Beyond Nudging: Debiasing Consumers Through Mixed Framing.

• Author: Godi, Matteo

NOTE CONTENTS INTRODUCTION 2036 I. ATTRIBUTE FRAMING 2040 A. The Behavioral Economics of Attribute Framing 2041 B. Mixed Framing: An Understudied Phenomenon 2045 II. MIXED FRAMING: A CASE STUDY 2050 A. The History of Agency Regulation of Percent 2052 Fat-Free Statements B. An Original Behavioral Survey: Attribute Framing and Percentage Statements 2055 III. DEBIASING THROUGH MIXED FRAMES: A NEW APPROACH TO REGULATING FOOD PRODUCTS 2062 A. A Policy Proposal: Mixed Frames as the Solution 2062 B. Shortcomings in the Current Food-Safety Regulatory 2064 Scheme 1. Misleading 2065 2. Technical Terms 2067 3. Bold Fonts 2068 C. Making Sense of the Goals of Consumer Regulation 2069 IV. FIRST AMENDMENT AND PATERNALISM 2071 CONCERNS A. First Amendment Challenges to Debiasing Through 2071 Mixed Framing B. What Mixed Framing Is Not: Paternalism and Nudging 2079 1. Traditional Paternalism 2080 2. Libertarian Paternalism and Nudging 2082 CONCLUSION 2084 I know... an economist... [who] would have said... that... a discount and a surcharge are the same thing economically... But we live in a world that not everybody is an economist, and many people think it's quite a different thing. --Justice Stephen Breyer (1) INTRODUCTION

Behavioral law and economics recently went to court. In Expressions Hair Design v. Schneiderman, New York merchants challenged a state law permitting discounts for cash purchases but banning equivalent surcharges for credit-card payments. (2) Both sides appealed to insights about consumer behavior. On the one hand, believing that individuals are affected by bounded rationality, (3) behavioral-economics scholars criticized the New York statute as an attempt to "bias[] consumers toward credit-card use." (4) On the other hand, law-and-economics scholars--endorsing the assumption that consumers are fully rational and act to maximize expected utility (5)--dismissed the impact of the two different pricing schemes. (6) Both sides understandably agreed that there was no quantitative difference between equivalent cash discounts and credit-card surcharges. But they disagreed on whether consumers' responses to the negative frame--that is, the credit-card surcharge, which primes the additional costs of credit-card use--differed from their reactions to the positive frame--the cash discount. In short, they disagreed on the framing effect. (7)

Ultimately, the U.S. Supreme Court tabled the debate. The Court implicitly accepted that merchants might have some reasons to prefer one kind of pricing scheme over the other. (8) The Court, however, did not address how the regulatory state ought to interact with behavioral economics. Rather, it remanded the case to the Second Circuit for a decision on whether the no-surcharge law is permissible, either as a regulation of commercial speech or as a disclosure requirement, or is instead an unconstitutional curtailment of First Amendment rights. (9)

Outside of the courts, in the consumer-protection literature, the battle lines have been similarly drawn. On the one hand, behavioral-economics scholars argue that, because there is no neutral way of conveying information, the government should improve people's welfare by influencing their decision-making through paternalistically framed regulatory disclosures. (10) Rational-choice theorists, on the other hand, object to most administrative regulations as attempts to interfere with rational people's autonomy to make decisions about their own utility" Scholars view this disagreement as a zero-sum game. (12) And entire books have tried to settle the debate, in vain. (13)

This Note argues that, in many settings, the dichotomy is overstated. Administrative regulations can take a different approach, one that is neither paternalistic nor libertarian. That is, they can offer dual, or mixed, frames. Mixed framing juxtaposes both positive and negative frames, providing seemingly redundant yet neutral information through mathematically equivalent frames. In other words, mixed framing is an improved, more complete disclosure requirement. The recent debate in the Expressions Hair Design litigation exemplifies this approach. In that case, a mixed frame would amount to something like this: "$10.00 (cash) / $11.00 (credit-card)," "$10.00 (cash) / 10% extra for credit card," or "$10.00 (cash) / $1 extra for credit card." The legality of one or more of these mixed-framing pricing schemes was the exact question that, on remand from the U.S. Supreme Court, the Second Circuit certified to the New York Court of Appeals. (14)

But courts are not the only ones who have yet to grapple with mixed framing. In fact, virtually no one has studied the concept. (15) And consumer-protection scholars have yet to fully acknowledge its regulatory potential--let alone notice that our administrative state currently relies on this tool. Surprisingly, indeed, we encounter this framing tool every day in our grocery stores. (16)

Mixed framing deserves more attention than it has received. Offering the first systematic empirical analysis of the impact of mixed framing on consumer products, (17) this Note recognizes that there is something intrinsically valuable about mixed frames. That "something" comes to light when focusing on the most elegant and mathematically clean example of mixed frames: "75% fat-free / 25% fat" food labels. Mixed frames can communicate more complete information without nudging consumers in a particular direction, thus appealing to the antipaternalist sentiment in favor of preserving consumer autonomy. And, in a proconsumer fashion, they can reduce the volatile and misleading effects of single frames. This Note's empirical analysis shows that mixed framing can have a discernible impact on consumer behavior and that percentage labeling of food products is an excellent area to introduce mixed-framing requirements for the disclosure of information.

By focusing on mixed framing, this Note takes a step back from the paternalism/libertarianism debate. In many instances, (18) more paternalistic approaches resulting from the application of behavioral insights to law and economics might be normatively appealing. (19) In others, however, they might be overkill. This Note is more agnostic about the identity of consumers' preferences or the way in which behavioral biases operate. To regulate efficiently and effectively, agencies often don't have to paternalistically settle those debates and burn political capital. Instead, in many circumstances, agencies should acknowledge the limits of human cognition isolated by behavioral economists, while adopting an approach that stays clear of both paternalism and libertarianism.

The concept of debiasing through mixed framing, this Note argues, is that novel approach. (20) In a nutshell, when regulating how manufacturers voluntarily elect to convey information to consumers through a percentage statement, the government should both facilitate the provision of complete, neutral information and maximize consumer autonomy by requiring the disclosure of the complementary frame--thus creating a mixed frame.

The Note proceeds as follows. Part I outlines the behavioral-economics theories of framing that will inform the rest of my analysis. Part II discusses the regulation of food labeling as a case study for this Note's novel approach to disclosure requirements. Through the first systematic attempt at uncovering the history of food regulation in the context of percentage statements, I focus on a rather unique phenomenon--mixed frames--that can be observed in the regulatory toolbox of one administrative agency. Because mixed frames have been understudied by behavioral economists and overlooked by legal scholars, I analyze the effects of mixed frames through an original between-subjects study. Relying on those empirical results, Part III maps out the analytic and normative case for why regulators, in cases like food labeling (where a single method of framing information must be chosen), should mandate the use of mixed frames. Part IV argues that, through a process of debiasing through mixed framing, agencies can promulgate disclosure requirements that minimize the risk of misleading and deceptive advertising tactics without running afoul of the First Amendment or falling into paternalistic restrictions on consumers' freedom. A brief conclusion follows, setting out a schematic agenda for further behavioral law-and-economics research on mixed framing.

1. ATTRIBUTE FRAMING

   The goal of this Part is to set the stage for a simple claim that should guide our approach to regulation: framing affects rational people's behavior. But not all frames are created equal, and some frames are more neutral than others. (21) In the face of the rational volatility of different frames, this Note will show that--both empirically and normatively--mixed framing emerges as the superior approach to administrative regulation of attribute disclosures.

   First, I survey the behavioral scholarship on attribute frames. Section LA simply makes the descriptive argument that the impact of frames on cognition cannot be ignored. It deliberately avoids delving into the potential biases that may lead to disparate behavior in response to positive, negative, or mixed frames. And it does not analyze whether framing effects obfuscate or promote one's true preferences. Those considerations are beyond the scope of this Note.

   Second, Section LB introduces the concept of mixed framing and discusses the few studies that have included a mixed-framing condition, mostly limited to the medical literature. Part II will update those studies and expand on their conclusions through the analysis of original empirical data on consumer preferences.

   1. The Behavioral Economics of Attribute Framing

      People do not always make choices solely based on their economic consequences. (22) Studies have repeatedly shown how individuals, depending on the way they process the information available to...