Etica & Politica / Ethics & Politics, XI, 2009, 2, pp. 10−17
Neuromarketing: Ethical and Political Challenges
Neil Levy
Florey Neuroscience Institutes, Melbourne, Australia,
Oxford Centre for Neuroethics
neil.nl.levy@gmail.com
ABSTRACT
Ethicists and ordinary people are typically more worried by interventions that alter agents’
mind by directly altering their brains than interventions than are focused on the environment,
and thereby indirectly change minds. I argue that the causal route to changing minds is not itself
important. Moreover, some of the most powerful techniques whereby behavior is altered
without the consent or knowledge of agents involve environmental manipulations: manipulations
of social space, for the benefit of those in the business of increasing consumption. I argue
that insofar as we are fixated on internal interventions, we overlook ways in which our autonomy
as agents is impaired. Once we recognize the power of environmental manipulations, however,
we should come to see social space as a legitimate target for political control.
0. Introduction
When philosophers and ordinary people worry about how agents can be manipulated,
they usually focus on internal goings-on. They focus on internal goings-on
in two ways: they worry more about manipulations which affect the brains of
agents than those that affect their environment, and they look for ways to (a)
prevent such internal manipulations and (b) strengthen the internal resources of
agents. There is good reason, sometimes, to worry about internal manipulations
and there are often good reasons to help agents build internal resources. But the
most effective and practicable manipulations of behavior, including buying behavior,
are most likely to be external manipulations; and the most effective ways
of strengthening self-control are likely to involve structuring the environment to
facilitate it. In this short paper, I shall argue that focusing on internal goings on
carries risks: risks of allowing marketers free reign to manipulate just where they
are most likely to be effective, and of distracting attention from ways of resisting
marketing pressures that are low-tech but effective.
1. Internal and External Manipulations of Behavior
Let me begin by very briefly recounting two recent experiments focusing on mind
and behavior. The first experiment examined the effects of the hormone oxytocin
on people’s behavior. It was found that administering the hormone (via a nasal
Neuromarketing: Ethical and Political Challenges
11
spray) increased people’s trusting behavior, as measured by their willingness to
cooperate in a bargaining game in which cooperation carries the risk of loss (Kosfield
et al. 2005). The second study was an imaging experiment, comparing the
brains of subjects drinking Coke and Pepsi, with and without the knowledge of
which brands they were drinking. When subjects were ignorant of the brands,
only the ventromedial prefrontal cortex was active. But when the brands were
identified, subjects exhibited additional – dorsolateral prefrontal and hippocampal
– responses to Coke alone (McClure et al. 2004).
Now, it is not so much the content of these experiments that interest me here
as the responses to them. In their book Affluenza, Clive Hamilton and Richard
Denniss label the Coke experiment ‘disturbing’ (Hamilton & Denniss 2005: 41-2).
It shows, they conclude, that ‘We have not so much been brainwashed into drinking
Coca-Cola: we have had our brains rewired to want it’ (42). Now Hamilton &
Denniss are surely right in thinking that due largely to Coke’s effective advertising
and cultural saturation strategy, the company has been successful in getting
people to make the association between the drink and a host of other things which
are in fact unrelated to it. But the fMRI results do not tell us that. What does tell
us that? Well, largely people’s behavior; their verbal behavior as well as their buying
behavior. What the fMRI results tell us that is that people have memories and
visual percepts associated with Coke. Surely, though, we didn’t need neuroscience
to tell us that (suppose that people failed to experience hippocampal activation in
the experiment. What would we conclude? Well, probably that the equipment was
faulty). Now, I am not saying that we can’t learn new and surprising things from
neuroscience. I am saying that we should not be impressed with images of brains
just because they are images of brains. The internal is not privileged. If anything,
it is the other way round: we look to behavior to validate neuroscience, not the
other way round.
Worries expressed over the oxytocin experiment are better founded. Some people
have expressed the worry that oxytocin might be used to bolster trust in
speakers at political rallies (Damasio 2005). It’s easy to see how this worry might
be extended to a sales context. Trust will greatly facilitate the job of a salesperson.
But notice that salespeople already engage in activities designed to facilitate
trust. Indeed, it is likely that they engage in actions which have the effect of facilitating
trust by way of increasing oxytocin (though of course they don’t think of
what they do in this way). Here’s an example from a sales handbook: tell the customer
that the product they’re looking at is somehow defective or unsuitable for
them. Since they think that if you’re trying to sell to them no matter what, you
will tell them that the product is good, their trust in you should rise. Now lead
them to the product you wanted them to buy all along. Here’s another technique,
which suggests itself once we know that oxytocin facilitates trust: stimulate it.
For instance, have kittens around, or pictures of kittens.
NEIL LEVY
12
My claim is that these entirely environmental manipulation should be just as
disturbing to us as internal. Yet people continue to believe that neuroscience has a
special and especially intimate ability to peer into the soul, or at least its closest
secular equivalent. Consider, in this context, another experiment. Weisberg and
colleagues gave subjects, both scientists and laypeople, explanations of events and
facts that were deliberately designed to be obviously flawed. They found that
both groups of subjects could identify faulty explanations. Yet when the subjects
were given the same explanations peppered with neuroscientific jargon, their critical
abilities left them, and they took bad explanations for good. We think brain
science is especially good science, and are credulous toward it. Perhaps this is because
the science is new; I think, though, that part of the explanation is that brain
science is an internal science. Because it looks at what is going on inside, it is
taken to have far greater depth than mere psychology. Once again, I am not saying
that brain science is not good science. I think it is still in its infancy, but its results
are already very exciting. I only want to say that we ought not be dazzled by
it, or overlook the other ways we learn about, and control, human behavior.
2. The Challenge of Neuromarketing
Now, what has this got to do with marketing? Recently, there has been a great
deal of talk, and worry about, neuromarketing. Neuromarketing would apply the
methods of neuroscience to moving products. For instance, rather than ask focus
groups what they think of a product, we could place them in fMRI machines and
measure their neural responses to products. Why do that? Well, perhaps subjects
are not always honest: they may tell people what they think they want to hear,
rather than what they really believe. Or perhaps even subjects who are trying to
be honest are unable to be sure what they think of a product. They may be deaf to
the signals that intimate that their infatuation with a new drink is likely to be
short-lived (or whatever the case may be). By peering directly into brains, the
hope of some (and the fear of others) is that we could cut through this messiness.
We could focus directly on signals of pleasure in subjects, for instance, avoiding
worries about insincerity. We could know more about subjects then they know
about themselves. In that way, we could very finely attune products and marketing
campaigns to target audiences and their desires, thus making products well
nigh irresistible.
I do not think we can rule out the possibility that these methods could prove
useful to marketers. But I think there is very good reason to be skeptical, for the
moment and for the foreseeable future. First, consider dishonesty. I see no reason
to deny that subjects might not exhibit neural signals of, say, pleasure, which
they can attempt verbally to dissimulate. fMRIs might be a way of detecting such
signals. It’s a pretty expensive way, though, and others that are cheaper are
Neuromarketing: Ethical and Political Challenges
13
available. We can observe subjects behavior covertly; we can measure physical responses
(psychophysiology is probably a more reliable guide of arousal than
fMRI). In any case, we can rest assured that most subjects attempt to be honest
most of the time. Similar points apply to ambivalence. For the moment, at least,
verbal response and overt behavior is a much better indicator of how subjects are
feeling than are brain scans.
Moreover, it is far from certain that fMRI will always detect dissimulation. Independent
testing of neuroscientific technologies specifically designed for lie detection
has not been very promising (Miyake, Mizutani & Yamamura, 1993); it would
be surprising if techniques developed for other purposes were more successful.
Moreover, when subjects are motivated to ‘beat’ neural lie detection technology
they appear able to succeed (Rosenfeld et al. 2004); successful countermeasures
are surprising simple and easy to implement. Investing in neuroscience to detect
dissimulation is likely to be ineffective; in any case, it is unlikely to be necessary.
For the most, I suspect, neuromarketing will prove an expensive waste of time.
It is also, and much worse, risky – not in itself, but insofar as it diverts attention
from external ways of manipulating behavior that are more effective, and therefore
far more in need of regulation. These ways of manipulating behavior build on
the techniques that marketers have been using for many decades now, of controlling
consumption by controlling space. They work, they are scientifically validated,
and they are likely to be refined in the future. We don’t worry about them
because they focus on the environment around us, rather than on our brains. But
we ought to be more concerned about them than the internal manipulations which
might, one day, prove practicable.
First, let me mention some of the techniques that marketers use today to encourage
consumption (that is, to encourage people to buy more than they need, as
well as to shift their buying preferences). It has long been known that how goods
are positioned affects consumer behavior. Putting the most popular goods in the
middle of aisles maximizes exposure to the products – if they are placed at the
ends, people may grab them and head to the checkouts without passing the products
they did not intend to buy. Similarly, essential items are often placed at the
back of the shop to increase traffic past non-essential goods. Products may be ordered
in terms of price, with more expensive goods placed where they will be encountered
earlier. The result is that cheaper goods are perceived as good buys, in
comparison with more expensive goods. More counterintuitively, pricing in multiple
lots can encourage greater buying, and once the buying pattern is established,
it tends to be maintained (Wansink et al. 1998).
Now, marketers might tell you that all these techniques are merely designed to
expand consumer choice. By making consumers walk past goods, they are made
aware of their existence. Special offers give people choices to buy in multiple lots,
thus expanding their options. It should be admitted that these responses have
some degree of plausibility. Certainly, these techniques don’t look particularly
NEIL LEVY
14
threatening. Unlike internal manipulations, they leave the agent, with her powers
of resistance and choice, unaffected. I believe that this distinction between the
agent and her environment both philosophically suspect and practically dangerous.
Let me now tell you about some techniques currently being explored to encourage
consumption behavior which work on agential powers from the outside.
These techniques build on a well-established paradigm in social psychology called
the ego-depletion hypothesis.
The ego depletion hypothesis is, roughly, this: willpower is a limited resource,
which we draw upon to resist temptation. That is, when we judge that we ought
to act in some way, but are tempted to act in a way that conflicts with our judgment,
we draw upon our willpower to overcome temptation. All too commonly, of
course, we fail to act as we judge we ought in these kind of circumstances: we continue
to smoke when we judge that we ought to give up, we eat or drink more
than we believe we ought to, we spend more than we think we should. But sometimes,
in what seems to be exactly the same circumstances, we manage to bring
our actions into line with our judgments. Moreover, some people seem to be better
at this than others. What explains our success at getting ourselves to do what we
judge we ought, when we succeed? And what explains why some people are better
at this than others? We often answer this question with just the word ‘willpower’.
Proponents of the ego depletion hypothesis say that we are exactly right.
Willpower, they argue, is resource-like inasmuch as we use it, we use it up (that
is what the word ‘depletion’ means in ego-depletion). Of course, we don’t use it up
permanently, just until we get an opportunity to replenish it. What replenishes it
is rest. This is sometimes referred to as the muscle model of self-control. How
much we can lift at any one time depends upon the state of our muscles. Muscular
strength is used up in the short term: if I have just done 20 push-ups, I find it very
difficult to do more. But this strength returns with rest. Moreover, muscular
strength increases with us; similarly, proponents of ego depletion suggest that
willpower can increase with regular exercise (Baumeister et al. 1998; Baumeister
2002).
Let me briefly sketch the evidence for the ego-depletion hypothesis. It comes
from a large set of studies. In these studies, subjects are divided into two groups.
One group is assigned a self-control task, which involves resisting a desire; the
control group is instead given a tiring task which does not require self-control to
anything like the same extent. For instance, in a typical study the first group was
ushered into a room filled with the smell of freshly-baked cookies, which were laid
out on plates in full view. The subjects were told that they were to participate in a
study on taste perception; one group of subjects would eat the cookies and another
would eat radishes. They, they were told, were the radish group. They were then
presented with plates of radishes, and left with instructions to eat several of them.
The experimenters withdrew and covertly observed the subjects. They were observed
to sniff and even handle the cookies, but despite the fact that they thought
Neuromarketing: Ethical and Political Challenges
15
they were unobserved, all of them dutifully ate radishes and not the cookies.
Meanwhile, the control group performed a task that did not require self-control,
but which was tiring – a typical task was performing a series of three digit multiplications
on paper.
Subjects from both groups were then tested on a common task, which does require
self-control. Typical tests include squeezing a handgrip for as long as possible,
persisting at an unsolvable anagram task, or keeping one’s hand immersed in
icy water. Subjects who have previously performed a self-control task – resisting
the cookies, or watching a funny video while keeping a completely straight face –
perform worse at the common task than do subjects in the control group, inasmuch
as they persist at the task for a significantly shorter time. Moreover, their
lack of self-control carries over to more naturalistic settings. For instance, dieters
who are ego-depleted eat more after the tests than do non-dieters (Vohs &
Heatherton 2000).
What apparently happens is that the subjects in the ego-depletion group –
those that had a self-control task prior to the common task – have less willpower
to draw on than subjects who have been given a different, yet tiring, task. How
ego-depletion works, precisely, is still rather mysterious; in recent work I have argued
that the resource depleted is not in fact a discrete self-control resource, but a
domain-general capacity: one that is involved in all effortful tasks, including deliberation
(Levy forthcoming(a)). Willpower is only one capacity that degrades
when it is called upon. What matters for our purposes, however, is that willpower
is subject to depletion: self-control tasks get progressively more difficult until the
agent gives in to temptation (succumbing, I have suggested, is mediated by judgment-
shift: once self-control resources are sufficiently depleted, the agents change
their mind concerning how they ought to behave).
Ego-depletion appears to have no phenomenology at all, at least in the shortterm
(there are studies reporting greater fatigue after the second task, but that is
after the effects of ego-depletion have already manifested; that is, after the damage
has been done). So there are no warning signs that signal to us that our willpower
is low. That makes ego-depletion a phenomenon ripe for exploitation by
those who would encourage us to consume more. If you want to make someone
buy a product that they desire, but which they prefer, all things considered, not to
purchase, you ought to ensure that their self-control resources are depleted when
they confront the option of purchasing it. This can be done by requiring potential
purchasers to engage in self-control tasks, which deplete their resources, before
they are presented with the option of purchasing. This seems to be relatively easy
to ensure: simply expose them to many tempting opportunities to consume. We
can expect them to find it progressively more difficult to resist repeated temptations.
Now, it may be that marketers are already doing just this, though they
don’t – quite – conceptualise what they’re doing in these terms. Think of a shopping
mall. Already these are intentionally designed to maximize consumer’s expoNEIL
LEVY
16
sure to temptations (escalators are positioned so that rather than just riding from
one floor to the next, we are required to continually walk around, thus ensuring
we pass shops on the way; similarly popular shops may be located so that we are
forced to pass speciality shops on the way). Once the ego depletion hypothesis is
deployed by marketeers, we can expect a more precisely targeted use of the strategy.
One possibility is to ensure that consumers are exposed to low value but
tempting items – eg. chocolate bars – early, to ensure that when they are exposed
to high-value items later, their self-control resources are at a low-ebb. Sellers of
large consumer items might find it profitable to get into retailing confectionary as
a side-line; credit card companies might find arrangements with such retailers
profitable as well.
To my knowledge, no one has tested the hypothesis that repeated shopping
opportunities are ego-depleting, all by themselves. However psychologists have
tested whether ego-depletion affects the propensity to consume, as well as whether
it affects subjects’ evaluation of consumer items (Vohs and Faber 2002). Egodepleted
subjects are more willing to buy, and will pay higher prices, at least in
the laboratory. Marketers seeking ways to control our behavior are far more likely
to use this kind of technique than to resort to the use of more invasive techniques.
They can structure the environment to produce the behavior they want, and they
can do so more effectively in this manner than they can using the tools of neuroscience.
Most people are not as worried about this kind of manipulation as they
are about neuroscientific technologies. But the differences are not important.
Both manipulate behavior, and they do so for purposes that are not those of the
people who are manipulated. Both can result in changes in beliefs, and in the actions
which express those beliefs. The way in which these alterations come about
doesn’t seem to matter, from a moral, or a political, point of view. If there are
good reasons to be concerned about the first, there are equally good reasons to
care about the second as well.
These considerations lend support to a conclusion that I have defended elsewhere
(Levy 2007; forthcoming (b)): the causal route by which behavior is modified
– whether it is largely internal or external to the skull – is irrelevant to its
ethical significance. The difference in degree of concern we see with regard to neuroscientific
results, which leads to anxiety over experiments like the Coca Cola
study mentioned earlier, but scant concern over ‘merely’ environmental manipulations,
is unjustified. There is no reason to think that because we can point to internal
goings on that autonomy is more threatened than by external goings-on.
Once we know how such goings-on can help to manipulate people, we should expect
to see them deployed. In response, we should demand control of our environments,
a control informed by the best science. One payoff might be our coming
to see many current advertising practices as a kind of mental pollution, just as
threatening to our autonomy as the kind of subliminal advertising widely feared,
and just as ripe for political control.
Neuromarketing: Ethical and Political Challenges
17
References
Baumeister, R.F., Bratslavsky, E., Muraven, M. & Tice, D.M. 1998. Egodepletion:
Is the active self a limited resource? Journal of Personality and Social
Psychology 74: 1252-1265.
Baumeister, R.F. 2002. Ego Depletion and Self-Control Failure: An Energy Model
of the Self's Executive Function. Self and Identity 1: 129-136.
Damasio, A. 2005. Human behaviour: Brain trust. Nature 435: 571-572.
Hamilton, C. & Denniss, R. 2005.Affluenza : when too much is never enough. Crows
Nest, NSW: Allen & Unwin.
Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U. & Fehr, E. 2005. Oxytocin
increases trust in humans. Nature 435: 673-676.
Levy, N. Forthcoming(a). Resisting Weakness of the Will. Philosophy and Phenomenological
Research.
Levy, N. Forthcoming(b). Neuroethics and the Extended Mind. In J. Illes and B.
Sahakian (eds) The Oxford Handbook of Neuroethics. Oxford University Press.
2007. Neuroethics: Challenges for the 21st Century. Cambridge: Cambridge University
Press.
McClure, S.M. Li, J., Tomlin, D., Cypert, K.S., Montague, L.M. & Montague. P.R.
2004. Neural Correlates of Behavioral Preference for Culturally Familiar Drinks.
Neuron 44: 379-387.
Miyake, Y., Mizutani, M., & Yamamura, T. (1993). Event-related potentials as an
indicator of detecting information in field polygraph examinations. Polygraph 22:
131-149.
Rosenfeld, J.P. Soskins,M., Bosh, G., & Ryan, A. 2004. Simple effective countermeasures
to P300-based tests of detection of concealed information. Psychophysiology
4: 205-219.
Vohs, K. D., & Heatherton, T. F. 2000. Self-regulatory failure: A resourcedepletion
approach. Psychological Science 11: 243–254.
Vohs, K.D., & Faber, R. (2002). Self-regulation and impulsive spending patterns.
In Advances in Consumer Research, Vol. 30, eds. Punam Anand Keller and Dennis
W. Rook, Provo, UT: Association for Consumer Research.
Wansink, B., Kent, R.J. & Hoch, S.J. 1998. An Anchoring and Adjustment Model
of Purchase Quantity Decisions. Journal of Marketing Research 35: 71-81.
Weisberg, D. S., Keil, F. C., Goodstein, J., Rawson, E., & Gray, J. (in press). The
seductive allure of neuroscience explanations. Journal of Cognitive Neuroscience
Neuromarketing: Ethical and Political Challenges
Neil Levy
Florey Neuroscience Institutes, Melbourne, Australia,
Oxford Centre for Neuroethics
neil.nl.levy@gmail.com
ABSTRACT
Ethicists and ordinary people are typically more worried by interventions that alter agents’
mind by directly altering their brains than interventions than are focused on the environment,
and thereby indirectly change minds. I argue that the causal route to changing minds is not itself
important. Moreover, some of the most powerful techniques whereby behavior is altered
without the consent or knowledge of agents involve environmental manipulations: manipulations
of social space, for the benefit of those in the business of increasing consumption. I argue
that insofar as we are fixated on internal interventions, we overlook ways in which our autonomy
as agents is impaired. Once we recognize the power of environmental manipulations, however,
we should come to see social space as a legitimate target for political control.
0. Introduction
When philosophers and ordinary people worry about how agents can be manipulated,
they usually focus on internal goings-on. They focus on internal goings-on
in two ways: they worry more about manipulations which affect the brains of
agents than those that affect their environment, and they look for ways to (a)
prevent such internal manipulations and (b) strengthen the internal resources of
agents. There is good reason, sometimes, to worry about internal manipulations
and there are often good reasons to help agents build internal resources. But the
most effective and practicable manipulations of behavior, including buying behavior,
are most likely to be external manipulations; and the most effective ways
of strengthening self-control are likely to involve structuring the environment to
facilitate it. In this short paper, I shall argue that focusing on internal goings on
carries risks: risks of allowing marketers free reign to manipulate just where they
are most likely to be effective, and of distracting attention from ways of resisting
marketing pressures that are low-tech but effective.
1. Internal and External Manipulations of Behavior
Let me begin by very briefly recounting two recent experiments focusing on mind
and behavior. The first experiment examined the effects of the hormone oxytocin
on people’s behavior. It was found that administering the hormone (via a nasal
Neuromarketing: Ethical and Political Challenges
11
spray) increased people’s trusting behavior, as measured by their willingness to
cooperate in a bargaining game in which cooperation carries the risk of loss (Kosfield
et al. 2005). The second study was an imaging experiment, comparing the
brains of subjects drinking Coke and Pepsi, with and without the knowledge of
which brands they were drinking. When subjects were ignorant of the brands,
only the ventromedial prefrontal cortex was active. But when the brands were
identified, subjects exhibited additional – dorsolateral prefrontal and hippocampal
– responses to Coke alone (McClure et al. 2004).
Now, it is not so much the content of these experiments that interest me here
as the responses to them. In their book Affluenza, Clive Hamilton and Richard
Denniss label the Coke experiment ‘disturbing’ (Hamilton & Denniss 2005: 41-2).
It shows, they conclude, that ‘We have not so much been brainwashed into drinking
Coca-Cola: we have had our brains rewired to want it’ (42). Now Hamilton &
Denniss are surely right in thinking that due largely to Coke’s effective advertising
and cultural saturation strategy, the company has been successful in getting
people to make the association between the drink and a host of other things which
are in fact unrelated to it. But the fMRI results do not tell us that. What does tell
us that? Well, largely people’s behavior; their verbal behavior as well as their buying
behavior. What the fMRI results tell us that is that people have memories and
visual percepts associated with Coke. Surely, though, we didn’t need neuroscience
to tell us that (suppose that people failed to experience hippocampal activation in
the experiment. What would we conclude? Well, probably that the equipment was
faulty). Now, I am not saying that we can’t learn new and surprising things from
neuroscience. I am saying that we should not be impressed with images of brains
just because they are images of brains. The internal is not privileged. If anything,
it is the other way round: we look to behavior to validate neuroscience, not the
other way round.
Worries expressed over the oxytocin experiment are better founded. Some people
have expressed the worry that oxytocin might be used to bolster trust in
speakers at political rallies (Damasio 2005). It’s easy to see how this worry might
be extended to a sales context. Trust will greatly facilitate the job of a salesperson.
But notice that salespeople already engage in activities designed to facilitate
trust. Indeed, it is likely that they engage in actions which have the effect of facilitating
trust by way of increasing oxytocin (though of course they don’t think of
what they do in this way). Here’s an example from a sales handbook: tell the customer
that the product they’re looking at is somehow defective or unsuitable for
them. Since they think that if you’re trying to sell to them no matter what, you
will tell them that the product is good, their trust in you should rise. Now lead
them to the product you wanted them to buy all along. Here’s another technique,
which suggests itself once we know that oxytocin facilitates trust: stimulate it.
For instance, have kittens around, or pictures of kittens.
NEIL LEVY
12
My claim is that these entirely environmental manipulation should be just as
disturbing to us as internal. Yet people continue to believe that neuroscience has a
special and especially intimate ability to peer into the soul, or at least its closest
secular equivalent. Consider, in this context, another experiment. Weisberg and
colleagues gave subjects, both scientists and laypeople, explanations of events and
facts that were deliberately designed to be obviously flawed. They found that
both groups of subjects could identify faulty explanations. Yet when the subjects
were given the same explanations peppered with neuroscientific jargon, their critical
abilities left them, and they took bad explanations for good. We think brain
science is especially good science, and are credulous toward it. Perhaps this is because
the science is new; I think, though, that part of the explanation is that brain
science is an internal science. Because it looks at what is going on inside, it is
taken to have far greater depth than mere psychology. Once again, I am not saying
that brain science is not good science. I think it is still in its infancy, but its results
are already very exciting. I only want to say that we ought not be dazzled by
it, or overlook the other ways we learn about, and control, human behavior.
2. The Challenge of Neuromarketing
Now, what has this got to do with marketing? Recently, there has been a great
deal of talk, and worry about, neuromarketing. Neuromarketing would apply the
methods of neuroscience to moving products. For instance, rather than ask focus
groups what they think of a product, we could place them in fMRI machines and
measure their neural responses to products. Why do that? Well, perhaps subjects
are not always honest: they may tell people what they think they want to hear,
rather than what they really believe. Or perhaps even subjects who are trying to
be honest are unable to be sure what they think of a product. They may be deaf to
the signals that intimate that their infatuation with a new drink is likely to be
short-lived (or whatever the case may be). By peering directly into brains, the
hope of some (and the fear of others) is that we could cut through this messiness.
We could focus directly on signals of pleasure in subjects, for instance, avoiding
worries about insincerity. We could know more about subjects then they know
about themselves. In that way, we could very finely attune products and marketing
campaigns to target audiences and their desires, thus making products well
nigh irresistible.
I do not think we can rule out the possibility that these methods could prove
useful to marketers. But I think there is very good reason to be skeptical, for the
moment and for the foreseeable future. First, consider dishonesty. I see no reason
to deny that subjects might not exhibit neural signals of, say, pleasure, which
they can attempt verbally to dissimulate. fMRIs might be a way of detecting such
signals. It’s a pretty expensive way, though, and others that are cheaper are
Neuromarketing: Ethical and Political Challenges
13
available. We can observe subjects behavior covertly; we can measure physical responses
(psychophysiology is probably a more reliable guide of arousal than
fMRI). In any case, we can rest assured that most subjects attempt to be honest
most of the time. Similar points apply to ambivalence. For the moment, at least,
verbal response and overt behavior is a much better indicator of how subjects are
feeling than are brain scans.
Moreover, it is far from certain that fMRI will always detect dissimulation. Independent
testing of neuroscientific technologies specifically designed for lie detection
has not been very promising (Miyake, Mizutani & Yamamura, 1993); it would
be surprising if techniques developed for other purposes were more successful.
Moreover, when subjects are motivated to ‘beat’ neural lie detection technology
they appear able to succeed (Rosenfeld et al. 2004); successful countermeasures
are surprising simple and easy to implement. Investing in neuroscience to detect
dissimulation is likely to be ineffective; in any case, it is unlikely to be necessary.
For the most, I suspect, neuromarketing will prove an expensive waste of time.
It is also, and much worse, risky – not in itself, but insofar as it diverts attention
from external ways of manipulating behavior that are more effective, and therefore
far more in need of regulation. These ways of manipulating behavior build on
the techniques that marketers have been using for many decades now, of controlling
consumption by controlling space. They work, they are scientifically validated,
and they are likely to be refined in the future. We don’t worry about them
because they focus on the environment around us, rather than on our brains. But
we ought to be more concerned about them than the internal manipulations which
might, one day, prove practicable.
First, let me mention some of the techniques that marketers use today to encourage
consumption (that is, to encourage people to buy more than they need, as
well as to shift their buying preferences). It has long been known that how goods
are positioned affects consumer behavior. Putting the most popular goods in the
middle of aisles maximizes exposure to the products – if they are placed at the
ends, people may grab them and head to the checkouts without passing the products
they did not intend to buy. Similarly, essential items are often placed at the
back of the shop to increase traffic past non-essential goods. Products may be ordered
in terms of price, with more expensive goods placed where they will be encountered
earlier. The result is that cheaper goods are perceived as good buys, in
comparison with more expensive goods. More counterintuitively, pricing in multiple
lots can encourage greater buying, and once the buying pattern is established,
it tends to be maintained (Wansink et al. 1998).
Now, marketers might tell you that all these techniques are merely designed to
expand consumer choice. By making consumers walk past goods, they are made
aware of their existence. Special offers give people choices to buy in multiple lots,
thus expanding their options. It should be admitted that these responses have
some degree of plausibility. Certainly, these techniques don’t look particularly
NEIL LEVY
14
threatening. Unlike internal manipulations, they leave the agent, with her powers
of resistance and choice, unaffected. I believe that this distinction between the
agent and her environment both philosophically suspect and practically dangerous.
Let me now tell you about some techniques currently being explored to encourage
consumption behavior which work on agential powers from the outside.
These techniques build on a well-established paradigm in social psychology called
the ego-depletion hypothesis.
The ego depletion hypothesis is, roughly, this: willpower is a limited resource,
which we draw upon to resist temptation. That is, when we judge that we ought
to act in some way, but are tempted to act in a way that conflicts with our judgment,
we draw upon our willpower to overcome temptation. All too commonly, of
course, we fail to act as we judge we ought in these kind of circumstances: we continue
to smoke when we judge that we ought to give up, we eat or drink more
than we believe we ought to, we spend more than we think we should. But sometimes,
in what seems to be exactly the same circumstances, we manage to bring
our actions into line with our judgments. Moreover, some people seem to be better
at this than others. What explains our success at getting ourselves to do what we
judge we ought, when we succeed? And what explains why some people are better
at this than others? We often answer this question with just the word ‘willpower’.
Proponents of the ego depletion hypothesis say that we are exactly right.
Willpower, they argue, is resource-like inasmuch as we use it, we use it up (that
is what the word ‘depletion’ means in ego-depletion). Of course, we don’t use it up
permanently, just until we get an opportunity to replenish it. What replenishes it
is rest. This is sometimes referred to as the muscle model of self-control. How
much we can lift at any one time depends upon the state of our muscles. Muscular
strength is used up in the short term: if I have just done 20 push-ups, I find it very
difficult to do more. But this strength returns with rest. Moreover, muscular
strength increases with us; similarly, proponents of ego depletion suggest that
willpower can increase with regular exercise (Baumeister et al. 1998; Baumeister
2002).
Let me briefly sketch the evidence for the ego-depletion hypothesis. It comes
from a large set of studies. In these studies, subjects are divided into two groups.
One group is assigned a self-control task, which involves resisting a desire; the
control group is instead given a tiring task which does not require self-control to
anything like the same extent. For instance, in a typical study the first group was
ushered into a room filled with the smell of freshly-baked cookies, which were laid
out on plates in full view. The subjects were told that they were to participate in a
study on taste perception; one group of subjects would eat the cookies and another
would eat radishes. They, they were told, were the radish group. They were then
presented with plates of radishes, and left with instructions to eat several of them.
The experimenters withdrew and covertly observed the subjects. They were observed
to sniff and even handle the cookies, but despite the fact that they thought
Neuromarketing: Ethical and Political Challenges
15
they were unobserved, all of them dutifully ate radishes and not the cookies.
Meanwhile, the control group performed a task that did not require self-control,
but which was tiring – a typical task was performing a series of three digit multiplications
on paper.
Subjects from both groups were then tested on a common task, which does require
self-control. Typical tests include squeezing a handgrip for as long as possible,
persisting at an unsolvable anagram task, or keeping one’s hand immersed in
icy water. Subjects who have previously performed a self-control task – resisting
the cookies, or watching a funny video while keeping a completely straight face –
perform worse at the common task than do subjects in the control group, inasmuch
as they persist at the task for a significantly shorter time. Moreover, their
lack of self-control carries over to more naturalistic settings. For instance, dieters
who are ego-depleted eat more after the tests than do non-dieters (Vohs &
Heatherton 2000).
What apparently happens is that the subjects in the ego-depletion group –
those that had a self-control task prior to the common task – have less willpower
to draw on than subjects who have been given a different, yet tiring, task. How
ego-depletion works, precisely, is still rather mysterious; in recent work I have argued
that the resource depleted is not in fact a discrete self-control resource, but a
domain-general capacity: one that is involved in all effortful tasks, including deliberation
(Levy forthcoming(a)). Willpower is only one capacity that degrades
when it is called upon. What matters for our purposes, however, is that willpower
is subject to depletion: self-control tasks get progressively more difficult until the
agent gives in to temptation (succumbing, I have suggested, is mediated by judgment-
shift: once self-control resources are sufficiently depleted, the agents change
their mind concerning how they ought to behave).
Ego-depletion appears to have no phenomenology at all, at least in the shortterm
(there are studies reporting greater fatigue after the second task, but that is
after the effects of ego-depletion have already manifested; that is, after the damage
has been done). So there are no warning signs that signal to us that our willpower
is low. That makes ego-depletion a phenomenon ripe for exploitation by
those who would encourage us to consume more. If you want to make someone
buy a product that they desire, but which they prefer, all things considered, not to
purchase, you ought to ensure that their self-control resources are depleted when
they confront the option of purchasing it. This can be done by requiring potential
purchasers to engage in self-control tasks, which deplete their resources, before
they are presented with the option of purchasing. This seems to be relatively easy
to ensure: simply expose them to many tempting opportunities to consume. We
can expect them to find it progressively more difficult to resist repeated temptations.
Now, it may be that marketers are already doing just this, though they
don’t – quite – conceptualise what they’re doing in these terms. Think of a shopping
mall. Already these are intentionally designed to maximize consumer’s expoNEIL
LEVY
16
sure to temptations (escalators are positioned so that rather than just riding from
one floor to the next, we are required to continually walk around, thus ensuring
we pass shops on the way; similarly popular shops may be located so that we are
forced to pass speciality shops on the way). Once the ego depletion hypothesis is
deployed by marketeers, we can expect a more precisely targeted use of the strategy.
One possibility is to ensure that consumers are exposed to low value but
tempting items – eg. chocolate bars – early, to ensure that when they are exposed
to high-value items later, their self-control resources are at a low-ebb. Sellers of
large consumer items might find it profitable to get into retailing confectionary as
a side-line; credit card companies might find arrangements with such retailers
profitable as well.
To my knowledge, no one has tested the hypothesis that repeated shopping
opportunities are ego-depleting, all by themselves. However psychologists have
tested whether ego-depletion affects the propensity to consume, as well as whether
it affects subjects’ evaluation of consumer items (Vohs and Faber 2002). Egodepleted
subjects are more willing to buy, and will pay higher prices, at least in
the laboratory. Marketers seeking ways to control our behavior are far more likely
to use this kind of technique than to resort to the use of more invasive techniques.
They can structure the environment to produce the behavior they want, and they
can do so more effectively in this manner than they can using the tools of neuroscience.
Most people are not as worried about this kind of manipulation as they
are about neuroscientific technologies. But the differences are not important.
Both manipulate behavior, and they do so for purposes that are not those of the
people who are manipulated. Both can result in changes in beliefs, and in the actions
which express those beliefs. The way in which these alterations come about
doesn’t seem to matter, from a moral, or a political, point of view. If there are
good reasons to be concerned about the first, there are equally good reasons to
care about the second as well.
These considerations lend support to a conclusion that I have defended elsewhere
(Levy 2007; forthcoming (b)): the causal route by which behavior is modified
– whether it is largely internal or external to the skull – is irrelevant to its
ethical significance. The difference in degree of concern we see with regard to neuroscientific
results, which leads to anxiety over experiments like the Coca Cola
study mentioned earlier, but scant concern over ‘merely’ environmental manipulations,
is unjustified. There is no reason to think that because we can point to internal
goings on that autonomy is more threatened than by external goings-on.
Once we know how such goings-on can help to manipulate people, we should expect
to see them deployed. In response, we should demand control of our environments,
a control informed by the best science. One payoff might be our coming
to see many current advertising practices as a kind of mental pollution, just as
threatening to our autonomy as the kind of subliminal advertising widely feared,
and just as ripe for political control.
Neuromarketing: Ethical and Political Challenges
17
References
Baumeister, R.F., Bratslavsky, E., Muraven, M. & Tice, D.M. 1998. Egodepletion:
Is the active self a limited resource? Journal of Personality and Social
Psychology 74: 1252-1265.
Baumeister, R.F. 2002. Ego Depletion and Self-Control Failure: An Energy Model
of the Self's Executive Function. Self and Identity 1: 129-136.
Damasio, A. 2005. Human behaviour: Brain trust. Nature 435: 571-572.
Hamilton, C. & Denniss, R. 2005.Affluenza : when too much is never enough. Crows
Nest, NSW: Allen & Unwin.
Kosfeld, M., Heinrichs, M., Zak, P.J., Fischbacher, U. & Fehr, E. 2005. Oxytocin
increases trust in humans. Nature 435: 673-676.
Levy, N. Forthcoming(a). Resisting Weakness of the Will. Philosophy and Phenomenological
Research.
Levy, N. Forthcoming(b). Neuroethics and the Extended Mind. In J. Illes and B.
Sahakian (eds) The Oxford Handbook of Neuroethics. Oxford University Press.
2007. Neuroethics: Challenges for the 21st Century. Cambridge: Cambridge University
Press.
McClure, S.M. Li, J., Tomlin, D., Cypert, K.S., Montague, L.M. & Montague. P.R.
2004. Neural Correlates of Behavioral Preference for Culturally Familiar Drinks.
Neuron 44: 379-387.
Miyake, Y., Mizutani, M., & Yamamura, T. (1993). Event-related potentials as an
indicator of detecting information in field polygraph examinations. Polygraph 22:
131-149.
Rosenfeld, J.P. Soskins,M., Bosh, G., & Ryan, A. 2004. Simple effective countermeasures
to P300-based tests of detection of concealed information. Psychophysiology
4: 205-219.
Vohs, K. D., & Heatherton, T. F. 2000. Self-regulatory failure: A resourcedepletion
approach. Psychological Science 11: 243–254.
Vohs, K.D., & Faber, R. (2002). Self-regulation and impulsive spending patterns.
In Advances in Consumer Research, Vol. 30, eds. Punam Anand Keller and Dennis
W. Rook, Provo, UT: Association for Consumer Research.
Wansink, B., Kent, R.J. & Hoch, S.J. 1998. An Anchoring and Adjustment Model
of Purchase Quantity Decisions. Journal of Marketing Research 35: 71-81.
Weisberg, D. S., Keil, F. C., Goodstein, J., Rawson, E., & Gray, J. (in press). The
seductive allure of neuroscience explanations. Journal of Cognitive Neuroscience
No comments:
Post a Comment