I thought I could keep my research separate from my Venezuela-blogging, but it's not really working out that way. Try this one, from a paper by Paul Sabatier (et al.) on "Perceptions and Misperceptions of Political Opponents":
We hypothesize that actors will evaluate their opponents' motives and behavior in more negative terms than will the rest of the policy community. These are really very straightforward arguments derived from theories dealing with own group bias and with cognitive balance/dissonance.
Most actors start with the assumption that they are right-thinking, virtuous and fair in their judgments (Harrison 1976). Thus anyone who disagrees with them must be mistaken about the facts, operating from the wrong value premises, or acting from evil motives. A fundamental tenet of balance/dissonance theories is that people find it very difficult to balance a positive self-image with a positive image of someone who disagrees with them (Festinger 1957; Abelson et al. 1968, Wicklund and Brehm 1976). The longer opponents persist in their "error" - i.e., resist our sound arguments - the more one begins to suspect their motives or otherwise regard them as dangerous and untrustworthy. This is particularly true if they persist in disagreeing with us on issues which we regard as salient (Lawrence 1976, Judd and Johnson 1981; Freeman and Hittle 1985.)
The dynamics of conflict creates tendencies for negative judgments to escalate over time. An important feature of policy conflicts is that the winners are often able to impose costs on the losers. If A - whom B already suspects of being misguided - imposes costs on B, B's view of A is likely to deteriorate further. As the competition escalates, B is tempted to take more and more questionable measures, but these can only be justified by portraying opponent A in more and more negative terms. Hence, in conflicts which are intense and of reasonably long duration, the dynamics of escalation tend to transform opponents from responsible adversaries into people with extreme and dangerous views (Coleman 1957). Thus opponents begin to impugn each other's motives more and more, and come to increasingly negative evaluations of the other's behavior.
One of the most striking characteristics of [the research] is that the "devil shift" has all the worst features of a positive feedback loop: the more one views opponents as malevolent and very powerful, the more likely one is to resort to questionable measures to preserve one's interests. But the more one does so, the greater the probability opponents will start perceiving one as a very wicked character, thus resorting to unscrupulous countermeasures, thus further confirming one's perception of them as "devils." Suspicion and conflict escalate, and it becomes very difficult to break the cycle.
-Paul Sabatier, Susan Hunger, Susan McLaughlin, "The Devil Shift: Perceptions and Misperceptions of Opponents" in The Western Political Quarterly, Vol. 40, No. 3 (Sept. 1987), 449-476.