Friday, October 17, 2014

When Pictures Are Worth A Thousand Words: An Empirical Approach

Guest Blogger

Christine Jolls

For the conference on Public Health in the Shadow of the First Amendment

Law, both within and beyond the domain of public health, is pervasively concerned with the availability of information.  A question of the first order is then what it means for an individual to “have” a given piece of information.  Consider someone who hears the following commercial:  “I admit it.  I thought of skimping.  Could you look into those big blue eyes and skimp on her?  So I buy KRAFT Singles.  Imitation slices use hardly any milk.  But KRAFT has five ounces per slice.  Five ounces.  So her little bones get calcium they need to grow.”  (The example, from a Federal Trade Commission decision, comes from an article by Professor Richard Craswell.)  Upon hearing this commercial, does the listener “have” the information that KRAFT slices contain more calcium than “imitation” slices?

One can imagine at least three ways of answering this sort of question.  The literal approach concludes, in the KRAFT case, that the listener does not “have” the information that KRAFT slices contain more calcium because the literal terms of the commercial are not logically inconsistent with KRAFT and “imitation” slices containing comparable amounts of calcium (which in fact they did; the “imitation” slices got their calcium from sources other than milk).  The legal-pragmatic approach asserts that the listener “has” the information that KRAFT slices contain more calcium because an authoritative legal decisionmaker – in this case, the Federal Trade Commission (FTC) – determined that a reasonable consumer would view the commercial as having this meaning.  Finally, the empirical approach suggests that the listener “has” the information that KRAFT slices contain more calcium to the extent that empirical evidence of the effects of KRAFT’s language, or “comparable” language in a related context, suggests that listeners often will in fact conclude from the commercial that KRAFT slices contain more calcium.  In the words of the FTC – which employs the empirical approach if it interprets a commercial to be ambiguous under its reasonable-consumer standard, as it found was not the case in In re KRAFT – relevant empirical evidence includes “reliable results from methodologically sound consumer surveys” and “generally accepted principles drawn from market research showing that consumers generally respond in a certain manner to advertisements that are presented in a particular way.”  (In re KRAFT (FTC 1991).)

The literal, legal-pragmatic, and empirical approaches to “informedness” are no less likely (and perhaps are more likely) to produce different conclusions when a communication combines words with graphics, as in the below example from Ontario of a sign warning about ticks carrying Lyme Disease:

In many areas of public health today, there is great interest in warnings that incorporate graphic elements of various types.

The literal approach in such cases may face difficulty in determining what a “literal” interpretation of a particular graphic element is.  In a prominent legal challenge to the Food and Drug Administration’s proposed tobacco warnings, for instance, the United States District Court for the District of Columbia displayed a form of literalism in responding to these warnings.  Speaking of the communication below, the court complained that the communication’s inclusion of autopsy staples on a deceased man’s chest suggested – notwithstanding the textual message “WARNING: Smoking can kill you” – that “smoking leads to autopsies” (p. 273 of the court’s opinion), and yet, the court remonstrated, “the government provides no support to show that autopsies are a common consequence of smoking.”  (R.J. Reynolds Tobacco Co. v. Food and Drug Administration (D.D.C. 2012).)

Neither the legal-pragmatic nor the empirical approach would succumb to this sort of interpretive strategy, but in the case of the former, pictorial material would tend to be downplayed, while text would often exert an outsize influence.  As Rebecca Tushnet has remarked in connection with the blank spaces Google Books substituted for books’ printed images, the blank spaces provide “a perfect if unintentional demonstration of how copyright, like much of law, thinks about images, which is to say it doesn’t think much about them at all, privileging the text….”  (Harvard Law Review, 2012.)

With respect to the empirical approach, the effect of graphic material would remain a firmly data-driven question.  Consider, for instance, another pictorial health warning (below) challenged in the tobacco litigation.  Results from a large-scale experimental study compared the likelihood of factually inaccurate perceptions of the risk of smoking among respondents who viewed this warning to the likelihood of factually inaccurate perceptions among respondents who viewed a warning containing only the textual component (“WARNING: Tobacco smoke can harm your children”).  Respondents’ level of agreement with the factually inaccurate statement “If I have smoked a pack of cigarettes a day for more than 20 years, there is little health benefit to me quitting smoking” was significantly lower among those who viewed the graphic-and-text warning than among those who viewed the text-only warning.  (Experimental Study of Graphic Cigarette Warning Labels: Final Results Report, Prepared for Center for Tobacco Products, Food and Drug Administration, available at!documentDetail;D=FDA-2010-N-0568-0008; Christine Jolls, 30th International Seminar on the New Institutional Economics-Behavioral Theory of Institutions, Journal of Theoretical and Institutional Economics, 2013.)

The empirical approach thus suggests that “informedness” in the tobacco context is higher with pictorial warnings than with text-only counterparts to such warnings.  This and similar empirical findings on the factual accuracy of people’s risk perceptions under alternate communication forms have an important yet under-recognized role to play in First Amendment litigation over such communication forms.

Christine Jolls is Gordon Bradford Tweedy Professor of Law and Organization at Yale Law School. You can reach her by e-mail at christine.jolls at


Older Posts
Newer Posts