Notes on (Huber et al. 1982) – Adding Asymmetrically Dominated Alternatives

Main Topic or Phenomenon Addressed

This paper examines how adding an asymmetrically dominated alternative (a “decoy”) to a choice set can systematically alter consumer preferences, specifically increasing the probability of choosing the item that dominates the decoy. This challenges fundamental assumptions about how consumers make choices when new alternatives are introduced to markets.

Theoretical Construct

Asymmetrically Dominated Alternative (Decoy): An alternative that is dominated by at least one item in the choice set but not by at least one other. The decoy is inferior to the “target” on all attributes but may be superior to the “competitor” on some attributes.

Regularity: A fundamental principle stating that the probability of choosing any item from set A cannot be less than the probability of choosing that same item from set B, where A is a subset of B. Formally: Pr(x;A) ≥ Pr(x;B) for all x ∈ A ⊂ B.

Similarity Hypothesis: The principle that a new alternative takes disproportionately more market share from products similar to it than from dissimilar products.

Key Findings

1. Regularity Violation: Adding asymmetrically dominated alternatives consistently violated regularity, with the target’s share increasing by an average of 9.2 percentage points across conditions.
2. Similarity Hypothesis Reversal: Contrary to the similarity hypothesis, decoys that were most similar to the target (range-increasing strategies) actually helped the target most, rather than hurting it.
3. Strategy Effectiveness Hierarchy:
   • Range-increasing strategies (R and R*): +13 percentage points
   • Range-frequency strategy (RF): +8 percentage points
   • Frequency strategy (F): +4 percentage points
4. Cross-Product Generalizability: Effects were consistent across six different product categories (cars, restaurants, beer, lotteries, film, TV sets).
5. Within vs. Between-Subject Effects: The decoy effect was stronger in between-subject comparisons (9 points) than within-subject (3 points), likely due to carryover effects.

Boundary Conditions and Moderators

Task Complexity: The authors note that increasing decision complexity (more alternatives or attributes) would likely reduce the decoy effect because:

• Dominance relationships become harder to detect
• Decision errors increase
• Cognitive limitations become more binding

Decoy Placement Strategy: The effectiveness of the decoy depends critically on its positioning:

• Range-extending decoys (extending the inferior dimension of the target) are most effective
• Frequency-increasing decoys (adding another level on the target’s superior dimension) are less effective
• Mixed strategies show intermediate effectiveness

Recognition of Dominance: The effect may be attenuated when dominance relationships are not readily apparent to consumers, particularly in the range-frequency condition where consumers must consider both dimensions simultaneously.

Building on Previous Work

This paper directly challenges core assumptions of established choice models:

Luce Choice Model (1959): Assumes proportional market share loss when new alternatives are added - this paper shows systematic violations through directional share gains.

Similarity-Based Models (Tversky 1972; Hauseman & Wise 1978): These models predict that similar alternatives hurt each other - this paper demonstrates the opposite can occur with dominated alternatives.

Previous Empirical Work: Earlier studies (Becker, DeGroot & Marshak 1963) found regularity held even when proportionality failed. This paper identifies specific conditions where regularity itself breaks down.

The paper extends theory by showing that the presence of dominated alternatives creates systematic distortions that previous models could not account for, requiring either model modifications or restricted applicability.

Major Theoretical Contribution

The paper demonstrates that fundamental axioms of rational choice theory can be systematically violated in predictable ways. It reveals that choice behavior is influenced not just by the attributes of alternatives themselves, but by the composition and structure of the entire choice set. This suggests that consumer preferences are more context-dependent and malleable than traditional economic and psychological models assumed.

The findings imply that choice models must account for how dominated alternatives can serve as “perceptual anchors” that shift the psychological representation of the choice environment, rather than treating all alternatives as equally relevant inputs to the decision process.

Major Managerial Implication

Companies can strategically increase demand for their products by introducing dominated “decoy” products that few customers will actually choose. The decoy’s primary function is to make the target product appear more attractive by comparison. Examples include:

• Retailers adding expensive items to make mid-priced options seem reasonable
• Service providers offering deliberately unattractive packages to steer customers toward preferred options
• Restaurants using high-priced entrees to increase sales of moderately-priced items

This represents a counterintuitive strategy where profitability can be increased by adding products that generate minimal direct sales but enhance the appeal of existing offerings.

Unexplored Theoretical Factors

Several potential moderators remain unexplored:

Individual Differences: Consumer expertise, need for cognition, maximizing vs. satisficing tendencies, and decision-making styles could influence susceptibility to decoy effects.

Temporal Factors: Time pressure, decision urgency, and the timing of decoy introduction relative to initial consideration sets.

Social Context: Presence of others, social proof, and cultural variations in decision-making norms.

Product Characteristics: Hedonic vs. utilitarian products, involvement level, purchase frequency, and attribute importance variations.

Choice Set Properties: Number of alternatives, attribute correlation patterns, and the presence of multiple decoys or dominated alternatives.

Cognitive Load: Working memory constraints, distraction, and mental fatigue during decision-making.

Framing and Presentation: Order effects, visual presentation format, and information display characteristics.

Reference

Huber, Joel, John W. Payne, and Christopher Puto (1982), “Adding Asymmetrically Dominated Alternatives: Violations of Regularity and the Similarity Hypothesis,” Journal of Consumer Research, 9 (1), 90–98.