Rubinstein bargaining model¶
An infinite-horizon alternating-offers bargaining game whose unique subgame-perfect division reflects the players' discounting and first-mover position.
Core Idea¶
The Rubinstein bargaining model is an infinite-horizon, two-player alternating-offers game. Costly delay makes rejection thresholds equal continuation values, and subgame perfection yields immediate agreement with a division shaped by patience and first-mover position. Continuation values are the load-bearing mechanism: rejecting today gives the responder a discounted proposer payoff tomorrow. This makes acceptance thresholds sequentially credible and distinguishes the result from an arbitrary Nash equilibrium supported by threats that would not be carried out. Discount factors make delay costly.
Scope of Application¶
The model applies to stylized bargaining with complete information, divisible surplus, alternating turns, and stationary time preferences. Use it only when the surplus, information, alternating protocol, infinite continuation, discount factors, and sequential credibility are represented; real negotiation is an application, not automatic identity.
- Game theory. Derives subgame-perfect division.
- Negotiation analysis. Studies delay and agenda power.
- Industrial organization. Models bilateral surplus division.
- Political bargaining. Explores proposer advantage under institutional turns.
- Mechanism comparison. Contrasts strategic and axiomatic bargaining.
Clarity¶
The model's force lies in continuation values, not endless conversation. An offer is credible because rejection changes time and proposer, and equilibrium must remain optimal after every possible history. The closest near miss sets the boundary: The ultimatum game is the closest near miss: one proposer makes an offer, but rejection ends the game rather than transferring proposal power indefinitely. A positive case must satisfy this test: Include the complete-information infinite alternating-offers game with costly delay and its subgame-perfect continuation logic.
Manages Complexity¶
An infinite game tree compresses to paired stationary offers and acceptance thresholds. That compression depends on complete information and geometric discounting; extensions can alter uniqueness and comparative statics. The central strategic detail–empirical idealization tradeoff is this: The model exposes credibility but omits many institutional moves. A second first-mover advantage–patient-player limit tension matters because Agenda power matters under discounting but shrinks as delay costs vanish.
Abstract Reasoning¶
Use three linked moves: specify the feasible surplus and each discount factor; write continuation values for the player who would propose next; set each responder's acceptance threshold equal to that continuation value. As a collapse test, the identity collapses if offers do not alternate, horizon is not effectively infinite, delay is costless, or threats need not be sequentially credible. A fourth check is to solve the paired stationary offers and verify sequential optimality. A final check is to interpret proposer advantage only within the model's assumptions.
Knowledge Transfer¶
Alternating proposal power and costly delay transfer to other formal bargaining settings. Literal transfer stops where information is incomplete, the horizon ends, surplus is indivisible, or institutions permit side actions not represented in the game. No canonical parent prime is currently asserted; broader structural comparisons remain related-prime analogies until separately adjudicated in the DAG. The model supplies a strategic mechanism for division. Subgame perfection selects credible continuation play.
Neighborhood in Abstraction Space¶
Rubinstein bargaining model sits in a crowded region of the domain-specific corpus (27th percentile for distinctiveness): several abstractions share nearly its structure, so a description that fits it tends to fit its neighbors too.
Family — Allocation, Ranking & Bargaining Models (11 abstractions)
Nearest neighbors
- Strong Nash equilibrium — 0.92
- Max-dominated strategy — 0.91
- Correlated equilibrium — 0.89
- Stable Roommates Problem — 0.88
- Obligationes — 0.88
Computed from structural-signature embeddings · 2026-10-08