Deposition
Reactance
In the matter on the record
Reactance is the motivational state of pushback (an amalgam of anger and counterarguing) aroused when a person perceives a behavioral freedom to be threatened or eliminated, driving them to restore that freedom by doing, wanting, or defending the threatened option more strongly.
Q. The claim on the record
First formalized by Jack Brehm in 1966, reactance theory holds that threatening someone's freedom to choose motivates them to reassert it, often by doing the opposite of what was urged (the "boomerang effect"). Modern work operationalizes reactance as an intertwined blend of anger and negative cognitions (Dillard & Shen, 2005), and meta-analyses confirm that high freedom-threatening message language reliably raises reactance (r approximately .20), which in turn reduces persuasion. The core message-framing effect replicates, but specific popular instances built on it (notably the "Romeo and Juliet effect") have failed direct replication.
A. For the defense
The core message-level effect is well supported: meta-analyses (Rains 2013, K=20, N=4,942; Li & Shi 2025, 28 articles comprising 33 studies, 146 effect sizes) consistently show freedom-threatening language raises reactance (r approx .20) and that reactance depresses persuasion. The intertwined operationalization replicates across topics. However, prominent specific applications are weaker: the 'Romeo and Juliet effect' (Driscoll, Davis & Lipetz, 1972) failed a same-measures replication by Sinclair, Hood & Wright (2014, N=396), which found parental interference predicted poorer, not stronger, relationships. Effect sizes for the basic message effect are small-to-moderate, and gain/loss framing shows essentially no reactance effect.
Exhibit A — Dillard & Shen (2005), Study 1 & 2
An 'intertwined' model in which reactance is a latent factor indicated by both anger and negative (counterarguing) cognitions best fit the data; the anger+cognition composite fully mediated effects of threat-to-freedom and trait reactance on attitude and intention. Established reactance as directly measurable. — James Price Dillard, Lijiang Shen, 2005
Model fit favored the intertwined model; anger and negative cognitions contributed roughly equally. Specific standardized path coefficients reported in paper; full mediation observed. · N=196 (flossing); N=200 (alcohol)Exhibit B — Li & Shi (2025) meta-analysis
High (vs low) freedom-threatening language reliably increased anger, negative cognitions, and overall reactance; reactance components in turn reduced persuasion. Gain-vs-loss framing had near-zero effects on reactance. — Zixi Li, Jingyuan Shi, 2025
High vs low threat language: anger r=.21 (k=28), negative cognitions r=.17 (k=25), overall reactance r=.20 (k=53). Reactance to persuasion: anger r=-.23 (k=42), negative cognitions r=-.18 (k=39). Gain/loss framing on reactance r=-.02 (k=12, n.s.). · 28 articles comprising 33 studies; 146 effect sizesExhibit C — Rains (2013) meta-analytic review
Supported reactance as an intertwined construct of anger and negative cognitions; freedom-threatening language increased anger, negative cognitions, and reactance. Found anger a relatively stronger indicator of reactance than negative cognitions. — Stephen A. Rains, 2013
Anger was a stronger indicator of reactance than negative cognitions (standardized loading lambda approx .62 vs .52). Anger-cognition correlation r approx .31. Specific freedom-threatening language effect sizes are reported in Rains (2013) tables; comparable r approx .17-.21 values are confirmed in the subsequent Li & Shi (2025) meta-analysis. · K=20 studies, N=4,942A. The mechanism in dispute
When a person perceives that a specific behavioral freedom is threatened or removed (e.g., a forceful directive, a ban, controlling language, or censorship), an aversive motivational state called reactance is aroused. The person is then motivated to restore the freedom, which manifests as increased desire for the threatened option, derogation of the source, counterarguing, and sometimes doing the opposite of what was urged (the boomerang effect). Magnitude scales with the importance of the freedom, the proportion of freedoms threatened, and the perceived legitimacy/forcefulness of the threat. Dillard & Shen (2005) and Rains (2013) operationalize the experienced state as an intertwined amalgam of anger (affect) and negative cognitions (counterarguing).
Q. On redirect
There has been a long-running debate over whether reactance is best modeled as purely cognitive (counterarguing), purely affective (anger), two parallel processes, or a single 'intertwined' affective-cognitive construct. Dillard & Shen (2005) and Rains (2013) favor the intertwined view, though some scholars argue anger carries more weight, and others treat reactance as partly an unmeasurable intervening motivational state per Brehm's original framing.
Matter of record — law / media
The Streisand Effect. In 2003 Barbra Streisand sued photographer Kenneth Adelman and Pictopia.com for USD 50 million to remove an aerial photo of her Malibu home (shot to document coastal erosion). The image had been downloaded only 6 times (2 by her own attorneys) before the suit; after media coverage of the lawsuit, over 420,000 people visited the site within a month. The suit was dismissed and Streisand was ordered to pay USD 177,000 in legal fees. Mike Masnick coined 'the Streisand Effect' on Techdirt on January 5, 2005. Widely cited as a real-world reactance/backfire case where attempted suppression amplified attention.
Matter of record — public health / marketing
Industry-sponsored anti-smoking ads boomerang among adolescents. Research found that tobacco-industry-sponsored anti-smoking advertisements (e.g., 'Think. Don't Smoke') produced reactance-consistent boomerang effects in adolescents: greater curiosity about and more favorable attitudes toward smoking, rather than reduced intent, illustrating reactance to perceived autonomy threats in health messaging.
Matter of record — public health
Extreme negative anti-smoking PSAs can backfire. Communication research reported that very graphic/disturbing anti-smoking PSAs can trigger defensive reactions strong enough that viewers stop processing the message, producing a 'boomerang' rather than persuasion, consistent with reactance to high-threat fear appeals.
The ruling — replication mixed
Sustained in part
Restore a sense of choice and reduce perceived threat: use autonomy-supportive language ('you're free to decide', 'it's up to you'), offer options rather than mandates, avoid commanding/dogmatic phrasing, and acknowledge the audience's freedom. Evidence shows gain/loss framing barely moves reactance, whereas explicit freedom-threatening language reliably raises it - so soften the threat, not just the topic.
Certified by Jack W. Brehm, 1966 — A Theory of Psychological Reactance. Reporter's note: Effect sizes for the basic message effect are small-to-moderate (r approx .17-.21), drawn largely from health-communication and lab-message studies, so generalization to high-stakes finance/aviation/law contexts is by analogy, not direct measurement. The documented real-world cases (Streisand effect, anti-smoking boomerangs) are cited from secondary/news sources and illustrate reactance plausibly but are not controlled experiments isolating reactance as the cause. The 'Romeo and Juliet effect' is a known failed replication and should not be treated as evidence for reactance. Brehm's original construct included an unmeasurable motivational state; the quoted definition is a standard paraphrase rather than a verified verbatim sentence from the 1966 text, since I did not retrieve the exact page. Wide prediction intervals in the 2025 meta-analysis indicate substantial heterogeneity across studies.
Cited authorities
- Boomerang effectchildThe observable behavioral outcome (doing the opposite of what was urged) often produced by reactance; reactance is the proposed mechanism.
- Forbidden fruit effectchildRestricted or banned options become more attractive specifically because access is limited - a direct consequence of reactance.
- Romeo and Juliet effectchildClaim that parental interference strengthens romantic attraction (Driscoll et al. 1972); a popular reactance application that FAILED same-measures replication (Sinclair et al. 2014).
- Streisand effectsiblingA reactance/backfire phenomenon in information suppression; attempts to censor amplify attention. Folk/coined term, not an experimental construct.
- Scarcity heuristiceasily-confusedBoth make restricted things more desirable, but scarcity is about perceived limited availability/value signaling, whereas reactance is about threatened freedom and autonomy.
- Status quo biaseasily-confusedBoth involve resistance to change, but status quo bias is preference for the default; reactance is motivated opposition to a perceived imposition.