Proposition 2, p. 3346 — price discrimination raises profit, buyer, low-type seller surplus and welfare; high types gain iff b > b̂ (then strict Pareto)
ProvedTwoSidedPD.Baseline.proposition2_welfareConsider a monopoly platform serving a mass one of buyers, who get a stand-alone value and a benefit per seller on the platform, and two categories of sellers of mass one each, who earn per buyer, with and . Outside options are uniform on . Assume Assumption 1:
Under uniform pricing the platform chooses to maximize ; under price discrimination it chooses to maximize . Both problems have a solution, and for every uniform optimum and every discrimination optimum:
- (i) the platform, buyers and low-type sellers are better off under price discrimination: , , ;
- (ii) total welfare is higher under price discrimination: ;
- (iii) high-type sellers are better off under price discrimination if and only if
- in this case price discrimination is a strict Pareto improvement over uniform pricing: the platform, the buyers, the low-type and the high-type sellers are all strictly better off.
This is the paper's main welfare result: unlike the classical one-sided analysis with linear demands, seller-side price discrimination by a two-sided platform raises total welfare and can benefit every group.
Formalization Note. Equilibria are global maximizers of over and of over (the objects the first-order conditions (7)–(11) solve), not points in . "Better off" means a strictly higher profit for the platform and strictly higher surplus for buyers and sellers, with surpluses defined by the integrals of App. A.2. is the standing assumption of §3.2 and footnote 10. Assumption 1(i) is used as printed (); the paper's bound (p. 3344) corresponds to , which is what interior buyer participation needs, so the printed assumption does not by itself keep participation in ; the statement does not claim it does. On the parameter region , so Real.sqrt is the ordinary square root.
import Mathlib import Definitions.Def_TwoSidedPD_Baseline_Setting
namespace TwoSidedPD.Baseline
theorem proposition2_welfare (v b θH θL : ℝ) (hL : 0 < θL) (hLH : θL < θH) (hb : 0 < b)
(hA : Assumption1 v b θH θL) :
((∃ x : ℝ × ℝ, IsUniformOpt v b θH θL x) ∧ (∃ y : ℝ × ℝ × ℝ, IsDiscrOpt v b θH θL y)) ∧
∀ (x : ℝ × ℝ) (y : ℝ × ℝ × ℝ), IsUniformOpt v b θH θL x → IsDiscrOpt v b θH θL y →
-- (i) platform, buyers and low-type sellers are strictly better off
(profitU v b θH θL x < profitD v b θH θL y ∧
buyerSurplusU v b θH θL x < buyerSurplusD v b θH θL y ∧
sellerSurplusU θL θH θL x < sellerSurplusDL θL y) ∧
-- (ii) total welfare is strictly higher
welfareU v b θH θL x < welfareD v b θH θL y ∧
-- (iii) high-type sellers are strictly better off iff b > b̂(θH, θL)
(sellerSurplusU θH θH θL x < sellerSurplusDH θH y ↔ bhat θH θL < b) ∧
-- in that case, a strict Pareto improvement
(bhat θH θL < b →
profitU v b θH θL x < profitD v b θH θL y ∧
buyerSurplusU v b θH θL x < buyerSurplusD v b θH θL y ∧
sellerSurplusU θL θH θL x < sellerSurplusDL θL y ∧
sellerSurplusU θH θH θL x < sellerSurplusDH θH y) := by sorry
end TwoSidedPD.Baseline
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.