7 Data and Five Reasons for War
Blattman (2023) proposes a five category typology of reasons for war. This typology is not really a theory of war but rather a framework for organizing many possible explanations for war. His typology, however, is premised on theory, specifically the classic “bargaining model of war,” similar to that proposed by Fearon (1995) in his now classic piece titled “Rationalist Explanations for War.” So while Blattman’s typology is itself not explicitly theoretical, it does have theoretical roots that in turn inform his choice of reasons for war.
When we think of why humans turn to violence to settle disputes or achieve political goals, we might assume the lens of a great-man-theory-of-history. Most histories of various wars are filled with narratives that emphasize the motivations and choices of individuals leaders, whether that leader is Abraham Lincoln or Napoleon Bonaparte. Other accounts of conflicts might alternatively emphasize the context in which leaders act. Factors like natural resources, rare minerals, public discontent, or external threats can make conflict seem inevitable. Depression era Germany is a great example of this kind of thinking. Suffering under harsh and embarasing reparations post-WWI while at the same time contending with the effects of a global economic depression, Germany’s aggressive turn toward its neighbors seems like a sure thing.
According to Blattman (2023), while the unique circumstances of any conflict are important, we run the risk of drawing misleading generalizations about why conflicts break out if we belabor the tawdry details of specific wars. If we want to think more clearly about why humans would choose to engage in sustained violent conflict, we need to take a 30,000 foot view. This view starts with a puzzle: war has always been costly, so why would anyone enter into war as long as there is a peaceful bargain that all would find preferable?
7.1 The Bargaining Model of War
This question is at the heart of the bargaining model of war. We begin by assuming that those in a position to choose to fight are rational actors, meaning that they have well-defined and consistent preferences and will always seek to take actions that give them the greatest benefit relative to the alternatives. The bargaining model then layers atop this basic assumption a simple but surprisingly generalizable scenario. The actors involved in any given conflict have a dispute over a prize worth some value. The prize can be monetary, territory, natural resources, or even some basket of many things put together. Whatever the prize, it is finite in value.
At the outset of a dispute, there is some status quo distribution of this prize between the actors. No matter the status quo, a key assumption of the bargaining model is that each actor would prefer to have all of the prize rather than a smaller share of it. That means both actors begin a dispute dissatisfied.
To change the status quo, actors have two choices. They can go to war over the prize or they can find some new distribution of the prize that both parties find acceptable. The question, of course, is how can the actors find a new bargain if any change to the status quo would necessarily mean that one actor will lose out on some of the prize while the other gets to enjoy the difference?
The answer to this question is that the alternative to peaceful settlement (war) will lead to the destruction of a certain share of the prize. Further more, no side of the conflict can know that it will win a war with certainty. War outcomes are probabilistic; not deterministic. That means the value of the prize from war must account for the probability of a side winning and the other losing. This expected value of war is found by multiplying this probability by the value of the prize. Since probabilities are between 0 and 1, that means the expected value of the prize with war is always less than the current value of total prize.
The cost of war combined with no guaranteed victory creates a “bargaining range” for the prize under contention. This is the set of possible new distributions of the prize between actors that will leave both better off relative to what they can expect to get from fighting. As long as this bargaining range exists—which is to say, as long as war is costly to both sides and no side can hope to win with certainty—a peaceful bargain is always better for both sides than war.
7.2 When the Bargaining Model Breaks Down
The bargaining model in its classic form makes war seem implausible and irrational. If a bargaining range exists, who in their right mind would want to fight? This question is party rhetorical because we know from history and even current events that groups or countries choose to go to war rather than find peaceful solutions all the time. Clearly, then, the bargaining model has some scope conditions that, if met, will lead to peace but, otherwise, will lead to war.
Blattman (2023) identifies five scenarios where the bargaining model can break down, making peace the less preferable option for at least one of the sides of a dispute. These are:
- Unchecked (leader) interests
- Intangible incentives
- Uncertainty
- Commitment problems
- Misperceptions
Each of the five sets of reasons for war deal with specific circumstances under which at least one of the conditions for a bargaining range fail to materialize. Unchecked interests addresses scenarios where there is a misalignment between the expected value and costs of war between leaders and the broader group. Sometimes this misalignment means that a head of state will benefit more from war than peace even though the broader citizenry will suffer. Intangible incentives follow when the actors involved have a dispute over something beyond a material prize. Such prizes can be of infinite value such as glory, revenge, holy war, and so on. Uncertainty follows when actors operate under a condition of imperfect information about the expected value of war or the resolve of the other size. Commitment problems follow from things like anticipated future changes in power that would lead to a change in the probability of victory in war. Such changes can mean that today’s bargaining range will not be tomorrow’s and this adds a new level of complexity to the bargaining problem. Misperceptions follow from bad information. This can lead two sides of a conflict to disagree over where the true bargaining range lies, adding a new dispute to the already preexisting one.
Any of the above scenarios can hinder the bargaining process by either narrowing the bargaining range or even inverting it such that there is an incentive for war. Furthermore, none of these scenarios are mutually exclusive. It is possible for two or more of them to operate at once.
7.3 Studying Bargaining Failure with Data
The goal of this book is not to litigate the appropriateness of this framework for understanding why war happens. Rather, it is to use this framework as a foundation for motivating applied data analysis examples. How can we operationalize these five reasons for war? How can we identify whether they really are correlated with a higher likelihood of conflict? We’ll seek answers to these questions in the coming chapters. We’ll start with unchecked interests and move on to intangible incentives, uncertainty, commitment problems, and misperceptions.