Costly Signaling, Crisis Bargaining, and Limited Kinetic Exchange

A Theoretical Framework for the US-Iran 2026 Context

Our analysis framework failed to forecast a return to the negotiation table as a high probability fork. This document highlight some applied game theory based on Fearon (1995) to the Iran conflict in order to integrate in future framework revisions.

TL;DR

  • War is bargaining failure under uncertainty, and limited kinetic exchanges are the mechanism by which uncertainty gets resolved. Fearon (1995) shows war is ex ante inefficient; Schelling, Powell, Slantchev, and Reiter show the bargaining range can only be reopened once private information about resolve, capability, and cost-tolerance is revealed — and battle is the most credible source of that revelation. Frameworks that treat limited strikes as “monotonic escalation” are wrong: they are type-revealing signals that frequently make settlement more, not less, likely afterward.
  • The US-Iran 2026 case fits the model precisely. The June 2025 Twelve-Day War, the February 28, 2026 strikes, and the April 7–8, 2026 Pakistan-mediated ceasefire (brokered by PM Shehbaz Sharif and Field Marshal Asim Munir, per Trump’s Truth Social statement and Carnegie analysis, April 2026) follow a textbook Fearon–Slantchev sequence: pre-crisis uncertainty → costly signaling exchange → convergence on revised priors → bargaining range reopens → MOU/framework. The current ceasefire is the post-revelation bargaining window the theory predicts.
  • Two structural risks dominate the next 30 days: a double principal-agent problem (Witkoff–Trump on the US side; Araghchi–Khamenei’s successor on the Iranian side) and the absence of a hard commitment device making war the explicit default if talks collapse. Without an exogenous focal point (Trump-Xi summit, Hajj, AUMF window) anchoring a deadline AND a binding “war if no deal” default, the framework predicts collapse back into chicken via either renewed strikes or Iranian breakout.

Applied Analytical Framework: US-Iran 2026

A. Limited Kinetic Exchange as Type-Revelation (the Mechanism Underweighted)

The Fearon-Slantchev-Reiter-Weisiger consensus: battle reveals what diplomacy cannot. Pre-strike, both sides bluff. Strikes generate non-manipulable information that no statement can produce.

Iran 2026 type-revelations from the kinetic record:

  • June 2025 (Twelve-Day War): Israel demonstrated air superiority over Iran; Iranian air defense fragility revealed. Per IAEA Verification and Monitoring Report GOV/2026/8 (September 2025), as of 13 June 2025 Iran’s stockpile of uranium enriched up to 60% stood at 440.9 kg (up from 408.6 kg per IAEA GOV/2025/24, May 17, 2025) — a publicly confirmed figure that updated US/Israeli priors on breakout time.
  • Feb 28, 2026 strikes: Decapitation of Khamenei and Larijani revealed willingness to escalate to leadership targeting; revealed US/Israeli intelligence penetration depth.
  • Iranian counter-strike on Al Udeid (June 23, 2025): Iran fired 14 short- and medium-range ballistic missiles at the base; Qatari and US Patriot batteries intercepted 13. Per CBS News and contemporaneous reporting, Iran gave advance warning to Qatar and the US hours before launch (confirmed by Trump). Classic Schelling brinkmanship — demonstrated capability while pre-coordinating limits on escalation. The pre-warning was itself the costly signal: “we can hit, but we will not press to nuclear levels.”
  • Hormuz closure and re-opening cycles (Mar–Apr 2026): revealed Iranian price-tolerance for sanctions vs. trade revenue.

Operative point for the user’s framework: Strikes that look like “escalation” produced more accurate priors on both sides. The April 7–8 Pakistan-mediated 10-point ceasefire framework — brokered by PM Shehbaz Sharif and Field Marshal Asim Munir per Trump’s own Truth Social statement and Carnegie Endowment analysis (April 2026) — would have been rejected as a starting point in January 2026; it became plausible only after the kinetic exchange exhausted information rents on both sides. This is the Slantchev Principle of Convergence operating in real time.

Per Mona Yacoubian, CSIS Middle East Program Director and Senior Adviser, “Iran’s War Strategy: Don’t Calibrate—Escalate” (CSIS, 2026): “Iran signaled its intent to widen and deepen the conflict from day one… Iran appears to have absorbed the lessons from previous conflicts.” Per FDD (Stricker et al., March 5, 2026): the strikes “Signal Resolve To End Tehran’s Nuclear Weapons Program.”

Test for the framework: if the analyst’s prior model treated the Feb 28 strikes as “monotonic escalation toward war,” it was wrong; if it treated them as type-revelation that would reopen bargaining, it correctly predicted the April 7–8 ceasefire.

B. The Double Principal-Agent Problem

Both sides have agents who cannot bind their principals.

  • US side: Witkoff/Kushner, with no technical expertise, demonstrably misrepresented Iran’s offers to Trump (Feb 28 collapse). Trump’s signaling is dispositional/personal, not institutional — Senate ratification is not on offer, so any deal is reversible by executive order (the JCPOA replay risk).
  • Iran side: Araghchi negotiates; Khamenei’s son (newly installed) and the IRGC ratify. The IRGC has revealed willingness to override diplomatic offers (June 2025 missile reserves dispute). Iranian parliament has not ratified the Additional Protocol.

Implication (Leventoglu-Tarar, Putnam): Any deal struck without explicit principal sign-off on both sides is fragile. The 30-day MOU window is exactly the structure that fails most often — short enough to avoid principal engagement, long enough to allow either side to defect.

Design fix: Require visible principal-level commitment ceremonies (Trump–Pezeshkian or Trump–Khamenei-successor) inside the MOU window. Make defection observable to both domestic audiences. This converts a sinking-cost game into a tying-hands game (Fearon 1997 logic).

C. Convergent External Deadlines as Schelling Focal Points

Functional focal points available in the next 60 days:

  • Hajj pilgrimage (mid-May 2026): Iranian regime cannot escalate kinetically without delegitimizing itself among Sunni neighbors; provides an exogenous quiet window.
  • AUMF debate window: any new US strike requires War Powers reporting; the 60-day clock from May 1 War Powers letter expires late June.
  • JCPOA termination already passed (October 18, 2025); E3 snapback formally triggered August 28, 2025; UN sanctions reimposed September 27, 2025: these are now sunk costs; cannot be undone unilaterally.
  • A possible Trump-Xi summit: would pull the issue into a great-power frame Iran cannot directly contest.

When focal points help (Schelling): when both sides need a face-saving stopping point and lack one. When they hurt (Leventoglu-Tarar): when premature focal-point salience locks both sides into incompatible public commitments before win-sets converge.

D. Commitment Device Design for the 30-Day MOU

Per Schelling and Powell: cooperation in a finite-horizon game requires that defection be made more costly than cooperation through a credible automatic punishment.

Design principles for a 30-day MOU:

  1. War as the explicit default if talks collapse. Not implicit. Written. Both sides publicly acknowledge the alternative is renewed strikes (US) or resumed enrichment to weapons grade (Iran). This creates Fearon (1994) audience costs on both sides for backing down.
  2. Verification milestones at days 7, 14, 21, 30. Each is a Schelling focal point. Each reveals incremental information (Slantchev convergence). Failure at any gate triggers escalation, not resumption.
  3. Pre-positioned military forces remain in theater. Per Slantchev (2005), this is both a sunk cost and a hands-tier — cheaper to use than to redeploy and re-mobilize. Conversely, withdrawing forces during the window would signal weak resolve and raise war risk per Slantchev’s endogenous-power logic.
  4. Third-party guarantor with skin in the game. Pakistan (current mediator), or Oman, or the E3 with snapback already activated. Per Kydd (2005), trustworthy hegemonic third parties enable cooperation between mistrustful adversaries.
  5. Avoid double public pre-commitment. Per Leventoglu-Tarar, when both sides publicly precommit to incompatible demands, the prisoner’s dilemma forecloses settlement. Keep the technical terms private; make the framework public.

E. Mutual Demonstration Without Operational Achievement

A frequently misread pattern: both sides “fail” militarily — US doesn’t destroy nuclear program (Iran retains the IAEA-verified 440.9 kg of 60% HEU), Iran doesn’t expel US forces — yet the fact of mutual demonstration itself reveals resolve types. Per Slantchev’s Principle of Convergence: information exhaustion, not victory, terminates fighting.

The Iran case parallel: US strikes did not “obliterate” the program; Iran did not close Hormuz permanently or sink a US warship. Both demonstrated capability and resolve up to thresholds; neither pressed beyond. The information content of further fighting is now low — exactly when the convergence principle predicts settlement.

F. Sequential Game Stages

The 2025–2026 Iran sequence maps to a recognizable formal structure:

  1. Stage 1 : Chicken (mutual defection trap): Pre-Feb 2026. Trump demands “unconditional surrender”; Iran refuses talks. Both bluff because each believes the other will swerve. Fearon (1995): when private information dominates and incentives to misrepresent are high, war becomes a positive-probability event.

  2. Stage 2 : Mutual demonstration (costly signaling exchange): Feb 28 – April 7–8, 2026. Strikes, counter-strikes, Hormuz closure. Information rents exhausted on both sides. Per Slantchev (2003): warfare ceases to be useful.

  3. Stage 3 : Coordination game (the current window): April 8 – present (May 7, 2026). Both sides know each other’s types; the question is which of multiple equilibria they settle on. This is where Schelling focal points become decisive.

The risk: returning to Stage 1 if Stage 3 produces no focal-point convergence. The opportunity: using the Stage 2 exchange as the foundation for a more durable deal than was possible pre-strike.

G. Where the Bargaining Model Could Fail Here

Per Reiter (2009) and Weisiger (2013), the model fails when commitment problems dominate:

  • If Israel reads Iran’s regime as dispositionally aggressive (Weisiger’s “dispositional commitment problem”), Israel will demand regime change — an unlimited war aim — regardless of what information the strikes revealed.
  • If Iran reads the US as committed to regime change (the Trump “unconditional surrender” rhetoric), Iran cannot credibly disarm because disarming is a one-shot move that eliminates its only deterrent.
  • Shifting power problem: Iran is currently weak (per Hossein Ajorlou, Allameh Askari International University, “The Strategic Multi-Layered Predicament of the Iran War,” Al Jazeera Centre for Studies, April 30, 2026: the US-Israeli campaign was premised on the framing of “an Iran weakened at home and abroad”); US/Israel may prefer war now to deal-making with a weak Iran that will recover. This is the classic Powell preventive-war logic.

These failure modes are why the 30-day MOU must include explicit commitment devices, not just information-revelation.