New Study: TV Food Ads Provoke Automatic Eating in Adults as Well as Children

One reason the public needs to know about limits to their free will

Posted Jul 21, 2009

In a new study published this month in the journal Health Psychology, TV food ads were found to significantly increase eating while viewing, in adults as well as children.  To go directly to a description of that study and a discussion of its implications for the obesity crisis, just jump down to the photo of the hamburger ad.

The following is another installment in an ongoing Psychology Today blog debate with Roy Baumeister concerning the existence of free will, for which the new study on automatic effects of TV ads is highly relevant.

In his reply to my blog entry The Will Is Caused, Not Free, Roy Baumeister echoed recent statements by Kathleen Vohs and Jonathan Schooler that "scientists should not tell the general public that determinism is a proven fact, because to do so may have socially undesirable effects."  Baumeister argued further that "tell[ing] the public that free will is an illusion... [is an] irresponsible misrepresentation of opinion as fact [which] has additional and damaging consequences, insofar as the unfounded opinion they advocate will promote antisocial behavior."

As I pointed out in Part 1 of my reply to Baumeister, it is not at all clear at present what, if any, consequences there might be in communicating to the general public scientific results that suggest limits to the scope of their free will.  Initial studies by Vohs and Schooler, and by Baumeister himself, have found increases in cheating and aggression in experimental participants, after they were told by the experimenters that science had proven that free will was an illusion.  In my prior post I noted that it is quite premature to draw strong conclusions from these few initial studies, as Baumeister and Schooler have, about any such negative social consequences, for several reasons: First, these initial studies only examined antisocial forms of behavior, and so only negative and no positive consequences could have been observed.  Secondly, I noted that we already have the historical precedent of evolutionary science and religion, in which substantial amounts of evidence in favor of the principle of evolution over the past 150 years has not shaken people's belief in supernatural causes such as the act of Creation -- so we probably don't need to be so afraid of informing the public about studies indicating their lack of free will.

Now, Baumeister has been careful in his recent posts to distinguish determinism from causality, and he and I seem to be converging, via this battle of the blogs, on the notion of relative freedom.  I see the body of social psychological research on unconscious or automatic causation of social judgment (e.g., stereotyping, forming impressions of others), social behavior, and social goal pursuits (such as cooperation, competition, achievement, and affiliation) as showing that the glass of free will is mostly empty, and Roy sees it as mostly full -- or at least fuller than I do.  But I cannot and do not conclude on the basis of that evidence alone that free will does not exist, just as Roy cannot and does not conclude from that evidence that the will is entirely free. 


Given this, I'm not sure what Baumeister, Schooler, and Vohs' position is concerning informing the public about experimental evidence about limitations in the scope of free will, or of situations and domains in which people believe they are exerting free will when in actuality they are not.  This is not the same as telling the public that determinism is an established scientific fact, or that free will is an illusion (in an absolute sense).  Rather, it consists of reporting to the public evidence of external and internal causes of their choices and behavior of which they are not aware and did not consciously intend.  In that spirit I want to alert you to a new study, just published this month in the journal Health Psychology, that illustrates perhaps the most important reason why it is irresponsible for us as scientists notto tell the general public about evidence of limits to the free scope of their will.

As Baumeister emphasized in his Tampa SPSP debate presentation, the belief in free will (as distinguished from free will per se) serves important motivational and subjective functions for the individual.  But it also can cause problems when it leads us to ignore or dismiss the possibility that there may be powerful influences on our behavior that we don't know about -- even for important behaviors such as who and what we vote for, or the kinds and amount of food we eat.  If we believe that we are the absolute captain of our soul, then we don't worry too much about these potential influences -- and thereby leave ourselves wide open and vulnerable to them. (See also the recent commentary on this point by Tom Clark of the Center for Naturalism.)

Television and other forms of advertising is expressly directed at getting us to do something that is in the best interests of the advertiser, but not necessarily our own.  We have already recognized this in the case of cigarette (tobacco smoking) advertising and as a consequence it has been banned now for many years.  In the new study, Jennifer Harris and Kelly Brownell of the Rudd Center for Food Policy and Obesity at Yale and I showed that passive exposure to food advertising on television may contribute to the ongoing obesity epidemic by automatically triggering eating behavior, right then and there while watching TV.  Experiment 1 focused on elementary school children because the Federal Trade Commission has reported that they see an average of 15 TV food ads per day and that fully 98% of these ads promote products high in fat and sugar.  We simulated the natural television viewing situation for our young participants by having them watch a 5 minute cartoon that contained a few 30 second food ads -- or, in the control condition, non-food ads.  While they watched the cartoon a bowl of goldfish crackers was made available to them.  As we had suspected, those children exposed to food ads during the cartoon ate significantly more of the snack food than did the children in the control condition.  Unexpected, at least to me, was the size of this effect: children consumed 45% more of the snack food when exposed to food advertising.

Now, we all know that children are not as able as adults to defend themselves against ads for toys, cereals, clothes, DVDs, etc., so perhaps this finding is not that surprising.  It certainly suggests that there is a direct and automatic effect of food ads on consumption behavior in children right then and there while they are watching television, not only on their preferences for certain brands or products for their parents to buy them.  And the sheer size of the effect strongly suggests that this automatic effect on consumption is indeed a contributor to the public health problem of obesity in children.

However, nearly all of the social psychological research on automatic and unconscious causes of human judgment and behavior over the years has been conducted on college-age or older adults, suggesting that adult television viewers might be just as vulnerable to the deleterious effects of food advertising as are children.  In our Study 2, we showed a group of adults a short television documentary that incidentally included either snack-food ads, nutritious-food ads, or no food ads.  After the program, they took part in what they thought was a separate study in which they taste-tested a range of healthy (e.g., fruits) and unhealthy snack foods.  We found that the adults who had been exposed to the snack food ads ate more of all types of food during the taste test compared to the other conditions.  Thus the automatic effect of snack food ads to increase the amount eaten while watching television holds for adult as well as child viewers, suggesting that TV snack and fast-food ads are a contributor to adult obesity as well.

In neither study was the amount eaten related to the participants' reported levels of hunger, and in careful questioning after the experiment was over, no one showed any awareness or appreciation that the amount of food they ate while watching the show was influenced by the ads they saw.  (In studies such as these, participants typically strongly resist such suggestions.)  These are unconscious effects, and so by definition one is not aware of them while they are happening. Because people do not experience these influences on their behavior, they have no chance at correcting or controlling them.  The only way then for the general public to know that there are interested agents out there (e.g., advertisers, government) exerting control over their behavior in these ways is for us as scientists to do these kind of studies and to report the findings publicly and as widely as possible. 

Bottom line:  we should be telling the public the truth about limits to their free will as revealed in experimental studies, and not decide for them what is good or bad for them to know. We may think we are doing them a favor by permitting them a positive illusion, but there are also negative consequences to such naivete -- namely, leaving oneself wide open to being controlled by others who are not so naive.

John Bargh and ACME Lab at Yale University conduct research on the unconscious causes of our preferences, motivations, and social behavior.  ACME research publications are available here