ABSTRACT
The confluence of political party polarization, interparty conflict, and contentious politics is nearly a universal axiom in Africa. Although the reasons for parties to protest and rally are many, pinpointing who actually participates is rather tricky. Building on previous literature regarding the influential power of political parties on member behaviour, this study proposes political party affiliation strength as metric for identifying whom within political parties is more and less likely to participate in in-person political protests and rallies. Using data on Kenya, Nigeria, and Tunisia, this study finds strong supporting evidence that party affiliation strength effects protest participation propensity mediated by party-government linkages. Those strongly affiliated to political parties that enjoy majority power in government are less likely to protest, and strong affiliates of non-majority power-holding parties are more likely to protest. Preliminary evidence presented here also demonstrates those with non-political party affiliation are significantly more likely to post online about political issues rather than participate in in-person contentious politics. This study’s findings help further bridge the gap between individual political psychology and macro-level political behaviour.
KEYWORDS:
Disclosure statement
No potential conflict of interest was reported by the author(s).
Notes
2. The distance from Kenya to Nigeria = 2,108 miles; Kenya to Tunisia = 2,971 miles, Nigeria to Tunisia = 1,708 miles.
4. Treatment effects models estimate experimental-type causal effects from observational data. Here, the treatment assignment (i.e., party affiliation strength) was not controlled but rather dictated by respondents. Given we are interested in the differential treatment effects of party affiliation strength on various types of participation, the model used here estimates the causal differential effects of party affiliation strength on the outcomes of interest, without covariates. This is referred to as a simple predictive comparison. This type of model assumes the effect of X on Y can be linear in its parameters; that observations are independent; X is measured without error (note that error in predictors of the latent classes does not bias estimates of differential effects); and error terms are normal within each latent class.
Additional information
Notes on contributors
Joel Blaxland
Joel Blaxland, PhD in Political Science (Temple University, 2018); Assistant Professor, Western New Mexico University (2018–); current interests: peace conflict processes, political psychology in violence-prone settings; book: Insurgency Prewar Preparation and Intrastate Conflict: Latin America and Beyond (Palgrave Macmillan, 2020)