Case Studies and Theory Development in the Social Sciences - Alexander L. George [173]
Homer-Dixon notes that a number of methods are available for testing hypotheses in environment-conflict research. Two are conventional quasi-experimental methods (correlational analysis of a large number of cases, and controlled case comparison). The third is process-tracing of the kind described by Alexander George and Timothy McKeown.700 Homer-Dixon defends his reliance on process-tracing by noting that “the stage of research strongly influences the method of hypothesis testing a researcher can use to best advantage.”701 He believes that process-tracing is advantageous particularly in the early stages of research on highly complex subjects. In these circumstances ″hypotheses are liable to be too crude to support testing that involves quantitative analysis of a large number of cases.” Research resources are used to best advantage “by examining cases that appear, prima facie, to demonstrate the causal relations hypothesized.″702
It is in this context that Homer-Dixon provides a detailed argument for selecting on the dependent as well as the independent variables. He recognizes that this could lead to criticisms of biased case selection, but defends the procedure by noting that process-tracing was used mainly on cases characterized as having both environmental scarcity and violent conflict (rather than cases in which environmental scarcity was neither a necessary nor sufficient cause of violent conflict). In response to criticism of his focus on cases embracing both environmental scarcity and violent conflict, Homer-Dixon argues that in the early stages of research, such a procedure is often the best and sometimes the only way to begin. For particular cases it can show whether or not the proposed independent variable is a cause of the dependent variable. That is, by making use of process-tracing, it answers the question of whether there are “any cases in which the independent variable is causally linked, in a significant and important way, to the dependent variable.”703
Homer-Dixon notes that in highly complex systems, such as the ecological-political systems he has studied, it is not likely that the proposed independent variable (environmental scarcity) will be a sufficient cause of the dependent variable (violent conflict). Rather, it will be necessary to identify and add “numerous and detailed scope conditions”—i.e., conditional generalizations. Without including adequate scope conditions, “a statistical analysis of the distribution of cases … will probably reveal little correlation, even though there might be important and interesting causal links between environmental scarcity and conflict” (i.e., a false negative).704
Homer-Dixon notes that in such circumstances “careful process-tracing, involving close examination of the causal process” operating in the cases in which both the independent variable and the dependent variable were present “will help identify the relevant scope conditions.”705 The author notes that researchers can then ask whether the scope conditions and intermediate variables identified via process-tracing were present, and why in other cases in which environmental scarcity existed violent conflict did not ensue. If these factors were present in a case, researchers could attempt to determine what other factors prevented environmental scarcity from causing violent conflict.706 Thus, Homer-Dixon suggests, researchers can develop from the findings presented more sophisticated hypotheses, and can test them using a broader range of methodologies, including cross-national statistical