···
–
I’ve been attempting to think about how one would approach defining
an
economic agent in inclusive and
unexceptional terms. ( “General” is a
perfectly good word, but its frequent misuse has had the effect of
reversing its meaning in ordinary usage. )
Right: a fact that’s “generally true” has come to mean
“sometimes true in some cases,” as opposed to the original
meaning which is “always true, in all cases.”
Given the assumptions with which orthodox theory begins-- that
consumer
preferences are the ultimate
starting point for economic analysis, the
advertising industry is a massive anomaly.
I think it might be worthwhile to back up a step from even that
apparently basic idea. What is a consumer preference, and how can it
affect how people behave? I know, one answer would be that a preference
is merely an observed tendency to buy one good rather than another, but
we can do better than that. What is it that can lead to such a tendency?
By that question, I don’t mean to ask what circumstances can lead to a
preference (for example, what sort of advertising), but what it is inside
a person that gives rise to an observable preference. Of course the
answer I would propose is that an observed preference is determined by
reference signals and control systems. This doesn’t answer the question
of what sets reference signals, a question to which the subject of
advertising may well be pertinent. It simply assumes that reference
signals are set in certain ways, and maybe that different control systems
have different loop gains, and asks what can be predicted about economic
behavior given those system parameters.
For example, suppose a person has a reference level for eating 2000
calories of food energy per day. What will happen to the vigor or amount
of the person’s efforts to obtain food as the amount obtained increases
toward the reference level? The amount obtained can be manipulated by
changing the amount of effort needed to get each calorie. As the
“price” in units of effort decreases, what do we predict will
happen?
We predict, on the basis of the simplest possible control model, that the
amount of food obtained (and consumed) will increase, and that the amount
of effort produced to obtain it will decrease. This would be true of a
control system with constant parameters; no parameters in the control
system need to change to produce this effect.
We also predict that for each person there is some particular amount of
obtained food at which all efforts to consume more of it would cease;
that would be the observable reference level, which reflects the internal
setting of the reference signal and the calibration of the perceptual
system directly or indirectly detecting food intake. In other words, we
predict that some level of food intake will be “enough.” We can
also predict that if food were donated in that amount (and of the right
kind), the amount of effort expended would drop to zero.
It seems to me that if data do not exist against which to check these
predictions, it should be easy enough to devise experiments for obtaining
data, from humans and from animals. I quite agree with Rick that
experimental data are essential for doing PCT economics, but of course if
no data exist this doesn’t mean we have to give up. It just means we have
to do the experiments, or persuade someone to do them.
What would be the position of conventional economics on this prediction?
Would it be possible to derive a prediction from the assumptions that are
made about how much of a good a person obtains, and how much effort the
person puts out to obtain it? If conventional economics does make a
prediction, it will either be the same as the PCT prediction or different
from it; furthermore, the actual behavior of people will either be like
what PCT predicts or not like it. Whatever the truth may prove to be, it
seems to me that this is an issue that can be settled one way or the
other.
The common element in advertising
is the premise that purchasing some
product is going to solve whatever problem you have.
Premise, or promise? It seems to me that advertising fulfils two quite
different functions. One is to notify people of the availability of goods
and services at certain prices. That is the “information”
function. It could indeed be that buying a widget would solve your
problem. The other is to distort the description of goods and services to
make it seem that they will satisfy more needs than they actually do
satisfy, or satisfy the same needs but with less effort, in order to
convince people that they should buy the advertised good, or buy it
rather than a competitor’s functionally identical good. That is the
“deception” function. In that case, buying the widget will NOT
actually solve your problem, Both functions work on the perceptions
of the consumer, one to provide an accurate picture of the product, the
other to provide a carefully-designed inaccurate picture.
So there is a PCT way of dealing with advertising that might well lead to
experiments. But suppose that we find that advertising does indeed have
effects on people’s preferences. What then? Will more advertising then
cause them to buy even more? Not necessarily. If I buy a potato-peeler
because I have seen that a good one is available at a price I consider
reasonable, does this mean that a stepped-up advertising campaign will
cause me to buy a second potato-peeler? Not likely. The concept of
“enough” ought to exist in economics if it doesn’t exist
now.
Reorganization may be fundamentally random in some sense,
but the solutions adopted it seems
to me are often selected from an
environment which is decidedly not random.
Of course. We reorganize randomly when there is no a priori basis
for guessing that one action would be more effective than another in
correcting an intrinsic error. One of the things we acquire from this
process is the ability to reason systematically about the world and take
actions based on understanding the world and making wise predictions
about it. If that ability proves reliable, we no longer need to
reorganize: the systematic system corrects the pertinent errors before
reorganization can come into play, removing the reason for
reorganization.
One of the aspects of this ability is the ability to translate
communications from others into perceptions even if we have not had those
perceptions ourselves. Descriptions become vicarious perceptions. We can
act to control those vicarious perceptions in good faith, taking what we
are told as true (within reason), just as we control the speed of a car
by controlling what the speedometer cdommunicates to us rather than
trying to estimate the speed from direct experience. Eventually, we will
come to distrust a faulty speedometer. We also come to distrust faulty
communications, those that deceive rather than inform. So ultimately,
advertising is self-defeating; this is probably why it is aimed to catch
the innocent and ignorant before they wise up.
Its not that the “Marbough
Man”
is a physical stimulus to takeup smoking. Instead, in the absence
of what
appears to be a better alternative, an adolecent selects as a role
model
for maturation the symbol and associated product which the
“Marbourgh Man”
appears to supply.
I’d add that the Marlboro Man implies that if you indulge in the same
behaviors he demonstrates, you will become like him in other respects:
strength, looks, sexiness, success, independence. That is deception, of
course, but if you are inexperienced and do not know better, you will
take this perception as the truth, and since it is true, you will do what
you are shown that it takes to be like that. The heart of the Marlboro
Man image is deception. It is not true that smoking Marlboros will endow
you with the Marlboro Man’s admirable attributes. But if you can be
persuaded it is true, and if you would like very much to have those
attributes, you will smoke the smoke and choke the choke. If it
were true, it would be quite rational to consider choosing present
glory over preventing possible future diabilities. The truth, of course,
is that you are subject to the future disabilities and you don’t end up
strong, sexy and so forth, either: you lose both ways.
Economic orthodoxy is absurdly
unable to generate anything plausible to say
about advertising. In contrast behaviorism does have a plausible
story in
explaination of why advertising works. After failing to persuade
people
that Billboards can’t compell people to do anything using a basic
control
theory argument, I had to rethink things. I don’t think People
believe
orthodox economics or behaviorism because they are
stupid.
Sometimes I agree with that. Sometimes I don’t. Anyway, a billboard that
tells the truth can persuade people (if not compel them) to do
things: 'FREE BEER THIS FRIDAY" would probably bring a lot of beer
drinkers in, at least until the word got out that the cover charge is
$50. It wouldn’t bring in many people who dislike beer.
It’s not that I’m opposed to a “general theory” based upon
control theory
of economic behavior. Instead
its that economic orthodoxy generates so
many anomolies that for the time being there isn’t any shortage of stuff
to
work on. I think for the time being, my “general theory” will
be-- if you
don’t understand why they’re doing it there’s probably a control
theory
explaination that explains it.
I really think we can do better than that. Uncovering anomalies isn’t
enough to persuade people they were wrong. They just assume there’s a
mistake somewhere that will fix it. The trick, I think, lies in catching
yourself taking things for granted, You’re used to the arguments that
economists use, but you’re also used to leaving out the concepts they
habitually leave out, like “enough.” I think your approach to
advertising in going the right way, but you have to watch the premises
you’re letting slip past.
One premise is that advertising is “effective.” I think the
most effective advertising ever done is the job the agencies have done in
convincing producers to spend enormous amounts of money on advertising.
Look, your competitor is spending ten million dollars on advertising –
obviously, you need to spend eleven million dollars to win back sales (we
reassure you, of course, that the shoddiness and high price of your
product couldn’t possibly explain why you sell less of it). Just watch
television for a couple of hours and ask yourself about the net effect of
commercials on your buying habits. Of course market analysis will show a
definite possible trend toward decreasing the disadvantageous position of
your product relative to saddle soap, but in fact you have to improve
your product or lower its net cost if you want to sell more of it for
long – which you already knew.
I think you will find that essentially all data about the effectiveness
of advertising has been produced by people whose fortunes depend on
convincing other people that it is effective. At least take that premise
with a grain of salt.
Best,
Bill P.