Skip to content

Stated–Revealed Preference Gap

Version
v5 · 2026-09-28 · History
Prime #
1464
Domain group
Social Sciences
Origin domain
Economics & Finance
Subdomains
Preference Measurement, Health Behavior, Organizational Behavior, Product Research → Economics & Finance
Also from
Political Science
Aliases
Stated vs Revealed Preference Gap, Say Do Gap, Report Choice Gap
Related primes
Stated Preference, Revealed Preference, Preference, Knowledge-Action Gap

Core Idea

The stated–revealed preference gap is a systematic divergence between a valuation an agent reports and the valuation an analyst would infer from the agent's behavior under a real choice opportunity. The report may say that privacy, preventive health, retirement saving, sustainable consumption, or a proposed product matters; the observed choice may point elsewhere.

The gap does not prove hypocrisy, irrationality, or the superiority of action over speech. Both sides are measurements produced under conditions. A stated preference is elicited through wording, response options, incentives, and social setting. A revealed preference is inferred from behavior under a particular menu, cost structure, default, information state, and time horizon. Reports can be cheap and socially shaped; choices can be constrained or deliberately engineered. The divergence is therefore evidence about the two access channels and contexts, not a license to pronounce either output the agent's true valuation without further testing.

This makes the prime a triangulation structure. Align the alternatives represented in both records, verify that the same agent and focal valuation are being compared, measure the direction and persistence of the mismatch, then intervene on each channel separately. The way the gap moves reveals which distortions are doing work.

How would you explain it like I'm…

Say vs. Do Puzzle

Sometimes a kid says 'I love carrots!' but at lunch picks the cookie. That doesn't mean the kid was fibbing. Maybe the cookie was closer, or everyone was watching when they talked about carrots. Grown-ups look at both the saying and the picking to figure out what's going on.

When Words and Choices Don't Match

The Stated–Revealed Preference Gap is when what someone says they care about doesn't match what they actually choose. A person might say saving money matters a lot, but then spend it anyway. That doesn't automatically mean they're lying or that actions are the 'real' truth. The question they were asked can shape what they say, and the situation they're in can shape what they do. Looking at how the gap changes helps figure out what's really going on.

The Say–Do Preference Gap

The Stated–Revealed Preference Gap is a consistent mismatch between what people report valuing and what their behavior suggests they value when facing a real choice — for example, people saying privacy matters a lot but clicking 'accept all'. The gap doesn't prove hypocrisy or that actions speak louder than words. Both are measurements taken under particular conditions: what people say is shaped by wording, answer options, and social pressure, while what people do is shaped by prices, defaults, available information, and time pressure. Either side can be distorted. The useful move is to treat the gap as a clue: check that you're comparing the same person and the same choice, measure which way it points and whether it lasts, then change one side at a time to see which distortion is causing it.

 

The Stated–Revealed Preference Gap is a systematic divergence between the valuation an agent reports and the valuation an analyst would infer from that agent's behavior under a real choice opportunity. Typical cases: people say privacy, preventive health, retirement saving, or sustainable consumption matters, while observed choices point elsewhere. The gap does not establish hypocrisy, irrationality, or the priority of action over speech. Both sides are condition-dependent measurements: stated preferences are elicited through wording, response options, incentives, and social setting; revealed preferences are inferred from behavior under a specific menu, cost structure, default, information state, and time horizon. Reports can be cheap and socially shaped; choices can be constrained or deliberately engineered by choice architecture. The divergence is therefore evidence about the two access channels, not a warrant to crown either as the true valuation. Methodologically it is a triangulation structure: align the alternatives in both records, confirm the same agent and focal valuation are compared, measure the gap's direction and persistence, and intervene on each channel separately to see which distortions move it.

Structural Signature

  • The preference-bearing agent — an individual or collective whose ordering is sought.
  • The focal alternatives or action — sufficiently aligned across report and choice to support comparison.
  • The stated estimate — an ordering or intention inferred from a solicited report.
  • The revealed estimate — an ordering inferred from behavior under an actual menu and consequence structure.
  • The shared comparison frame — common alternatives, time horizon, and evaluative target.
  • The systematic divergence — rankings, choices, or magnitudes differ beyond random error.
  • The report-channel distortions — wording, hypothetical bias, social desirability, articulation, or strategic response.
  • The choice-context distortions — constraints, missing options, defaults, friction, salience, timing, or present bias.
  • The triangulation test — vary one channel or context while holding the other as stable as possible and observe which estimate moves.

What It Is Not

  • Not hypocrisy by definition. The report can be sincere while behavior is contextually defeated or constrained.
  • Not proof that behavior reveals the true preference. A defaulted or high-friction choice may say more about the environment than the agent.
  • Not proof that the report is the true preference. A low-cost hypothetical answer can be socially desirable or poorly predicted.
  • Not Knowledge-Action Gap. That narrower prime requires correct knowledge, sincere intention, an appropriate action, and failed translation. This gap requires only aligned report and choice estimates that diverge.
  • Not Elicitation Channel Contribution alone. A channel effect can distort one record without any opposing behavioral measure.
  • Not temporal preference reversal. Time can cause the divergence, but the general gap also arises from constraints, framing, information, and measurement.

Broad Use

  • Digital privacy: concern reported in surveys versus permissive data disclosure.
  • Health behavior: stated treatment, exercise, diet, or vaccination preferences versus adherence.
  • Retirement and finance: desired saving versus enrollment, contribution, or portfolio choices.
  • Politics: declared priorities versus turnout, voting, donation, or migration behavior.
  • Sustainability: environmental concern versus travel, energy, and purchasing decisions.
  • Organizations: espoused safety or quality priorities versus enacted routines and resource allocation.
  • Product research: claimed interest versus sign-up, payment, adoption, retention, and use.

Clarity

The prime clarifies that “preference” is not directly observed in either channel. Analysts infer it from representations. A report is a representation produced by elicitation; behavior is a representation produced by a choice environment. Their disagreement is thus not an embarrassment to discard but a diagnostic object. It can reveal which contexts activate or suppress a valuation and whether the alternatives were truly comparable.

It also separates the general pattern from named domain instances. Privacy Paradox is the digital-disclosure species. Intention-behavior gaps in health, under-saving in retirement, and survey-versus-vote differences are co-instances of the more general report-choice divergence, not metaphorical “privacy paradoxes.”

Manages Complexity

The gap organizes a sprawling list of say-do inconsistencies into a two-channel audit.

  1. Align the object being valued and the time horizon.
  2. Audit the stated side for hypothetical, social, strategic, and instrument effects.
  3. Audit the revealed side for menu restrictions, prices, defaults, friction, information, and timing.
  4. Intervene on one side at a time.
  5. Treat residual disagreement as evidence of context-dependent or unstable valuation rather than forcing a single answer.

This replaces moral diagnosis with a tractable measurement and design problem.

Abstract Reasoning

Let \(P_s\) be an ordering inferred from a stated-preference channel \(C_s\), and \(P_r\) an ordering inferred from behavior in choice context \(C_r\). A gap exists when an aligned comparison function \(D(P_s,P_r)\) remains nonzero across repeated measurements or materially changes the predicted choice.

The analyst does not observe a context-free \(P\). Instead,

\[ R_s=f(P,C_s), \qquad R_r=g(P,C_r). \]

Identification proceeds by perturbing \(C_s\) and \(C_r\). If wording changes move \(R_s\) while the choice context is stable, report-channel contribution is implicated. If reversing a default or equalizing friction moves \(R_r\) toward the report, the earlier behavior was not a clean preference readout. If the mismatch persists under both controls, context-dependent preferences, self-prediction error, or multiple evaluative systems become more plausible.

Knowledge Transfer

The transfer is exact across human and organizational domains: obtain a report, observe a consequential choice, align the alternatives, measure the mismatch, audit both channels, and vary the context. Privacy researchers contribute consent-architecture tests; health psychology contributes intention-behavior and adherence designs; retirement economics contributes default reversals; product research contributes staged commitment ladders. These methods transfer because they manipulate the same report-versus-choice roles.

Examples

Retirement enrollment

Employees report that retirement saving is important and endorse a target contribution, yet many remain unenrolled when participation requires active paperwork. Automatic enrollment sharply changes behavior without changing the available plan. The result shows that the earlier non-enrollment was partly a default- and friction-shaped revealed estimate, not a transparent rejection of saving.

Product adoption

Survey respondents say they would pay for a feature, but few purchase when it launches. Before declaring the survey false, the analyst aligns price, implementation, timing, switching cost, and available substitutes. A refundable deposit or pre-order inserted at research time raises the report's consequence level and tests whether the gap came from hypothetical bias.

Structural Tensions

  • Which channel deserves priority: costly behavior can be constrained; sincere reports can be hypothetical.
  • Stable versus context-constructed preference: the gap may expose distortion or show that no single context-free ordering exists.
  • Correction versus manipulation: changing defaults to align behavior with reports can serve or override the agent.
  • Alignment versus false comparison: report and choice may concern different time horizons, bundles, or information states.
  • Individual versus aggregate gap: population averages can diverge even when different people generate the two records.

Structural–Framed Character

The Stated–Revealed Preference Gap is a framed prime on the structural–framed spectrum: calling a mismatch a preference gap has already cast the situation in the vocabulary of valuation. The prime is a systematic divergence between the valuation an agent reports and the valuation an analyst would infer from that agent's behavior when a real choice was available.

It is unintelligible without human practice — it needs a reporting agent, a question, a real menu, and a consequential choice — and it inherits an economics lexicon (elicitation, revealed preference, incentive compatibility) that comes along wherever it is used: shoppers who rank sustainability highly and then buy on price, or survey respondents who endorse preventive screening and skip the appointment. Both sides are measurements made under conditions, so the prime declines the charge of hypocrisy; still, the espoused-versus-enacted framing carries some charge of its own. On the remaining diagnostics it reads framed.

Structural Core vs. Domain Accent

Remove privacy, health, politics, retirement, and product language. What remains is two access channels to one focal valuation, a shared comparison frame, a persistent mismatch, and an intervention that varies each channel's conditions. The domain accents determine which reports, choices, and frictions appear; they do not change the relational structure.

Relationships to Other Abstractions

Local relationship map for Stated–Revealed Preference GapParents appear above the current abstraction, mutual partners to the right, and children below. Node labels state whether each abstraction is prime or domain-specific; colors identify relation types.Stated–RevealedPreference GapPRIMEPrime abstraction: Comparison — includesComparisonPRIMEPrime abstraction: Revealed Preference — includesRevealedPreferencePRIMEPrime abstraction: Stated Preference — includesStatedPreferencePRIMEDomain-specific abstraction: Privacy Paradox — is a kind ofPrivacy ParadoxDOMAIN

Current abstraction Stated–Revealed Preference Gap Prime

Parents (3) — outgoing typed relations

  • Stated–Revealed Preference Gap includes Comparison Prime

    The gap contains a comparison that aligns the stated and revealed estimates over the same focal alternatives and reads off their systematic divergence.

  • Stated–Revealed Preference Gap includes Revealed Preference Prime

    The gap contains the choice-based estimate inferred from what the agent does when alternatives and consequences are present.

  • Stated–Revealed Preference Gap includes Stated Preference Prime

    The gap contains the report-based estimate whose output is compared with behavior under a real choice opportunity.

Children (1) — abstractions linked here by typed relations

  • Privacy Paradox Domain-specific is a kind of Stated–Revealed Preference Gap

    Privacy Paradox is the digital-data-disclosure species of the general gap between reported valuation and behavior under a consequential choice context.

Hierarchy paths (5) — routes to 4 parentless roots

  • Stated–Revealed Preference Gap → includes Comparison

Neighborhood in Abstraction Space

Stated–Revealed Preference Gap sits among the more crowded primes in the catalog (16th percentile for distinctiveness): several abstractions describe nearly the same structure, so a description that fits it will tend to fit its neighbors too — transporting it usually means disambiguating within this family rather than landing on it exactly.

Family — Framing & Preference Construction (14 primes)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25

Solution Archetypes

No catalogued solution archetypes reference this prime yet.