8
(e.g., Guatamala, Honduras, Panama, and Peru).
7
Consequently, preferences for
democracy or the international order should not be viewed as concomitant. In addition, it
is important to note that the distribution of conflict differs between regime type and status
quo preferences. While the modal category for disputes categorized according to regime
type is the mixed-dyad, the modal category for status quo preferences is the anti-status
quo dyad. Thus, regime type and status quo preferences of disputants can be said to be
distinct in both the theoretical and statistical domains.
DATA AND MEASUREMENT
Data on IGO intervention was obtained from the Project to Assess Non-Violent
and Direct Action (PANDA). This data set codes direct action by an actor vis-à-vis a
target from 1984 to 1994. These data were combined with all available dyadic
militarized disputes found in the Correlates of War Militarized Interstate Dispute data set
(N=338) which lasted more than one day (Jones, Bremer, and Singer, 1996). Thus, the
unit of analysis is the conflict-dyad/year.
Dependent Variable: IGO Action
The PANDA data set constructs four separate categories for IGO actions. These
include appeal, coercion, manipulation, and force. However, due to the limited
occurrence of non-forceful actions for some groups of IGOs I consider only three levels
of action in the multivariate analyses for the United Nations where 0= no action, 1=
action short of force (i.e., appeal, coercion, and manipulation) and 2= the use of force by
an IGO in a dyadic dispute. A low frequency of interventions necessitates a simpler
categorization for democratic/pro-status quo and authoritarian/anti-status quo IGOs. For
these latter two types of organizations the dependent variable is a dummy variable where
0= no action and 1= action by an IGO.
7
The relationship between regime type and preferences for the international status quo have been examined
in detail by Lemke and Reed (1996). In this paper, pro-status quo states and democracies correlate at 0.52.
Convention