Why Wars Begin: The Bargaining Failures Behind Conflict

If war is almost always worse than a deal, why do states fight anyway?

Egyptian President Anwar Sadat, U.S. President Jimmy Carter, and Israeli Prime Minister Menachem Begin during the 1978 Camp David negotiations. Bargaining offers states a way to settle competing interests without bearing the costs of war. Image: Bill Fitz-Patrick / Jimmy Carter Library, via Wikimedia Commons. Public domain. Color treatment by the author.

Introduction

States know war incurs costs. Russia fights under extensive sanctions and mounting fiscal pressure, while Ukraine faces growing tensions over military recruitment. Fighting between the United States and Iran has pushed oil prices higher, and Iran’s currency, the rial, has fallen to record lows.

Ask the average Joe whether war is bad, and the answer seems obvious. Yet states still go to war. Why?

Political scientists have wrestled with that question for decades. If war imposes SIGNIFICANT costs on both sides, there should usually be some negotiated settlement that both would prefer to fighting. So why do states sometimes fail to reach a settlement before a dispute escalates into war?

Bargaining and War

Actors commonly bargain over goods. Bargaining is an interaction in which two or more actors must decide how to distribute something of value. Increasing one actor’s share of the good decreases the share available to the other.

Figure 1. The Pareto frontier shows the set of efficient deals between two states. Any point on the green line fully divides the available value between them, while points below the line are inefficient because both sides could still do better. [Image created by the author using ChatGPT.]

The graph shows all the ways two states could divide something they both value (say, territory). Moving right gives State A more, while moving left gives State B more. The green line is the Pareto frontier. Every point on it represents an efficient bargain where all available value has been distributed. The red point falls below that line, which means both states could improve their position without making the other worse off.

Figure 2. The feasible set shows the full range of possible agreements between two states over the same disputed good. Any point along the bar represents a different deal, with outcomes closer to one end favoring State A and outcomes closer to the other favoring State B. [Image created by the author using ChatGPT.]

This bar shows the same concept represented differently. States have an incentive to achieve the best possible deal; each wants to get as much of the bar as possible. In bargaining, a deal doesn’t necessarily have to be ideal for either side; it just has to be preferable to the available alternatives.

However…

States aren’t choosing between different bargains in a vacuum. Another option also exists: war.

Bargaining Range

Figure 3. A possible bargain over disputed territory. State A and State B each prefer to control the entire island, but any negotiated settlement divides the territory between them. The proposed boundary determines each state’s share, illustrating how bargaining allocates a contested good between actors with opposing ideal outcomes. [Image created by the author using ChatGPT.]

Here is a more picturesque example. The problem clearly outlines that both states want as much of the island as possible. While both could agree on a possible deal, the problem becomes more interesting once either state decides to reject the proposed border and fight for a better outcome.

State A, for instance, might expect to win 60% of the island, but a conflict would cost it the equivalent of 10% of the island’s total value. Similarly, State B expects to retain 40%, but if war costs it 10%, then fighting is really only worth 30%.

Figure 4. Expected outcome of war. State A anticipates controlling roughly 60% of the disputed island after fighting, while State B expects to retain about 40%. This expected division becomes each side’s baseline for deciding whether a negotiated settlement is preferable to war. [Image created by the author using ChatGPT.]

This leads to the bargaining range. The bargaining range is a set of deals that both parties in a bargaining interaction prefer over the reversion outcome, which can be many things, but in this instance, is war.

Figure 5. The top panel uses those reduced war values to identify the bargaining range: the set of territorial settlements that both State A and State B prefer to war. The bottom panel shows the expected outcome of war and how war costs reduce the value of fighting for each state. [Image created by the author using ChatGPT.]

State A should therefore accept at least 50%, whereas State B should accept any division where it receives at least 30%. Because of the incurred war costs, State A can receive as much as 70%, so any division that gives State A between 50–70% leaves both sides better off than war. In the academic literature, this is what’s called the bargaining range.

This image shows a clear boundary between good deals and military conflict. All a state should have to do is stay within the bargaining range. So why does bargaining fail?

Why Bargaining Fails

Wars isn’t a simple phenomena, and they’re heavily situational state-by-state. However, there are three mechanisms that can explain bargaining failures: incomplete information, commitment problems, and perceived indivisibility.

Iraq and the WMD Problem. Incomplete information arises as a situation where actors in a strategic interaction lack information about other actors interests and/or capabilities. They may underestimate actor’s resolve, that is, their willingness to endure costs in order to acquire a particular good; overestimate their own capabilities; or misinterpret intentions.

Washington D.C. judged that Iraq had continuing Weapons of Mass Destruction (WMD) programs which included chemical and biological weapons, while also acknowledging that intelligence agencies lacked specific information on important aspects of those programs. Later on, a Senate Intelligence Committee review concluded those statements were either overstated or undersupported by the current information. Many analysts in intelligence agencies interpreted this ambiguous evidence through a biased lens that Iraq did possess active WMD programs.

The U.S. government believed its opponent possessed capabilities that it didn’t actually have, so the expected costs and benefits of any bargaining range became distorted.

Of course, Iraq wasn’t simply an intelligence failure. Governments have incentives to conceal their capabilities surrounding critical infrastructure like weapons programs to gain leverage and make uncertainty harder to assess. This example demonstrates one mechanism which undermines bargaining, where policymakers calculated risks of compromise and war using mistaken estimates of the other side’s capabilities. These incorrect estimates are, like the other examples, an illustration of the described mechanism.

U.S. Air Force crew chiefs prepare a B-2 Spirit during Operation Epic Fury, March 17, 2026. The bomber represents the enduring U.S. strike capabilities at the center of the U.S.–Iran commitment problem. Image: U.S. Air Force / U.S. Central Command Public Affairs, via Wikimedia Commons. Public domain. Color treatment by the author.

United States & Iran: Credibility/Commitment. Credibility is believability. Actors may make credible threats, one which the target believes could be plausibly carried out; whereas commitment is a commitment or promise that the recipient believes will be honored.

The war between the U.S. and Iran can be illustrated with a simple analogy: the prisoner’s dilemma. Two prisoners are put in separate rooms and given choices by the prosecutor. If both cooperate, both receive two years of prison time. If one defects while the other cooperates, the defector gets the best outcome, one year. Because each fears being exploited, both may choose to defect. By doing so, each would receive ten years. They end up worse off than if they had trusted each other.

In international politics, this helps explain why agreements can fail despite cooperation benefitting both sides, as actors worry the other may cheat first.

More specifically, Iran can agree to nuclear restrictions and reopen the Strait of Hormuz in exchange for sanctions relief as well as reduced U.S. military pressures. But once Iran has made those concessions, what guarantees that Washington will continue to honor its side of the bargain? Similarly, nothing guarantees that Iran will comply with not building nuclear weapons if the U.S. commits to its side of the bargain.

A note should be made here that this is a present-day war with progressing events, however, this does illustrate how commitment problems make bargaining exceptionally more difficult when neither side trusts each other to commit to a deal they have incentive not to follow through with.

Armed Pashtun tribesmen move toward Kashmir during the 1947–48 India–Pakistan conflict, marking the beginning of the long-running territorial dispute. Image: Unknown photographer / Pakistan Army, via Wikimedia Commons. Public domain in Pakistan; U.S. Color treatment by the author.

Kashmir, India-Pakistan: Perceived Indivisibility. The Judgement of Solomon is a story from the Book of Kings. Two women both claim a baby is theirs. Neither being able to reach a settlement, bring it to King Solomon. The king says the solution is simple: split the baby in half so each gets a share. One woman is satisfied, but the other screams at the thought. For a baby cannot be divided, without harming it.

This illustrates the final concept, indivisibility.

While it is easy to say that territorial disputes can be settled 50/50, or at the very least within the bargaining range, states may claim indivisibility for historical, religious, or ethnic reasons. Kashmir is one example.

Since the partition of British India in 1947, India and Pakistan have contested the political status of the former “princely” state of Jammu and Kashmir. Today, the territory remains divided by the Line of Control, but that division has not produced an agreed final settlement. Kashmir has thus become a hotspot for competing ideas of sovereignty, national identity, history and political legitimacy. And while bargaining may enable both states to acquire, say, 60% to one, and 40% to the other, neither is required to accept such deals under claims of indivisibility.

This is also strategic in the sense that if states cannot bargain for such reasons, then they don’t have to lose a percentage of the territory. So Kashmir may not necessarily be physically and literally indivisible, but rather politically and strategically indivisible.

Incomplete information makes it harder for states to figure out where the bargain is acceptable; commitment problems create trust issues; and indivisibility makes compromise harder to achieve.

Conclusion

If war is so costly, why do states still fight? The answer lies largely in bargaining failures. States generally have a range of agreements they would prefer to war, however, those agreements can become untenable when governments misunderstand each other, cannot credibly commit to promises, or treat the disputed objects as indivisible.

War can thus begin when bargaining processes fail.

**If you enjoyed this article, follow me for more writing on geopolitics and international relations, and follow The Geopolitical Economist for more analysis on global affairs.**

The Geopolitical Economist is now at geopoliticaleconomist.com: analysis, research and a free weekly newsletter. Subscribe at geopoliticaleconomist.com/newsletter.

Writers: pitch us at geopoliticaleconomist.com/write-for-us.