The present article investigates contemporaneous and lagged effects of
democracy, coercive capacity, and political conflict on repressive be
havior. As designed, 51 countries from 1948 to 1982 are examined with
an Almon distributed lag model on yearly data (N = 1820). From the emp
irical investigation, both short- and long-term relationships are foun
d to be significantly related to the rate at which censorship and poli
tical restrictions are applied. Past values of democracy are found to
affect repression negatively for five years and past values of disside
nt behavior are found to affect repression positively for seven years.
I conclude that a relatively complex memory structure, encompassing b
oth short- and long-term explanations, must be employed when one attem
pts to understand why stales use repressive behavior.