Scope Neglect

Explain why elicited valuations stay flat across hundredfold changes in magnitude — the count is replaced by a single affect-laden prototype, so the number is understood but never priced in.

Core Idea

Scope neglect is the empirical regularity, within judgment and decision-making, that the valuation a person attaches to an outcome is strikingly insensitive to the outcome's magnitude. Asked how much they would pay to save 2,000, 20,000, or 200,000 birds from oil-soaked deaths, respondents return willingness-to-pay figures statistically indistinguishable across a hundredfold range in scope. The number that should drive the answer — how many — slides past the judgment unmetabolized; what the valuation tracks instead is a single mental image, one oiled seabird, the prototype of the category. The affective response to that prototype, not the count behind it, sets the price.

The mechanism is representational and affective, not arithmetical. Encoded as a discrete, emotionally charged image, the cardinal quantity has no place to land: feelings are roughly bounded, and the distress evoked by one suffering animal versus ten thousand is far smaller than the ratio of the counts, because the imagery does not multiply. Kahneman and Frederick call this attribute substitution — a hard target attribute (scope-weighted value) is unconsciously replaced by an easier one (the prototype's affect), and the easy answer reported as if it answered the hard question.

Two finer commitments make it precise. Embedding: willingness to pay for a part (one region's lakes) is close to that for a superset (all the country's lakes), since both summon the same prototype — so valuations cannot be additive in scope. Evaluability: scope becomes usable only when magnitude is presented in comparable form, where the larger number does decision-relevant work; in single evaluation scope is nearly inert, in joint evaluation sensitivity partly returns. Scope neglect is thus not that people cannot count, but that under affect-heavy single-stimulus valuation the count is the wrong cue at the wrong time, and the prototype wins.

This is not general magnitude underweighting: Weber–Fechner compression keeps response logarithmic (still rising); scope neglect in its sharp form produces a flat curve — a decoupling specific to elicited valuation of affect-laden goods.

Structural Signature

Sig role-phrases:

• an affect-heavy outcome category summoned by a single prototype (S) — the oiled bird, the starving child stands in for the class.
• a valuation request in a magnitude-bearing currency (S) — willingness to pay/accept or a rating that ought to scale with scope.
• cardinal magnitude presented without comparative scaffolding (S) — a raw count offered in single, isolated evaluation.
• attribute substitution swaps scope-weighted value for prototype affect (V) — the hard target attribute replaced by the accessible affective one.
• the response curve comes out flat (or near-flat) in magnitude (V) — equal valuations across order-of-magnitude scope differences.
• part and superset valuations embed rather than add (V) — willingness to pay for the whole barely exceeds that for a contained part.
• scope sensitivity recoverable by joint / trade-off-forcing evaluation (V) — explicit comparison restores partial proportionality.

What It Is Not

• Not innumeracy. Respondents giving flat valuations can state that 200,000 is a hundred times 2,000. The defect is not comprehending the magnitude but recruiting it — the number is understood, then unused.
• Not measurement noise. A noisy elicitation gives an uncertain slope; scope neglect gives one reliably near zero, replicating across studies and designs — a named effect, not a complaint about survey quality.
• Not risk aversion or budget constraint. The flatness is not "$80 is all they can spare" — it appears at every income level and for non-monetary ratings with no budget at all.
• Not a law that valuation never responds to scope. Under joint evaluation, conjoint trade-offs, or repeated exposure, sensitivity partly returns; neglect is conditional on affect-heavy, single-stimulus, no-comparison elicitation.
• Not the part valued more than the whole. Ordinary embedding gives part ≈ whole; a vivid part valued above its superset is a stronger, separate pathology (sign-reversed scope effect).

Scope of Application

Scope neglect is a workhorse diagnostic across the valuation and risk subfields of JDM, and its reach is within that domain, not across substrates.

• Contingent valuation in environmental economics — the home turf. Surveys for protecting a species or watershed routinely fail the scope test (that willingness to pay rises with the quantity of the good); embedding (valuing "all the lakes" barely above "this lake") was the worry the NOAA Blue Ribbon Panel forced on the field, and scope sensitivity became its own validity criterion.
• Charitable giving / the identifiable-victim literature. Donations are flat, sometimes decreasing, in victim count when the appeal describes a group rather than one named individual — "psychophysical numbing."
• Risk perception and public-health communication. Risks of "thousands" versus "millions" elicit similar responses when the message centers a prototype victim; probability and magnitude are both neglected.
• Public-goods valuation and policy support. Support often fails to scale with stated reach: helping 100,000 draws support not much greater than helping 10,000, both keyed to the same prototype beneficiary.
• Jury awards. Punitive damages show outrage well-calibrated to blameworthiness but poorly to dollar magnitude (Kahneman, Schkade & Sunstein's "shared outrage, erratic awards").

Clarity

Naming scope neglect makes an otherwise-invisible failure mode legible. Without the label, a flat willingness-to-pay curve looks like a survey that "didn't work," inviting ad hoc explanations — bad sampling, hypothetical bias, protest responses. With it, the analyst asks a sharp structural question: is the elicitation delivering magnitude in a usable form, or harvesting prototype-keyed affect? That separates a genuine preference (the public really does value the larger program only slightly more) from a substitution artifact (the real preference masked because scope never entered) — and distinguishes three things survey practice blurs: comprehension (the respondent knows the number), valuation (the number drives the price), and expression (the format lets it show). Localizing the breakdown to the middle stage tells the designer that fixes aimed at the other two cannot move a slope whose problem is substitution.

Manages Complexity

The stated-preference literature is a sprawl of anomalies — embedding, part–whole bias, ordering effects, probability insensitivity, the single–joint disparity. Scope neglect compresses a large slice into one regularity: valuation under affect-heavy single evaluation is flat in scope. A meta-analyst can then treat "scope sensitivity" as a single measurable quantity — the slope of valuation on magnitude — rather than re-deriving the pathology for every good and survey, and reason from two parameters (how affect-heavy the representation, whether the format forces comparison) to predict when the anomaly appears — collapsing a high-dimensional "which surveys misbehave" problem into a low-dimensional structural one.

Abstract Reasoning

Scope neglect licenses three reasoning moves, all from the substitution mechanism. Diagnostic: facing a flat valuation, infer the respondent answered an easier affect-based question, and predict the co-occurring signature (embedding, single–joint disparity, probability insensitivity). Interventionist: to recover sensitivity, change the form of magnitude so the number acquires decision-relevant work — comparisons rather than absolute counts, conjoint trade-offs, a de-affectivized prototype; each is a prediction that the slope should steepen, and how much measures how much flatness was substitution. Boundary-drawing: a figure passing a scope test can enter cost–benefit analysis; one that fails cannot be read as scope-weighted value, and multiplying it by a large quantity yields a benefit estimate with no defensible relation to preference. The concept is a validity gate on a class of inferences.

Knowledge Transfer

Within JDM the diagnostic transfers cleanly and literally. Any elicitation asking for a numerical valuation of an affect-heavy outcome — environmental goods, charitable causes, health risks, public programs, legal damages — should be scope-tested, and the same fixes (joint evaluation, trade-off forcing, prototype de-emphasis) apply; the currency changes but the structure and remedies do not.

To adjacent fields transfer is real but partial, resting on shared mechanism. Marketing's price-evaluability work (Hsee) is the same attribute-substitution machinery on product attributes; behavioral public policy borrows scope neglect directly when warning that survey "public valuations" may be scope-insensitive.

Honesty requires marking where transfer becomes mere analogy. "Scope insensitivity" is invoked loosely for any response failing to scale with an input — evocative, but not the same mechanism (no elicited valuation, no affect-keyed prototype, no embedding to test). Outside elicited valuation of affect-laden outcomes the phenomenon dissolves into its neighbors (affect heuristic, prototype reasoning, identifiable-victim effects, logarithmic perception) or fails to reproduce: the within-domain transfer is substantive, the cross-domain one metaphor.

Examples

Canonical

The defining demonstration is Desvousges et al. (1992), popularized by Kahneman and Knetsch. Three independent groups were asked their willingness to pay to cover oil ponds so migratory birds would not die, told the measure would save 2,000, 20,000, or 200,000 birds per year. Mean willingness to pay was essentially constant — roughly $80, $78, $88 — far smaller than the hundredfold scope difference and not distinguishable from zero. The valuation was keyed to the prototype, one bird drowning in oil; the count, though understood, never entered the price. It became the textbook anchor for scope neglect and the spur for the scope test.

Mapped back: The prototype-summoned category is "birds dying in oil," represented by one drowning bird; the valuation request is willingness to pay; the cardinal magnitude without scaffolding is the bare count in isolated evaluation; attribute substitution swaps the rescue's scope-weighted value for prototype affect; the flat curve is the $80–$88 plateau across a 100× range.

Applied / practice

A wildlife nonprofit surveys willingness to pay for a migratory-bird program to feed a cost–benefit case. The first design asks three random groups, in isolation, for a single figure for "saves 20,000 / 100,000 / 500,000 birds annually." Means come back at $52, $55, $54 — flat. Multiplying $54 by 500,000 implies an enormous benefit, but the flatness is a red flag: the survey measured prototype affect, not scope-weighted value, so the aggregate would be an artifact. The team redesigns around joint evaluation and forced trade-offs ("A saves 100,000 birds for $40/year; B saves 500,000 for $150 — which, and at what price would you switch?"). In explicit comparison the larger magnitude finally does decision-relevant work, and willingness to pay recovers a near-proportional slope — which, not the flat single-evaluation figure, can defensibly enter the model.

Mapped back: The prototype-summoned category is again "birds saved"; isolated single evaluation produces the flat curve via substitution; the recoverability clause fires when the redesign supplies comparative scaffolding (joint evaluation + trade-off), steepening the slope — the structural prediction, confirmed in practice.

Structural Tensions

T1: Flat versus merely compressed. The sharp claim is a flat curve; the softer reality in many datasets is a compressed but positive slope, so whether a study counts as "neglect" rests on a partly conventional threshold — and a permissive reading lets ordinary diminishing sensitivity masquerade as the dramatic null. Diagnostic: is the fitted slope statistically flat, or merely compressed but positive — and is the "neglect" verdict resting on a permissive threshold?

T2: Mechanism versus measurement artifact. Critics (the CV defense) argue some insensitivity is an artifact of hypothetical elicitation — protest responses, warm glow — not genuine substitution. Is it a fact about judgment or bad instruments? The answer decides whether better incentives or better representation is the fix. Diagnostic: does the insensitivity survive real-incentive (non-hypothetical) elicitation, or does it dissolve once protest responses and warm glow are stripped out?

T3: Bias versus true preference. A flat slope can mean the respondent failed to weigh scope (a bias to correct) or genuinely values larger and smaller programs almost equally (a symbolic preference to respect, where quantity legitimately does not matter). The slope alone cannot tell the two apart; conflating them overrides real values or launders an artifact. Diagnostic: once scope is made salient, does the respondent revise toward sensitivity (a bias to correct), or hold the quantities near-equal on reflection (a symbolic preference to respect)?

T4: Single versus joint as the "correct" frame. Joint evaluation restores sensitivity, but real consumption is often single (one charity, one risk at a time), so the insensitive single figure may be ecologically valid and the joint figure an artifact of unnatural comparison. The central remedy (force comparison) presupposes an answer it cannot fully justify. Diagnostic: is the real decision made one option at a time (single evaluation ecologically valid), or side by side (joint the natural frame)?

T5: Affect as cause versus correlate. The standard story makes affect-heavy representation the cause, but insensitivity appears for affect-neutral goods and some affect-laden goods do show sensitivity. Is affect doing the work or co-varying with the real driver (low evaluability)? This bears on whether de-affectivizing the prototype is the right intervention. Diagnostic: does the insensitivity track affect specifically (fading for affect-neutral goods, present for affect-laden ones), or does it track low evaluability regardless of affect?

T6: Domain effect versus instance of general substitution. Scope neglect is studied as its own effect, yet its mechanism is a special case of the broader attribute-substitution account behind the affect heuristic, identifiable-victim effects, and prototype reasoning. Treating it as autonomous buys diagnostic precision but obscures that it is one face of a general structure — and resolving toward "general" is what keeps it below the prime bar. Diagnostic: resolve toward the general attribute-substitution structure (shared with the affect heuristic, identifiable-victim effects, prototype reasoning) when asking what travels cross-domain; toward the named effect when diagnosing a specific scope-insensitive valuation in situ.

Structural–Framed Character

Scope neglect sits toward the framed pole, decisively short of the structural pole of primes like logarithmic perception. Its content is not a substrate-neutral form but an empirical regularity about human judgment, bound to a human practice — the elicitation of valuations. Vocabulary travels weakly: "willingness to pay," "embedding," "scope test," "contingent valuation" are stated-preference economics' lexicon and lose their referents off-substrate. Human-practice-bound is high: it presupposes a judging agent with affect, a prototype, and an elicitation context — remove the valuer and there is nothing to neglect. Institutional origin is real: the effect is inseparable from the apparatus (CV surveys, the NOAA panel, the scope test) that defines it. Evaluative weight is low — described as a regularity, not praise or blame — though not value-free, since "neglect" implies the slope ought to have been steeper.

The one structural-looking feature — a flat response as a function of magnitude — tempts a cross-domain reading, but abstracted from valuation it is just compressive magnitude encoding; what makes it scope neglect (substitution of prototype affect for scope-weighted value under single-stimulus elicitation) is framed and practice-bound. The concept is mixed-but-framed: a thin structural skeleton wearing thick human-practice clothing.

Structural Core vs. Domain Accent

This section decides why scope neglect is a domain-specific abstraction and not a prime.

What could lift into a prime. Strip the valuation context and one skeleton remains: a system's output is insensitive to an input's magnitude over a wide range, because the input is re-represented in a bounded or compressed code before driving the output. That is substrate-portable — already in the catalog as logarithmic_perception_and_encoding, with the flat case its degenerate (saturated) end. The embedding relation also has a structural shadow: any system keyed to category membership rather than count shows part ≈ whole. These are the legitimate structural core — precisely the features scope neglect instantiates (next section), not what makes it distinctive.

What is irreducibly domain-bound. The substance the literature studies is the causal story: an affect-heavy prototype substitutes for the scope-weighted value under single-stimulus valuation, recoverable by joint evaluation, diagnosable by a scope test — every term presupposing a human valuer, an emotional response, imagery-summoned representation, an elicitation practice. The neglected "magnitude" is not a sensory intensity but a number the respondent was told; the "flatness" is not a transducer saturating but a judgment swapping one attribute for another.

Why this does not clear the prime bar. The bar requires the mechanism — not a surface resemblance — to recur across genuinely distinct substrates, and scope neglect fails twice. Its three apparent "domains" (environmental valuation, charitable giving, health-risk communication) are not distinct substrates but instances of one elicitation context; and where it reaches further the reach is analogy, not mechanism — an organization responding identically to small and large problems "neglects scope" only metaphorically, with no affect-keyed prototype and no embedding to test. What remains outside elicited valuation — "it didn't scale with the input" — is already covered literally by logarithmic_perception_and_encoding and the attribute-substitution family. It clears the domain-specific bar comfortably for JDM; it fails the prime bar's cross-domain mechanistic recurrence.

Instantiates / Related Primes

A domain instance of two confirmed catalog primes, with a third as a partial parent. (All three slugs verified to exist at prime_abstractions/v2/<slug>.md.)

• salience (confirmed). A salience phenomenon in the valuation arena: the prototype (the oiled bird, the named child) is the high-salience item that captures processing capacity before scope-weighted analysis runs. Salience's three commitments map directly — bottom-up (from the stimulus, not the goal of valuing scope), relational (dominance depends on framing), competitive (when the prototype wins, the count loses). Scope neglect is what salience does when the dominating item is an affect-laden prototype and the suppressed one is cardinal magnitude.

• logarithmic_perception_and_encoding (confirmed). Scope neglect instantiates this structure at its limiting case: log encoding compresses a wide-range magnitude so equal steps mean equal ratios; push compression to its ceiling (a bounded affective response barely growing as the count climbs orders of magnitude) and the curve goes concave to flat. The prime supplies "response is a compressed function of input magnitude"; scope neglect adds why it is total (affect, not logarithms).

• diminishing_returns (confirmed) — partial / weaker parent. The affective side resembles diminishing marginal sensitivity ("psychophysical numbing"); a looser fit, borrowing its shape (concavity, flattening) more than its production-function content.

I rejected anchoring as a parent (a neighbor, not a prime this instances — neglect persists with no anchor). The seed gestures at an "affect heuristic" prime; no such prime exists in v2 (nearest is the solution archetype affect_evidence_separation), so I do not assert it, though the affect heuristic is the relevant mechanism.

Relationships to Other Abstractions

Current abstraction Scope Neglect Domain-specific

Parents (1) — more general patterns this builds on

• Scope Neglect is part of Attribute Substitution Domain-specific

  Scope neglect contains attribute substitution as its defining internal judgment operation: scope-weighted value is silently replaced by prototype affect, producing the flat magnitude response.

Hierarchy path (1) — routes to 1 parentless root

• Scope Neglect → Attribute Substitution → Relevance Substitution

Not to Be Confused With

• Anchoring. Judgments pulled toward an initial number, then insufficiently adjusted — the stickiness of a reference value, versus scope neglect's inertness of a magnitude. Separable: neglect appears with no anchor supplied; anchoring appears for otherwise-weighted magnitudes. Tell them apart by whether an initial number distorts (anchoring) or the scale is ignored (neglect).

• Framing effect. Framing changes valuation by changing the description of a fixed outcome (gain vs loss, 90% survival vs 10% mortality); scope neglect holds framing fixed and varies the quantity, finding valuation unmoved. Sensitivity to how it is described versus insensitivity to how much it contains. (They interact — an affect-heavy frame deepens neglect — but the manipulated variable differs.)

• Affect heuristic. The general substitution of a good/bad feeling for analytic evaluation. Scope neglect is one specific, measurable consequence — what it produces when the substituted-away attribute is scope. The heuristic is the mechanism; scope neglect is the narrow diagnostic symptom.

• Denominator / base-rate / sample-size neglect. Statistical-reasoning biases: misjudging a ratio by ignoring the denominator, or ignoring base rates in inference. Scope neglect is the valuation cousin — the ignored quantity is beneficiaries in a willingness-to-pay judgment, not a denominator in a probability one; the structure rhymes, but the home task is valuation and the response is often flat rather than systematically miscalculated.

• Identifiable-victim effect. The closest neighbor, arguably an upstream cause: making one victim identifiable strengthens the prototype that drives the substitution. It differs in being specifically one named individual versus many, whereas scope neglect is the general flatness across magnitude. (Note the catalog prime identifiability is a different concept — parameter recoverability in statistics — not the identifiable-victim effect, despite the shared word.)

References

Light pass; real citations only, gaps marked honestly.

• Desvousges, William H., et al. Measuring Nonuse Damages Using Contingent Valuation. RTI Monograph 92-1 (1992). The 2,000 / 20,000 / 200,000 oiled-birds experiment. (Reissued 2010, RTI Press — DOI not verified.)
• Kahneman, Daniel, and Jack L. Knetsch. "Valuing Public Goods: The Purchase of Moral Satisfaction." J. Environmental Economics and Management 22, no. 1 (1992): 57–70. Embedding; "purchase of moral satisfaction."
• Kahneman, Daniel, Ilana Ritov, and David Schkade. "Economic Preferences or Attitude Expressions?" J. Risk and Uncertainty 19 (1999): 203–235. Valuations as attitude expressions.
• Frederick, Shane, and Baruch Fischhoff. "Scope (In)sensitivity in Elicited Valuations." Risk, Decision and Policy 3, no. 2 (1998): 109–123. Names and consolidates the phenomenon.
• Kahneman, Daniel, and Shane Frederick. "Representativeness Revisited: Attribute Substitution in Intuitive Judgment." In Heuristics and Biases, eds. Gilovich, Griffin & Kahneman, 49–81. Cambridge Univ. Press, 2002. (Chapter; no DOI.)
• Hsee, Christopher K. "The Evaluability Hypothesis…" OBHDP 67, no. 3 (1996): 247–257. The joint-versus-separate reversal.
• Slovic, Paul. "'If I Look at the Mass I Will Never Act': Psychic Numbing and Genocide." Judgment and Decision Making 2, no. 2 (2007): 79–95. The numbing / giving face.
• Kahneman, Daniel, David Schkade, and Cass R. Sunstein. "Shared Outrage and Erratic Awards." J. Risk and Uncertainty 16, no. 1 (1998): 49–86. Extends to jury awards.
• Gap: the NOAA "Blue Ribbon" panel report (Arrow, Solow, et al., Federal Register 58, 1993) is cited from memory; exact reference not verified.

Neighborhood in Abstraction Space

Scope Neglect sits in a sparse region of the domain-specific corpus (97^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Unclustered & Miscellaneous (309 abstractions)

Nearest neighbors

• IKEA Effect — 0.80
• Attribute Substitution — 0.80
• Halo Effect — 0.79
• Engel curve — 0.79
• Focusing Effect (Focusing Illusion) — 0.79

Computed from structural-signature embeddings · 2026-07-12