Controlled Experiments from Outside Economics Show
Direction of Causality
While we may sometimes be able to infer direction of causality by studying
economic history, we need also to recognize that controlled experiments outside
of economics have shown narratives’ effects on human behavior.
In the field of marketing, Jennifer Edson Escalas notes, self-referencing
occurs when the viewer of an advertisement relates a product to his or her
personal experiences. But not all self-referencing is equally effective in changing
buyer behavior. Using controlled experiments, Escalas has compared analytical
self-referencing (an explanation of why you need the product) to narrative self-
referencing and narrative transportation (which presents a story that causes an
individual to imagine himself or herself to be another person, using the word I
rather than you). Escalas found that the narrative transportation is more effective,
especially when the analytical case for the product is weak.8
In journalism, Marcel Machill and his coauthors, noting evidence that viewers
of television news retain little of the news they hear, presented an actual TV
news report on the dangers of air pollution to a control group. They also
presented a variation of the report to the experimental group in the form of a
story with a protagonist, a baker with health problems caused by air pollution, in
an unfair struggle against antagonists who benefited from the polluting activities.
The experimental presentation of the news was retained better.9
In education, Scott W. McQuiggan and his coauthors have found motivational
benefits of narrative-centered learning. Each eighth-grade student in the
experimental group played a virtual-reality computer game in the role of a young
Alyx, whose father, in the fictitious story, is the head of a team of research
scientists on Crystal Island. A mysterious grave disease has afflicted some of the
scientists, including Alyx’s father. Alyx is determined to find out why. Playing
involves interacting in dialogues with other simulated people. In the process, the
student learns about microbiology, about bacteria, viruses, fungi, and parasites.
The study documents an advantage in learning relative to the control group with
regard to “self-efficacy, presence, interest, and perception of control.”10
In health interventions, Michael D. Slater and his coauthors studied how to
persuade people to eat more fruits and vegetables. They concluded from
experiments that didactic presentations of evidence on nutrition were not

effective. Audience response was stronger to narrative messages when the
audience identified with persons portrayed in the message. In health
interventions, these results underscore the need for carefully pretesting the story
and choosing the right persons to convey the message.11
In philanthropy, Keith Weber and his coauthors (2006) asked subjects to read
a message involving organ donation before asking them to sign an organ donor
card. The content of the message (narrative versus statistics) was manipulated.
Results indicated that narrative messages were more effective than statistical
messages.
In law, Brad E. Bell and Elizabeth F. Loftus (1985) conducted a controlled
experiment in which subjects took on the role of jury members. The goal was to
determine the jury members’ response to vivid prosecutions and nonvivid
prosecutions. For example, the vivid prosecution included the irrelevant line that
the accused, at the time of the crime, accidentally “knocked over a bowl of
guacamole dip onto the white shag carpet.” That irrelevant but vivid mental
image helped obtain a conviction from the experimental jury.
In sum: economics can learn from other social sciences, including psychology
(especially social psychology), sociology, anthropology (especially cultural or
historical anthropology), and history (especially cultural and intellectual history
or histoire des mentalités). Because controlled experiments about whole
economies are not readily available to economists, it is all the more important
that we specify and understand the building blocks of economic narratives.
Stories are one key building block.