|
Gregory
S. Berns M.D.
Professor of Psychiatry and Behavioral Sciences, Emory University
School of Medicine
"Neuroeconomics and
the Social Brain"
Consider this Neuroeconomics 101. It's probably a field about three
or four years old in general. That doesn't do justice to the fact
that there have been a number of behavioral economists talking about
this for a decade. It has taken a critical mass of neuroscientists
and economists to come together and attack the problems that are
interesting not only to this audience here today, but to people
in general. We're interested in how the brain makes decisions. In
one sense, all that really matters is the end result, but that's
a bit shortsighted, because we can learn a great deal from how the
brain makes decisions. I don't have all the answers, because this
is really just the beginning of the field. I will give you a quick
tour of some of the work that we have done in our lab and in other
labs, as well.
One way to think about the brain is that it's a part of several
systems that are nested with one another. The questions that we
ask depend first on the level that you're looking at. Decision-making
questions can be framed at the level of the genes that we all entered
this world with, and how they might bias or impact our decisions.
In psychology and behavioral science, there tend to be periods where
the field embraces first nature and then nurture. In the 1950s,
it was all about Skinner and behavioralism, and a lot of people
really didn't understand genes or the role they might play. Instead,
they believed that we come into the world as amorphous beings, which
are shaped and molded. That's shifted towards the biological model
in recent years, particularly along with the sequencing of the human
genome. We're finding more and more that our genes govern a lot
of our behavior. In very few cases, there's no one gene that's going
to govern a behavior all by itself.
The process of medicine and genomics has become bogged down a bit
by the fact that genes interact with one another, and it's a lot
more complicated than we imagined. Right now we're not at a level
where we can point to the genes that govern what decisions people
make, but what we can do is understand that genes code the instructions
for the hardware that ultimately makes the decisions. The hardware
is the brain. We don't understand the software so well, but we're
beginning to understand the hardware.
So genes are in cells, and the brain is made up of cells, and out
of this black box that we call the brain, somehow we get decisions.
These can be everything from the mundane to the sublime-it doesn't
matter what the decision is. Instead, it's interesting that we have
the ability to look into this black box and try to figure out how
the decisions are made. Finally, at more of a cultural and market
level, we find the effects of collective decisions. It becomes more
interesting at this level than at the individual level. This level,
of course, leads back to genes, because the ultimate biological
decision we make is who to mate with. This closes the circle.
Let's look at some examples of decision making. Here's a picture
of people buying stuff in a convenience store. They seem rather
intent on what one person is doing. He's buying a lottery ticket.
What's interesting about lotteries is that they're irrational on
the objective level. You can make some sort of argument that it's
a wash because the jackpot roughly matches the odds of winning.
How many people here have bought a lottery ticket? And when you
did so, did you know you weren't going to win? So why do it? This
is an example of a question that economists have struggled with.
There are a variety of explanations for it. A common explanation
is that people don't understand the odds, and they really think
they're going to win. That's a classic explanation. That wasn't
the case with me, however, when I bought a lottery ticket. I knew
I wasn't going to win.
There's another possibility: When you buy a lottery ticket, you're
actually buying an experience. On the occasions I've bought a lottery
ticket, I'd try to buy it as early as possible in the week so that
I could have as many days as possible of fantasizing about winning!
[laughter] It's like buying a ticket for a ride. However, the people
in the picture of the convenience store probably wouldn't agree
with that assessment. There's also another interesting aspect of
this: the expectation of the outcome. It's easy to fantasize about
the outcomes, and that's what motivates the behavior in the first
place. Most people find it difficult to win $100 million in one
fell swoop. It's the fantasy that motivates the behavior. The outcome
is quite different.
Here's a picture of some winners of a $314 million lottery. The
man's name is Andrew Whittaker, and it was the largest lottery won
in the US at the time. But look at the expressions on the faces
of the people in the picture. You can see that Mr. Whittaker isn't
smiling-certainly not as happy as you might expect. Even his wife
isn't smiling. The saddest is his daughter, who ended up dead a
year later of a probable drug overdose. This gentleman has had nothing
but bad luck after winning the lottery. The moral of the story is
that in decision making things are not generally what you expect
them to be. It's the expectation that drives decision making more
than anything else. There are often breakdowns between our expectations
and the outcomes, even though we often fool ourselves or convince
ourselves that the decisions we make are governed by the power of
the outcome.
So who cares? We care because expectations are not something that
we get to know just by asking people. So possibly, neuroeconomics
can get into and explore this aspect of decision making and the
intervening time between when the decision is made and the outcome
experienced.
Here's a more salient example. The name "Google" immediately
conjures up all sorts of thoughts and expectations. Depending on
your situation and context, your expectations may be different.
You could have an expectation of its value as a search engine. Or
you might be interested in the mapping and direction-finding feature.
Or you might have an expectation concerning the performance of the
company over time. Or you might be using Google as a way to make
a decision about buying, holding or selling a stock. What does that
mean? Specifically, for this audience, I thought about this problem
and pulled up a buy/hold/sell recommendation sheet through Google,
and the interesting thing is that I can look at the sheet and have
different reactions to it. I can look at the fundamentals and scratch
my head. There's a P/E ratio of 60 and a price-to-book ratio. I
see different "buy" recommendations, and most of the analysts
are recommending a "buy" on this particular stock. Then
I go back and look at the price history chart and scratch my head
some more. I'm not an expert in this sort of thing, so I tend to
use experts to look at it for me.
I want you to keep this example in mind. Ask yourself whether it's
possible that everyone who analyzes the Google stock comes to the
same buy/hold/sell conclusion. The information is basically the
same. It's possible that analyst A comes to a "buy" recommendation
and analyst B doesn't come to a "buy" recommendation.
Part of this comes out of the idea that markets are efficient. Not
everyone has equal access to information. Generally, it's probably
true that in the aggregate markets are efficient except in certain
circumstances. That's what I'd like to focus the rest of the talk
on-how to tell when we're in circumstances that represent the exception
to the efficiency assumption.
Neuroeconomics has several components. First, I'll talk about the
actual apparatus being used. I'll talk mostly about brain scanning
or functional MRI. Next, I'd like to give some concrete examples
of experiments that have been done in two classes. The first is
microeconomic behavior, including valuation and decision making.
The second class is macroeconomic behavior: markets and herd behavior.
One of the core concepts that has come out of economics is the
idea of utility. It's a very old idea that says that we make decisions
to maximize our expected utility. Utility is just another word for
well-being that people get out of an object or an action. There
are different ways to measure it. You can ask someone how much pleasure
they get out of winning the lottery vs. how much pleasure they get
out of expecting to win the lottery. Those are two different answers.
Or you can forget about asking the person altogether and just see
what they do, under the assumption that whatever the person does
at a given moment is maximizing their expected utility. This is
rational decision making. We know there are numerous instances of
anomalous behavior outside of this model. There are other ways to
assess utility, such as someone's willingness to pay for something
and how much they're willing to pay. Then there are neurobiological
methods.
Which of these approaches is the right answer? We don't really
know-that's at the heart of the science and the inquiry.
Let's talk a bit about the brain. I'm going to show some brain
images, and there are two types. One is an anatomical image, which
is a 3-D snapshot of the brain. If you've had an MRI or seen an
MRI, then you know what I'm talking about. The other is more useful,
and it's called a functional image. It's like a movie. When we do
functional MRIs, we scan the brain very rapidly, about once every
second. While the person is in the scanner, we have them do various
simple tasks. In this way, we can link up what's happening in the
brain with what they're doing.
There are three ways to slice the brain: coronal, sagittal (or
side view), and axial. After you look at these for a while, you
can navigate your way through the brain in three dimensions. Sometimes
we do 3-D renderings of the brain. We can look at what's happening
on the surface, but this is somewhat limiting, so we usually resort
to using these different types of slices to see what's going on.
Usually we overlay the functional data on top of an anatomical
image. The colored parts indicate activity in some region of the
brain. Active is a relative term; the brain never switches off.
Surprisingly, these changes are quite small. Usually the amount
of activity change is less than 1%. This requires special software
to pick up these changes from within the noise that surrounds the
data.
A picture of the scanner reveals the physical constraints that
we are faced with. The subject lies on his or her back in a tube.
We can project computer images for them to react to while they're
in the scanner, but the experience is not a natural setting. There
are new scanners that are a little more user friendly. The experience
is quite isolating, and some people get very claustrophobic.
Let me now acquaint you with some of the regions of the brain that
we look at. The amygdala is one that you heard last night in Robert
Sapolsky's presentation. It's not that large and seems to have a
lot to do with fear and stress. Another area that comes into play
in neuroeconomics is the ventral striatum, and I'll talk about it
later.
A couple words about where the technology came from. It began in
Italy in 1850. The father of the lie detector, or polygraph, Angelo
Mosso, found a way to monitor autonomic systems like blood pressure
and heart rate. There was a bricklayer who suffered an accident
and cracked his skull. He always had a defect in his skull-covered
with skin-but you could put your hand on it and touch his brain
through his skin. Mosso put this very crude blood pressure device
on the hole and made him do various things and measured the pressure
in this part of the brain. He noticed that when the man was doing
tasks that required a lot of mental effort, the pulsations in this
part of the brain started increasing. One time he was measuring
these and the bells went off in the town square. At this point,
the pulsations started going off like crazy, and he asked the man
what he was thinking about. He replied that he remembered that he
hadn't gone to church that week. This is the basis for modern brain
imaging. Parts of the brain that are working send out signals to
recruit more blood. That's all there is to it.
Technically, MRI is like a microwave. The magnet is quite powerful.
It flips all of the hydrogen atoms in your body so they line up
with the magnetic field. They behave in peculiar ways, like tops.
They spin at a certain frequency. MRI sends a radio signal into
the body at the right frequency, and the hydrogen atoms start spinning
even more. Then you turn off the radio pulse, and the atoms relax
and become like an antenna from which you pick up a signal. There
are some tricks in the software, but that's the basic principle.
The trick that allowed MRI to be discovered goes back to Linus
Pauling. It turns out that blood exists in two different forms.
There's either oxygen in it or not (arterial or venus). Because
there's iron in blood, it distorts magnetic fields depending on
whether or not there's oxygen attached to it. When there's oxygen
attached to a blood molecule, it doesn't distort the magnetic field,
and when there's no oxygen in a blood molecule, it does distort
the field. We take advantage of that with brain imaging. We pick
up changes in oxygen. Parts of the brain that are more active require
more oxygen. A functional brain image looks like an amorphous blob.
How do we bring this into decision making? The rational model involves
some systems in the brain concerned with perception-what you see
and what you think you see. Other systems include memories and drives
like food, safety and sex. Then there are actions. There are probably
a fairly small number of actions that we do compared to all these
other processes. At any moment in time, we combine the perception,
memory and drives to form a prediction of expected utility. Then
the ideal rational process optimizes this combination and chooses
a decision that becomes an action. But there's a problem with this.
It's not a linear process. In fact, there is feedback involved.
In the process of making the decision, for example, you may re-evaluate
what you're looking at (your perception). As you're making a decision
about a stock, for example, you might see things different or remember
something different. These processes feed back on one another, and
eventually the feedback loops settle down and you make a decision.
But it's quite difficult to map directly onto the brain.
One of the core concepts I'm working on is the idea of utility
or valuation. How does the brain predict value? Presumably all the
decisions we make are done in the interest of maximizing value.
This leads us to talk for a moment about dopamine. Dopamine has
been known for about 50 years. It was originally thought of as a
pleasure chemical. Science has advanced quite a bit from that point
of view, though. What I'm showing you now is a time series of a
dopamine cell firing. Spikes are accumulated into histograms over
time.
This time series is from a monkey that is undergoing a Pavlovian
test. This is the simplest form of learning that we know of: associative
learning. A light comes on, and a few seconds later the monkey gets
a squirt of juice. The squirt of juice is a form of reward. If there's
no light, and you just squirt the juice in his mouth, then you get
bursts of dopamine. But if you start pairing the juice with the
light, the dopamine is released when the light goes off, not when
the juice is delivered. If you show the light and don't squirt the
juice, then you get the spike at the light and a little depression
when the juice doesn't arrive. Dopamine is not specifically about
pleasure, then, but about expectations and predictions. The field
evolved quite a bit after learning this. The brain is geared to
look into the future, and dopamine plays a big role in this.
Here's an experiment where the researcher had people in a scanner
working simple tasks with different amounts of money. He wanted
to know at which point the brain was most active: when people actually
get the money, or some time before that. The answer has to do with
anticipation. One part of the brain is more active in anticipation
than in receipt of the reward. There are other parts of the brain
in the prefrontal cortex that seem to respond more to the reward
itself. We have different systems that seem to compute different
aspects of the valuation problem.
The phenomenon is not specific to money either. Another test had
to do with beautiful faces (male and female) and having them rated
on attractiveness. You can do this in a scanner and see how the
brain responds to the same test. On a paper and pencil test, men
won't truthfully rate how they view other men. The brain doesn't
do that, however. You can focus in on the same parts of the brain
where dopamine is most active. You can compare what happens in the
brain when viewing a beautiful female vs. an average female and
what happens when you view a beautiful male vs. an average male.
The same parts of the brain are distinguishing the aesthetic features
of the faces. So it does seem that there are these common systems
that happen to deal with valuation.
Now I'll shift gears and talk about a different aspect of valuation:
time. In almost every decision that people make, there's an element
of time. People have to make a decision in advance of the outcome,
otherwise they're just reacting. This is where the issue of discounting
comes in. Samuelson drew the original curve, which is an exponential
function. This was the first idea of how to think about valuation
and time. He assumed that time degrades value at a constant rate.
Mathematically it's very simple, but this is clearly not how people
value time. It was just a mathematical suggestion. It's a very entrenched
idea about time, though.
Alternatives have been proposed to account for behavioral anomalies.
One of the anomalies is called temporal myopia. Things that are
right in front of us are worth more than things that are far away
in time. Near-term myopia becomes overwhelming. It explains a lot
of behavioral problems, ranging from why people have a hard time
saving to dieting. We make plans to diet, but then those plans go
out the window when the meal and the immediate gratification are
right there in front of us. If this is so pervasive across the animal
world as a phenomenon, there must be a biological basis for it.
In fact there is.
We got interested in this. Here is a Norman Rockwell painting called
"Freedom From Fear." I like this picture because there
are different emotions. The father's face seems to reveal a mixture
of happiness and dread and uncertainty about the future. What goes
on in the brain in the intervening time between when you have to
make a decision and when the outcome actually occurs? Time becomes
a factor in peculiar ways.
One way to get at the answer is to look at how pain works. There's
a general idea that pain comes in different forms, but it may go
through the same pathways, regardless of whether it's physical or
psychological. The pain matrix in the brain has been well described.
There are many aspects of pain, though. One has to do with the intensity
of whatever is causing the pain. The bigger it is, the worse it
is. As you increase the intensity of that stimulation, not only
does it hurt more, but you pay more attention to it. As attention
increases, the worse it hurts. This describes a positive feedback
loop. Then there are emotional responses.
We did an experiment on pain. We gave people pain stimuli-little
electric shocks. We were interested in what would happen if people
had to wait for the pain, and we gave them choices. For example,
would you prefer to wait for a 50 volt shock for a minute, or would
you rather have a 100 volt shock right now? I'll show you what this
feels like through a video of a volunteer. She gets two shocks:
one at the beginning-and you'll see her flinch-and one that she
has to wait for. Just pay attention. [Video shows the first shock
and then the volunteer starts waiting and says, "I hate waiting!"
Finally the second shock arrives.]
It's uncomfortable to even watch. It underscores the fact that
time is palpable and affects decision making. Next we asked her
to make a decision and choose between a 60% shock in 27 seconds
or a 90% shock in 3 seconds. You can compute how sensitive a person
is to time by how they're willing to trade a specific strength of
pulse for a particular time frame. Most people don't like to wait-we
call those people dreaders. Only a subset of those people really
doesn't like to wait- we call them extreme dreaders. The true incidence
in society is probably much higher because the people who volunteer
for an experiment like this are probably not big dreaders to begin
with.
We produced curves for a person's preference for level of voltage
vs. a delay of a specific period of time. They show the effect of
time. They look very similar to discount curves. But what can the
brain tell us about why some people have different curves? We can
identify the parts of the brain that are responding to pain and
measure what's going on in there while people are waiting.
Pain is kind of funny because it's hard to imagine what pain is
like. If I ask you to imagine a painful experience, you can imagine
the circumstances around it, but you don't call up a memory of the
pain itself, which is probably a good thing. For example, women
can remember circumstances around childbirth, but that doesn't mean
they're experiencing a memory of the pain of labor. But when you
cue a person and tell them they're going to be shocked, that's a
very different situation. The same parts of the brain that experience
the shock simulate the shock while you're waiting. There's a lead-up
time to when the shock occurs. They're basically exponential growth
curves like the discount curve. These curves differ depending on
whether a person is a mild or an extreme dreader. We find that what
distinguishes these two types of people is the initial response.
Some people ramp up rapidly and sustain the activity in the pain
centers of the brain, and others do not. The ones who ramp up rapidly
are the extreme dreaders. The dreaders just don't get worked up
about it until they get closer to the time of the shock. This phenomenon
affects how each type of person makes decisions. Some will choose
to get it over with even if it means more pain. When I describe
this experiment, most people know who they are.
Let me say a few things about game theory. Here's Dr. Strangelove,
but the real one was John von Neumann. He was tasked by Eisenhower
to solve the "first strike problem" in the 1950s. The
problem was, should we nuke Russia first or wait for them to nuke
us first? This led to a set of problems in game theory called the
Prisoner's Dilemma. It's been studied thousands of times. The game
goes like this. You have a partner and you have the choice of cooperating
with the other person or defecting against them. You don't know
what that person will do-you reveal your responses together at the
same time. For each combination of responses, there's a payout.
If you both cooperate, you both get two dollars. If you cooperate
and your partner defects, you get nothing, while he gets three dollars
(and vice versa). If you both defect, you get a little bit of money,
say one dollar. What's the optimal solution? From the rational point
of view, the safest thing to do is defect, but it's not the optimal
solution. In real lab situations, most people cooperate. There are
a variety of explanations for that, and we wanted to know what's
going on in the brain when people cooperate. In the brain, cooperation
increases activity in the reward system and the valuation system.
It's as if that outcome is more valuable than the actual money or
reward. This means there are social currencies as well as monetary
currencies. In any activity, we know that social currencies are
very valuable, perhaps more so than the actual outcome.
You can do other games like that. There's a game called Ultimatum.
Two people play. You're offered $10. You make any offer to split
the pot, and your partner accepts or rejects the offer. If he rejects
the offer, neither of you get anything. If he accepts the offer,
you execute the split you proposed. It's been studied a thousand
different ways in different cultures. The fair thing to do is to
offer 50% of the pot. The rational thing to do is to offer a minimum
amount. The other person should accept it, no matter what percentage
it is, because it's money they would otherwise not have. There's
no rational reason to refuse it. In practice, though, people will
reject offers less than 20%. They will accept offers between 20%
and 50%, but they won't like it. We can observe what happens in
the brain when people receive offers that are unfair: the pain matrix
lights up. It's painful when someone senses that something is unfair.
This leads to my final topic: social norms. What happens when you
start increasing the number of individuals in decision making? A
very famous experiment in social psychology was conducted by Solomon
Asch in the 1950s. He experimented on undergraduates at colleges
like Princeton. He brought people into a room and showed them pictures
like this. [The picture shows four lines: a control line and three
lines of different lengths labeled A, B and C.]
The task was simple: to pick the line segment from among the three
on the right that was the same length as the line on the left. It's
pretty easy. Except he brought the subjects in and everyone sat
around a table. If there were eight people, seven of them were plants
and only one of them was the real subject. The other seven would
all pick the wrong answer. Asch then documented how often the subject
went along with everyone else in the group. Surprisingly, a large
number of people would go along with the group. There are several
explanations for it, and it depends a bit on the incentives you
give. The most important question is: Are they doing it because
they want to go along with the group because it's difficult to stand
alone? That seems to be the consensus. Asch pointed out the possibility
that perhaps the group changed the perception of the subject. Certainly
in cases that are more complex than the line example I'm showing
you, that seems possible. If you're not sure what you're seeing,
you may rely on other people's perceptions.
We got interested in this experiment. We wanted to understand bubbles,
but we didn't think we could tackle the problem of bubbles head
on because we realized there was this problem of perception. Think
back to Google. It's not clear that the rise in the value of Google
was because people arrived at the same conclusion independently
and decided that Google was worth $400. Or did some people look
at the chart thinking, "Well everyone else gets this value,
so that's what it must be." Does that actually change how they
viewed the fundamentals?
So we did a version of the Asch experiment. I'll show you a clip
from Prime Time. They wanted to recreate the experiment in their
studio. I was very skeptical about doing it, and warned them that
it was an experiment and might not work on TV. Fortunately it did-surprisingly
well. Here's the clip and then I'll explain the science behind it.
[Clip plays:] "Like birds in a flock. Like sheep in a pasture.
Like salmon swimming upstream. We follow, sometimes at our own peril.
But why? That's what we set out to discover. We gathered a group
of unsuspecting people for a test of what we call 'visual perception.'
But our Prime Time lab is really after a question far more revealing:
What makes people follow each other? It's the question that Dr.
Gregory Berns at Atlanta's Emory University answered in a recent
groundbreaking paper. It's so compelling that we set up a demonstration
recreating his work.
"The actual test is simple: to take geometric shapes and compare
them to see if they're alike or different. First our volunteers
have to write down their answers to 10 questions privately. Then
they have to give the series of answers out loud for everybody to
hear. But this verbal test comes with a twist. Most of the participants
are in on the experiment, except for Tony. He's being set up to
see if he will follow the pattern. When the group gives a right
answer, he agrees. But even when the shapes are vastly different,
Tony still gives the wrong answer. He follows the pack. What's really
going on? Unwittingly, Tony has demonstrated Dr. Berns's point precisely
by going along with the wrong answer. The group's influence on Tony
profoundly altered the results. He went from 90% on his written
test to just 10% when he heard everyone's answers. Tony's not alone.
We changed the hot seat several times. Eric got a 100% on the written
exam, but when he followed the others he got 60%. A few people followed
the courage of their convictions, giving the correct answer even
when the group did not." [Clip ends]
This is how people behave. Do people do this consciously? Do they
go along with the social network, or does their perception actually
change? The actual experiment was actually worse than what you just
saw, because the participants were in the scanner and had to look
at the shapes and give their answers. Simultaneously, you see pictures
of four people who you just met and they're giving their responses.
People got about 10% to 15% wrong when they solved the problems
themselves. But when the group is giving wrong answers, the error
rates jump up to about 40%. That's quite a big jump.
What happens in the parts of the brain that govern perception?
Perception is more or less in the back of the brain. When people
go along with the group and give the wrong answer, these places
in the back of the brain light up. The activity is not in the frontal
lobes. Now, what happens when people stand up to the group? The
amygdala lights up, which indicates the fear of standing alone.
Even when people are feeling they're right, it's still hard for
them to do. Our perceptions are sometimes colored by other people.
It's not an issue of courage or integrity. You actually start to
revise what you're looking at, and you see it differently. When
you get further away from the physical reality of what you're looking
at and it's more of a judgment decision, then you can imagine how
overpowering some of these perceptions can become.
To summarize, I've given you some examples of micro- and macroeconomics.
The real interesting thing would be to look at policymaking. Currently,
neuroeconomics lies somewhere between the areas of subjective states
and behavior.
Q&A Session
Question: [unintelligible] Delay preference as it relates
to transience or permanence…
Answer: It hasn't been studied as much, primarily because
the longer the unpleasant event is, the worse it is. But still…
[unintelligible] This is an evolving area. There's this overriding
idea that you have these algorithms in the brain that you apply
to all outcomes, but you actually have different discount functions
for different decisions. An example is road rage, where people just
snap and there's no future-everything is instantaneous. But that's
very difficult to study. It probably has a lot to do with a specific
domain. What's interesting about money is that we probably don't
have a specific system in our brains for money decisions. How we
discount money may depend upon which other systems are involved:
food systems, reproductive systems. They all have different discount
functions.
Question: Many of these studies have been performed on a
small subset of the population-mainly poor, young college students.
How dangerous is it to apply the results to the general public?
Answer: You're right that we use small samples, but we do
our best these days to get not so biased in terms of using college
students. We make a genuine effort to reach out to people outside
of the university. Most of our studies have an average age closer
to 30. We try to get people in their 40s to balance things out.
There is an encouraging trend in neuroscience moving to larger sample
sizes. You used to be able to publish a study with 10 or 12 samples.
There are a couple of studies where researchers have studied hundreds
of subjects, but we always have a problem of selection bias. We
have to rely on volunteers.
Question: You showed the people who were standing alone
vs. those who were influenced by the group as having different brain
actions. Is there evidence that people can train themselves to have
a different response?
Answer: In that context, I don't know. In other contexts
where people have done tasks and learned something new, it seems
clear that there is a corresponding change in brain activity. The
question I think you're asking is "Can we use the brain imaging
to try to train people?" That's an interesting question. My
sense is yes. There's been a development in this area that's real-time
feedback. There's a company in Palo Alto that's developing this
idea. It's a kind of biofeedback. The images I showed you are fed
back to the subject in real time. You monitor a part of the brain,
feed it back and have them learn from it. It's been done successfully
with pain.
Question: In your studies, does the stature of the people
in relation to the subject matter? For instance, can college professors
influence students in their decision making?
Answer: When we did the study, all of the people in the
room were peers. They weren't authority figures. Authority figures
do tend to amplify the effect of going along with the crowd.
Question: How might payoffs affect the results? The upside
is just the satisfaction of being right, but the downside is this
feeling of not being with the group. Even for rational people, that
causes a little pain. What if there were payoffs for people guessing
right. You could then calculate how much it would be worth to be
out of sync with the rest of the group.
Answer: That's right. You can get rid of the temptation to
go along with the group by giving the right incentives. Then the
issue is what to do in the fuzzy area. Incentives make a difference.
This just points out that the social context serves as an incentive,
as well.
Question: What are the implications on committee decision
making? What are the pitfalls in making a decision as a group?
Answer: That's a basic management question. How many people
have been on committees where they have some task and then they
vote on it? The benefit of a committee is bringing together people
with different views and then looking at the aggregate of that information.
If the committee is structured so that votes are taken publicly,
then you've lost a lot of the advantage of the committee due to
the social pressure. Also, there are problems when people talk before
the meeting asking each other how they're going to vote on something.
Committees need to minimize the social pressure, and there are a
variety of ways to do that, like closed ballots and mechanisms where
people are not forced to stand up in front of everyone else and
say why they think everyone else is wrong.
Question: You said that women came across as being more
susceptible to influence. Do you think the differences between men
and women were significant, and if so, what studies have been done
to understand that difference?
Answer: We never published that specific result. As a baseline,
women tended not to do as well on the test. That's been documented
on a lot of psychological tests, but it's not to say that it's a
biological cause. Even when you factor out the baseline performance,
there was still a difference in the rate at which women were going
along with the group versus men. We never pursued that very hard,
and I don't know if it had to do with confidence at all. The women
actually thought they were worse than they really were in going
along with the group. It is statistically and practically significant.
It underscores the power of these group effects.
The
comments, opinions and any forward predictions presented about any
particular security, the economy and "the market" are
based on the analysis of the speaker. These are not necessarily
the opinion of, and should not be construed as a recommendation
on the part of Legg Mason Capital Management or any of its affiliates.
|
