On this page
A buyer says that a product is worth 120. The same buyer chooses a 90-priced alternative when budget, access, implementation risk, and a buying committee enter the decision. The two observations are not contradictory. They describe different objects.
Revealed preference is an inference about preference from an observed choice under a declared choice set and constraint structure. It is not a direct measurement of willingness to pay.
The conjoint article owns hypothetical multi-attribute choice. This page owns the record around a choice that actually occurred and the limits on what that choice can reveal.
What is revealed preference?
Separate the evidence objects:
| Object | Observation | Boundary |
|---|---|---|
| Stated preference | What someone says they prefer or would buy | Prompt, wording, hypothetical context |
| Purchase intent | A declared likelihood or willingness | No transaction is required |
| Choice data | Selection from a shown or available alternative set | Alternatives, prices, timing, and access matter |
| Transaction | An executed exchange | Budget, authority, fulfillment, and terms remain relevant |
| Revealed preference | An interpretation of observed choice | It is an inference under assumptions |
| Willingness to pay | Maximum value under a defined decision frame | It is not observed merely because a choice occurred |
Table 1What is revealed preference?
Source: Table from this essay. Sources and interpretation are given in the article.
The final two rows are especially easy to collapse. Choosing A over B at the prices shown can support a relative choice statement. It does not identify the highest price the buyer would have paid for A, the value under a different alternative set, or the value after a service failure.
Why does context change the choice?
Urban, Weinberg, and Hauser review premarket forecasting for really-new products and show why direct responses about purchase need a wider forecasting structure. Aydin, Kwong, Ji, and Law survey market-demand estimation and distinguish direct questions from indirect choice methods (Urban et al., 1996; Aydin et al., 2014).
Simonson and Tversky show that the surrounding choice set can alter preference through tradeoff contrast and extremeness aversion. An observed choice therefore carries information about the available alternatives, not only about the chosen product (Simonson & Tversky, 1992).
The sources do not produce a universal discount, willingness-to-pay uplift, or pricing rule. They make the choice set part of the record.
What does a choice record look like?
The comparison below is synthetic. A value of 1 means the field is recorded in that row and 0 means it is missing. The values do not represent a buyer, preference score, or conversion.
Figure 1The revealed-preference choice record
The fields are illustrative. Observed choice narrows an inference only when the conditions around the choice remain visible.
Source: Author's schematic framework grounded in Urban et al. (1996), Aydin et al. (2014), and Simonson and Tversky (1992); presence coding is not preference data.
The observed row is richer, not automatically truer. If the buyer could not access an alternative, the choice may reveal availability or procurement constraints. If the price was negotiated, it reveals a choice at that term, not at the public list price.
Which questions can a team answer?
With a complete record, a team may state:
- the buyer selected A over the available alternatives;
- A was selected at the recorded price and terms;
- the choice occurred under the recorded budget, authority, timing, and access conditions;
- the observed choice is consistent with a declared relative preference interpretation.
The team may not state, without further assumptions:
- the buyer would pay any higher price;
- the buyer preferred A outside the shown choice set;
- the transaction represents the whole market;
- the choice proves a stable preference or a causal price effect.
The choice record is therefore a bridge from behavior to an inference. The assumptions belong in the sentence, not in an invisible model.
What is revealed preference not?
It is not a survey answer, a willingness-to-pay fact, a market-demand curve, or a causal estimate. It is not a universal property of a product independent of alternatives, price, budget, timing, access, or institutional constraints.
Observed choice is valuable because it is anchored to behavior. It remains conditional on the choice set and the constraints that made that behavior possible.
The Gabor-Granger article keeps stated price acceptance separate from observed choice.
References
- Aydin, R., Kwong, C. K., Ji, P., & Law, H. M. C. (2014). Market demand estimation for new product development by using fuzzy modeling and discrete choice analysis. Neurocomputing, 142, 136-146. DOI
- Simonson, I., & Tversky, A. (1992). Choice in context: Tradeoff contrast and extremeness aversion. Journal of Marketing Research, 29(3), 281-295. DOI
- Urban, G. L., Weinberg, B. D., & Hauser, J. R. (1996). Premarket forecasting of really-new products. Journal of Marketing, 60(1), 47-60. DOI
