A Provocation: You Won’t Persuade The Public. Target Elites Instead.
Plus: Microtargeting works, but has diminishing returns (and isn't always necessary); two new studies on climate messaging; a new poll finds voters strongly favor Community Safety Departments.
In this edition:
SPOTLIGHT: A Provocation: You Won’t Persuade The Public. Target Elites Instead.
Microtargeting Works, But Has Diminishing Returns—And Isn’t Always Necessary To Persuade An Audience.
Two new studies from the Yale Program on Climate Change Communication.
A poll showing: Voters Strongly Favor Community Safety Departments—A Third Public Safety Branch That Sits Alongside The Fire and Police Departments.
New empirical research or polling you’re working on, or know of, that’s interesting? We’d love to hear from you! Please send ideas to hi@notesonpersuasion.com.
SPOTLIGHT: A Provocation: You Won’t Persuade The Public. Target Elites Instead.
Two of the hottest themes in the philanthropic and social profit worlds right now are narrative change—and deep frustration with how narrative change efforts are going.
The focus on narrative change makes sense. Public support for an issue makes it easier for a public official to back it. That’s because political leaders want to keep their jobs, and backing unpopular ideas makes doing so harder.
The frustration makes sense, too. Durably shifting public opinion by focusing on persuading a mass audience is extremely difficult and—almost always—cost-prohibitive.
Dollars and time could be put to better use by focusing less on persuading mass audiences and more on persuading the people who make policy decisions and shape the media and cultural context in which those decisions are made.
Pursuing Hearts & Minds Wastes Dollars & Time.
A common strategy for these mass persuasion efforts is to focus first on persuading some large segment of the public—often called the “hearts and minds” approach. Think of this strategy as a funnel: A message enters into the broad base of the funnel, which is the mass audience. In turn, public support generates media coverage and shifts priorities for advocacy groups and political parties. Ultimately, on the narrow end of the funnel, these shifts push public officials to support your issue.
The problem with the hearts and minds approach is that durably persuading mass audiences is extremely difficult and—in most cases—cost prohibitive. Here are some of the latest findings on persuasive effects from the issue advocacy, electoral, and public health contexts:
Issue Advocacy: A recent randomized field experiment testing the efficacy of advocacy-group created television ads on two different issues—immigration and transgender discrimination—found that the three immigration advertisements had “minimal impacts on public opinion.” The transgender discrimination advertisement performed slightly better, yielding a modest persuasive effect while it aired. But even that small effect completely dissipated a day after the ads stopped airing.
Elections: A groundbreaking meta-analysis of 49 general election field experiments found that “the best estimate for the persuasive effects of campaign contact and advertising—such as mail, phone calls, and canvassing—on Americans’ candidate choices in general elections is zero.” A different meta-analysis from the Analyst Institute arrived at more optimistic, though still sobering, conclusions, finding that “during presidential general elections, likely effects of persuasion programs vary from office-to-office, with average effects ranging from 0.7 percentage points for presidential races to 3 percentage points for local/municipal races.”
Public Health: A $40 million randomized control trial of Facebook ads served to tens of millions of adults across thirteen U.S. states featured doctors and nurses urging travelers to stay home for the Thanksgiving (and, later, Christmas) holiday in order to reduce the spread of Covid-19.
The ad campaign resulted in a nearly 1% reduction in distance traveled in the days leading up to the holiday and a 3.5% reduction in Covid-19 infections.
Spending millions of dollars for single-digit percentage public opinion gains makes sense in the context of an election with an anticipated razor thin vote margin, or when it comes to public health-related behavioral changes such as convincing people to get a Covid-19 vaccine. However, modest shifts in public opinion in the context of issue advocacy usually provides a less clear return on investment despite the fact that interest groups spend hundreds of millions of dollars on paid television ads alone to shape public opinion on various issues.
Why is mass persuasion so hard?
Capturing attention is very hard. Before a television ad can persuade, the viewer must have the television turned on and tuned to the right channel, be in the room, and not be scrolling through their phone. Social media—where people scroll through feeds while shopping at the grocery store or procrastinating at work—hardly offers a more favorable persuasion environment.
Once you gain attention, the task doesn’t get easier. The same message that moves some people toward your issue can move others in the opposite direction. Moreover, if an issue scans as partisan that also inhibits persuasion. Partisans tend to adopt the ideological preferences of their party as opposed to adopting a party that fits their ideological preferences.
Even when your message moves a person, the shift often is temporary—think hours and days, not weeks and months. The rapid decay comes from both forgetting and the reality that your target audience might hear your perspective today, and then tomorrow hear the opposition perspective—through a paid ad, at the barber shop, or while reading the news.
Bottom Line: Even when mass persuasion works, the results are small, short-lived, and extremely expensive. These costs are worth bearing in close elections or for public health campaigns where tiny and temporary reductions in undesirable behavior literally saves lives. But small and rapidly decaying shifts in public opinion on an issue rarely warrant the same expenditure of time and money. For example: If you had the $40 million dollar budget that researchers spent on the Covid-19 persuasion campaign, would you get more for your money marginally and temporarily shifting public support on your biggest issue priority or funding a $1 million dollar lobbying effort in each of forty different jurisdictions?
Focus More Dollars And Time On Elite Audiences
Don’t start by trying to persuade a mass audience. Instead, flip the persuasion funnel: Focus first on the narrowest part of the funnel—persuading the comparatively tiny groups of people making policy decisions and shaping the media and cultural context in which those decisions are made. This elite audience communication can persuade decision-makers to support the policy while influential messengers send the signal to those decision-makers that the issue is popular and important. In aggregate, this elite audience also helps to move the dial on public opinion, which in turn, helps to pressure more decision-makers to support your issue.
There are three benefits of elite persuasion:
It’s more efficient. Trying to change public opinion directly—to borrow an analogy from Seth Godin—is like dropping a few drops of purple dye into the ocean—the water stays the same color. But dropping purple dye into a swimming pool is different—the water will turn purple. Lawmakers, journalists, and other elite audience members are a very small group relative to the general population, but they have a massive effect on political common sense. In other words, mass persuasion is the ocean; elite persuasion is the swimming pool.
They make the decisions. Concrete policy wins build momentum for an idea, making it more likely that elite audiences engage with it and other decision-makers back it. Actual widespread public support is helpful to achieve these policy wins, but decision-makers can enact policies knowing that support doesn’t exist yet or based on their belief that their constituents support it.
Elite audiences do the mass persuasion work for us. Elite audiences—and especially the media-shapers and relevant cultural influencers—bring the status, reach, and repetition that better positions them to move the dial with mass audiences.
Where to start:
Start with the decision-makers that can enact your idea. Decision-makers often enact policy shifts based on their perception that it’s smart, effective, and popular regardless of whether public opinion has actually shifted. The goal is to shape the information context in which these officials make decisions. The relevant audience includes their staff, advisors, and others in their personal and professional life. You want these decision-makers to feel how everyone, everywhere in their social sphere is talking about these issues.
Media, political and cultural elites are the next most important audiences. These are the people who most shape the information pool that the public official swims in.
Three Caveats:
Persuading mass audiences is not the only reason to engage with them.
You might target mass audiences for small dollar donations. Success is easier to measure in this context—are you raising more dollars than you are spending on running the ads to solicit the donations?
You might leverage members of the public as part of an effort to persuade elite audiences. For example, driving constituent calls to a lawmaker can signal public support for an issue. In fact, these types of signals can matter as much as actual public opinion data.
You might run an advertisement that appears to be targeting a mass audience, but in reality the audience is the decision-maker or other elites. For example, a seven-figure paid media spend can send a signal to party elites about the importance of an issue or the resources that the people backing it are willing to spend.
Mass movements can shift public awareness and opinion. For example, the George Floyd protests appear to have built support for the idea that civilian experts can better serve some of the functions that law enforcement fill today. Investing in creating the conditions that spark mass movements is a complementary strategy, but the low-probability, high impact nature of mass mobilization makes it a complement to—not a replacement for—elite focused persuasion campaigns.
Even elite audiences need repetition. Persuading elite audiences is less about a single, perfectly crafted message and far more about a steady drip of information. It’s often the passive, repetitive flow of information that increases the salience and importance of ideas and shapes opinions and behavior over time.
Given the same level of resources, you are better off communicating a bunch of times to a tiny number of influential people than a few times to a massive number of people.
Takeaway: Focus more of your persuasion efforts on elite audiences—lawmakers, media, cultural influencers. It’s a vastly smaller pool of people; they have the power to make and shape policy outcomes; and their status and reach make them better-positioned to shift public opinion. Be extremely cautious before focusing dollars or time on persuading mass audiences. It’s very expensive to move the needle on public opinion; and, when you do, the effects are small and they evaporate rapidly.
Wait, before you read further: We’d love to hear your reaction to this spotlight: What nuance did we miss? What do we get wrong? Did we get something right, but miss the best evidence that supports our position? Let us know: hi@notesonpersuasion.com.
THREE ITEMS WORTH YOUR TIME:
1. Microtargeting Works, But Has Diminishing Returns—And Isn’t Always Necessary To Persuade An Audience.
The words “Cambridge Analytica” sparks a visceral reaction in politics watchers. For some, the ability to draw on vast consumer, demographic, and psychographic data to tailor persuasive messages evokes fear of mind-control and threatens to undermine democratic norms.
To others, the ability to draw on vast data to tailor messages is a welcome advance in the sophistication of political campaigns and speaks to the ability to educate voters in the terms most likely to appeal to their interests and sensibilities.
Yet, despite the hype, “little research has directly estimated the persuasive advantage of microtargeting over alternative campaign strategies,” as MIT’s Ben Tappin and colleagues—Chloe Wittenberg, Luke B. Hewitt, Adam J. Berinsky, and David G. Rand—wrote in an article published recently in the Proceedings of the National Academy of Science.
Filling the gap, the authors examined three different strategies for delivering persuasive messages across two separate single-issue advocacy campaigns (immigration and universal basic income):
Microtargeting, which means “showing people whichever message is predicted to be best for them personally, based on their demographic and psychological traits.”
Single-best-message, which means “showing everyone the message that performed the best on average in a pre-test.”
Naive, which means “showing people a message selected randomly from a full set of messages” crafted to increase support for the topic in question.
The study found that the “microtargeting strategy exceeded the persuasive impact of alternative messaging strategies by an average of 70% or more” averaged across the two issued areas (immigration and universal basic income). Importantly, the authors found that almost all of the persuasive effects flowed from single factor targeting (e.g. partisanship—sending one message to Democrats and another to Republicans) and “found no evidence that targeting messages by more covariates amplified [the] persuasive impact[.]”
But this kind of “targeting”—aiming specific messages to a sub-group of the population likely to best resonate with that message— is hardly the stuff of Cambridge Analytica lore. Indeed, as one of the study’s authors told MIT News, “targeting is often going to be a good idea, and if you’re not doing that, you may be leaving persuasive power on the table… At the same time, it’s clearly not mind control.”
A second recent study authored by Jon Green and colleagues—James Druckman, Matthew Baum, David Lazer, Katherine Ognyanova, Matthew Simonson, Jennifer Lin, Mauricio Santillana, Roy Perlis—and published in the British Journal of Political Science, explains that “counter to common portrayals, the politicization of science does not preclude using broad messages that resonate with the entire population,” which means that “generalized messages also can matter.”
The authors tested persuasive messages in the context of “extreme COVID-19 vaccine resistance” because, as the authors write, “COVID-19 reactions epitomized politicization, with partisanship becoming such a driver of health decisions that it was pernicious enough to threaten the health of citizens.” The authors tested a range of persuasive messages (variously “based on science cues, moral frames, or descriptive norms”), and found that all but one of them decreased vaccine resistance related to the control group. However, the authors found “very little evidence of heterogeneous effects” when “exploring whether messages had differential effects on particular subgroups of our sample…” Thus:
“...while there are significant differences in baseline attitudes concerning COVID-19 vaccination by political identities, there are not significant differences in sensitivity to pro-vaccine messages… [R]ather than working to identify a variety of successful messages, each tailored to a particularly responsive subgroup, our results suggest it can be more efficient to identify the most persuasive messages overall and broadcast them widely.”
Context: Insofar as these two studies show surprisingly homogeneous persuasive effects across sub-groups, both could be understood in the context of Alexander Coppock’s seminal work— “Persuasion In Parallel”, which he describes as a “challenge to the dominant view that persuasive information can often backfire because people are supposedly motivated to reason against information they dislike.” Instead, “drawing on evidence from a series of randomized controlled trials,” Coppock shows both that a “backfire response is rare to nonexistent” and “most everyone updates in the direction of information, at least a little bit.” Meaning, “even messages we don’t like can move us in the right direction.”
Takeaway: Whether to use a single-best-message or to micro-target often depends on available time, money, and sophistication of a given initiative. Presidential candidates using a single best message approach likely will want to fire their political consultants. However, local campaigns, public health efforts, and other contexts with tighter timelines, smaller teams and leaner budgets, should be heartened to know that a single best message approach could be effective. Moreover, when using targeting strategies, it would behoove most campaigns to remember the possibility of rapidly diminishing returns when weighing the data, staffing, and opportunity cost of segmenting target persuasion audiences based on multiple factors.
2. Two new studies from the Yale Program on Climate Change Communication:
Climate Change in the American Mind: The Yale Program on Climate Change Communication, in partnership with the George Mason Center for Climate Change Communication, released a new survey of 1,011 adults nationally. The full results are worth a read, but here is one finding that captured our attention:
“A majority of Americans understand that global warming will cause harm…Half or more Americans think global warming will cause either ‘a great deal’ or ‘a moderate amount’ of harm to future generations of people (71%), plant and animal species (70%), the world’s poor (69%), people in developing countries (68%), people in the United States (64%), people in their community (55%), or their family (52%). Many Americans also think they themselves (47%) will be harmed.”
“Messages about harms of fossil fuels increase support for renewables” (h/t Hilary Moglen, Principal, Rally): In partnership with the University of Cincinnati's Center for Public Engagement with Science, the Yale Program on Climate Change Communication conducted a study, which found that four separate video “messages about harms of fossil fuels increase[d] support for renewables.” The main purpose of the study was to determine whether leveraging ethical or moral appeals would increase support for renewable energy (in this study, it didn’t increase support more than other kinds of messages). We’ll get to moral foundations theory and messaging in an upcoming newsletter, but for now we highlight this study for a different purpose—it’s one of the rare studies that measures the durability of treatment effects. From the authors:
“Most studies on persuasion only measure immediate effects – that is, how attitudes and opinions are affected right after persuasive messages are presented. But it is critical to also understand how durable these changes are … Accordingly, we measured participants’ opinions at three different times: immediately after seeing the message, about 10 days later, and then finally after another 10 days. Our findings showed that all four messages – whether moralized or not – had durable persuasive effects on people’s support for a transition to renewable energy. Across the four different messages, between 32% and 48% of the original treatment effect was still present after three weeks.”
There are important questions here around the definition of durability and real world applications of these findings. “Durability” defined as “32% and 48% of the original treatment effect was still present after three weeks” is very valuable in the context of academic research; or, say, a ballot measure where the ad spending is proximate to the election. But, as discussed in our spotlight, the value shifts when we are talking about more drawn out electoral or issue campaigns. In other words, this “durability” is impressive, but depending on the real world application—your mileage may vary.
3. Voters Strongly Favor Community Safety Departments—A Third Public Safety Branch That Sits Alongside Fire and Police Departments.
Albuquerque, New Mexico recently launched the nation’s first Community Safety Department, an umbrella department that centralizes the city’s unarmed crisis response programs. For example:
The department’s “behavioral-health workers respond to calls, mostly from 911, about nonviolent crises involving mental health, homelessness, or substance use” and the team is “trained to connect people from some of the city’s most vulnerable populations with professional help…In doing so, they also reduce those residents’ interactions with local law-enforcement agencies, which in recent years have had the second-highest fatal-shooting rate among major American cities.” Since its creation, Albuquerque Community Safety has responded to more than twenty thousand calls, and “less than one per cent have required eventual police involvement.”
To gauge whether Americans would support or oppose an Albuquerque style community safety department where they live, Safer Cities conducted a poll of 1,704 registered voters nationally.
75% of voters, including most Republicans and most Democrats, support their city creating a community safety department that would function as a separate and coequal city department alongside the police and fire departments. Among Democrats, support sits at 88% (51% “strongly support”).
After watching a local television news segment on the Albuquerque Community Safety department, voter support for creating a community safety department where they live spiked to 88%. You can watch the news segment here.
These results are part of a much deeper dive that Safer Cities conducted on public opinion around Community Safety Departments. Here are the full results.
Related: By a 73 point margin, voters indicate that “when someone is in a mental health or suicide crisis” they “should receive a mental health response” as opposed to a “police response,” according to a recent poll commissioned by the National Alliance on Mental Illness (NAMI).
New empirical research or polling you’re working on, or know of, that’s interesting? We’d love to hear from you! Please send ideas to hi@notesonpersuasion.com.