COERCION AND PROTEST - AN EMPIRICAL-TEST IN 2 DEMOCRATIC-STATES

Theory: The theory of protest under varying levels of coercion forms t he context for an investigation of the data on protest coercion in Ger many and Northern Ireland for 11 years (1982-92), aggregated weekly. H ypotheses: The standard inverted-U hypothesis is tested against compet ing unstable (protest and coercion diverge and oscillate); backlash (c oercion increases protest); and adaptation (protesters change tactics after coercion) hypotheses. Methods: Three forms of the biological pre dator-prey model are estimated with two- and three-stage least squares and supplemented with a Bayesian updating test. Results: The predator -prey mechanism fits the German data well, even in a context of low co ercion. The results cast doubt on the inverted-U hypothesis, support t he backlash hypothesis and strengthen the evidence that protesters ada pt. Northern Ireland's terror-based protest and coercion did not confo rm as well to the predator prey model; but protesters did adapt in a s eparate test of Bayesian updating.