Alliances are costly to form and to terminate, and yet alliances change frequently. Scholars typically attribute these decisions to static factors, such as the power balance, and retrospective ones, such as past power shifts. We highlight another factor: prospective changes, particularly anticipated military strength shifts. We analyze a three-country bargaining model of alliances and war that incorporates forward-looking power dynamics. The model, unlike those restricting players to set roles, flexibly allows players to ally in any arrangement. We find that alliance arrangements that are optimal when power is static are often suboptimal when power fluctuates. Maintaining prior alliances despite expected power shifts may even lead to preventive war. States thus strategically look to the future to identify optimal alliances in the present. Quantitative analyses corroborate the expectation. As the anticipated size of power shifts increases, alliance changes become more common. Accordingly, states navigate expected changes in the international landscape by rearranging current alliance commitments that can help minimize the risk of conflict. When power balances are in flux, malleable institutional arrangements may prove preferable to rigid ones.

中文翻译：

---

权力转移阴影下的联盟动态

联盟的形成和终止成本很高，但联盟经常变化。学者们通常将这些决定归因于静态因素，例如权力平衡，以及回顾性因素，例如过去的权力转移。我们强调另一个因素：预期的变化，特别是预期的军事实力变化。我们分析了一个包含前瞻性权力动态的联盟和战争的三国谈判模型。与那些限制玩家设定角色的模式不同，该模式灵活地允许玩家以任何安排结盟。我们发现，当权力静止时最优的联盟安排在权力波动时往往不是最优的。尽管预期的权力转移，但维持先前的联盟甚至可能导致预防性战争。因此，各国从战略上着眼于未来，以确定当前的最佳联盟。定量分析证实了预期。随着权力转移的预期规模增加，联盟变化变得更加普遍。因此，各国通过重新安排有助于将冲突风险降至最低的当前联盟承诺来应对国际格局的预期变化。当权力平衡不断变化时，可塑的制度安排可能比僵化的制度安排更可取。