Chapter 33: Reversals

Type: Chapter 33

Category: Psychology

Difficulty: Intermediate

Reading time: 6 min

Core idea

Consider two legal settlements. Settlement A: victim gets $200,000. Settlement B: victim gets $100,000 plus the defendant must fund a $1M public safety program. When people evaluate these individually, they rate Settlement B more favorably — the public safety component feels meaningful. When they evaluate them jointly, Settlement A is preferred — $200,000 cash clearly dominates.

The preference reversal: joint evaluation produces one preference; individual evaluation produces another. Neither is simply “correct” — each mode of evaluation makes different features salient.

Kahneman uses this to make a broader point about how evaluation mode shapes what we notice and how we judge. Joint evaluation favors analytical comparison — which option is better on comparable dimensions? Single evaluation favors emotional response — how does this option feel? These are different questions, and different modes activate different cognitive processes.

Why it matters

Why reversals occur

In single evaluation (evaluating one option without comparison), System 1 produces an overall affective response. The response is influenced by the most emotionally salient feature of the option — which may not be the feature most relevant to the actual decision.

In joint evaluation (evaluating two or more options simultaneously), System 2 becomes more active. Comparison forces a common dimension of assessment — which feature explicitly differs between options? The comparison highlights whatever is quantitatively different and compressible into a scale, which may be a different feature than what dominated in single evaluation.

This means: different evaluation modes can rank the same options differently, not because of inconsistency or error, but because each mode genuinely highlights different features.

Evaluation mode in policy and law

Reversals have significant consequences for legal and policy decisions. Legal damages often involve single evaluation (a jury deliberates about one plaintiff’s harm), while policy analysis typically involves joint evaluation (comparing programs on common metrics). The same harm or benefit can appear larger or smaller depending on the mode.

In Kahneman’s example, when environmental damage cases were evaluated jointly, the amount of damages awarded tracked the number of birds harmed (more birds = more damages). When evaluated singly, the size of the damages was influenced by how emotionally powerful the individual case was — independent of the number of birds.

Neither mode is always right

There is no single correct mode of evaluation. Joint evaluation imposes common dimensions that may not be appropriate for all decisions — it tends to overweight quantitative comparisons that can be scaled, and underweight qualitative differences that resist direct comparison. Single evaluation may overweight emotional salience and underweight less vivid but more consequential features.

Author’s argument: A society that wants rational policy should consider what mode of evaluation is most appropriate for each type of decision — not just whether the evaluation was done carefully.

Key takeaways

Key takeaways

• Preference reversals: the same options evaluated jointly vs. separately can produce reversed preferences — not due to inconsistency, but because different evaluation modes highlight different features.
• Single evaluation activates System 1: emotional response, affective salience, and non-comparative intuition drive the judgment.
• Joint evaluation activates System 2: comparison, quantitative scaling, and feature-by-feature analysis dominate.
• The settlement example: Settlement B (cash + safety program) preferred in single evaluation; Settlement A (cash only) preferred in joint evaluation — the comparison dimension changes.
• Legal and policy implications: damage awards and policy choices made in single evaluation mode may not match preferences under joint evaluation of the same options.
• Neither mode is always right: joint evaluation over-weights quantifiable comparisons; single evaluation over-weights emotional salience. The choice of mode is itself a value judgment.

Mental model

click to zoom

Preference reversal — how evaluation mode changes which features are salient and which option is preferred.

Read it as: Single evaluation activates emotional, affective responses to individual options — features that feel meaningful dominate. Joint evaluation activates comparative, analytical responses — features that can be quantitatively compared dominate. When these processes highlight different features, preferences reverse. The reversal is not a mistake; it is the expected output of two genuinely different evaluation processes applied to the same options.

Practical application

Designing decision processes

When designing how a group will evaluate options — vendor selection, policy choice, performance review — the choice between joint and single evaluation is consequential. Joint evaluation produces more analytical, comparison-based decisions; single evaluation produces more affective, intuition-based ones. Neither is universally correct. Match the evaluation mode to the decision type: use joint for decisions where explicit comparison is appropriate; use single for decisions where overall impression matters.

Practical implications:

• Salary offers: presenting a salary offer in isolation activates emotional response to the offer itself. Presenting it alongside a comparison set (“market rate” or “competing offer”) activates joint evaluation and changes what is salient.
• Philanthropic appeals: individual identifiable victims produce stronger emotional responses (and more donations) than statistical descriptions of equivalent harm — single evaluation activates emotional salience, statistical framing activates joint evaluation.
• Product pricing: presenting a product alone activates the affective value of the product; presenting it in a product lineup activates price-quality comparisons. Luxury brands often sell in single-evaluation contexts to avoid unflattering price comparisons.

Example

A nonprofit solicits donations for two programs: Program A helps 1,000 children in one country; Program B helps 2,000 children in two countries. When donors evaluate each program individually, Program A receives roughly equal donations to Program B — the emotional response to “1,000 children helped” feels just as compelling as “2,000 children helped” in isolation. When donors evaluate both programs jointly, Program B receives significantly more — the direct comparison of 1,000 vs. 2,000 makes the scale immediately visible.

The single-evaluation mode responded to the emotional salience of helping children; the joint-evaluation mode responded to the quantitative comparison. Which mode should the nonprofit use? It depends on whether scale or story is the more relevant dimension for their decision context.

Related lessons

Related

Chapter 34: Frames and Reality How framing changes which aspects of a situation are salient — the broader mechanism of which evaluation mode is one specific instance. Chapter 26: Prospect Theory The reference-dependent value function that makes different reference points highlight different aspects of options. Framing Effects The full concept — how the presentation of information changes which features are salient and which options are preferred.