This Is How We Craft Messaging to Appeal to Different Political Audiences
Summary of Our Journey to Validating Existing Messaging Models
As you know, for the past 2 years we’ve been investigating ways to reach young people with the ideas of liberty. Rather than starting by reinventing the wheel, we chose two promising political messaging models to test and attempt to validate: the SCARF model by David Rock and the 3-Axis Model (3AM) model by Arnold Kling.
In the end, we’ve validated Arnold Kling’s 3-Axis Model and seen first-hand how our engagement with various political groups soars when we use its messaging techniques to engage audiences unlike us.
Abandoning SCARF Measurement
Before we get into the model we ultimately validated, let’s talk more about the SCARF model that we abandoned.
The underlying concept of SCARF is essentially an observation by psychologist David Rock that people are (typically) motivated by the desire to improve five specific aspects of their lives: Status, Certainty, Autonomy, Relatedness, or Fairness. For example, an individual might work to achieve a promotion or a better job in order to increase their level of social status, or they may check the weather in the morning to have more certainty and make better day-to-day decisions, or perhaps they choose to join a youth group in order to find more friends and improve their sense of relatedness to others. Rock suggests that by better understanding this behavioral model, leaders can more effectively motivate the people they work with and create more functional teams.
While this model is a useful tool when thinking about management theory and can even be helpful as a way to think about character motivations in narrative writing, we found it to be very difficult to implement consistently in message testing scripts for several reasons.
The Problem with SCARF
The most immediate problem is that none of the SCARF motivations are mutually exclusive, nor are they objectively defined, and categorizing them in a way that would result in valid data is next to impossible.
People are frequently motivated by several goals at once (ie. an applicant might seek a promotion in order to increase status, autonomy, and certainty). Additionally, if someone is motivated by multiple SCARF goals, they may not have equal weight, significance, or meaning, so tagging these traits accurately or in a way that is genuinely quantifiable presents a significant challenge both to the media creator and the data scientist.
SCARF as a Limitation of the Message Testing Medium
Another issue with incorporating SCARF successfully comes as a limitation of the message testing medium itself.
Our testing strategy is to create fictional narrative stories, produced as short videos and show them to targeted audiences while measuring their response. These stories should (1) be centered around an individual protagonist and (2) highlight a specific relatable problem they have, which (3) connects to a FEE lesson or value we want to teach.
Each video must also be under 5 minutes in length (preferably closer to 3) and feature specific language and framing devices that we believe will be most effective with either Libertarians, Progressives, or Conservatives based on Arnold Kling’s 3-Axis Model of political languages.
As a writer and producer, simply meeting those objectives effectively is already a significant burden and in order to have any hope at success we also need to do our best to tell compelling and engaging stories.
Among other things, this means adhering to best practices in narrative storytelling, and one standard principle of good narrative writing is never to have a character directly state or exposit their internal emotional state. As the adage goes, “Show, don’t tell”.
Unfortunately, this only makes including usefully clear SCARF-model motivations more difficult.
We can’t simply have a protagonist announce that they are motivated by status or relatedness because doing so degrades the quality of storytelling to a point where our videos would lose verisimilitude and be rejected as poor quality narratives by our intended audiences, and there’s not enough time in 3–5 minutes to present a set of clearly defined SCARF motivations in addition to all of the other requirements we have to meet.
In the end, measuring SCARF in addition to specific messages, topics, and 3AM framing proved not to be possible given our current testing strategy, though we may revisit emotional motivation triggers at a later time.
How We Write for the 3AM
Writing to suit specific audiences is always, first and foremost, an exercise in empathy. It requires seriously attempting to understand the way an audience thinks and feels, and then producing content with that understanding in mind.
Arnold Kling’s 3-Axis Model helps us do this more effectively by providing effective heuristics for how Progressives, Conservatives, and Libertarians think. According to Kling, each of these three groups has a different narrative lens through which they view the world. Each narrative is defined by a particular ideological axis.
“A progressive will communicate along the oppressor-oppressed axis, framing issues in terms of the (P) dichotomy. A conservative will communicate along the civilization-barbarism axis, framing issues in terms of the © dichotomy. A libertarian will communicate along the liberty-coercion axis, framing issues in terms of the (L) dichotomy.”
For the purposes of our message tests, what this means is that as we draft content targeted to each of these audiences, we have to adapt our stories in order to conform to their unique cultural/temperamental languages.
Of course, first we need to come up with an effective story concept.
How We Devise Effective Story Concepts
Recall that each story should have a:
- Compelling protagonist, facing a…
- Relatable problem, which…
- Connects to a FEE value.
In order to accomplish this task, in a sense we have to work backwards.
Our pre-existing audience research (Market Lab, 2017) told us that the top issues for our target demographic (14–26 year old Americans) were career uncertainty, inequality, health care, the cost of higher education, and environmental protection. This informs our understanding of what would constitute a “relatable problem”. Simultaneously, FEE’s mission is to educate people on the economic, ethical, and legal principles of a free society, and we’ve identified five core benefits that we want to explore in our message tests: justice, community, prosperity, personal fulfillment, and problem-solving. These are the values we need to connect those problems to.
From there, we generate what we might call a story archetype, or perhaps a “master narrative”, written with the express goal of helping our audience understand how a free society might be a solution to one of their core concerns. Structurally, these stories are similar to fables in that they are realistic fiction driving towards a particular moral.
The first iteration is usually in the form of a simple logline that describes the basic plotline and characters, such as:
“Trinh Le opens a new dry-cleaning business in Boston only to have it taken by eminent domain so wealthier developers can build a new hotel.”
In this case, our goal is to connect the concern over inequality (ie. two-tiered property rights rules that advantage wealthy developers over small business owners), and the free society benefit of community that is lost under an illiberal regime.
If done right, our story will show our target audience that FEE cares about inequality (at least in the context of a legal principle) just like they do, that more legally equal society is one facet of what we mean by a “free society”, and that a free society has positive benefits such as promoting strong, cohesive communities.
However, coming back to Arnold Kling, we assume (but are testing the assumption) that our audience is actually split into at least three political “tribes.”
Writing the Story Concept into Three Languages
Based on this assumption and on our existing experience trying to communicate these ideas, we know that showing everyone the same video using identical framing and language is — at best — unlikely to effectively appeal to all individuals of different tribes. And at worst, the wrong messaging can actually backfire and put people off to understanding our ideas.
Showing a Conservative a message presented in Progressive language/framing can trigger that Conservative’s ideological defenses and make them less likely to accept the content, and vice versa.
So, using Kling’s 3AM as a guide, we split our master narrative into three distinctive narratives targeted separately to Progressives, Conservatives, and Libertarians.
Messaging to Progressives
For Progressives, we play up the oppressor vs. oppressed aspects of the story.
For instance, with Trinh’s Dry Cleaning, we use both the language in the narration and the on-screen imagery to emphasize the size of the corporation receiving her property and the callousness of the government willing to use eminent domain on their behalf. We highlight collective identity differences such as race and age which are suggestive of stereotypical power dynamics that are a frequent topic of criticism among progressives. We also depict Trinh and her community as largely powerless in the face of this obstacle — which of course, most people in real life are when battling eminent domain abuse, so there’s a lot of truth to this aspect of the story.
Messaging to Conservatives
However, for Conservatives, we focus on different attributes.
Instead of making the imagery and narration about how Trinh is a lower-income minority woman going up against an evil corporation, we present the story as one where she’s a hard working business owner who has built a name for herself in her community. We put the onus of eminent domain more squarely on the shoulders of bureaucrats and politicians, and highlight the destruction of Trinh’s neighborhood as bulldozers come in to wreck all the homes and businesses on her block, all in the name of redevelopment. The story for conservatives becomes about the outsider coming in and harming an existing community.
Messaging to Libertarians
And of course, for Libertarians, we play up Trinh’s autonomy as a business owner, and the government’s role in unjustly taking away her rightful property and the opportunities she created for herself and her neighborhood.
The beats and characters are essentially the same in each story, but what we choose to emphasize changes to more closely fit the tribal languages and unique concerns of our intended audience groups.
Messaging Can Be Verbal and Non-Verbal
Also, it’s critical to understand that our means of emphasis is not merely done by changing the words spoken in each video. Most of people’s perception of any story comes from the way that story is framed, and like real life, most of the framing in a video is a product of its non-verbal elements — the overall color pallet, the specific imagery choices, the music, the sound design… It all matters.
The meta-contextual choices also matter — the race and gender of the main and supporting characters, the specific setting, the gender or cultural identity of the voice over narrator, etc.
The more we carefully consider all of those elements holistically, the more our message will be tailored to the intended audience and the more effective it will be at communicating the ideas we hope to convey to that audience. To this end, we researched news sources, political party platforms, and think tank content to draw specific topics/messages/language choices to better understand how each of the 3 tribes communicate. Moreover, we draw from Jonathan Haidt’s research on Moral Foundations to leverage how to speak to the moral taste buds of each of the three tribes. We’re working to incorporate as much of this type of information as we have.
The better we understand and implement distinctive differences in our messaging targeted to our specific audiences, the easier it is to make intelligent messaging decisions and succeed in our ultimate goal of communicating our ideas with people from a wide range of viewpoints.