# Propositions from ALL-CAPS

**Citation:** Yonathan A. Arbel & Andrew Toler, ALL-CAPS, 17 J. Empirical Legal Stud. 862 (2020).

**Source:** [prepublication working-paper PDF](https://works.battleoftheforms.com/papers/ssrn-3519630/paper.pdf)

**Review status:** 32 model-drafted, source-checked; 0 human-reviewed. Page references use the printed pagination and, separately, the 1-based PDF page number.

## 1. Legal reliance on blocks of capital letters is unsupported contract lore rather than an evidence-based method of improving consumer consent

**Location:** Introduction, printed pp. 3-7 (PDF pp. 3-7)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 3–7, that courts and legislatures inherited all-caps from typewriter-era practice and repeatedly treat it as proof that onerous terms were conspicuous, read, and understood. That policy developed without direct evidence of improved consent and despite early legibility research pointing toward slower reading. This is significant because an untested typographic convention determines whether consequential waivers and disclaimers bind consumers. It connects to contract lore, conspicuousness, informed consent, legal formalism, typography, and evidence-based law.

**Evidence anchor:** The introduction traces the practice to assumption rather than proof and identifies negative historical evidence on reading speed.

**Boundary:** The authors challenge standard blocks of all-caps, not every possible use of capitalization or every nonbehavioral contractual formality.

**Connections:** contract lore; conspicuousness; informed consent; legal formalism; typography; evidence-based law

**Record:** `ssrn-3519630-p01` · `machine-drafted-source-checked`

## 2. Consumer inattention to fine print permits firms to shroud onerous terms, creating the no-reading problem that conspicuous disclosure is supposed to solve

**Location:** Introduction, printed pp. 3-4 (PDF pp. 3-4)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 3–4, that consumers usually know salient price and quantity terms but remain ignorant of arbitration provisions, automatic renewal, damages limits, warranty waivers, and other fine print. Firms can cut costs by hiding unfavorable terms that do not affect demand and may actively make them difficult to process. This is significant because the conspicuousness project responds to a genuine consent failure even if its chosen typography fails. It connects to rational ignorance, shrouded attributes, standard forms, asymmetric information, boilerplate, and consumer consent.

**Evidence anchor:** The introduction describes widespread ignorance of nonsalient clauses and firms’ incentives to hide unfavorable terms.

**Boundary:** Consumers may know some terms from context or experience, and the paper does not claim that full reading is always efficient or sufficient.

**Connections:** rational ignorance; shrouded attributes; standard forms; asymmetric information; boilerplate; consumer consent

**Record:** `ssrn-3519630-p02` · `machine-drafted-source-checked`

## 3. Conspicuous-disclosure rules make typography a gateway to enforcement of especially consequential consumer terms

**Location:** Introduction, printed pp. 4-6 (PDF pp. 4-6)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 4–6, that statutes and courts often condition enforcement of warranty disclaimers, liability releases, arbitration clauses, automatic subscriptions, and other important terms on conspicuous presentation. Capitalization is frequently treated as strong or dispositive evidence of conspicuousness. This is significant because formatting can determine compensation after wrongful death or serious property loss even when it did not change the consumer’s understanding. It connects to UCC section 2-316, warranty disclaimer, liability waiver, arbitration, automatic renewal, and procedural assent.

**Evidence anchor:** The introduction catalogs statutes and cases that rely on capitalization to render onerous provisions enforceable.

**Boundary:** All-caps is sometimes one factor among several, and courts can reject capitalized terms when placement or surrounding formatting defeats notice.

**Connections:** UCC section 2-316; warranty disclaimer; liability waiver; arbitration; automatic renewal; procedural assent

**Record:** `ssrn-3519630-p03` · `machine-drafted-source-checked`

## 4. Courts and statutes widely equate capitalization with conspicuousness even though the supporting empirical record is absent and legibility evidence is mostly negative

**Location:** Conspicuous Disclosure and Asymmetric Information, printed pp. 11-16 (PDF pp. 11-16)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 11–16, that conspicuousness rules are a dominant response to uninformed contracting and that all-caps has become their principal mechanical implementation. Yet the authors find no empirical validation for improved consent. Typography studies dating to 1928 instead associate capitals with slower reading, plausibly because uniform letter height removes ascenders and descenders. This is significant because doctrine ignored the best available evidence while imposing serious consumer consequences. It connects to mandatory disclosure, conspicuous type, reading speed, ascenders and descenders, judicial notice, and regulatory evidence.

**Evidence anchor:** Part I documents legal endorsement, finds no supporting empirical study, and reviews mostly negative legibility research and professional skepticism.

**Boundary:** Older psychology studies used nonlegal text, and slower reading could theoretically promote deliberation rather than reduce understanding.

**Connections:** mandatory disclosure; conspicuous type; reading speed; ascenders and descenders; judicial notice; regulatory evidence

**Record:** `ssrn-3519630-p04` · `machine-drafted-source-checked`

## 5. A computational study of standard forms from 500 high-traffic websites measures capitalization in contracts governing hundreds of millions of consumer relationships

**Location:** All-Caps in Action: Methodology, printed pp. 17-19 (PDF pp. 17-19)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 17–19, that the prevalence of formatting matters to the proof demanded for its effectiveness. They analyze forms from 500 popular websites selected through Alexa traffic rankings and use a Python script to classify capitalization at word, sentence, paragraph, and header levels, defining a capitalized paragraph by an eighty-percent threshold. This is significant because it supplies systematic evidence about legal typography in use rather than relying on memorable cases. It connects to computational legal studies, standard-form contracts, web traffic, text classification, prevalence measurement, and corpus methods.

**Evidence anchor:** Part II describes the 500-contract sample, Alexa selection, and automated capitalization-detection rules with manual verification.

**Boundary:** Header and paragraph boundaries lack unique definitions, and popular website terms may not represent every offline or small-firm contract.

**Connections:** computational legal studies; standard-form contracts; web traffic; text classification; prevalence measurement; corpus methods

**Record:** `ssrn-3519630-p05` · `machine-drafted-source-checked`

## 6. All-caps is pervasive in major consumer contracts, with roughly seventy-seven percent containing a capitalized paragraph

**Location:** All-Caps in Action: Findings, printed pp. 18-20 (PDF pp. 18-20)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 18–20, that about seventy-seven percent of the sampled standard forms contain at least one all-caps paragraph; approximately 17.4 percent of detected headers and nine percent of words are capitalized. Because the sample includes firms such as Amazon, Facebook, Uber, and Dell, these forms structure a vast number of individual relationships. This is significant because even modest comprehension harm is multiplied across widespread contractual use. It connects to boilerplate prevalence, consumer scale, uppercase paragraphs, contract design, online services, and cumulative harm.

**Evidence anchor:** Table 1 reports capitalization rates and the authors explain the reach of forms used by leading websites.

**Boundary:** The automated estimates can be under- or overinclusive, and observed prevalence does not by itself establish motivation or effect.

**Connections:** boilerplate prevalence; consumer scale; uppercase paragraphs; contract design; online services; cumulative harm

**Record:** `ssrn-3519630-p06` · `machine-drafted-source-checked`

## 7. The near absence of all-caps from persuasive marketing is evidence against its effectiveness when firms genuinely want consumers to understand a message

**Location:** Industry Findings, printed pp. 19-20 (PDF pp. 19-20)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 19–20, that advertising uses many fonts, sizes, colors, and layouts but almost never long blocks of capitals. When firms have direct financial incentives to make attractive information comprehensible, professional guidance tells them to avoid all-caps. The convention appears principally in legal text used to bind consumers. This is significant because firms’ revealed design choices cast doubt on the claim that capitals communicate effectively. It connects to revealed preference, advertising design, copywriting, persuasive communication, legal genre, and disclosure incentives.

**Evidence anchor:** The manuscript contrasts widespread contractual capitals with professional marketing practice that avoids capitalized blocks.

**Boundary:** Marketing and legal disclosure serve different purposes, so absence from advertising is suggestive rather than a controlled causal test.

**Connections:** revealed preference; advertising design; copywriting; persuasive communication; legal genre; disclosure incentives

**Record:** `ssrn-3519630-p07` · `machine-drafted-source-checked`

## 8. Every plausible behavioral justification for all-caps predicts improved recall, whether the mechanism is attention allocation, readability, warning, or slower deliberation

**Location:** All-Caps and Consumer Consent, printed pp. 20-21 (PDF pp. 20-21)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 20–21, that capitalization might hypothetically direct a limited attention budget, make letters accessible, operate as a fire-siren warning of onerousness, or slow reading enough to promote reflection. Despite different mechanisms, each functional account predicts that the highlighted obligation should be remembered more accurately than ordinary text. This is significant because recall supplies a common, falsifiable outcome for otherwise vague judicial intuitions. It connects to salience, attention budgets, readability, warning signals, deliberation, and testable hypotheses.

**Evidence anchor:** Part III enumerates four functional theories and explains why each should generate better retention of the capitalized term.

**Boundary:** Recall is not the entirety of consent, and formalities may be defended on evidentiary, cautionary, or channeling grounds not directly measured.

**Connections:** salience; attention budgets; readability; warning signals; deliberation; testable hypotheses

**Record:** `ssrn-3519630-p08` · `machine-drafted-source-checked`

## 9. All-caps should not satisfy conspicuousness without a meaningful behavioral effect because the error cost is enforcement of unknown and onerous terms

**Location:** All-Caps and Consumer Consent, printed pp. 21-22 (PDF pp. 21-22)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 21–22, that a purely formal defense is normatively inadequate where capitalization supplies the but-for basis for enforcing a wrongful-death release or comparable limitation. If the type choice does not measurably improve understanding, it cannot plausibly solve nonreading merely by shifting responsibility to the consumer. This is significant because the severity of enforcement error determines how demanding the evidentiary burden should be. It connects to error-cost analysis, formal requirements, burden shifting, meaningful assent, liability releases, and consumer remedies.

**Evidence anchor:** The authors state that capitalization should not count as conspicuous absent meaningful effect because enforcing unknown onerous terms creates large error costs.

**Boundary:** The paper does not empirically test every possible nonbehavioral function, but argues that none articulated is compelling enough for high-stakes enforcement.

**Connections:** error-cost analysis; formal requirements; burden shifting; meaningful assent; liability releases; consumer remedies

**Record:** `ssrn-3519630-p09` · `machine-drafted-source-checked`

## 10. A randomized experiment isolates capitalization by presenting 570 American respondents with an identical fifteen-paragraph subscription contract and varying one cancellation paragraph

**Location:** Primary Experiment: Methodology, printed pp. 22-27 (PDF pp. 22-27)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 22–27, that a controlled design can overcome confounding in field observation. Participants recruited online read a two-page TideTunes contract adapted from a music-service agreement; one group saw normal print and the other saw the automatic-billing and cancellation paragraph in capitals. They then chose among five cancellation rules without returning to the text. This is significant because randomization makes formatting the key systematic difference between groups. It connects to randomized controlled trials, MTurk, automatic renewal, experimental contracts, treatment assignment, and recall measurement.

**Evidence anchor:** The methods section details the sample, control and treatment contracts, cancellation question, attention safeguards, and timing measures.

**Boundary:** The online laboratory setting cannot reproduce every sales interaction, negotiation, product, or real-world incentive.

**Connections:** randomized controlled trials; MTurk; automatic renewal; experimental contracts; treatment assignment; recall measurement

**Record:** `ssrn-3519630-p10` · `machine-drafted-source-checked`

## 11. The experiment uses demographic comparison, geolocation, digital fingerprints, attention checks, reading time, and multiple answer choices to improve response validity

**Location:** Primary Experiment: Methodology, printed pp. 22-27 (PDF pp. 22-27)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 22–27, that online samples require quality controls. The study compares sample demographics with the United States, uses Positly screening, verifies unique domestic participants through digital fingerprints and geolocation, includes attention checks, records an average 102-second reading period, and offers five plausible answers to reduce guessing. This is significant because the credibility of a null or harmful effect depends on subjects actually engaging with the contract. It connects to data quality, sample representativeness, attention checks, geolocation, statistical power, and construct validity.

**Evidence anchor:** The authors report recruitment checks, demographic skews, power analysis, observed reading time, and participant feedback indicating engagement.

**Boundary:** The sample remained more educated and somewhat more male, white, and lower-income than the general population, and external validity remains limited.

**Connections:** data quality; sample representativeness; attention checks; geolocation; statistical power; construct validity

**Record:** `ssrn-3519630-p11` · `machine-drafted-source-checked`

## 12. Capitalizing the cancellation paragraph produced no improvement at all in respondents’ ability to identify the contract’s cancellation rule

**Location:** Primary Experiment: Results, printed pp. 27-30 (PDF pp. 27-30)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 27–30, that the primary experiment yielded the same proportion of correct answers in the all-caps and low-caps conditions. A non-inferiority analysis also rejects the hypothesis that capitalization supplied a policy-relevant benefit of ten percentage points or more. Thus the result is not merely a failure to obtain conventional statistical significance; the data affirmatively undermine a meaningful benefit. This is significant because the dominant legal convention failed its central comprehension test. It connects to null results, non-inferiority testing, statistical power, contract recall, conspicuousness, and evidence-based doctrine.

**Evidence anchor:** The paper reports identical correctness rates across the two capitalization conditions and rejects a benefit of at least ten percentage points.

**Boundary:** The test concerns recall of one cancellation paragraph in an online experimental contract rather than every term, font, or transaction.

**Connections:** null results; non-inferiority testing; statistical power; contract recall; conspicuousness; evidence-based doctrine

**Record:** `ssrn-3519630-p12` · `machine-drafted-source-checked`

## 13. All-caps did not act as a fire-siren warning that alerted readers to the existence or importance of the tested cancellation term

**Location:** Primary Experiment: Results, printed pp. 29-30 (PDF pp. 29-30)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 29–30, that the data provide no support for the fire-siren theory. Readers exposed to the capitalized paragraph were no more likely to retrieve the automatic-renewal and cancellation rule than readers who saw ordinary type. A visual signal cannot justify enforcement as a warning if it does not change the relevant response. This is significant because conspicuousness doctrine often assumes that visual distinctiveness itself creates notice. It connects to salience, warning design, attention capture, inquiry notice, automatic renewal, and consumer cognition.

**Evidence anchor:** The authors interpret unchanged recall as inconsistent with the claim that capitalization operates like an alarm or warning signal.

**Boundary:** The experiment measured answer accuracy rather than eye movements or subjective awareness, so it does not show that no participant ever noticed the formatting.

**Connections:** salience; warning design; attention capture; inquiry notice; automatic renewal; consumer cognition

**Record:** `ssrn-3519630-p13` · `machine-drafted-source-checked`

## 14. All-caps substantially harmed older readers: respondents over fifty-five made errors at roughly twice the rate of peers who saw low-caps text

**Location:** Primary Experiment: Age Effects, printed pp. 30-32 (PDF pp. 30-32)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 30–32, that capitalization has a sharply heterogeneous age effect. Among respondents older than fifty-five, about sixty percent in the all-caps condition answered incorrectly, compared with about thirty-one percent in the low-caps condition—a twenty-nine-point difference. The estimated effect for younger participants was negligible. This is significant because a supposed consumer-protection device may selectively burden a vulnerable group rather than merely fail on average. It connects to aging and vision, heterogeneous treatment effects, accessibility, disparate impact, consumer vulnerability, and inclusive design.

**Evidence anchor:** Regression and descriptive results show a twenty-nine-point error increase for participants over fifty-five, with little difference among younger readers.

**Boundary:** The age analysis is a subgroup result from one experiment and the authors offer possible perceptual and cultural explanations rather than establishing the mechanism.

**Connections:** aging and vision; heterogeneous treatment effects; accessibility; disparate impact; consumer vulnerability; inclusive design

**Record:** `ssrn-3519630-p14` · `machine-drafted-source-checked`

## 15. A null average benefit combined with substantial harm to older readers shifts the justificatory burden to defenders of all-caps

**Location:** Primary Experiment: Implications, printed pp. 30-32 (PDF pp. 30-32)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 30–32, that the conventional presumption should reverse. The primary study finds no general comprehension gain and evidence of substantial harm among older participants, while capitalization can determine whether onerous clauses are enforced. On that evidence, continued reliance requires an affirmative justification rather than inertia. This is significant because it reframes a familiar drafting practice as an intervention that must earn legal legitimacy. It connects to precaution, evidentiary burdens, status quo bias, age-sensitive regulation, contract enforcement, and empirical legal studies.

**Evidence anchor:** The authors combine the lack of overall benefit, age-related harm, and high legal stakes to argue that defenders must justify continued use.

**Boundary:** The proposed burden shift rests on the authors’ normative assessment of the evidence and enforcement stakes, not on a universally accepted statistical rule.

**Connections:** precaution; evidentiary burdens; status quo bias; age-sensitive regulation; contract enforcement; empirical legal studies

**Record:** `ssrn-3519630-p15` · `machine-drafted-source-checked`

## 16. A second experiment tests whether all-caps becomes useful under severe time pressure by randomizing capitalization across three short contracts with strict reading limits

**Location:** Exploratory Study: Time Pressure, printed pp. 32-34 (PDF pp. 32-34)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 32–34, that any attention-allocation advantage should be most visible when readers have little time. Eighty-one participants encountered three contracts with key clauses randomly shown in capitals or ordinary type and faced respective reading limits of twenty-three, twenty, and fifteen seconds, producing 240 usable responses. This is significant because the design gives the salience theory a deliberately favorable stress test. It connects to bounded attention, time scarcity, randomized experiments, contract skimming, cognitive load, and ecological validity.

**Evidence anchor:** The authors describe three timed contract tasks, randomized casing, eighty-one subjects, and 240 retained responses.

**Boundary:** The small sample and artificial time limits make this an exploratory test rather than a definitive estimate for real contracting environments.

**Connections:** bounded attention; time scarcity; randomized experiments; contract skimming; cognitive load; ecological validity

**Record:** `ssrn-3519630-p16` · `machine-drafted-source-checked`

## 17. Even under strict time pressure, all-caps did not improve accuracy and ordinary type performed somewhat better in two of the three tests

**Location:** Exploratory Study: Time Pressure, printed pp. 34-35 (PDF pp. 34-35)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 34–35, that the timed study supplies no evidence of a capitalization advantage. Low-caps participants did somewhat better in the second and third contracts, although only one comparison approached conventional significance and the study was underpowered. The authors therefore treat the pattern cautiously while emphasizing that it again fails to rescue the fire-siren theory. This is significant because the format remains ineffective under the condition thought most likely to favor it. It connects to time pressure, exploratory evidence, statistical uncertainty, attention allocation, replication, and contract design.

**Evidence anchor:** Across three timed tests capitalization never improved accuracy; low-caps did better in two, with limited statistical precision.

**Boundary:** The sample is small, one difference only approaches significance, and the direction and magnitude should not be generalized without replication.

**Connections:** time pressure; exploratory evidence; statistical uncertainty; attention allocation; replication; contract design

**Record:** `ssrn-3519630-p17` · `machine-drafted-source-checked`

## 18. Readers rated an all-caps contract substantially harder to read and somewhat harder to understand than the same contract in ordinary type

**Location:** Exploratory Study: Reading Difficulty, printed pp. 35-38 (PDF pp. 35-38)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 35–38, that subjective experience points in the opposite direction from the convention’s justification. In a 102-person study, respondents randomly assigned the fully capitalized contract rated it about twenty-two percent harder to read, a statistically significant difference, and about thirteen percent harder to understand, a difference that was not statistically significant. This is significant because conspicuousness should facilitate access to information rather than impose perceptual friction. It connects to readability, subjective difficulty, typography, cognitive fluency, user experience, and informed consent.

**Evidence anchor:** The study reports a significant twenty-two-percent increase in perceived reading difficulty and a nonsignificant thirteen-percent increase in perceived comprehension difficulty.

**Boundary:** Difficulty ratings are self-reported, and the smaller understanding difference is statistically uncertain.

**Connections:** readability; subjective difficulty; typography; cognitive fluency; user experience; informed consent

**Record:** `ssrn-3519630-p18` · `machine-drafted-source-checked`

## 19. All-caps readers took about thirteen percent longer to finish, a directionally important but statistically inconclusive result

**Location:** Exploratory Study: Reading Time, printed pp. 38-39 (PDF pp. 38-39)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 38–39, that capitalization may slow readers rather than promote efficient review. Participants assigned all-caps took about 94.7 seconds on average, compared with 83.4 seconds for low-caps, but large variance prevented a statistically confident conclusion. The authors note that reading-time data may also mix careful reading with skimming or disengagement. This is significant because slower processing would compound the costs of already unread standard forms. It connects to reading speed, processing fluency, measurement error, skimming, transaction costs, and accessibility.

**Evidence anchor:** Mean completion time was 94.7 seconds for all-caps and 83.4 seconds for low-caps, an approximately thirteen-percent nonsignificant gap.

**Boundary:** The difference is not statistically significant, the variance is large, and elapsed time does not reveal whether a participant read carefully or disengaged.

**Connections:** reading speed; processing fluency; measurement error; skimming; transaction costs; accessibility

**Record:** `ssrn-3519630-p19` · `machine-drafted-source-checked`

## 20. A four-arm experiment compares ordinary type, all-caps, a boxed clause, and bolding only the key sentence in a recreational-vehicle liability disclaimer

**Location:** Exploratory Study: Alternative Interventions, printed pp. 39-41 (PDF pp. 39-41)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 39–41, that comparative design can ask whether capitalization fails because disclosure design is futile or because lawyers chose the wrong intervention. The 241-person experiment presents a recreational-vehicle rental agreement and randomizes the liability disclaimer among low-caps, all-caps, a box, and selective bolding of the dispositive sentence. Questions separately test on-road and off-road injuries. This is significant because it benchmarks the legal default against plausible alternatives in the same text. It connects to comparative treatment design, liability waivers, selective emphasis, visual borders, comprehension testing, and legal design.

**Evidence anchor:** The paper details four randomized formatting conditions and two outcome questions concerning different accident locations.

**Boundary:** The study tests one recreational-vehicle contract and selected implementations of each visual treatment, not every possible version.

**Connections:** comparative treatment design; liability waivers; selective emphasis; visual borders; comprehension testing; legal design

**Record:** `ssrn-3519630-p20` · `machine-drafted-source-checked`

## 21. Selectively bolding the key sentence produced the lowest error rates and substantially outperformed ordinary text on the main liability question

**Location:** Exploratory Study: Alternative Interventions, printed pp. 41-43 (PDF pp. 41-43)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 41–43, that targeted bolding is a promising alternative to blanket capitalization. On the on-road-liability question, the bold condition produced a twenty-seven-percent error rate, compared with fifty-seven percent for low-caps, fifty-two percent for all-caps, and forty-eight percent for the box. On the off-road question, bold again had the lowest error rate—thirteen percent, versus twenty-one, thirty, and twenty-four percent—although that comparison was not statistically significant. This is significant because a simple design change can materially improve comprehension where the canonical method cannot. It connects to selective salience, information hierarchy, liability disclaimers, effect size, visual design, and consumer understanding.

**Evidence anchor:** The bold condition had the lowest error rates on both questions and a statistically significant advantage on the principal on-road question.

**Boundary:** Only the first outcome yields a statistically significant advantage, and the authors do not claim that bolding will work across every contract or population.

**Connections:** selective salience; information hierarchy; liability disclaimers; effect size; visual design; consumer understanding

**Record:** `ssrn-3519630-p21` · `machine-drafted-source-checked`

## 22. The bolding result is an existence proof that contract presentation can improve understanding, not a universal prescription for bold text

**Location:** Exploratory Study: Alternative Interventions, printed pp. 41-44 (PDF pp. 41-44)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 41–44, that selective bolding should be treated as an existence theorem rather than a finished regulatory formula. It shows that readers can respond to a well-targeted intervention, which makes the persistent failure of blanket capitals more informative. Yet only one bolding comparison was statistically significant, other attempted interventions did not succeed, and effective design is likely contextual. This is significant because it preserves the possibility of useful disclosure without converting one experiment into new drafting lore. It connects to proof of concept, replication, context-sensitive design, behavioral interventions, external validity, and regulatory humility.

**Evidence anchor:** The manuscript characterizes the bold result as evidence that an effective intervention exists while noting failed alternatives and limited statistical precision.

**Boundary:** The study does not establish that bolding is optimal, that its effect will replicate, or that it works for different clauses and reader populations.

**Connections:** proof of concept; replication; context-sensitive design; behavioral interventions; external validity; regulatory humility

**Record:** `ssrn-3519630-p22` · `machine-drafted-source-checked`

## 23. Judicial reliance on all-caps has concrete high-stakes consequences, including denial of personal-injury recovery and enforcement of class-action waivers

**Location:** Policy Implications, printed pp. 45-47 (PDF pp. 45-47)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 45–47, that typography is not a harmless drafting eccentricity. Courts have invoked capitalized text when enforcing releases that foreclose compensation for serious injuries and clauses that prevent collective litigation. A person’s substantive legal position may therefore turn on a convention that the experiments find ineffective or harmful. This is significant because the costs of a false inference of notice are borne in lost remedies, not just awkward reading. It connects to exculpatory clauses, personal injury, class-action waivers, procedural rights, meaningful assent, and access to justice.

**Evidence anchor:** The policy discussion identifies cases in which capitalized clauses supported denial of compensation or enforcement of procedural waivers.

**Boundary:** The cited cases illustrate consequences but do not measure how often capitalization alone changes outcomes across all jurisdictions.

**Connections:** exculpatory clauses; personal injury; class-action waivers; procedural rights; meaningful assent; access to justice

**Record:** `ssrn-3519630-p23` · `machine-drafted-source-checked`

## 24. A formal all-caps safe haven gives firms incentives to satisfy the appearance of disclosure while preserving consumer ignorance

**Location:** Policy Implications, printed pp. 45-48 (PDF pp. 45-48)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 45–48, that the legal convention can align form with strategic nondisclosure. A firm obtains an enforcement advantage by capitalizing onerous language even if the format does not improve comprehension and may make reading harder. The contrast with marketing is revealing: when firms want audiences to understand favorable claims, they generally avoid long capitalized blocks. This is significant because a rule intended to discipline drafters may instead give them a cheap way to manufacture consent. It connects to safe harbors, strategic compliance, shrouded attributes, firm incentives, revealed preference, and consumer exploitation.

**Evidence anchor:** The authors combine firms’ legal gains from capitalization with its absence from persuasive marketing to argue that the convention can facilitate strategic formal compliance.

**Boundary:** The incentive account is inferred from legal structure, usage patterns, and experimental evidence; it does not directly observe firms’ internal motives.

**Connections:** safe harbors; strategic compliance; shrouded attributes; firm incentives; revealed preference; consumer exploitation

**Record:** `ssrn-3519630-p24` · `machine-drafted-source-checked`

## 25. Courts and lawmakers should stop assigning all-caps any special weight in deciding whether a term was conspicuous or knowingly accepted

**Location:** Policy Implications, printed pp. 45-50 (PDF pp. 45-50)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 45–50, that the legal system should abandon its reliance on capitalization. The case rests on mutually reinforcing considerations: no affirmative empirical foundation supported adoption; prior reading research predicts difficulty; the authors’ main and exploratory studies show no benefit and possible harm; industry avoids capitals when comprehension matters; and erroneous enforcement can impose grave losses. This is significant because the remedy is not merely better drafting advice but removal of an evidentiary shortcut from contract doctrine. It connects to doctrinal abandonment, conspicuousness, evidentiary weight, judicial notice, statutory reform, and contract legitimacy.

**Evidence anchor:** The authors synthesize historical, experimental, industry, and error-cost arguments in favor of giving all-caps no doctrinal significance.

**Boundary:** Removing special weight from capitals does not itself determine whether a term is enforceable under every other doctrine or factual circumstance.

**Connections:** doctrinal abandonment; conspicuousness; evidentiary weight; judicial notice; statutory reform; contract legitimacy

**Record:** `ssrn-3519630-p25` · `machine-drafted-source-checked`

## 26. Replacing all-caps with another universal formatting safe haven would recreate the same vulnerability to manipulation and contextual mismatch

**Location:** Beyond a New Mechanical Safe Haven, printed pp. 50-51 (PDF pp. 50-51)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 50–51, that reform should not simply canonize a new font treatment. Any mechanical rule can be gamed by surrounding the nominally emphasized term with distracting text, overusing the required signal, or ignoring how the term functions in the document and transaction. Conspicuousness is relational and contextual, not an intrinsic property of a single typographic command. This is significant because it cautions against rebuilding the failed formalism under a different visual label. It connects to rules and standards, regulatory arbitrage, information overload, contextual interpretation, salience dilution, and legal formalism.

**Evidence anchor:** The manuscript explains how mechanical compliance can be diluted or manipulated and why effective notice depends on the surrounding presentation.

**Boundary:** Context-sensitive review may increase uncertainty and adjudication costs compared with a bright-line formatting rule.

**Connections:** rules and standards; regulatory arbitrage; information overload; contextual interpretation; salience dilution; legal formalism

**Record:** `ssrn-3519630-p26` · `machine-drafted-source-checked`

## 27. Contract design should separate a salient notice marker or heading from a readable low-caps explanation of the legal term

**Location:** Beyond a New Mechanical Safe Haven, printed pp. 51-52 (PDF pp. 51-52)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 51–52, that two distinct disclosure functions need not be forced into the same typography. A brief heading, icon, or selectively bold sentence can alert a reader that an important issue exists, while the operative body remains in ordinary type that is easier to process. Blanket capitalization tries to supply both notice and explanation and may accomplish neither. This is significant because separating attention capture from substantive communication opens a more coherent design space. It connects to layered notice, information architecture, headings, icons, plain language, and progressive disclosure.

**Evidence anchor:** The authors suggest using a distinct attention signal while preserving readable ordinary-case text for the actual explanation.

**Boundary:** The paper offers a design principle rather than validating a particular combination of markers and explanatory text across settings.

**Connections:** layered notice; information architecture; headings; icons; plain language; progressive disclosure

**Record:** `ssrn-3519630-p27` · `machine-drafted-source-checked`

## 28. Failure of the leading conspicuousness device supports skepticism about disclosure mandates, while success of selective bolding shows that disclosure is not necessarily futile

**Location:** Implications for Disclosure, printed pp. 52-53 (PDF pp. 52-53)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 52–53, that the evidence cuts in two directions in the broader debate over mandatory disclosure. The entrenched intervention fails despite extraordinary legal confidence, which reinforces concern about untested disclosure mandates. But selective bolding improves performance in one study, indicating that presentation can sometimes help readers. This is significant because the right lesson is empirical discrimination among interventions, not categorical faith in or rejection of disclosure. It connects to mandatory disclosure, behavioral law and economics, policy experimentation, evidence-based regulation, debiasing, and consumer information.

**Evidence anchor:** The authors contrast the repeated failure of capitalization with the more encouraging result for targeted bolding.

**Boundary:** The evidence base remains limited to several experiments and cannot resolve the effectiveness of disclosure regimes generally.

**Connections:** mandatory disclosure; behavioral law and economics; policy experimentation; evidence-based regulation; debiasing; consumer information

**Record:** `ssrn-3519630-p28` · `machine-drafted-source-checked`

## 29. Onerous terms should be presumptively unenforceable unless the drafting firm can show that its disclosure method effectively communicated them

**Location:** Performance-Based Conspicuousness, printed pp. 53-54 (PDF pp. 53-54)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 53–54, that conspicuousness should become performance-based. Instead of courts assuming that a prescribed format creates notice, a firm seeking to enforce an onerous term should bear the burden of showing that its chosen presentation actually communicated the term effectively. The default would be nonenforcement when that proof is absent. This is significant because it replaces a typographic proxy with an outcome-oriented accountability rule. It connects to burden shifting, performance standards, unconscionability, informed consent, evidentiary proof, and default rules.

**Evidence anchor:** The policy section proposes presumptive nonenforcement of onerous terms unless firms demonstrate that their disclosure technique works.

**Boundary:** The paper does not fully specify acceptable study designs, thresholds, litigation procedures, or costs for proving effective communication in every case.

**Connections:** burden shifting; performance standards; unconscionability; informed consent; evidentiary proof; default rules

**Record:** `ssrn-3519630-p29` · `machine-drafted-source-checked`

## 30. A performance rule would use firms’ existing A/B-testing capacity to generate innovation and cumulative knowledge about effective contract communication

**Location:** Performance-Based Conspicuousness, printed pp. 53-55 (PDF pp. 53-55)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 53–55, that firms are better positioned than courts or consumers to test competing presentations at scale. Companies routinely A/B test marketing and interface choices; if enforceability required effective communication, that capacity could be redirected toward contract terms. Successful practices could diffuse, while drafters would have skin in the game rather than receiving a free safe harbor. This is significant because the reform harnesses private experimentation to build an evidence base that doctrine currently lacks. It connects to A/B testing, regulatory incentives, innovation, information production, compliance design, and organizational learning.

**Evidence anchor:** The authors argue that firms already test communications and would develop and share better practices if effectiveness affected enforceability.

**Boundary:** Testing capacity varies across firms, and experimentation raises implementation, comparability, privacy, and distributional questions not fully developed here.

**Connections:** A/B testing; regulatory incentives; innovation; information production; compliance design; organizational learning

**Record:** `ssrn-3519630-p30` · `machine-drafted-source-checked`

## 31. Disclosure should be evaluated for vulnerable subgroups and may properly minimize misunderstanding rather than maximize average comprehension

**Location:** Performance-Based Conspicuousness, printed pp. 54-55 (PDF pp. 54-55)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 54–55, that average treatment effects are an inadequate target when an intervention harms identifiable readers. Because the primary study shows a large adverse effect among older participants, a disclosure method that modestly raises the mean could still be unacceptable if it imposes concentrated misunderstanding. Regulators might therefore minimize error across groups or set protective floors. This is significant because accessibility and distribution belong inside the definition of effective communication. It connects to heterogeneous effects, vulnerability, social welfare functions, distributional analysis, universal design, and anti-discrimination norms.

**Evidence anchor:** The authors use the older-reader result to argue for attention to heterogeneous performance and objectives that constrain misunderstanding among vulnerable groups.

**Boundary:** The paper identifies the normative choice but does not select a complete welfare function or precise subgroup-protection threshold.

**Connections:** heterogeneous effects; vulnerability; social welfare functions; distributional analysis; universal design; anti-discrimination norms

**Record:** `ssrn-3519630-p31` · `machine-drafted-source-checked`

## 32. Contract law should abandon untested all-caps formalism, demand proof from firms, and pursue targeted methods that genuinely help consumers understand consequential terms

**Location:** Conclusion, printed pp. 55-56 (PDF pp. 55-56)

Professor Yonathan A. Arbel and Andrew Toler claim, in “ALL-CAPS” on manuscript pages 55–56, that capitalization should lose its privileged role in contract enforcement. Across prevalence evidence, a large primary experiment, exploratory tests, reader reports, and an alternative-format study, capitals do not improve comprehension and may harm older readers, while targeted highlighting shows qualified promise. The authors would shift proof to firms and encourage empirically validated design. This is significant because it turns an examination of typography into a general program for evidence-based consent. It connects to contract modernization, consumer protection, empirical legal reform, legal design, performance regulation, and meaningful choice.

**Evidence anchor:** The conclusion calls for abandoning legal reliance on all-caps, shifting the burden of proof, and developing targeted evidence-backed alternatives.

**Boundary:** The proposed regime requires further work on proof standards, testing costs, implementation, and transferability across transactions.

**Connections:** contract modernization; consumer protection; empirical legal reform; legal design; performance regulation; meaningful choice

**Record:** `ssrn-3519630-p32` · `machine-drafted-source-checked`