• Choose a category
• All
• Classic-Fiction

By Root 713 0

interpretations of the same cases. On the issue of case selection, there is always the danger that case study researchers’ subjective biases and commitments to certain theoretical propositions will lead them to select cases that over-confirm their favorite hypotheses (a different and potentially more serious problem than that addressed in standard discussions of selection biases in statistical studies, which result in truncated samples and under-confirmation of hypotheses).129 Biased case selection can also arise from the fact that evidence on certain cases is more readily accessible than that on others and from the tendency for historically important cases to be overrepresented relative to studies of obscure—but theoretically illuminating—events. Miriam Elman argues, for example, that democratic peace case studies overemphasize cases involving the United States and that they have focused excessively on the study of the Fashoda Crisis and the Spanish-American War compared to possible exceptions of the democratic peace. She also maintains that democratic dyads have been over-studied relative to mixed and nondemocratic dyads.130 On the other hand, for some theory-building purposes mixed dyads are less interesting, and existing studies of wars in mixed and nondemocratic dyads may help fill this gap. Studies that show that states have initiated wars despite inferior military capabilities, for example, call into question assertions that military imbalances alone help explain cases of successful crisis management by democracies. Still, Elman is justified in arguing that more dedicated case studies of mixed and even nondemocratic dyads are needed for comparative research designs like Ray’s study of the Spanish-American War and the Fashoda Crisis.

Yet the substantial convergence among supporters and critics of the democratic peace on which cases deserve study demonstrates that case selection is not an arbitrary process. Several cases have been mentioned by numerous scholars as possible deviant cases, or exceptions to the democratic peace, including the War of 1812, the U.S. Civil War, conflicts between Ecuador and Peru, the Fashoda Crisis, the Spanish-American War, and Finland’s conflict with Britain in World War II. Many of the fourteen other possible exceptions to the democratic peace listed by Ray have also been cited by more than one author or subjected to more than one case study.131 The initial focus on “near wars” between democracies and “near democracies” that went to war was appropriate for the first wave of case studies of the interdemocratic peace, as it offered tough tests of such a theory. As researchers accumulate adequate studies of these cases, they can branch out into more comparisons to mixed and nondemocratic dyads, as Elman has begun to do.

As researchers conduct multiple studies of particular cases, how can they reconcile or judge conflicting interpretations of the same cases? Olav Njølstad emphasizes this problem in case study research, noting that differing interpretations may arise from several sources. First, competing explanations or interpretations could be equally consistent with the process-tracing evidence, making it hard to determine whether both are at play and the outcome is overdetermined, whether the variables in competing explanations have a cumulative effect, or whether one variable is causal and the other spurious. Second, competing explanations may address different aspects of a case, and they may not be commensurate. Third, studies may simply disagree on the “facts” of the case.

Njølstad offers several useful suggestions on these problems.132 These suggestions include: identifying and addressing factual errors, disagree ments, and misunderstandings; identifying all potentially relevant theoretical variables and hypotheses; comparing various case studies of the same events that employ different theoretical perspectives (analogous to paying careful attention to all the alternative hypotheses in a single case study); identifying additional testable and observable implications of competing