Lemma 1 — price effect vs. selection effect: the induced review falls in iff
ProvedPricingStars.lemma_1dynamic-pricingdynamic-programmingeconomicsoperations-researchp2o-gran-per-chapterp2o-plan-paperp2o-strengthenedp2o-v1rating-systems
Let , a true quality and an aggregate rating be given, and write for the induced average review (8) at price .
- For every price , (9) holds: .
- At every price, the marginal selection effect equals , the marginal direct price effect equals , and (10).
- The direct price effect dominates the selection effect, , if and only if ; for the selection effect dominates, .
- If , a price increase strictly decreases the induced average review: is strictly decreasing.
- If , a price increase strictly increases it: is strictly increasing.
The lemma predicts the sign of the correlation between prices and reviews and drives the direction of the seller's dynamic price distortion.
Differences from the source:
- reading: 'the direct price effect dominates the selection effect' → |−κ| > |1 − κ|, the magnitudes of the two marginal effects of (10); reason: p.10 names −κ the marginal direct price effect and 1 − κ the marginal selection effect, and compares them
- generalization: the three claims of Lemma 1 → also (9) for every p and the derivatives of (10) at every p; reason: these are the displays the proof ('follows from the preceding discussion') rests on; the lemma's claims are included unchanged
- reading: the linear-quadratic model: q = 2(ψ̄ − p(1−κ)) (7) and the review (9), with no boundary solutions specified (footnote 5) → demand, cutoff and review are the paper's affine formulas on all of ℝ², prices are arbitrary reals, nothing is clamped to [0,1]; reason: the paper solves exactly this unconstrained problem (its remarks that flow profits are bounded, pp.12 and 24, do not hold for these formulas and are not used); with clamping the value function is not quadratic and (28)–(29) are not derived
Preamble
import Mathlib import Definitions.Def_PricingStars_Model
Formal statement
namespace PricingStars
/-- Lemma 1 (Price & Selection Effect) (Stenzel–Wolf–Schmidt, Pricing for the Stars, CRC TR 224
DP No. 143 (Jan 2020), p.11 (PDF 13); with (8)–(10), p.10 (PDF 12)). For `κ ∈ [0,1)`, true
quality `θ` and aggregate rating `ψ̄`:
1. (9): the induced average review is `θ + 1 − ψ̄ + p(1 − 2κ)` for every price `p`;
2. (10): at every price the marginal selection effect `d ω_e(ω̃(ψ̄, p))/dp` equals `1 − κ`, the
marginal direct price effect `d(−κp)/dp` equals `−κ`, and the induced review has derivative
`1 − 2κ`;
3. the direct price effect dominates the selection effect (`|−κ| > |1 − κ|`) iff `κ > 1/2`, and the
selection effect dominates (`|1 − κ| > |−κ|`) when `κ < 1/2`;
4. if `κ > 1/2` the induced review is strictly decreasing in the price;
5. if `κ < 1/2` it is strictly increasing in the price.
**Formalization Note** "Dominates" compares the magnitudes of the two marginal effects of (10).
The model is the paper's unclamped linear one. -/
theorem lemma_1 (θ κ ψ : ℝ) (hκ0 : 0 ≤ κ) (hκ1 : κ < 1) :
(∀ p : ℝ, inducedReview θ κ ψ p = θ + 1 - ψ + p * (1 - 2 * κ)) ∧
(∀ p : ℝ, HasDerivAt (fun x => avgTaste (cutoff κ ψ x)) (1 - κ) p ∧
HasDerivAt (fun x => -κ * x) (-κ) p ∧
HasDerivAt (inducedReview θ κ ψ) (1 - 2 * κ) p) ∧
(|1 - κ| < |-κ| ↔ 1 / 2 < κ) ∧
(κ < 1 / 2 → |-κ| < |1 - κ|) ∧
(1 / 2 < κ → StrictAnti (inducedReview θ κ ψ)) ∧
(κ < 1 / 2 → StrictMono (inducedReview θ κ ψ)) := by sorry
end PricingStars
Source
Stenzel–Wolf–Schmidt, Pricing for the Stars: Dynamic Pricing in the Presence of Rating Systems, CRC TR 224 DP No. 143 (Jan 2020), Lemma 1, p.11 (PDF 13), with (8)–(10), p.10 (PDF 12)
Human review
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.