ISPS Experiments Workshop: Kevin Russell, Yale

Kevin Russell
Graduate Student in Political Science, Yale University

"A Survey Experiment Concerning ANC Politics and Constitutional Compliance"

Note: I intend to conduct this experiment in November or December. The motivation and concepts are fairly well developed in other parts of my dissertation. The nuts and bolts of the actual survey experiment still need some work. I would very much appreciate feedback on making the scenarios credible, addressing external validity concerns, sample size and sample selection (given costs), number of treatment groups, and measuring the dependent variable.

Abstract: It is perhaps a truism that even the most idealistic of political parties if entrenched in office for too long with little threat of being displaced will eventually abuse that position and subvert the public interest. However, if they commit abuses over time, one would eventually expect a backlash, since a population will perceive abuses as wrong and they will have downstream negative consequences on governance. How much transgression will a population put up with? What other political factors shape popular perception of transgressions? How does this vary across different segments of the population? Using the case of South Africa, this survey experiment tests three plausible hypotheses for how the ANC can retain its legitimacy despite some abuses of power: (1) by providing material benefits; (2) by appealing to a greater moral purpose; (3) by assigning blame outside the ANC or to “a few bad apples.” These ANC political arguments, which they make regularly, will be presented to respondents who will also have received one of three scenarios of hypothetical ANC transgressions (though mirroring real events) to see how much each argument “mitigates” any negative attitudes toward transgressions. In addition to the relative effectiveness of these hypotheses, I test the hypothesis that each of the political arguments will mitigate the negative effect of transgression, especially among ANC supporters, but that this mitigation will disappear at high levels of transgression because people do ultimately care about procedural justice applied to their leaders. The conclusion of these tests has important implications for the development of rule of law in South Africa or any country with a new constitution and a strong political group positioned to take advantage of a historical or political claim to legitimacy.