Chapter 5 presents a theory of borders and rivalry onset. When a border region contains attributes capable of affecting state power, a commitment problem is more likely to develop.The commitment problem undermines the settlement process by threatening the stability of any settlement agreement that might be signed, and therefore, encourages the involved states to eschew reaching an agreement in the first place, leaving borders unsettled and disputed. States then invest in foreign policy tools to compete over the insecure border, in the hopes of gaining a future bargaining advantage that will allow them to overcome the commitment problem and achieve a favorable and durable settlement.The competition produces a rivalry relationship between the states. We derive three hypotheses connecting unsettled borders, power endowments, and rivalry relations.In order to address alternative theoretical explanations, four additional hypotheses are presented relating rivalry initiation to regime type, power relations, alliances, and ethnic identify claims.

Chapter 6 evaluates the relationship between border settlement and the onset of interstate rivalry.We estimate a series of models based on existing rivalry research.These models provide a baseline comparison to our fully specified models, which include unsettled borders and power endowments.Several key findings emerge.Strong support is found for all the expectations derived from the commitment problem framework. Unsettled borders are associated with an increased likelihood of rivalry initiation. Power endowments feature disproportionately more often in contiguous dyad-years with unsettled borders relative to settled borders. Neighboring states are more likely to form rivalries when contesting territory with power endowments. Mixed and relatively weak evidence is found for alternative explanations. Democratic neighbors are less likely to form rivalries but the results are not consistent across models. States closer in power are more likely to form rivalries but again the results are inconsistent. No relationship is found between allied states and rivalry formation, and little evidence is found that disputed ethnic identify claims are related to rivalry initiation.

Chapter 7 evaluates rivalry termination expectations.The rivalry process helps states overcome the commitment problem but not necessary through war, as traditionally expected.States instead use the rivalry process to consolidate power so as to disincentivize the revision of an eventual agreement. Given the difficulties of overcoming the commitment problem, we would expect these rivalries to be of longer duration and more violence prone.We derive a series of predictions from this argument. Conflict management techniques should be somewhat effective at helping rivals resolve border disagreements within rivalry but only in the absence of power endowments. The exception is legalized dispute resolution techniques, which may have features that help states overcome commitment problems. Border settlement within rivalry will facilitate rivalry termination but rival states bargaining over territorial borders that contain power endowments will be less likely to terminate. Relations between these rivals will generally improve after border settlement.We also derive hypotheses based whether the neighbors are democracies, share an alliance, power relations, and presence of ethnic kin.

Chapter 8 empirically evaluates the hypotheses proposed in Chapter 7. The first set of hypotheses examines the effectiveness of conflict management efforts to settle the border within the context of rivalry. The empirical patterns are consistent with our expectations. Negotiation and mediation generally increase the likelihood of border settlement but this relationship does not hold when power endowments are present. Legal approaches generally help neighbors settle borders with and without power endowments, but are generally more effective in the absence of power endowments. We then examine the relationship between border settlement, power endowments, and rivalry termination. The probability of rivalry termination increases with border settlement but termination is less likely when power endowments are present. Rivalry relationship transforms once border settlement occurs but the rivalry does not immediately terminate. Crises and disputes are less severe and of shorter durations.We find little evidence that democratic neighbors, allied, closer in parity, or the presence of ethnic kin in the border region affect the odds of rivalry termination.