Hypotheses of No Relationship

Hypotheses of no relationship exist across the subfields of political science. They typically argue that some previously analyzed variable does not actually matter in the presence of some other variable. To give a very famous example from international relations, a prominent study on the effectiveness of economic sanctions found that when states apply economic sanctions to other states, the targeted state gives in to the sanctioner’s demands 34% of the time (Huffbauer, Schott, & Elliott 1990). By this measure, then, economic sanctions appear to be fairly effective at obtaining the sanctioner’s preferred outcome. This strong of a success rate would justify policymakers’ continued use of sanctions as a tool of statecraft. Another scholar, though, was not as convinced. Because Huffbauer, Schott, and Elliott had made their dataset publicly available, Robert Pape (1997) was able to re-analyze the data. He argued that sanctions actually had no effect on most of the supposedly successful cases. Instead, the threat or use of military force ultimately caused the target’s concessions. Pape carefully traced the sequence of events in each case that Huffbauer, Schott, and Elliott claimed as a sanctions success. He found that in 18 of the 40 supposed ‘success’ cases, the primary cause of concessions was the threat or use of military force. Sanctions were the primary cause of concessions in only 5 of Huffbauer, Schott, and Elliott’s 40 ‘sanctions success’ cases, suggesting that no relationship existed between sanctions and outcomes.