(A.2) for a subunit : expected payoff under equals the expected conditional expectation
ProvedIncentivesInTeams.Conglomerate.appendix_A2_subunitConsider a conglomerate model whose component weights are probability weights, a joint strategy , a subunit , a strategy , and a subunit . Let denote the information functions when is played. Then
where is subunit 's factor of the head's conditional expectation under (model file): the average of over the states of whose message under is the one recorded in .
This is equation (A.2) of the paper for a subunit , with its right-hand side in the factorized form. Together with the head's analogue it gives (A.1).
Formalization Note. Under the strategy of subunit and the head's strategy are those of , so the left-hand side is the -th payoff component under . Information follows the one-exchange protocol of the model file.
import Mathlib import Definitions.Def_IncentivesInTeams_Conglomerate_Model
namespace IncentivesInTeams.Conglomerate
theorem appendix_A2_subunit {ι : Type*} [Fintype ι] [DecidableEq ι] {S₀ : Type*} [Fintype S₀] {S : ι → Type*}
[∀ i, Fintype (S i)] {Z₀ : Type*} {Z M₀ M : ι → Type*} {D₀ : Type*} {D : ι → Type*}
(T : Model S₀ S Z₀ Z M₀ M D₀ D) (hT : T.WeightsOK)
(βs : JointStrategy S₀ S Z₀ Z M₀ M D₀ D) (i : ι)
(b : SubStrategy (S i) (Z i) (M₀ i) (M i) (D i)) (hb : b ∈ T.B i) (j : ι) (hji : j ≠ i) :
T.expect (T.subPayoff (βs.update i b) j) =
T.expect (fun s => T.subFactor βs j ((βs.update i b).headInfo s)) := by sorry
end IncentivesInTeams.Conglomerate
Read-back
What the Lean code literally says, in plain math · claude-opus-5-5
Setting. is a finite set of subunits. and each are finite. The model's weights and are assumed to be probability weights, with product and . The payoffs are and .
A joint strategy has head functions and subunit functions , as follows:
- , , ;
- ;
- and act on .
Its information maps are and . Subunit 's payoff under is
For fixed and , the factor is
Here , with value when the denominator is .
Theorem. Take a joint strategy , a subunit , a subunit- strategy , and a subunit . Let be with subunit 's strategy replaced by . Then
In words: is built from but evaluated at the head information produced under .
is not required to lie in any strategy set, and no positivity of the conditioning events is assumed. Where such an event has weight , takes the value .
Degenerate cases. The weight hypothesis forces all state sets to be nonempty. If has exactly one element, no exists and the statement is vacuous.
Confirmed by the mission captain (proposal self-audit).
Confirmed by the moderator at approval.