Chapter 24: Framing, Priming, and Persuasion

On a Tuesday afternoon in the seventh week of the Garza-Whitfield Senate race, Adaeze Nwosu pulls up her chair next to Sam Harding's workstation and poses a question she's been thinking about since the weekend. "I want to understand what the coverage is actually doing to people," she says. "Not just how much coverage there is. What's it making them think about, and what's it making them think?"

The distinction she's drawing—between volume and meaning—is the subject of this chapter. The previous chapter examined the media ecosystem: the channels, platforms, and organizations through which political information flows. This chapter examines the content of that information and the psychological mechanisms through which content shapes political cognition. Framing, priming, and persuasion are not just academic concepts; they are the mechanisms that connect media coverage to public opinion, and understanding them is essential for any serious political analyst.

Adaeze has built ODA's framing analysis tools over three years, iterating through multiple approaches before arriving at a methodology she believes is both theoretically grounded and practically applicable. We will follow that methodology through this chapter, applying it to the Garza-Whitfield race as a sustained example.

24.1 The Architecture of Meaning: What Framing Is and Is Not

Beyond Content: The Shape of Information

The first and most common misunderstanding about framing is conflating it with spin or bias. Framing is not simply the selection of favorable or unfavorable facts; it is a more fundamental cognitive operation. Framing refers to the selection and salience structure imposed on information—which aspects of an event or issue are emphasized, which are omitted, what interpretive schema is invoked to make sense of the facts.

Robert Entman, whose 1993 essay "Framing: Toward Clarification of a Fractured Paradigm" is probably the most cited definitional treatment, proposed that framing involves: selecting some aspects of a perceived reality and making them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation. Notice how many cognitive operations are bundled in this definition: defining what the problem is, explaining why it exists, evaluating who or what is responsible, and suggesting what should be done about it.

These four functions—problem definition, causal interpretation, moral evaluation, and treatment recommendation—can be analyzed independently. A single news story may have explicit problem definition but implicit causal framing; a campaign advertisement may foreground moral evaluation while leaving problem definition unstated. Entman's framework provides a checklist for systematic framing analysis.

💡 Frames Are Cognitive Shortcuts

Cognitive psychology provides the mechanism for why framing works: frames activate pre-existing mental schemas that organize how we process new information. When a news story frames immigration as an "invasion," it activates the schema of warfare and national defense. When the same phenomenon is framed as "families seeking asylum," it activates the schema of humanitarian need and legal rights. The facts about migrants—their numbers, their origins, their legal status—may be identical in both stories, but the schemas activated are different, and those schemas shape the interpretation of all subsequent information.

Two Types of Framing

The political communication literature distinguishes two fundamentally different types of framing that operate through different mechanisms:

Equivalence framing (also called valence framing) occurs when logically equivalent information is presented in different verbal or numerical formats, producing different judgments. The classic demonstration comes from Daniel Kahneman and Amos Tversky's "Asian Disease Problem" (1981): when a public health intervention is described as saving 200 of 600 lives (gain frame), most people prefer it. When the identical intervention is described as resulting in 400 deaths (loss frame), most people prefer an alternative. The factual content is identical; the framing changes the decision.

In political contexts: characterizing tax cuts as "keeping $1,000 of your money" versus "government revenue reduction of $1,000" represents equivalence framing. Both describe the same fiscal reality; the first activates a property rights frame, the second a fiscal policy frame. The politically consequential finding is that equivalence framing works—the frame shifts preferences even when the underlying facts are identical and the framing is logically irrelevant to the decision.

Emphasis framing (also called issue framing) occurs when different aspects of a multifaceted issue are emphasized, rather than logically equivalent information presented differently. The "death tax" vs. "estate tax" example is emphasis framing: both terms refer to the same tax on inheritance, but "death tax" emphasizes the timing (triggering at death) while "estate tax" emphasizes the subject (inherited wealth). These frames don't present logically equivalent facts; they foreground different morally salient features of the same policy. Research by Luntz, Berinsky, and others finds that the "death tax" framing significantly increases opposition to the inheritance tax even among voters who would not be affected by it.

The distinction matters for analysis because equivalence framing has stronger theoretical grounding (it directly demonstrates irrational cognitive inconsistency) but narrower empirical scope (relatively few real political choices involve strictly equivalent alternatives). Emphasis framing is more pervasive in real political discourse but is harder to separate from legitimate substantive disagreement about which policy features are most important.

📊 The "Pro-Life" / "Anti-Abortion" Asymmetry

Survey research has consistently found that the framing of abortion preference questions affects expressed opinion. When respondents are asked whether they identify as "pro-life" or "pro-choice," response distributions differ from when they are asked whether they "support" or "oppose" abortion rights—and both differ from responses to specific policy questions (parental notification, gestational limits, public funding). The label "pro-life" activates a moral frame centered on protection; "anti-abortion" activates a regulatory restriction frame. These frames shift the self-identification numbers by meaningful margins. Survey researchers in this area face genuine methodological challenges: there is no "neutral" frame for abortion; every question wording invokes some interpretive schema.

24.2 Iyengar's Landmark Research: Episodic vs. Thematic Framing

The Framework

Shanto Iyengar's 1991 book "Is Anyone Responsible? How Television Frames Political Issues" remains one of the most important empirical demonstrations of framing effects in American political communication. Iyengar conducted a series of experiments in which subjects were randomly shown either episodic or thematic versions of television news stories about major political issues—poverty, crime, terrorism, and others.

Episodic framing presents political issues through the lens of specific events, individuals, or cases. A news story about poverty that profiles a specific family struggling to pay rent; a story about crime that follows the story of a particular victim and perpetrator; a story about homelessness that focuses on an individual person's experience. Episodic framing is the dominant style of American television news: it is visually engaging, emotionally compelling, and relatively easy to produce.

Thematic framing presents political issues in terms of structural conditions, social trends, and policy contexts. A news story about poverty that discusses unemployment rates, minimum wage policy, and the structure of the safety net; a story about crime that addresses sentencing policy, incarceration rates, and crime trends over time. Thematic framing requires more abstract presentation and is less visually compelling, but provides more context for understanding causation and policy responsibility.

The Consequences Iyengar Found

Iyengar's experimental results were striking. Subjects who were shown episodic framing of political issues consistently attributed responsibility for those issues to the individuals involved—the poor family, the criminal, the terrorist. Subjects who were shown thematic framing of the same issues were more likely to attribute responsibility to societal conditions and government policy.

This finding has profound implications for political accountability. If most people's exposure to poverty comes through episodic profiles of struggling individuals, they will tend to explain poverty as resulting from individual failure or bad luck—and will be less likely to hold political institutions accountable for poverty outcomes. If their exposure comes through thematic coverage of structural labor market changes and safety net adequacy, they will tend to explain poverty as a policy outcome—and will be more likely to evaluate politicians on their poverty policy record.

Iyengar's experiments used random assignment (which establishes causation rather than mere correlation), but they were laboratory experiments using constructed media stimuli—which raises questions about external validity. Has subsequent research confirmed these effects in more naturalistic settings? The answer is generally yes: field studies and quasi-experimental designs have found episodic framing effects in the directions Iyengar predicted, though effect sizes vary with issue type, audience characteristics, and the salience of prior information.

🔴 Critical Thinking: The Politics of Format

The dominance of episodic framing in television news is not politically neutral. Episodic framing consistently attributes responsibility to individuals rather than systems, to private behavior rather than public policy. This has the structural effect of making political institutions appear less responsible for social outcomes than they actually are—which benefits political actors who prefer to limit government accountability. This is not necessarily a deliberate conspiracy; it is the emergent political consequence of a television format that is driven by visual storytelling conventions and commercial pressures. But the political effect is real regardless of intent.

24.3 Classic Language Frames in American Politics

The Deliberate Architecture of Political Language

Political consultants and communications strategists have long understood—often intuitively before political science had systematic frameworks for it—that language choices shape political outcomes. Frank Luntz is the most prominent practitioner of deliberate political language framing: his 2007 book "Words That Work" and his extended work for Republican political clients (his 1994 "Contract With America" language consulting is a landmark) demonstrate the practical sophistication of applied framing.

Some of the most well-documented political language frames include:

"Death tax" vs. "estate tax": The estate tax, which applies only to inherited estates above a relatively high threshold (over $12 million for individuals in 2023), affects a tiny percentage of American families. Yet polling consistently shows opposition to the "death tax" far exceeding opposition to the "estate tax"—despite near-universal agreement that the two terms refer to the same policy. The "death tax" frame works by emphasizing the temporal trigger (death) in a way that implies universality (everyone dies) while suppressing the threshold information (only large estates are affected) and the equity dimension (the tax applies to unearned inherited wealth).

"Undocumented immigrants" vs. "illegal aliens": The same population is described either through a civil/administrative lens (lacking documents, which are bureaucratic requirements) or a criminal lens (violating law, which is a moral category). "Illegal aliens" additionally uses the term "alien," which in contemporary vernacular carries overtones of foreignness and non-humanity beyond its technical legal meaning. Research finds that these framing choices significantly affect support for immigration enforcement versus immigration reform policies.

"Enhanced interrogation" vs. "torture": During the post-9/11 period, the Bush administration's use of "enhanced interrogation techniques" rather than "torture" was a deliberate framing choice designed to avoid the legal and moral consequences of the latter term—consequences that included prohibition under the UN Convention Against Torture. Survey research showed that public support for "enhanced interrogation" substantially exceeded support for "torture" even when the specific techniques being described were identical.

"Climate change" vs. "global warming": Research (particularly by Leiserowitz and colleagues at Yale) has found that these terms activate somewhat different responses. "Global warming" emphasizes temperature increases and is more viscerally vivid; "climate change" is broader and more scientifically accurate but less emotionally engaging. Interestingly, Frank Luntz's 2003 memo advising Republican clients to use "climate change" rather than "global warming" on the ground that the former "sounds less severe"—advice that appears to have been adopted—demonstrates how framing strategy and political consequences interact.

⚖️ The Ethics of Deliberate Framing

When political actors deliberately choose language frames to shape public opinion, what ethical obligations apply? There are at least three competing positions: (1) Language choice is simply advocacy, and all political actors are entitled to frame issues favorably—voters can discount this just as they discount obvious advocacy. (2) Framing that exploits cognitive biases (like the death tax's manipulation of mortality salience) is qualitatively different from ordinary advocacy and crosses an ethical line. (3) The practical line matters: framing that is deceptive about factual matters (calling a tax that affects .1% of estates a "death tax" implies it affects all deaths) differs from framing that legitimately foregrounds certain values. Analysts should be clear about which of these positions they hold and why.

24.4 Agenda-Setting: What to Think About

The Media's First Power

Maxwell McCombs and Donald Shaw's 1972 study "The Agenda-Setting Function of Mass Media" is one of the most replicated findings in political communication research. McCombs and Shaw interviewed Chapel Hill, North Carolina voters during the 1968 presidential campaign and correlated the issues voters described as most important with the issues most heavily covered by the news media. The correlation was strong: the issues the media covered most were the issues voters named as most important.

This correlation does not tell us anything about the direction of causation—voters might be demanding coverage of the issues they care about, rather than learning from media what to care about. Subsequent experimental and longitudinal research established the causal direction: media coverage, on issues where people have limited prior information and experience, shapes the perceived importance of issues in the public's agenda. The effect is strongest for public affairs that are "unobtrusive" (distant from people's direct experience) and weakest for "obtrusive" issues (directly experienced).

The agenda-setting effect represents the first of what political scientists call the "three orders" of media effects: - First-order effects (agenda-setting): The media shapes what people think is important. - Second-order effects (attribute agenda-setting): The media shapes which attributes of issues and candidates are considered most important. - Third-order effects (framing): The media shapes how people interpret and evaluate political objects.

These are not always cleanly separable in practice, but the conceptual distinction is useful for analysis: a news story can simultaneously be setting an agenda (making an issue salient), setting an attribute agenda (emphasizing certain issue dimensions), and framing the issue (providing an interpretive structure).

Modern Agenda-Setting in a Fragmented Environment

A key question for contemporary political analysis is how agenda-setting works in the fragmented media ecosystem described in Chapter 23. The original McCombs and Shaw research was conducted in a media environment with limited outlets and high audience concentration. Does agenda-setting still operate when audiences are scattered across hundreds of outlets?

The honest answer is: yes, but differently. Research on "intermedia agenda-setting" finds that agendas set by elite legacy media (the New York Times, Washington Post, wire services) continue to influence what other media cover—but this transmission is now more contested and takes longer to propagate through the fragmented ecosystem. Social media creates alternative agenda-setting processes through hashtag campaigns, viral content, and platform trending mechanisms that can elevate issues to salience independently of elite media attention.

The 2016 election provided a stark demonstration: certain topics (notably Hillary Clinton's email server) received coverage patterns that most media critics believed were dramatically disproportionate to their policy significance—suggesting agenda-setting that rewarded controversy and novelty over substantive relevance.

24.5 Priming: What's on Your Mind Shapes How You Judge

The Mechanism

Priming, in the political communication context, refers to the influence of recently encountered political information on subsequent political evaluations. The concept draws on cognitive psychology's established finding that mental concepts that have been recently activated—"primed"—are more accessible for subsequent cognitive operations. Applied to politics: if a voter has recently been exposed to extensive coverage of crime, then crime-related considerations will be more cognitively accessible when that voter evaluates a political candidate or votes in an election.

The political science formalization of priming comes primarily from Iyengar and Kinder's 1987 book "News That Matters," which demonstrated experimentally that network news content primed viewers to use the issue domains covered most heavily as the criteria by which they evaluated presidential performance. Subjects who were shown news that emphasized economic coverage evaluated the president more on economic grounds; subjects shown foreign policy coverage evaluated the president more on foreign policy grounds.

Accessibility and Applicability

Political scientists distinguish two components of priming effects:

Accessibility: The degree to which a consideration is cognitively available at the moment of evaluation. Recent, vivid, or repeated exposure increases accessibility. Accessibility is the "what's on your mind" component—have you recently been thinking about this issue?

Applicability: The degree to which an accessible consideration is seen as relevant to the evaluation task at hand. Even if a consideration is highly accessible, it will not influence evaluations unless the person sees it as applying to the political target being evaluated. For example, if a voter has been primed to think about foreign policy, foreign policy considerations will be more accessible—but they will only influence candidate evaluation to the degree that the voter sees foreign policy as relevant to judging that candidate.

This accessibility-applicability distinction matters for analysis because it explains why priming does not affect all voters equally. Highly knowledgeable voters who have strong prior assessments of candidate competence across multiple dimensions may resist priming because new considerations are absorbed into an already-organized evaluative structure. Less-informed voters, whose candidate evaluations are more loosely organized, may be more susceptible to priming because new considerations more easily dominate their evaluations.

📊 Applied to Garza-Whitfield

For ODA analysts, the priming implications of media coverage of the Garza-Whitfield race are direct. Sam's media monitoring shows that coverage in the state's traditional media has heavily emphasized two issue domains over the past three weeks: crime (following a high-profile robbery case in the state's largest city) and economic anxiety (driven by plant closure announcements). Both issue domains favor Tom Whitfield in ways that may be independent of Whitfield's actual policy positions.

Crime coverage primes voters to evaluate Senate candidates on public safety grounds—an evaluation dimension where Whitfield's "law and order" messaging is designed to perform well. Economic anxiety coverage primes voters to think about economic security—but research on economic voting suggests economic anxiety voting tends to punish incumbents and reward outsiders, which also benefits Whitfield against Garza, who is associated with the current political establishment in the state.

"The question Adaeze keeps asking," Sam explains in their analysis memo, "is whether we can measure whether coverage is doing this intentionally, incidentally, or in ways nobody in media is tracking. The answer is: usually incidentally. Crime stories are being covered because crimes are happening. The plant closing is real. But the cumulative priming effect is structurally advantageous to Whitfield regardless of intent."

24.6 Persuasion: Direct Effects of Media on Opinion Change

The Minimal Effects Era

The history of political communication research includes a paradox: the academic consensus for much of the mid-twentieth century was that media had "minimal effects" on political opinion—a conclusion that, at first glance, seems to contradict everything political campaigns believe about media. How did this consensus develop, and how has it changed?

The minimal effects paradigm emerged from Paul Lazarsfeld, Bernard Berelson, and Hazel Gaudet's landmark study "The People's Choice" (1944) and the subsequent "Voting" (1954), both based on panel surveys in Erie County, Ohio during the 1940 and 1948 presidential campaigns. The research found that most voters' intentions remained stable throughout the campaign; where opinion change occurred, it was more likely to reflect interpersonal influence than media influence directly; and the media was more effective at reinforcing existing views than changing them.

These findings were accurate for their context—a media environment dominated by newspapers and radio, in a political era with strong party identification and limited media competition. The minimal effects conclusion does not straightforwardly generalize to the contemporary media environment.

The Rehabilitation of Media Effects

Starting in the 1970s and accelerating through the 1980s and 1990s, political communication researchers began demonstrating media effects that the minimal effects paradigm had missed or underestimated. The accumulation of experimental evidence—particularly from Iyengar and Kinder's agenda-setting and priming research—established that media had genuine and measurable effects on political cognition. The debate shifted from "do media effects exist?" to "how large are they, for whom, and under what conditions?"

Contemporary understanding of media persuasion effects includes several well-established findings:

Effects are largest on the less-informed and less-engaged. Voters with strong prior political identities and high political knowledge are more resistant to persuasion—they have well-organized prior beliefs that new information must overcome. Voters with weak political identities and low prior information are more susceptible—new information can more easily shape evaluations that are not yet well-anchored. This implies that media campaigns are most effective at the margins of the electorate: among persuadable low-information voters.

Repetition increases persuasion up to a point. Political advertising research (reviewed in Chapter 25) finds that repeated exposure to campaign messages increases effectiveness, but only up to a wear-out threshold at which additional exposures produce diminishing or negative returns.

Emotional appeals are more persuasive than rational arguments under conditions of low motivation. Elaboration Likelihood Model (Petty and Cacioppo) predicts that persuasion proceeds through either "central route" (careful argument evaluation) or "peripheral route" (heuristics and emotional cues) depending on the recipient's motivation and ability to process the message. Political messages designed for low-engagement voters are more effective when they appeal to peripheral cues—emotional tone, candidate appearance, group identity—than when they marshal complex policy arguments.

Threatening or fear-evoking content produces motivated reasoning. When political content threatens strongly held values or identities, recipients are motivated to counterargue and dismiss the content rather than accept it. This means highly partisan content from the other side may, paradoxically, produce "backlash" effects—increasing rather than decreasing commitment to prior positions.

💡 The Persuasion Paradox

Here is a puzzle that deserves attention: if partisan media primarily reinforces rather than persuades (confirming the views of already-convinced audiences), and if strongly identified partisans are more resistant to persuasion, then who is being persuaded by the enormous flood of political media? The answer appears to be: not the highly engaged partisan media consumers, who constitute most of the visible political discussion online, but the less-engaged, lower-information voters who consume political media incidentally and have less anchored prior beliefs. This implies that the most consequential media persuasion occurs in environments—local television, social media feeds incidentally encountered—that receive less analytical attention than high-intensity partisan media.

24.7 The Two-Step Flow and Opinion Leaders

Lazarsfeld's Second Major Contribution

Alongside the minimal effects findings, Lazarsfeld's research produced a second influential concept: the two-step flow of communication. Rather than finding that mass media affected voters directly, Lazarsfeld found that information flowed from media to opinion leaders—highly engaged, politically active individuals who consumed media heavily—and then from opinion leaders to their less-engaged social networks through personal communication.

The two-step flow model had important implications: it suggested that the politically consequential media effects were mediated through social influence, not direct audience effects. Political communicators interested in shaping opinion should invest in reaching and activating opinion leaders, who would then extend the reach of media messages through trusted personal channels.

Opinion Leaders in the Digital Era

The two-step flow concept has been substantially revised—and partially vindicated—in the social media era. On Twitter/X, YouTube, and political podcast platforms, a new class of digital opinion leaders (influencers, political commentators, podcast hosts with large followings) mediate between news events and public interpretation in ways that directly parallel Lazarsfeld's original description. When a political event occurs, it is not just processed by individual news consumers independently; it is interpreted, contextualized, and emotionally framed by influential voices whose interpretations reach millions of followers.

The key structural difference from the original two-step flow is scale and speed. Traditional opinion leaders (the politically engaged neighbor, the union steward, the community pastor) reached dozens or hundreds of people through sustained personal relationships. Digital opinion leaders reach millions instantly, through parasocial relationships that lack the reciprocity and accountability of genuine personal ties. The influence channel is faster, broader, and less embedded in local community context.

24.8 ODA's Framing Analysis Tools

The Methodology

Over three years of iteration, Adaeze and Sam developed ODA's framing analysis methodology through five stages:

Stage 1: Unit of analysis definition. Each news story or broadcast segment is treated as a unit. For longer articles, sections may be coded separately. The unit of analysis choice matters: coding articles as wholes misses variation within articles; coding paragraphs creates an unwieldy number of units.

Stage 2: Frame identification. The ODA coding scheme identifies seven candidate-level frames and six issue-level frames from the literature and adapted through iterative application to political coverage. Candidate frames include: Competence (does the story evaluate the candidate's qualifications?), Character (does it assess honesty, integrity?), Leadership (does it assess the ability to govern?), Conflict (is the story structured as a competitive battle?), Horse-race (is it primarily about who's winning?), Empathy (does it assess the candidate's connection to voters?), and Contrast (does the story primarily compare the two candidates?). Issue frames follow Entman's four-component scheme: What is the problem? What caused it? Who is responsible? What should be done?

Stage 3: Tone coding. Separate from frame, each unit is coded for tone: favorable to Garza, favorable to Whitfield, neutral/balanced, or unfavorable to both. Tone and frame are analytically separable—a story can be framed as horse-race but tone-neutral; a story can be framed around competence and tone-favorable to one candidate.

Stage 4: Reliability testing. Before the main coding run, two coders independently code a random sample of stories (approximately 15 percent of the corpus). Inter-rater reliability is calculated using Cohen's kappa. ODA's minimum reliability standard for main categories is κ ≥ 0.70; for more ambiguous subcategories, κ ≥ 0.60. Stories below reliability threshold are sent back for coder discussion and reconciliation before the protocol is finalized.

Stage 5: Automated coding extension. Once the human coding protocol is reliable, ODA trains a supervised machine learning classifier using the human-coded stories as training data. The classifier extends coding to the full corpus (thousands of stories) at far lower cost than full human coding. The automated classifier is validated against held-out human-coded examples; classification accuracy for the main frame categories is approximately 78 percent, compared to human-to-human agreement of 85 percent.

📊 The Reliability Challenge

Framing analysis faces a fundamental challenge: frames are in the mind of the researcher as well as in the text. Two trained coders with different political orientations may genuinely disagree about whether a story is framing an issue episodically or thematically—because the story is ambiguous, not because one coder is wrong. ODA's response to this challenge is: make the coding protocol as explicit as possible (providing specific textual examples for each frame category), use rigorous reliability testing, and report reliability statistics prominently in all publications so readers can assess the confidence warranted by the findings.

Applying the Tools to Garza-Whitfield

Seven weeks into the race, Sam's framing analysis of 2,341 news stories from monitored outlets produces several findings that Adaeze considers analytically significant:

Frame distribution: 48 percent of stories are primarily horse-race framed (who's winning the race), 22 percent are conflict framed (structured as competition or attacks), 14 percent are competence framed, 9 percent are character framed, and 7 percent are policy framed. The dominance of horse-race framing is consistent with prior research on Senate race coverage.

Asymmetric competence framing: When stories are framed around candidate competence, Garza receives more competence-affirming coverage (her background as state Attorney General is frequently cited as evidence of qualifications) while Whitfield receives more questions about competence but also more "outsider authenticity" framing that treats his lack of government experience as a positive in some outlets.

Crime issue framing breakdown: Of stories framing crime as a campaign issue, 67 percent use episodic framing (specific crime incidents linked to campaign messaging), 33 percent use thematic framing (crime trends, policy analysis). The episodic dominance is structurally advantageous to Whitfield's "law and order" message.

Geographic variation in framing: State capital media frames the race more around policy and governance; suburban market media frames it more around personality and horse-race; rural market media frames it most heavily around cultural identity and values.

"What this tells us," Adaeze explains in a presentation to advocacy partners, "is that the same race looks completely different depending on where you're sitting in the state. Voters in the capital are reading about policy contrast. Voters in rural areas are reading about values and identity. These audiences are being set up to evaluate the candidates using completely different criteria."

24.9 Framing Effects: What the Experimental Evidence Shows

How Frames Shape Interpretation

The experimental evidence on framing effects is extensive and generally consistent: frame changes that seem minor from the perspective of factual information can produce substantial changes in policy preferences and political evaluations. Several landmark studies deserve attention.

Iyengar's crime poverty framing experiment (already discussed): Episodic frames produce individualistic causal attributions; thematic frames produce systemic attributions. Effect sizes were substantial—subjects shifted causal attributions by 15-30 percentage points depending on frame condition.

Nelson and Oxley's Klan rally framing (1997): Subjects read a news story about a Ku Klux Klan rally framed either as a free speech issue or as a public safety issue. Free speech framing increased tolerance for the rally (because the relevant evaluative principle became free speech rights); public safety framing decreased tolerance (because the relevant evaluative principle became community safety). The identical event, framed through different value lenses, produced opposite preference responses.

Druckman and Nelson's immigration framing experiments: Framing immigration as an economic issue (immigrants compete with native workers) produced more restrictive immigration preferences; framing immigration as a humanitarian issue (immigrants are families fleeing hardship) produced more permissive preferences. Crucially, this effect moderated when subjects were able to discuss the issue with others who held different views—interpersonal deliberation reduced framing effects by creating competing considerations.

Who Gets Framed Out

ODA's framing analysis incorporates a specific protocol that Adaeze insisted on after her second year: frame absence analysis. In addition to coding what frames are present in coverage, coders note which frames are conspicuously absent given the available information.

For the Garza-Whitfield race, frame absence analysis reveals several systematic absences. Coverage of the economy almost exclusively frames economic issues through the lens of the state's manufacturing sector—job losses, plant closings, union membership. This frame is accurate for that population but systematically absent from framing of the service sector, gig economy workers, and informal economy participants who constitute a growing share of economic life in the state. The frame absence means these workers' economic experiences are not being made cognitively accessible as relevant criteria for evaluating Senate candidates' economic platforms.

Similarly, coverage of crime as a campaign issue systematically frames victims as residents of suburban and rural areas. The state's largest city has experienced both declining violent crime rates in some neighborhoods and increasing rates in others, but the coverage frame does not distinguish—the "crime frame" is applied to the state as a whole in ways that obscure the dramatically unequal distribution of victimization by neighborhood and race.

🔴 Structural Framing and Who Disappears

Frame absence analysis reveals a consistent pattern in political news coverage: people whose experiences don't fit available narrative frames become invisible in the media's political reality. Coverage requires a frame, and frames were developed to cover certain kinds of people in certain kinds of situations. When political events involve people who don't fit the standard frames—farmworkers covered by immigration frames rather than labor frames, Native Americans covered by reservation/tribe frames rather than state political frames, domestic violence victims covered by crime frames rather than policy frames—coverage either uses a poorly-fitting frame that distorts the story or doesn't cover the story at all. The cumulative effect is that some political experiences are systematically represented in the media's political world while others are systematically absent.

24.10 Persuasion in the Garza-Whitfield Context: What the Coverage Is Doing

Synthesizing the Mechanisms

Having traced the theoretical mechanisms of framing, priming, and persuasion, we can now synthesize their implications for the Garza-Whitfield race. The ODA analysis tells a coherent story:

The media ecosystem (Chapter 23) determines who is exposed to what information. The framing of that information (this chapter) shapes how that information is interpreted. The priming effects of coverage determine which criteria voters use to evaluate the candidates. The persuasion effects of repeated exposure gradually shift the opinion distribution among persuadable voters.

In the Garza-Whitfield race, the cumulative media environment is doing the following:

1. Agenda-setting: Making crime, the economy, and the horse-race question ("who's ahead?") the salient issue domains.

2. Priming: Making crime-security considerations and economic-anxiety considerations cognitively accessible when voters think about the Senate race, which favors the candidate (Whitfield) who has stronger messaging in those domains.

3. Episodic framing: Presenting crime and economic issues through individual-incident frames that reduce systemic/policy attribution, making it easier for Whitfield to offer emotional responses ("I'll be tough on crime") without being evaluated on the policy substance of those responses.

4. Horse-race framing dominance: Crowding out the policy-focused coverage that would disproportionately benefit Garza, whose strongest ground (law, constitutional governance, policy expertise) is poorly served by horse-race and conflict frames.

This is not a conspiracy. No editor is deliberately constructing this framing environment to benefit Whitfield. It is the emergent consequence of dozens of individual journalistic decisions, each locally rational, that collectively produce a coverage pattern with real consequences for how voters evaluate the candidates.

What Campaigns Can Do

Campaigns and their analysts cannot control media framing, but they can respond to it. The research suggests several campaign communication strategies that address framing effects:

Strategic pre-loading: Expose voters to frames you prefer before opposition frames become dominant. Garza's campaign's early advertising (described in Chapter 25) focused heavily on her background as Attorney General—establishing a competence frame before horse-race and conflict frames dominated coverage.

Frame repetition across channels: Consistent framing of your own message across television, digital, and earned media creates a reinforcing frame environment. Inconsistency across channels creates cognitive dissonance that reduces frame effects.

Exploiting frame absence: If coverage systematically ignores issue dimensions where your candidate is strong, campaigns can create newsworthy events that force coverage of those dimensions. Garza's campaign created a "policy week" event series specifically designed to generate thematic issue coverage in a race dominated by horse-race frames.

Activating counter-considerations: When unfavorable frames dominate, campaigns activate considerations that counter those frames. Garza's response to the crime-heavy coverage cycle was not to fight the crime frame but to introduce a parallel economic accountability frame—adding a consideration set that competes with crime-security for cognitive accessibility.

24.11 Measurement: Putting Framing Analysis into Practice

Building a Framing Codebook

For analysts interested in conducting their own framing analysis, the core methodological challenge is developing a codebook that is both theoretically grounded and practically applicable. A functional framing codebook includes:

Clear definitions: Each frame category should be defined precisely enough that two independent coders, reading the same story, will usually agree on the coding. Avoid definitions that rely on subjective interpretation ("a story that is biased against X") in favor of behavioral descriptions ("a story in which X is explicitly compared unfavorably to Y").

Mutually exclusive and collectively exhaustive categories: Frame categories should be designed so that every story fits into at least one category (exhaustive) and the categories don't systematically overlap (mutually exclusive). In practice, some overlap is inevitable and should be acknowledged.

Decision rules for ambiguous cases: The codebook should include explicit rules for resolving the most common ambiguous cases—the story that seems to fit two frame categories, the story that uses elements of multiple frames. Decision rules should be developed inductively through pilot coding before the main coding run.

Training examples: Each frame category should include 3-5 exemplary stories that clearly exemplify the frame, plus 2-3 borderline cases with explanations of how the decision rule applies.

⚠️ The Coder Training Trap

A common mistake in framing analysis is training coders to agree with the lead researcher rather than to reliably apply the codebook. If reliability is tested only after coders have been trained by discussing cases together, coder agreement may reflect shared understanding of the lead researcher's intentions rather than genuine intersubjective reliability. Best practice: develop the codebook independently, train coders separately, test reliability before any reconciliation discussions, and document the reliability statistics for each coding round.

Automated Frame Detection

Machine learning approaches to frame detection have advanced considerably. Supervised classification using models like BERT or RoBERTa, fine-tuned on human-coded examples, can now achieve accuracy in the 75-85 percent range for main frame categories—comparable to inter-rater reliability between human coders. These approaches are most reliable for well-defined, frequent frames in domains with substantial training data; they perform less well for novel frames, highly context-dependent coding, and low-resource languages.

For practical political analysis at the campaign or advocacy organization level, a hybrid approach is most feasible: human coding for a representative sample that establishes reliability and validates the classification model, automated coding for the full corpus, and human review of cases where the classifier's confidence is low.

24.12 The Ethical Dimensions of Framing Analysis

When Framing Knowledge Is a Weapon

The same understanding of framing effects that enables academic researchers to analyze media content also enables political strategists to deliberately construct frames designed to exploit cognitive biases. Frank Luntz's career is largely built on this application: systematic testing of which language frames move public opinion, followed by strategic deployment of those frames in political communications.

This raises a genuine ethical question: is there a meaningful ethical distinction between understanding framing effects (the analyst's role) and deliberately exploiting them (the strategist's role)? One position: no distinction—all political communication involves framing, and analysts who develop framing science are responsible for its applications. Another position: the distinction lies in transparency. Academic framing analysis is published, subject to peer critique, and available to all actors. Proprietary strategic framing is deliberately concealed from the audiences it targets.

⚖️ The Frank Luntz Question

Frank Luntz has argued that his work simply identifies language that people respond to—language that connects political ideas to existing values and experiences. His critics argue that his work deliberately exploits cognitive biases to manipulate public opinion in ways that bypass rational political evaluation. Both descriptions are partially accurate. The "death tax" frame exploits mortality salience in ways that are not relevant to any rational evaluation of the policy. But the frame also genuinely connects the policy to values (property rights, intergenerational transfer) that are legitimately relevant to policy evaluation. Political framing rarely consists of pure manipulation or pure legitimate advocacy; it is usually both simultaneously, and the ethical assessment requires case-by-case judgment.

Adaeze's Standard

Adaeze has developed what she calls a "transparency test" for ODA's own framing choices in its communications: "Before we choose how to frame our findings, we ask: if the audiences we're trying to reach knew exactly what framing choice we're making and why, would they feel that choice was being used to inform them or to manipulate them? If the answer is 'manipulate,' we go back to the drawing board."

This test is imperfect—it relies on the researchers' ability to accurately model their audiences' epistemic standards—but it represents a genuine attempt to hold applied framing analysis to an ethical standard that goes beyond "whatever is most effective."

24.13 Frame Contests: How Frames Compete and Change

The Dynamic Ecology of Frames

A common misconception about framing in political communication is that frames are static: that a given issue has an established frame that shapes interpretation uniformly over time. The reality is that frames compete and evolve. At any given moment, multiple frames for a political issue may be circulating simultaneously, each backed by different political actors, media organizations, and social movements. The dominance of any particular frame at a given time is the product of this ongoing competition—and competition outcomes can change.

The sociology of knowledge provides a useful perspective on how frames compete. Berger and Luckmann's "social construction of reality" framework, applied to political frames, suggests that frames gain and maintain dominance through a combination of institutional backing (which organizations promote this frame?), resource advantages (which actors have the resources to consistently deploy this frame through media, advertising, and political communications?), and social resonance (does this frame connect to widely shared cultural values and experiences?). No single factor determines frame dominance; all three interact.

In the political arena, frame competition is one of the central activities of political actors. Political campaigns are, in significant part, frame competition contests: each campaign attempts to establish the frame through which the race is evaluated—what the election is "about." When Whitfield's campaign frames the Senate race as fundamentally about public safety, and Garza's campaign frames it as fundamentally about economic security and healthcare, they are competing not just over voters but over the terms of evaluation that voters use.

Frame Resonance and Cultural Context

Why do some frames win and others lose? William Gamson and colleagues' research on social movements and frames emphasizes the concept of "frame resonance"—the degree to which a frame connects to the broader cultural narratives and values of its intended audience. A frame that successfully links a political position to widely shared cultural values, personal experiences, or established narratives has a structural advantage over a frame that requires people to adopt new interpretive schema.

The "law and order" frame that structures Whitfield's public safety message resonates because it connects to: anxiety about personal safety (experiential resonance), a longstanding American cultural narrative about the frontier, self-reliance, and protection of family and community (cultural resonance), and the schema of criminal prosecution with which many Americans have some familiarity (cognitive accessibility). The frame does not require people to learn new concepts; it activates familiar ones.

The "economic security" frame that Garza's campaign is attempting to deploy is harder to activate because it requires a more novel cognitive link: connecting healthcare costs and economic stability to Senate-level policy, which requires a more sophisticated understanding of federal legislative action than the direct "tough cop punishes criminals" narrative Whitfield offers. This is not to say the economic security frame is wrong or ineffective; it is to say that its activation requires more cognitive scaffolding—which is a strategic liability in a media environment that rewards simplicity and emotional directness.

🔗 Connection to Chapter 25

This analysis of frame resonance directly connects to advertising strategy. Campaigns that understand frame resonance can design advertising messages that activate frames strategically—not just choosing issue topics, but choosing the cultural connections that make frame activation easy and automatic. The advertising chapter examines how message testing identifies which frame activations are most effective for which voter segments.

When Counter-Framing Works

Research on counter-framing—attempts to replace a dominant frame with a competing one—identifies several conditions under which it is more or less likely to succeed:

Competing frames are more effective when they pre-load before the dominant frame is established. Iyengar's research suggests that frames encountered first have an advantage in organizing subsequent information. Counter-framing after a dominant frame is established requires working against an already-organized interpretive schema, which is cognitively costly.

Counter-frames work better when they draw on alternative but equally resonant cultural narratives. Attempting to replace the crime/fear frame with a purely statistical counter-narrative ("crime rates are actually declining") is less effective than activating an equally resonant alternative frame ("government accountability means holding the powerful accountable, not just the powerless"). The second counter-frame doesn't dispute the first's activation of safety concerns; it introduces a competing application of the safety value.

Deliberation moderates frame effects. Druckman and Nelson's finding (Section 24.9) that interpersonal deliberation reduces framing effects suggests that contexts that enable discussion of competing frames—community meetings, editorial board endorsement processes, candidate forums—can reduce the advantage of whichever frame dominates media coverage. This has direct implications for campaign debate strategy.

24.14 Global Dimensions: Framing Across Political Contexts

Frames Are Not Culturally Universal

The framing research reviewed in this chapter is primarily American, conducted with American research subjects on American political issues. It would be a mistake to assume that the specific frames and their effects generalize across political cultures. Frame resonance is inherently cultural: a frame that activates familiar schema in one cultural context may activate very different schema in another, or may activate nothing at all.

Several cross-national framing studies illustrate the cultural specificity of frame effects. Research comparing issue framing in individualistic versus collectivist cultural contexts (building on Hofstede's cultural dimensions) finds that individual responsibility frames are more resonant in individualistic cultures (such as the United States) and community/social responsibility frames are more resonant in collectivist cultures. The practical implication: the episodic/individual framing dominance in American television news reflects not just commercial incentives but also cultural values that make individual-centered narratives more resonant in the American context.

🌍 Comparative Framing: Immigration in Different Contexts

Immigration framing in American political media tends to oscillate between economic competition frames (immigrants compete with native workers) and security/sovereignty frames (borders and national identity). In European contexts, immigration is more frequently framed through labor market integration, social welfare system sustainability, and cultural cohesion frames that reflect different institutional and cultural contexts. Neither framing is inherently correct; each reflects the political debates most salient in its national context. Cross-national comparative framing analysis is a growing research area that illuminates both the cultural specificity of political communication and the structural pressures that produce convergent framing patterns across diverse contexts.

24.15 The Measurement of Framing Effects: Survey and Experimental Methods

The Experimental Foundation

The strength of the framing effects literature comes primarily from experimental research—studies where participants are randomly assigned to receive different versions of political information, with differences in framing being the only systematic variation between conditions. This experimental logic enables causal inference: if subjects who received the episodic-framed story attribute responsibility differently than subjects who received the thematic-framed story (and all else is equal through random assignment), then the difference in framing must be causing the difference in attribution.

However, experimental framing research has persistent external validity concerns. Most experiments expose participants to a single piece of media content under controlled conditions that differ from natural news consumption in important ways: real-world news consumption is cumulative (you read dozens of stories over weeks), contextual (you have prior information that interacts with the frame), and social (you discuss what you've read with others who provide alternative framings). A single-exposure laboratory experiment captures the mechanics of framing effects but may exaggerate their magnitude and persistence relative to real-world conditions.

The field of political communication has increasingly addressed these concerns through longitudinal survey designs that track opinion change over time alongside media consumption, and through survey experiments embedded in nationally representative samples that provide greater external validity than undergraduate laboratory samples. The cumulative evidence from these methodological advances has generally confirmed the direction of laboratory framing effects while finding smaller magnitudes—consistent with the expectation that real-world deliberative and media environment context moderates laboratory-derived effects.

Measuring Frames in Survey Instruments

An applied challenge for analysts conducting public opinion research is that question wording in surveys is itself a form of framing. When a survey asks "Do you support or oppose government funding for abortions?" it activates a government spending frame. When the same policy is described as "public funding for women's reproductive healthcare," it activates a healthcare frame. The responses to these questions—which are legally describing the same policy—may differ substantially.

This is not merely an academic problem for survey designers; it is a routine challenge in political polling. Campaigns commission polls that use favorable question wording to produce favorable topline results; opposition research involves finding survey results that show unfavorable question wording can generate unfavorable numbers for a candidate. Analysts reading public polling should always examine question wording rather than just reported toplines, because the question wording contains implicit framing choices that may substantially determine the reported opinion distribution.

Sam Harding's ODA polling protocols include a deliberate practice of running split-sample question-wording experiments: the same question on a politically sensitive issue is asked with two different wordings in different sample halves, and the difference in responses is reported as a "framing sensitivity estimate." If 67 percent support a policy under one wording and 52 percent support it under alternative wording, the policy's "true" level of public support is somewhere between those numbers—and the 15-point gap documents how consequential the framing choice is for the reported figure.

⚠️ Reading Poll Numbers: Always Check the Question

A common error in political journalism and campaign analysis is reporting poll toplines without examining question wording. "54 percent of voters support immigration reform" is not a fact about the policy; it is a fact about responses to a specific question wording administered to a specific sample at a specific time. Before citing any poll number in strategic or analytical contexts, analysts should locate the actual question wording, assess what frame it activates, and consider how alternative wordings might produce different numbers. Polling is framing, and framing shapes polling results.

Behavioral Measures Beyond Survey Response

Because framing effects in surveys are themselves susceptible to the measurement-shapes-reality problem—survey questions frame the issue in asking about it—researchers have developed complementary measurement approaches that use behavior rather than self-report as the outcome variable.

Implicit Association Tests (IATs) measure the speed and accuracy of responses to category pairings, revealing automatic associations that respondents may not be consciously aware of or may be unwilling to report. Applied to political framing, IATs can test whether a frame has shifted automatic associations (e.g., connecting "immigration" with "threat" versus "opportunity") even when stated opinions don't change.

Eye-tracking studies measure where subjects look while reading a news story or watching a video—revealing which elements capture attention and which are processed most deeply, with implications for which frame elements are actually being registered.

Information search behavior in web-based experiments can reveal how framing affects what additional information people seek: after exposure to a crime-framing story, do people seek out criminal justice statistics (thematic extension) or more crime incident details (episodic extension)? The search behavior reveals the direction of cognitive engagement in ways that survey response cannot.

These behavioral measures are more expensive and logistically complex than survey experiments, which is why most applied political framing research still relies on survey self-report. But they represent an important supplement for analysts interested in the deeper cognitive mechanisms underlying framing effects.

Summary

Framing, priming, and agenda-setting represent the cognitive mechanisms through which media content shapes political reality. Framing—both equivalence framing that exploits cognitive biases and emphasis framing that foregrounds certain issue dimensions—shapes how voters interpret political information, attribute responsibility, and evaluate policy options. Priming makes certain considerations cognitively accessible, determining which criteria voters use to evaluate candidates and leaders. Agenda-setting shapes which issues people consider important enough to bring to bear in political evaluation.

The episodic versus thematic frame distinction, established by Iyengar's research, is particularly consequential: a media environment dominated by episodic coverage systematically reduces the attribution of political responsibility to institutions and policies, with structural consequences for political accountability.

ODA's framing analysis methodology—combining systematic human coding, reliability testing, and automated classification extension—demonstrates how these academic concepts can be applied in practice. The analysis of the Garza-Whitfield race reveals how multiple, individually unremarkable coverage decisions accumulate into a framing environment that advantages certain political actors and certain interpretive frameworks over others.

The next chapter turns from how messages are framed to how they are paid for and delivered: political advertising, where campaigns exercise their greatest direct control over the political information environment.

Key Terms

Agenda-setting: The media's power to shape the public's perception of which political issues are most important.

Applicability effects: The degree to which an accessible consideration is seen as relevant to a political evaluation task.

Emphasis framing: Framing in which different aspects of a multifaceted issue are foregrounded, rather than logically equivalent information presented differently.

Episodic framing: Presentation of political issues through the lens of specific events, individuals, or cases; tends to produce individualistic causal attributions.

Equivalence framing: Framing in which logically equivalent information is presented in different verbal or numerical formats, producing different judgments.

Framing: The selection and salience structure imposed on political information that shapes interpretation, causal attribution, moral evaluation, and treatment recommendations.

Opinion leaders: Highly politically engaged individuals who mediate between mass media and less-engaged members of their social networks in the two-step flow model.

Priming: The influence of recently encountered political information on subsequent political evaluations through increased cognitive accessibility of related considerations.

Thematic framing: Presentation of political issues in terms of structural conditions, social trends, and policy contexts; tends to produce systemic causal attributions.

Two-step flow: The model of political communication in which media effects are mediated through opinion leaders who then influence their social networks through interpersonal communication.

24.16 The Analyst's Role: Framing Effects and Professional Responsibility

The research in this chapter presents political analysts with an uncomfortable mirror. The same body of work that documents how media frames shape voter interpretation also applies, without modification, to the frames embedded in polling questions, in data visualizations, in briefing documents, and in the analytical reports that analysts produce for campaigns, advocacy organizations, and the media.

An analyst who selects which issue domains to include in a tracking poll is making a framing choice. An analyst who chooses to present crime data as a rate trend (thematic framing) rather than as a recent high-profile incident (episodic framing) is making a framing choice. An analyst who recommends that a campaign's messaging should emphasize economic anxiety is recommending a priming strategy — one that will, if successful, shift which considerations voters weigh when evaluating the candidate.

These are not neutral technical decisions. They are choices with political consequences, and they are choices that analysts make, often invisibly to the people who receive their work.

24.16.1 The Analyst as Frame Setter

Political analysts sit at a peculiar junction in the information environment. Unlike journalists, who face formal norms of objectivity and face public accountability for their framing choices, analysts typically work within organizations — campaigns, polling firms, advocacy groups — where their framing choices are invisible to the public. The frame embedded in a campaign poll's question wording is not disclosed when the topline results are published. The frame embedded in a strategic briefing document shapes a campaign's message strategy without external scrutiny.

Adaeze's "transparency test" from Section 24.12 — asking whether the audiences would feel informed or manipulated if they understood the framing choice and why it was made — is one approach to holding this power accountable. A complementary approach is the "reverse frame" exercise: before finalizing an analytical recommendation, consider how a researcher with the opposite ideological prior would frame the same data. If the two framings produce opposite conclusions from the same underlying evidence, that divergence is a signal that the framing choice is load-bearing and should be made explicitly rather than embedded as a hidden assumption.

24.16.2 The Limits of Analytical Neutrality

Some analysts respond to the framing problem by aspiring to frame-free analysis — presenting only raw data, avoiding interpretive language, letting the numbers speak for themselves. This aspiration is well-intentioned but impossible to achieve. Every choice about what data to collect, how to present it, and what comparisons to draw is a framing choice. A pie chart and a bar chart of the same data prime different interpretations. A scale that runs from 0 to 100% and a scale that runs from 45% to 55% around a threshold make the same finding look like a large majority or a small plurality.

The honest response to the framing problem is not the pretense of neutrality but the practice of transparency: making framing choices explicit, acknowledging the alternatives, and giving analytical audiences enough information to evaluate whether the framing is appropriate for their purposes.

⚖️ The Analyst's Standard The framing research in this chapter suggests that analysts — precisely because they understand the mechanisms through which framing shapes political interpretation — bear a heightened obligation to use that understanding responsibly. Knowing that episodic framing depresses systemic attribution of responsibility imposes a responsibility to consider carefully whether to use episodic framing in one's own analytical products. Knowing that priming effects make recently encountered considerations more cognitively accessible imposes a responsibility to consider which considerations one's briefings make accessible, and for whom. The analyst who uses this knowledge strategically without acknowledgment is doing something meaningfully different from the analyst who uses it transparently, with attention to its effects. The distinction matters, and students entering this field should develop a clear answer to where they stand.