September 1, 2009

The missing costs of advertising

Posted in Rationality and consumer choice at 3:07 pm by Eamon Aghdasi

CAUTION: Viewing Newport ads may be hazardous to your perception of reality

CAUTION: Viewing Newport ads may be hazardous to your perception of reality

For a while now I’ve been wondering about whether or not certain types of advertising distort consumer choices and welfare. We all know that a lot of advertising sometimes bends the truth and attracts potential buyers indirectly (for example, linking the desire for beautiful women to the desire for beer). But is this actually bad? A lot of us jump to conclude that the answer is yes, based simply on what we see in the news or in our own neighborhoods. The sub-prime mortgage crisis, which among other things succeeded in exposing apparently dishonest practices in lending, certainly added fuel to this fire.

Advertising’s power to sway consumer choices is nothing new. If you’re interested in this sort of thing, I strongly recommend a documentary called The Century of the Self, which chillingly chronicles the nearly century-old history of persuasive advertising and marketing. At its center is the figure of Edward Bernays, a nephew of Sigmund Freud and one of the fathers of public relations. Bernays believed strongly that the actions of the masses could be controlled for economic or political purposes, through the clever use of advertising messages that tapped into normal human desires and fears. In Bernays’s most famous intervention in 1929, he arranged for a handful of attractive young women to march side-by-side in New York’s Easter Parade and, simultaneously, light up what they called “torches of freedom” as a demonstration of women’s rights. Amid extensive newspaper coverage and the resulting dialogue it generated, the simple plan helped turn smoking among women from an unfeminine habit into a symbol of equal rights and female independence.

The skeptical economics student and devout believer in the rationality of human beings may still not see the cause for alarm. After all, aren’t human beings rational actors capable of maximizing their own welfare via their own choices? No matter what advertising messages the individual is exposed to, he or she sensibly processes all the information and acts in a way that maximizes welfare. In this world, advertising is simply the provision of information and a useful tool for signaling, and the only downside is the apparent waste that a “good” product has to go through to prove itself and create a separating equilibrium from the rest.

Unfortunately, the reality is that economics really doesn’t make this claim about human nature, and the concept of “rationality” is more a mathematical arrangement that allows the platform of microeconomics to work in any useful way. In fact, economists (along with psychologists and others) have created an entire branch of their own social science — behavioral economics — to find the places where human beings tend not to behave rationally, as classical economics assumes for the sake of creating models.

Is there any basis in economics to question the potentially detrimental effects of advertising, then? I think there is. Economists have asked this question as far back as the 1920s, when the economist Dorothea Braithwaite wrote alarmingly about the new persuasive techniques that had started to be used. More recently, Bertrand et al (2005) used randomized trials to show that psychological “cues” can make us more likely to accept loan offers. One of the more eyebrow-raising findings of this paper is that among South African male recipients of experimental loan offers, replacing the photo of a man with that of a woman on the promotional letter was equivalent to lowering the interest rate by 4.5 percentage points, in terms of likelihood of acceptance.

Edward Bernays, father of public relations and great crumudgeon of the 20th century

Edward Bernays, father of public relations and great crumudgeon of the 20th century

Simultaneously, there’s a lot of behavioral research recently that suggests that when people make decisions, there are multiple sets of preferences or tastes at play, depending on the mood or psychological state of the individual. This research includes McClure (2004), Bernheim and Rangel (2004), and Ariely (2006), among others. This research suggests especially that when you bring an individual into the “heat of the moment” through some outside stimulus, that person’s choices could be radically different than had that stimulus never been present.

I feel strongly that this reality elucidated by behavioral economics — that individuals don’t in fact have one set of preferences, and that in fact they can temporarily switch to a different set of preferences — has huge implications for how we see the value of advertising as a concept. If everyone is rational and has a single set of preferences, life is neat and easy; whatever the individual chooses maximizes his or welfare, given his or her budget constraints. But if there are two sets of preferences, and we’re still defining welfare through the original set of preferences, then the thing that brings the individual into the second set and causes him or her to make different choices is dangerous, because it no longer maximizes welfare. In other words, the thing that switches you to “heat of the moment” decision-making can make you worse off.

How might this work in a model? Imagine there are two goods, A and B, and these goods’ true values are XA and XB, respectively. Let’s say that good A can be advertised but good B can not (like many things in life which are not owned privately, for instance). Imagine also that the effect of advertising is that in the “heat of the moment”, a good’s apparent value is raised by the quantity s. A single consumer must choose between the two goods.

It turns out that there are three scenarios that arise:

Scenario 1: XA ≥ XB

In this case, the value of the advertised good A is higher, and the consumer chooses that good. The option of advertising makes no difference in the choice of the good or the value gained by the consumer (XA).

Scenario 2: XA + s ≤ XB

In this scenario, the value of good B is so much greater than that of A that even if the owner of good A advertises, he can not convince the consumer that A is of higher value than B. He thus chooses not to advertise (and saves his money), and once again the option of advertising in this case makes no difference towards what good has been chosen or to the value gained by the consumer (XB).

Scenario 3: XA ≤ XB ≤ XA + s

Here is where things get interesting. Good B has a true value that is greater than good A, but the owner of A can make it appear that A is more valuable through advertising. He thus chooses to advertise, and the consumer picks good A over B, even though the real value of A is lower. Most importantly, the value gained by the consumer is lower because of the existence of advertising, and its effect of causing him or her to pick a less valuable option.

The real critical question here, once again, is whether or not advertising really does work this way. Can it pull us away from “rational” thinking, where true welfare is defined, into a non-rational emotional state where advertising rules? And even if that is true, is it necessarily the case that welfare should only be defined by the first set of preferences, and not the second?

I would love to see this tested with an experiment, and I actually don’t think it would be too difficult. You could simply present individuals with the option of buying one of two different goods. In the control group, you’d just put the goods in front of the individual and let him or her decide. In the second group, you could expose the individual to some sort of advertising for good A before the decision.

If advertising works, you’d see a statistically significant difference in the proportions of individuals in the two groups choosing either good (the proportion choosing A in the variable group should be higher). But all this says is that advertising has an effect, which we knew already and doesn’t really prove anything. The more revealing step would be to contact individuals some time later to ask if they’d like to change their decisions, and switch the good they chose for the good they turned down. If the proportion of those who chose good A and would like to switch to good B is significantly higher for the variable group, this might reveal that in fact some of the s, the added appeal from advertising, has worn off as the individual came back to reality. In plain words, it might mean that amid advertising, more people end up unhappy with their decisions when they’re finally allowed to reflect clearly on what they chose.

What are your thoughts? Is the underlying premise and theory sensible? What about the experiment? Share your thoughts…

Scenario 1: XA ≥ XB


  1. DRDR said,

    I think these are certainly interesting questions, though I’m not too familiar with the advertising/marketing academic literature. My advisor actually wrote the chapter on the economics of advertising for the latest Handbook of Industrial Organization, you should take a look here, though I know him more for his work on international trade.

    I recall some of the behavioral literature has shown that people will generally have different utility when they’re actually experiencing consumption vs. their memory of the experience (and even use a different set of utilities when making choices!!), whether advertising is involved or not. It seems tricky to disentangle what welfare is exactly and what’s optimal in this context — which type of utility is more important? I believe Ariely in his book mentioned that people subjected to Coke advertising actually get greater utility from drinking Coke (based on brain scans) when they know they’re drinking it. If this advertising only increases one of those two types of utility I mentioned, does that necessarily mean that the advertising is bad? I guess it depends on how you weight the types of utility I mentioned.

  2. Ryan Nichols said,

    I’m not really sure about your “more revealing step”. I am no marketing expert, but at that point I think it’s safe to say these are the few possible outcomes:

    1) The user is happy with A, and now through habit and familiarity will elect to continue using the product

    2) The user is unhappy with A, perhaps having discontinued its use, and will gladly try another similar product that may offer greater satisfaction

    3) The user may be either happy or unhappy with A, but given exposure to some potentially differentiating characteristics (read: advertising) of B, is willing to make the exchange

    The flaw in the theory lies in the determination of the value of Xa and Xb. If the value of each is easily determined, then advertisement is irrelevant. If I put two stacks of money in front of you, one of $1 bills and one of $100 bills, even if I insist that $1 bills are the sexiest thing on the planet and will offer you countless hours of pleasure at your favorite gentleman’s club, I strongly suspect you will still take the stack of $100 bills.

    At the other end of the spectrum, we assume that goods A and B are pure substitutes. In this case, advertising expenditures can certainly make a difference — but as you point out, we already know this.

    And so once again the Marketing discipline foils the efforts of Economists (behavioral and otherwise) to explain it. The answers lie in differentiation, personal preferences, and cultural influences — and explain why tenured Marketing professors have not yet had their positions usurped by their Economics counterparts.

  3. Kareen El Beyrouty said,

    Hi Eamon, great post! I was thinking about two things: (1) the brain is plastic and constantly changing and (2) maybe preferences are never absolute and always relative.

    The first idea is a basic tenet of neuroscience. It is very possible that advertising provides us with new information and thus helps us to make decisions. Sometimes I find that I decide NOT to consume something if I find the advertising really distasteful (yes, advertising can swing the decision both ways!).

    The second idea basically says that putting a photo of a pretty woman in a bar ad gives viewers a standard with which to compare the taste of the beer. You can’t TASTE the beer, so you approximate it based on what you SEE.

  4. anisa said,

    hey! i didnt know you had a blog! this is a very interesting post.
    im personally interested in looking at these dynamics beyond a consumption event – when we displace other value systems informing decision-making with a materialistic impulse, thereby driving people to prioritize consumption (and accumulation) of material wealth over other arenas of activity. if you ever come across literature about “the thing that switches you to “heat of the moment” decision-making can make you worse off”, let me know! im all ears.

Leave a Reply to Ryan Nichols Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: