 185–196. doi:10.1016/j.conb.2008.08.003
Ellsberg, D. (1961). Risk, ambiguity, and the savage axioms. Quarterly Journal of Economics, 75(4), 643–669.
Glimcher, P. W., & Fehr, E. (2014). Neuroeconomics: Decision making and the brain
(2nd ed.). Amsterdam, Netherlands: Elsevier/Academic Press.
Gul, F., & Pesendorfer, W. (2008). The case for mindless economics. In A. Caplin &
A. Schotter (Eds.), The foundations of positive and normative economics: A handbook.
Oxford, England: Oxford University Press.
Hampton, A. N., Bossaerts, P., & O’Doherty, J. P. (2008). Neural correlates of
mentalizing-related computations during strategic interactions in humans. Proceedings of the National Academy of Sciences of the United States of America, 105(18),
6741–6746. doi:10.1073/pnas.0711099105
Harbaugh, W. T., Mayr, U., & Burghart, D. R. (2007). Neural responses to taxation
and voluntary giving reveal motives for charitable donations. Science, 316(5831),
1622–1625. doi:316/5831/1622 [pii] 10.1126/science.1140738
Hunt, L. T., Kolling, N., Soltani, A., Woolrich, M. W., Rushworth, M. F., & Behrens,
T. E. (2012). Mechanisms underlying cortical activity during value-guided choice.
Nature Neuroscience, 15(3), 470–476 S471–473. doi:10.1038/nn.3017
Kable, J. W., & Glimcher, P. W. (2007). The neural correlates of subjective value during
intertemporal choice. Nature Neuroscience, 10(12), 1625–1633.
Kable, J. W., & Glimcher, P. W. (2009). The neurobiology of decision: Consensus
and controversy. Neuron, 63(6), 733–745. doi:S0896-6273(09)00681-3 [pii] 10.1016/
j.neuron.2009.09.003
Kosfeld, M., Heinrichs, M., Zak, P. J., Fischbacher, U., & Fehr, E. (2005). Oxytocin increases trust in humans. Nature, 435(7042), 673–676. doi:nature03701 [pii]
10.1038/nature03701
Lee, D. (2008). Game theory and neural basis of social decision making. Nature Neuroscience, 11(4), 404–409. doi:10.1038/nn2065
12
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Lee, D., Seo, H., & Jung, M. W. (2012). Neural basis of reinforcement learning
and decision making. Annual Review of Neuroscience, 35, 287–308. doi:10.1146/
annurev-neuro-062111-150512
Levy, D. J., & Glimcher, P. W. (2011). Comparing apples and oranges: Using
reward-specific and reward-general subjective value representation in the brain.
Journal of Neuroscience, 31(41), 14693–14707. doi:31/41/14693 [pii] 10.1523/
JNEUROSCI.2218-11.2011
Levy, D. J., & Glimcher, P. W. (2012). The root of all value: A neural common currency for choice. Current Opinion in Neurobiology, 22(6), 1027–1038. doi:10.1016/
j.conb.2012.06.001
Louie, K., & Glimcher, P. W. (2010). Separating value from choice: Delay discounting
activity in the lateral intraparietal area. Journal of Neuroscience, 30(16), 5498–5507.
doi:10.1523/JNEUROSCI.5742-09.2010 30/16/5498 [pii]
McCabe, K., Houser, D., Ryan, L., Smith, V., & Trouard, T. (2001). A functional imaging study of cooperation in two-person reciprocal exchange. Proceedings of the
National Academy of Sciences of the United States of America, 98(20), 11832–11835.
doi:10.1073/pnas.211415698 211415698 [pii]
Newsome, W. T., Britten, K. H., & Movshon, J. A. (1989). Neuronal correlates of a
perceptual decision. Nature, 341, 52–54.
Neylan, T. C. (1999). Frontal lobe function: Mr. Phineas Gage’s famous injury. Journal
of Neuropsychiatry and Clinical Neurosciences, 11(2), 280–281.
O’Doherty, J. P., Dayan, P., Friston, K., Critchley, H., & Dolan, R. J. (2003). Temporal
difference models and reward-related learning in the human brain. Neuron, 38(2),
329–337.
Padoa-Schioppa, C., & Assad, J. A. (2006). Neurons in the orbitofrontal cortex encode
economic value. Nature, 441(7090), 223–226.
Plassmann, H., O’Doherty, J. P., & Rangel, A. (2010). Appetitive and aversive goal values are encoded in the medial orbitofrontal cortex at the time
of decision making. Journal of Neuroscience, 30(32), 10799–10808. doi:10.1523/
JNEUROSCI.0788-10.2010
Platt, M. L., & Glimcher, P. W. (1999). Neural correlates of decision variables in parietal cortex. Nature, 400(6741), 233–238.
Platt, M. L., & Huettel, S. A. (2008). Risky business: The neuroeconomics of decision making under uncertainty. Nature Neuroscience, 11(4), 398–403. doi:nn2062
[pii] 10.1038/nn2062
Savage, L. J. (1954). The foundations of statistics. New York, NY: Wiley.
Schultz, W., Dayan, P., & Montague, P. R. (1997). A neural substrate of prediction and
reward. Science, 275(5306), 1593–1599.
Sharp, C., Monterosso, J., & Montague, P. R. (2012). Neuroeconomics: A bridge for
translational research. Biological Psychiatry, 72(2), 87–92. doi:10.1016/j.biopsych.
2012.02.029 S0006-3223(12)00165-5 [pii]
Shizgal, P., & Conover, K. (1996). On the neural computation of utility. Current Directions in Psychological Science, 5(2), 37–43.
Strait, C. E., Blanchard, T. C., & Hayden, B. Y. (2014). Reward value comparison
via mutual inhibition in ventromedial prefrontal cortex. Neuron. doi:10.1016/
j.neuron.2014.04.032
Neuroeconomics
13
Sutton, R. S., & Barto, A. G. (1998). Reinforcement learning: An introduction. Cambridge,
MA: MIT Press.
Tom, S. M., Fox, C. R., Trepel, C., & Poldrack, R. A. (2007). The neural basis
of loss aversion in decision-making under risk. Science, 315(5811), 515–518.
doi:315/5811/515 [pii] 10.1126/science.1134239
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty—heuristics and
biases. Science, 185(4157), 1124–1131.
Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology
of choice. Science, 211(4481), 453–458.
Von Neumann, J., & Morgenstern, O. (1944). Theory of games and economic behavior.
Princeton, NJ: Princeton university press.
FURTHER READING
Glimcher, P. W. (2011). Foundations of neuroeconomic analysis. New York, NY: Oxford
University Press.
Glimcher, P. W., & Fehr, E. (2014). Neuroeconomics: Decision making and the brain (2nd
ed.). Amsterdam, Netherlands: Elsevier/Academic Press.
IFAT LEVY SHORT BIOGRAPHY
Ifat Levy is a faculty member in Yale School of Medicine, conducting neuroeconomics research in humans. She uses behavioral methods together with
MRI to investigate decision making under uncertainty in healthy individuals,
as well as in disorders including obesity and mental illness.
URL: http://bbs.yale.edu/neuroscience/people/ifat_levy.profile
RELATED ESSAYS
Models of Revealed Preference (Economics), Abi Adams and Ian Crawford
Coevolution of Decision-Making and Social Environments (Sociology),
Elizabeth Bruch et al.
Mental Models (Psychology), Ruth M. J. Byrne
Choice Architecture (Psychology), Adrian R. Camilleri and Rick P. Larrick
Heuristic Decision Making (Political Science), Edward G. Carmines and
Nicholas J. D’Amico
Culture and Cognition (Sociology), Karen A. Cerulo
Self-Fulfilling Prophesies, Placebo Effects, and the Social-Psychological
Creation of Reality (Sociology), Alia Crum and Damon J. Phillips
Regulatory Focus Theory (Psychology), E. Tory Higgins
Behavioral Economics (Sociology), Guy Hochman and Dan Ariely
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
14
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Models of Duality (Psychology), Anand Krishna et al.
Against Game Theory (Political Science), Gale M. Lucas et al.
Neural and Cognitive Plasticity (Psychology), Eduardo Mercado III
Embodied Knowledge (Psychology), Diane Pecher and René Zeelenberg
Attention and Perception (Psychology), Ronald A. Rensink
Event Processing as an Executive Enterprise (Psychology), Robbie A. Ross
and Dare A. Baldwin
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen
-
Neuroeconomics
IFAT LEVY
Abstract
In recent years, researchers in economics, psychology, and neuroscience have joined
forces in the study of decision-making processes to form the new discipline of
neuroeconomics. Neuroscientists turned to theories in economics and psychology
to make sense of the increasing amounts of neurobiological data. At the same time,
economists and psychologists turned to neuroscience for mechanistic constraints on
their theories. Neuroeconomics studies tackle a host of topics, from financial choices
through reinforcement learning to social decision making. Combining behavioral
techniques with brain imaging in humans and electrophysiological recordings in
animals, as well as complementary techniques, this interdisciplinary research has
already generated new insights about the neural architecture of decision making.
The neural mechanisms of some of the behavioral decision processes are increasingly
understood, but many challenges remain. Extending neuroeconomics research to
psychiatric disorders and incorporating new research tools are promising avenues
for future studies.
THE HISTORY OF NEUROECONOMICS
INTRODUCTION
How does one choose between an apple and an orange? We can ask this
question at several different levels. At the behavioral level, we may be interested in accurately predicting individual choices. At the algorithmic level,
we can explore the neural mechanisms that underlie the observed choice
behavior. Yet another level of inquiry is examining the mental states that give
rise to the observed choice behavior. For centuries, researchers in economics,
neuroscience, and psychology have studied decision making as largely separate disciplines. Researchers in each discipline have employed very different
strategies to study one of these aspects of decision making.
Since the 1930s, neoclassical economists have essentially strove to predict
human choice behavior based on rigorous mathematical models. These models were typically “as if” models, in the sense that they were not required
to accurately describe the algorithm implementing the choice process, but
Emerging Trends in the Social and Behavioral Sciences. Edited by Robert Scott and Stephen Kosslyn.
© 2015 John Wiley & Sons, Inc. ISBN 978-1-118-90077-2.
1
2
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
rather to correctly predict behavior. As Paul Samuelson and others have come
to realize, making a small number of simple and reasonable assumptions,
or “axioms” about choice behavior, such as “if a person prefers apples to
oranges she will not also prefer oranges to apples,” allows us to describe the
decision maker’s behavior “as if” they are trying to maximize some “utility” function. On the basis of these axioms, Von Neumann and Morgenstern
(1944) developed the expected utility (EU) theory, a model for choice between
uncertain outcomes, which was later extended by Savage (1954) to take into
account the decision maker’s subjective estimation of outcome probability
(subjective EU).
At the same time that these economists were seeking to predict behavior,
neuroscientists were aiming to reveal the mechanisms underlying the same
behavior. For decades neurological research essentially depended on brain
lesions in humans and animals. In this approach, behavioral deficits are correlated with the particular brain injury in an attempt to infer the function of
the damaged neural system. Perhaps the most famous example is the 1848
case of Phineas Gage, a young railroad worker whose brain was penetrated
by a steel rod (Neylan, 1999). Despite his remarkable physical recovery following the accident, Gage has exhibited substantial changes in personality
and decision making, providing the first evidence in humans for the involvement of the prefrontal cortex, the part of the brain most affected by the accident, in decision-making processes. While brain lesions were very helpful
in identifying general associations between brain structures and particular
sensory, motor, or cognitive functions, they could provide little information
about the neural algorithms that are implemented in each of these structures.
Advances in both economic theory and neurobiological techniques during
the second half of the twentieth century led researchers in both of these disciplines to look to psychology, and eventually toward each other, in what
turned out to be the birth of neuroeconomics.
On the economics side, economists began to note examples of choice behavior which was not compatible with neoclassical economic theory. In several
cases, human choices violated one or more of the core axioms of the theory.
Allais was the first to describe such violation in what is known as the “Allais
paradox” (Allais, 1953), followed by Ellsberg, who described the “Ellsberg
paradox” (Ellsberg, 1961). In a series of seminal studies published in the 1970s
and 1980s Kahneman and Tversky have documented numerous substantial
behavioral deviations from EU theory, demonstrating that these deviations
were the rule rather than the exception (Tversky & Kahneman, 1974, 1981).
The work of Kahneman and Tversky suggested to many economists and psychologists that economic models could benefit from psychological data and
insights. These economists and psychologists formed the discipline of behavioral economics, a union of economics and psychology.
Neuroeconomics
3
While these processes were taking place in the economics world, neuroscientists were also experiencing considerable transformations in their
research. The introduction of novel data collection techniques markedly
increased the ability of neuroscientists to investigate the neural circuits
underlying higher cognitive functions. The first advance was made in the
1960s, when recording of neural activity in the brains of awake behaving animals became available. The next development occurred in the 1980s–1990s,
when noninvasive imaging methods in humans, in particular positron
emission tomography (PET) and functional magnetic resonance maging
(fMRI) became widely used. These tools enabled neuroscientists to examine
dynamic changes in neural activity while humans and other animals were
engaged in complex behavior. The ability to record the activity of single neurons in awake behaving animals, and to image the activity of populations of
neurons in humans, allowed much more than a simple correlation between
neural activation and observed behavior. The analysis tools for making use
of these new techniques and the large quantities of data that they generated,
however, were not well developed. Similar to economists, neuroscientists
have also begun to look to psychology, in order to use models of cognitive
function in designing their experiments and analyzing their data. Instead of
correlating a damaged brain area with impaired cognitive function, neuroscientists could now look for correlations between neural activity and hidden
variables of the mental models. These studies gave rise to the new discipline
of cognitive neuroscience, a union of neuroscience and psychology.
The formation of behavioral economics and cognitive neuroscience prepared the ground for what followed. Starting in the mid-1990s researchers in
each of these disciplines, who were studying decision making, began to look
to the other discipline. Several cognitive neuroscientists considered the use
of economic theory as a normative theory against which they could examine
their neurobiological data. At the same time, a few economists contemplated
employing the mechanistic and algorithmic properties of the human nervous
system as constraints on their models. Both of these processes set the stage
for the emergence of the new discipline of neuroeconomics.
EMERGING FIELD
Shizgal and Conover were probably the first to explicitly apply economic
theory to neurobiological data (Shizgal & Conover, 1996). In a review paper,
these authors used economic theory to describe the neurobiology of choice
in rats pressing a lever to directly stimulate reward-related neurons in their
brains. Shortly afterwards, Platt and Glimcher (1999) published a highly
influential study in which they showed that neurons in monkey parietal
cortex encoded both the probability and the magnitude of expected juice
4
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
rewards, compatible with a role for these neurons in encoding the EU of
the rewards. A similar approach was adopted in a human study by a group
of experts in economics, neuroscience, and psychology. Breiter, Aharon,
Kahneman, Dale, & Shizgal (2001) combined neuroimaging with economic
theory to study the neural responses to expectation and experience of monetary gains and losses. The experimental design was based on Kahneman
and Tversky’s prospect theory (Kahneman was also one of the authors),
testing two principles of the theory: that the evaluation of an uncertain
option depends on whether it is presented as a gain or a loss, and that losses
loom larger than gains of the same magnitude.
An explicit call for neuroscientists and economists to join forces in the
study of decision making was made shortly afterwards by Glimcher in a
2003 book. Glimcher asserted that economics could provide a normative
theory for the study of the neurobiology of higher cognitive function. In
the next few years, a growing number of neurobiological papers in humans
and other animals relied on economic theory in the design and analysis
of their experiments. These papers examined the neural mechanism of a
variety of behaviors, including reinforcement learning (Lee, Seo, & Jung,
2012), intertemporal choice (Kable & Glimcher, 2007), decision under risk
and uncertainty (Platt & Huettel, 2008), valuation of goods of different types
(Levy & Glimcher, 2012), social decision making (Lee, 2008) and a variety of
decision biases. Some of these studies are described in the next section.
While neuroscientists widely appreciated the potential contribution of
economics to neuroscience and quickly embraced the emerging field, the
economic community was slower in its acceptance of neuroeconomics,
and is still divided regarding the usefulness of neurobiological insights to
economics theory. Perhaps the most famous opposition to neuroeconomics
in the economics community was made by Gul and Pesendorfer (2008).
These scholars argued that the goal of economic theories was to make
predictions about behavior and that the actual machinery by which choice
is accomplished must remain irrelevant to economists. In recent years, however, a growing number of economists have been advocating for considering
neurobiological data in the development of economic theories. Camerer and
colleagues first made the case for neuroeconomics from the economics side
(Camerer, Loewenstein, & Prelec, 2005), arguing that the neural mechanisms
of decision making should provide constraints on possible theories of
decision making and may direct future studies in economics. In other words,
these scholars proposed a shift from the “as if” models to models that
use neural data in order to describe the actual decision mechanisms. As
described later, substantial empirical support for this approach now exists.
In parallel to the first instances of interdisciplinary research combining
economics, psychology, and neuroscience, several meetings and conferences
Neuroeconomics
5
were held to bring together scientists from these different disciplines. These
meetings eventually led to the formation of the Society for Neuroeconomics,
which has been holding annual meetings since 2005, featuring the most
recent studies in neuroeconomics. In 2008, the Society published a volume
entitled “Neuroeconomics: Decision-Making and the Brain,” which was
edited by Glimcher, Camerer, Poldrack, and Fehr, and authored by many
central scholars in the field. A new and largely modified second edition
of this volume was published in 2014 (Glimcher & Fehr, 2014), summarizing nearly all of the most recent advances in neuroeconomics. This book
serves as an excellent introduction to the discipline, as well as a handbook
for researchers in the field and a textbook for students. Many of these
researchers and students belong to specialized centers for neuroeconomics,
which were open in many universities around the world. These centers
provide both training to emerging neuroeconomists and support for established scientists, who, together with scholars in traditional departments
for neuroscience, psychology, and economics, continue to investigate the
behavior and neurobiology of decision making. Although neuroeconomics
has only been around for a little over a decade, researchers in this discipline
have made substantial progress in understanding various aspects of the
neural architecture of decision making. Many questions, however, remain
open, and it will be interesting to witness how the field evolves within the
next few years. Some of these accomplishments and open questions are
described here.
CUTTING-EDGE RESEARCH AND CENTRAL FINDINGS
Decision theories generally agree that the decision-making process consists
of first assigning a “subjective value” to each available option and then
choosing the most valuable option (Kable & Glimcher, 2009). These values
need to be learned, stored, and modified when circumstances change.
Typically, intentions and preferences of other agents also need to be taken
into account. Numerous recent studies have employed a neuroeconomics
approach to study each of these aspects of decision making. Owing to space
limitations I briefly survey some of the most notable studies—the reader is
encouraged to turn to the further readings for additional information.
THE NEURAL REPRESENTATION OF VALUE
To be able to choose between an apple and an orange, the decision maker
needs to estimate the value of each piece of fruit using some “common
currency.” One of the most solid findings that came out of neuroeconomics
research is the identification of a valuation system in the brain, which
6
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
encodes values precisely in that manner. Accumulating evidence from a
large number of studies strongly supports the existence of such a system. A
common feature of most of these studies is that they used the choices that
each participant made to infer her unique preferences, and then searched
for “psychometric-neurometric matches” or correspondence between the
behavioral measurement and the neural measurement. This approach, which
was first employed in studies of perceptual decisions (Newsome, Britten,
& Movshon, 1989), has proved to be very useful in studying economic
or “value-based” decisions. Recording neural activity in the orbitofrontal
cortex in monkeys, Padoa-Schioppa and Assad (2006) identified neurons
whose activity was compatible with a common-currency representation
of value. The monkeys made choices between pairs of different juices at
varying quantities. On the basis of these choices, Padoa-Schioppa and
Assad estimated the subjective value of each juice type at each quantity,
and showed that a substantial number of their recorded neurons encoded
that “offer value” on a common scale. fMRI studies soon yielded similar
results. In one study, Kable and Glimcher examined how the value of
immediate and delayed rewards is encoded in the brain by having subjects
make choices between possible gains of different monetary amounts at
different times of receipt (Kable & Glimcher, 2007). Critically, Kable and
Glimcher used the participants’ behavior to infer the subjective values that
options of different delays and amounts held for each individual participant. They were then able to show that activity in three brain areas—the
medial prefrontal cortex, the striatum (part of the basal ganglia) and the
posterior cingulate cortex—is correlated with this measure of subjective
value. This finding was of importance not just for neuroscientists but also
for economists, because it was not compatible with a prominent economic
theory of intertemporal choice. Thus, this study is a good example for the
potential of neurobiological data to falsify existing economic theories and
to generate constraints for new theories. Two of the same brain areas, the
medial prefrontal cortex and the striatum, were also indicated as a common
currency valuation system in another study published the same year. In that
study, Tom, Fox, Trepel, & Poldrack (2007) compared the effect of anticipated
rewards (monetary gains) and anticipated punishments (monetary losses)
on neural activation in the whole brain. Activity in the medial prefrontal
cortex and the striatum was modulated by the magnitude of both rewards
and punishments. When participants anticipated a larger reward, activity in
these two brain areas increased. When participants anticipated a larger loss,
activity in the same brain areas decreased. Moreover, activation patterns
reflected the individual’s idiosyncratic aversion to losses, as estimated from
their behavior. These two initial studies in humans were followed by a surge
of studies examining various aspects of the neural representation of value.
Neuroeconomics
7
Subsequent studies have shown that the same brain areas encode the value
of a host of different rewards and punishments, including appetitive (Levy
& Glimcher, 2011) and aversive (Plassmann, O’Doherty, & Rangel, 2010) food
items, as well as other consumables (Chib, Rangel, Shimojo, & O’Doherty,
2009). A recent meta-analysis (Bartra, McGuire, & Kable, 2013) of over 200 of
these studies confirmed this notion of a single, common-currency, valuation
system.
THE LEARNING OF SUBJECTIVE VALUES
Where do these values encoded in the medial prefrontal cortex and the
striatum come from? One of the main sources of information about value
is learning by experience, another area where research has made great
strides. Substantial research has pointed to a role of the neurotransmitter
dopamine in this type of learning (Dayan & Niv, 2008). In a seminal work,
Schultz, Dayan, & Montague (1997) have shown that a theoretical algorithm
from computer science (Sutton & Barto, 1998) provides a surprisingly
good description of the activity of dopaminergic neurons in the monkey
brain stem. The firing rate of these neurons seemed to signal the difference
between the obtained and expected juice rewards, or the “reward prediction
error,” a signal that can drive learning. At the start of the experiment, juice
was generally unexpected, and so whenever the monkey received a drop
of juice, the dopaminergic neurons increased their firing rate. With time,
however, the monkey learned that rewards were associated with visual
or auditory cues that preceded them. He learned, for example, that after
hearing a certain tone, he was likely to receive a reward. The intriguing
observation of Schultz and his colleagues was that in the course of learning
the dopaminergic neurons shifted their response from the reward to the
reward-predictive cue. After several presentations of the cue followed
by the reward, the appearance of the cue fully predicted the subsequent
reward, which therefore did not generate a “reward prediction error” any
longer. The cue itself was now the unpredicted rewarding event, and it
was in response to these predictive cues that the dopaminergic neurons
now increased their firing rate. Similar reinforcement learning signals have
now been observed by many groups in several brain areas in both animals
and humans. For example, fMRI studies have identified neural signals that
reflect prediction errors in the striatum, a major target area of dopaminergic
projections (O’Doherty, Dayan, Friston, Critchley, & Dolan, 2003). This type
of learning is known as “model-free”—discrepancies between expected
and actual outcomes directly affect future expectations. More recent studies
have also turned to examine “model-based” types of learning, in which
the animal or the human constructs a cognitive model of the decision
8
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
situation, taking into account internal and external factors beyond the
simple stimulus–outcome associations. These studies have identified neural
correlates of both model-free and model-based learning (Daw, Gershman,
Seymour, Dayan, & Dolan, 2011), and suggest a promising avenue for
probing individual differences in value learning.
SOCIAL DECISION MAKING
Decisions are seldom made in isolation. An increasing number of neuroeconomics studies incorporate social aspects in their experimental designs.
Behavioral economics provides a convenient framework for many of these
studies in the form of game theory. Von Neumann and Morgenstern, the
fathers of the EU theory, also laid out the basis for game theory, in which
decisions are affected by choices made by many players with competing
interests. These decisions are complicated, because they require the decision
maker to infer the beliefs and intentions of other behaving agents, and
these beliefs and intentions in turn also depend on the decision maker’s
actions. Although humans and animals typically deviate in their behavior
from the precise predictions of the theory, game theory, just like EU theory,
provides a useful benchmark for quantifying these behavioral deviations
and searching for their neural correlates. In monkeys, these studies usually
employ simple paradigms, such as the well-known rock-paper-scissors game.
In one example, using the slightly simpler game of matching pennies, Lee and
colleagues (Lee, 2008) have examined how monkeys’ choices are affected
by changes in the strategies of their opponent. In matching pennies, each of
two players chooses one of two available options. One player wins if both
players make the same choice, the other wins if the choices are different.
The monkey played the game against a computer opponent, whose game
strategy was systematically manipulated by the experimenters. What they
changed, essentially, was the degree to which the computer exploited the
monkey’s choice history in making its own choices. The researchers found
that monkeys were, in fact, able to adapt their behavior, making their
choice patterns more random, and thus more difficult to exploit, with each
increase in the computer’s level of sophistication. While the monkeys were
playing that game, the experimenters recorded the activity of neurons in
the dorsolateral prefrontal cortex and found that those neurons encoded
the monkey’s past choices and rewards (Barraclough, Conroy, & Lee, 2004),
providing information that could be used to update the monkey’s estimate
of future rewards.
In humans, more complex game-theory-based paradigms are frequently
used, examining levels of trust, cooperation, and preferences for the
well-being of others. McCabe, Houser, Ryan, Smith, & Trouard (2001) were
Neuroeconomics
9
the first to use game theory in an fMRI experiment, in which participants
had to decide whether to trust another player. Results showed that those
participants who tended to trust their opponents had higher neuronal
activation in regions of the medial prefrontal cortex while playing against
humans compared to playing against computers. A similar trust game was
used a few years later by Kosfeld, Heinrichs, Zak, Fischbacher, & Fehr (2005),
but, critically, in their study the brain levels of oxytocin, a neuropeptide that
is thought to play a role in social attachment, were increased before they
made their decision. Kosfeld and colleagues found that those participants
treated with oxytocin were subsequently more trusting, compared to a
control group. This study is probably the earliest example demonstrating
the potential role for neuroscientific data in shaping economic theory.
Game-theory paradigms are also used to probe mentalizing processes,
namely, the ability of humans to understand the mental states of others and
to predict their behavior based on this understanding (Hampton, Bossaerts,
& O’Doherty, 2008). Finally, a few studies have directly examined charitable giving, revealing neural activity in reward-related brain structures
in response to charitable donations, even when those were mandatory
(Harbaugh, Mayr, & Burghart, 2007). While one should be careful of making
inferences from neural activation about behavior, this result is consistent
with an account of “pure altruism,” suggesting that people are capable of
experiencing rewarding sensations in response to the good fortune of others.
OPEN QUESTIONS AND FUTURE DIRECTIONS
The field of neuroeconomics continues to expand. Many questions have been
at least partially answered, but many more are still waiting to be solved. In
the next few years, neuroeconomics will continue to look for the missing
pieces for a comprehensive understanding of the neural architecture of decision making. At the same time, it is likely to broaden its scope to additional
questions. In particular, a promising avenue for neuroeconomics research is
to apply its rigorous analytical tools to the study of impaired decision processes, which are common in mental disorders.
A COMPLETE MODEL OF CHOICE
One of the major goals of neuroeconomics is to obtain a complete neural
model of the choice process. While we know a great deal about how values
are encoded in the brain, it is less clear how those values are used to produce
choice. How are values compared to each other to select the option of the
highest value? How does this selection in turn guide action? Recent findings
10
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
suggest that at least part of the choice process takes place in the medial prefrontal cortex, the same brain area whose activity also represents the values
of available options. Single-unit recordings in monkeys (Strait, Blanchard, &
Hayden, 2014) and magnetoencephalography (MEG) recordings in humans
(Hunt et al., 2012) identified signals in this brain area whose magnitude was
correlated with the difference between the values of available options, an
essential computation for choice. Choice-related signals were also identified,
however, in the activity of neurons in the posterior parietal cortex (Louie &
Glimcher, 2010). The posterior parietal cortex is ideally suitable to accommodate a choice process. First, it is spatially organized such that activity of
specific neurons is linked to specific spatial locations, and can therefore conveniently encode the subjective values of items in these spatial locations.
Second, a choice generated in this brain region can be directly communicated to the appropriate motor brain areas, to generate the required motor
action that will implement the choice. More research is needed to paint a full
picture incorporating the different components of the choice process. This
picture will likely also involve additional brain areas that have been shown
to have a role in valuation and choice, such as the posterior and anterior cingulate cortices, the amygdala and the insula.
THE NEUROECONOMICS OF PSYCHIATRY
The bulk of neuroeconomics research focuses on healthy individuals, in an
attempt to illuminate the neural mechanisms underlying decision-making
processes in the intact brain. Recently, however, a few researchers have begun
to apply neuroeconomics techniques to the study of mental illness (Sharp,
Monterosso, & Montague, 2012). This extension of the scope of neuroeconomics research from the normal to the abnormal seems obvious. Current
psychiatric research is largely based on self-report questionnaires. While certainly useful, such questionnaires may be biased, as they require the individual to reflect upon behavior. Instead of asking participants what they would
do, neuroeconomics techniques allow researchers to observe what participants actually do, and to study the neural mechanisms of this actual behavior.
Neuroeconomics is likely to offer novel insights on mental disorders, which
could direct more personally tailored interventions.
REFERENCES
Allais, M. (1953). Le comportment de l’homme rationnel devant le risqué: Critique
des postulates et axioms de l’ecole americaine. Econometrica, 21, 503–546.
Barraclough, D. J., Conroy, M. L., & Lee, D. (2004). Prefrontal cortex and decision
making in a mixed-strategy game. Nature Neuroscience, 7(4), 404–410. doi:10.1038/
nn1209 nn1209 [pii]
Neuroeconomics
11
Bartra, O., McGuire, J. T., & Kable, J. W. (2013). The valuation system: A coordinatebased meta-analysis of BOLD fMRI experiments examining neural correlates of
subjective value. NeuroImage, 76, 412–427. doi:10.1016/j.neuroimage.2013.02.063
S1053-8119(13)00218-8 [pii]
Breiter, H. C., Aharon, I., Kahneman, D., Dale, A., & Shizgal, P. (2001). Functional
imaging of neural responses to expectancy and experience of monetary gains and
losses. Neuron, 30(2), 619–639.
Camerer, C., Loewenstein, G., & Prelec, D. (2005). Neuroeconomics: How neuroscience can inform economics. Journal of Economic Literature, 43(1), 9–64.
Chib, V. S., Rangel, A., Shimojo, S., & O’Doherty, J. P. (2009). Evidence for a common
representation of decision values for dissimilar goods in human ventromedial prefrontal cortex. Journal of Neuroscience, 29(39), 12315–12320. doi:29/39/12315 [pii]
10.1523/JNEUROSCI.2575-09.2009
Daw, N. D., Gershman, S. J., Seymour, B., Dayan, P., & Dolan, R. J. (2011). Modelbased influences on humans’ choices and striatal prediction errors. Neuron, 69(6),
1204–1215. doi:S0896-6273(11)00125-5 [pii] 10.1016/j.neuron.2011.02.027
Dayan, P., & Niv, Y. (2008). Reinforcement learning: The good, the bad and the ugly.
Current Opinion in Neurobiology, 18(2), 185–196. doi:10.1016/j.conb.2008.08.003
Ellsberg, D. (1961). Risk, ambiguity, and the savage axioms. Quarterly Journal of Economics, 75(4), 643–669.
Glimcher, P. W., & Fehr, E. (2014). Neuroeconomics: Decision making and the brain
(2nd ed.). Amsterdam, Netherlands: Elsevier/Academic Press.
Gul, F., & Pesendorfer, W. (2008). The case for mindless economics. In A. Caplin &
A. Schotter (Eds.), The foundations of positive and normative economics: A handbook.
Oxford, England: Oxford University Press.
Hampton, A. N., Bossaerts, P., & O’Doherty, J. P. (2008). Neural correlates of
mentalizing-related computations during strategic interactions in humans. Proceedings of the National Academy of Sciences of the United States of America, 105(18),
6741–6746. doi:10.1073/pnas.0711099105
Harbaugh, W. T., Mayr, U., & Burghart, D. R. (2007). Neural responses to taxation
and voluntary giving reveal motives for charitable donations. Science, 316(5831),
1622–1625. doi:316/5831/1622 [pii] 10.1126/science.1140738
Hunt, L. T., Kolling, N., Soltani, A., Woolrich, M. W., Rushworth, M. F., & Behrens,
T. E. (2012). Mechanisms underlying cortical activity during value-guided choice.
Nature Neuroscience, 15(3), 470–476 S471–473. doi:10.1038/nn.3017
Kable, J. W., & Glimcher, P. W. (2007). The neural correlates of subjective value during
intertemporal choice. Nature Neuroscience, 10(12), 1625–1633.
Kable, J. W., & Glimcher, P. W. (2009). The neurobiology of decision: Consensus
and controversy. Neuron, 63(6), 733–745. doi:S0896-6273(09)00681-3 [pii] 10.1016/
j.neuron.2009.09.003
Kosfeld, M., Heinrichs, M., Zak, P. J., Fischbacher, U., & Fehr, E. (2005). Oxytocin increases trust in humans. Nature, 435(7042), 673–676. doi:nature03701 [pii]
10.1038/nature03701
Lee, D. (2008). Game theory and neural basis of social decision making. Nature Neuroscience, 11(4), 404–409. doi:10.1038/nn2065
12
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Lee, D., Seo, H., & Jung, M. W. (2012). Neural basis of reinforcement learning
and decision making. Annual Review of Neuroscience, 35, 287–308. doi:10.1146/
annurev-neuro-062111-150512
Levy, D. J., & Glimcher, P. W. (2011). Comparing apples and oranges: Using
reward-specific and reward-general subjective value representation in the brain.
Journal of Neuroscience, 31(41), 14693–14707. doi:31/41/14693 [pii] 10.1523/
JNEUROSCI.2218-11.2011
Levy, D. J., & Glimcher, P. W. (2012). The root of all value: A neural common currency for choice. Current Opinion in Neurobiology, 22(6), 1027–1038. doi:10.1016/
j.conb.2012.06.001
Louie, K., & Glimcher, P. W. (2010). Separating value from choice: Delay discounting
activity in the lateral intraparietal area. Journal of Neuroscience, 30(16), 5498–5507.
doi:10.1523/JNEUROSCI.5742-09.2010 30/16/5498 [pii]
McCabe, K., Houser, D., Ryan, L., Smith, V., & Trouard, T. (2001). A functional imaging study of cooperation in two-person reciprocal exchange. Proceedings of the
National Academy of Sciences of the United States of America, 98(20), 11832–11835.
doi:10.1073/pnas.211415698 211415698 [pii]
Newsome, W. T., Britten, K. H., & Movshon, J. A. (1989). Neuronal correlates of a
perceptual decision. Nature, 341, 52–54.
Neylan, T. C. (1999). Frontal lobe function: Mr. Phineas Gage’s famous injury. Journal
of Neuropsychiatry and Clinical Neurosciences, 11(2), 280–281.
O’Doherty, J. P., Dayan, P., Friston, K., Critchley, H., & Dolan, R. J. (2003). Temporal
difference models and reward-related learning in the human brain. Neuron, 38(2),
329–337.
Padoa-Schioppa, C., & Assad, J. A. (2006). Neurons in the orbitofrontal cortex encode
economic value. Nature, 441(7090), 223–226.
Plassmann, H., O’Doherty, J. P., & Rangel, A. (2010). Appetitive and aversive goal values are encoded in the medial orbitofrontal cortex at the time
of decision making. Journal of Neuroscience, 30(32), 10799–10808. doi:10.1523/
JNEUROSCI.0788-10.2010
Platt, M. L., & Glimcher, P. W. (1999). Neural correlates of decision variables in parietal cortex. Nature, 400(6741), 233–238.
Platt, M. L., & Huettel, S. A. (2008). Risky business: The neuroeconomics of decision making under uncertainty. Nature Neuroscience, 11(4), 398–403. doi:nn2062
[pii] 10.1038/nn2062
Savage, L. J. (1954). The foundations of statistics. New York, NY: Wiley.
Schultz, W., Dayan, P., & Montague, P. R. (1997). A neural substrate of prediction and
reward. Science, 275(5306), 1593–1599.
Sharp, C., Monterosso, J., & Montague, P. R. (2012). Neuroeconomics: A bridge for
translational research. Biological Psychiatry, 72(2), 87–92. doi:10.1016/j.biopsych.
2012.02.029 S0006-3223(12)00165-5 [pii]
Shizgal, P., & Conover, K. (1996). On the neural computation of utility. Current Directions in Psychological Science, 5(2), 37–43.
Strait, C. E., Blanchard, T. C., & Hayden, B. Y. (2014). Reward value comparison
via mutual inhibition in ventromedial prefrontal cortex. Neuron. doi:10.1016/
j.neuron.2014.04.032
Neuroeconomics
13
Sutton, R. S., & Barto, A. G. (1998). Reinforcement learning: An introduction. Cambridge,
MA: MIT Press.
Tom, S. M., Fox, C. R., Trepel, C., & Poldrack, R. A. (2007). The neural basis
of loss aversion in decision-making under risk. Science, 315(5811), 515–518.
doi:315/5811/515 [pii] 10.1126/science.1134239
Tversky, A., & Kahneman, D. (1974). Judgment under uncertainty—heuristics and
biases. Science, 185(4157), 1124–1131.
Tversky, A., & Kahneman, D. (1981). The framing of decisions and the psychology
of choice. Science, 211(4481), 453–458.
Von Neumann, J., & Morgenstern, O. (1944). Theory of games and economic behavior.
Princeton, NJ: Princeton university press.
FURTHER READING
Glimcher, P. W. (2011). Foundations of neuroeconomic analysis. New York, NY: Oxford
University Press.
Glimcher, P. W., & Fehr, E. (2014). Neuroeconomics: Decision making and the brain (2nd
ed.). Amsterdam, Netherlands: Elsevier/Academic Press.
IFAT LEVY SHORT BIOGRAPHY
Ifat Levy is a faculty member in Yale School of Medicine, conducting neuroeconomics research in humans. She uses behavioral methods together with
MRI to investigate decision making under uncertainty in healthy individuals,
as well as in disorders including obesity and mental illness.
URL: http://bbs.yale.edu/neuroscience/people/ifat_levy.profile
RELATED ESSAYS
Models of Revealed Preference (Economics), Abi Adams and Ian Crawford
Coevolution of Decision-Making and Social Environments (Sociology),
Elizabeth Bruch et al.
Mental Models (Psychology), Ruth M. J. Byrne
Choice Architecture (Psychology), Adrian R. Camilleri and Rick P. Larrick
Heuristic Decision Making (Political Science), Edward G. Carmines and
Nicholas J. D’Amico
Culture and Cognition (Sociology), Karen A. Cerulo
Self-Fulfilling Prophesies, Placebo Effects, and the Social-Psychological
Creation of Reality (Sociology), Alia Crum and Damon J. Phillips
Regulatory Focus Theory (Psychology), E. Tory Higgins
Behavioral Economics (Sociology), Guy Hochman and Dan Ariely
Cultural Neuroscience: Connecting Culture, Brain, and Genes (Psychology),
Shinobu Kitayama and Sarah Huff
14
EMERGING TRENDS IN THE SOCIAL AND BEHAVIORAL SCIENCES
Models of Duality (Psychology), Anand Krishna et al.
Against Game Theory (Political Science), Gale M. Lucas et al.
Neural and Cognitive Plasticity (Psychology), Eduardo Mercado III
Embodied Knowledge (Psychology), Diane Pecher and René Zeelenberg
Attention and Perception (Psychology), Ronald A. Rensink
Event Processing as an Executive Enterprise (Psychology), Robbie A. Ross
and Dare A. Baldwin
How Form Constrains Function in the Human Brain (Psychology), Timothy
D. Verstynen