Democratic peace theory posits that democratic regimes are in general about as conflict- and war-prone as non-democratic regimes; on the other hand, democratic regimes have rarely gone to war against one another (Axelrod 1984, Maoz & Abdulali 1989, Maoz & Russett 1993, Lake 1992, Merrit & Zinnes 1991, Weart, 1998). In this research paper, it is posited that another regime type, namely the federation as a regime type, may be equally robust in predicting or post-predicting for the occurrence of war. In proposing to contrast the case of democratic peace with a federal one, I plan to eventually employ statistical approaches which will show whether federalism along with other important factors--most notably
(a) autocratic versus democratic regime types--,
(b) affective variables (Geva & Mintz 1993),
(c) perceived utility (Bueno de Mesquite & Lalmon), or
(d) perceived national interests and structural factors, such as (e) level economic development (Gartzke 1998)-"each also plays positive or negative roles in a state’s willingness to go to war or to avoid major international conflict.
It should be noted, however, that concerning war and democratic peace theory, research on pre-20th Century war has necessarily required a relaxation of the definition of democracy to mean (1) periodic, (2) competitive elections, or that (3) the powerful can be kicked out of power, and (4) that a body of citizens hold equal rights, regardless of their class or status. Using such definitions, Rummel (1999) notes that “Weart, (1998) and others, finds that as far back in history as classical Greece, democracies rarely, if at all made war on each other. Weart, however, concludes that using a relaxed definition of democracy, democracies fight each other ‘not at all’." The claim that democracies never fight teach other is certainly called into question by both Rummel and Lake (1992).
This federalist paper is a foray into the research for testing a new federal peace hypothesis, specifically contrasting both David Lake’s (1992), “Powerful Pacifists: Democratic State’s and War”, and Maoz and Russett’s (1993) "Normative and Structural Causes of Democratic Peace, 1946-1986" democratic peace hypotheses with a similar federal theory. Lake found that “democracies, constrained by their societies from earning rents, will devote greater absolute resources to security, enjoy greater social support for their policies, and tend to form overwhelming counter coalitions against expansionist autocracies (Lake: 24).” In addition, he indicated that all these political-economic factors place democracies in the position of likely winning the wars that they fight in. Similarly, Maoz and Russett claim that democratic peace was equally as robust when dyadic years were involved in the analysis. Dyadic year theory, and analysis of the state actors in the interstate system, claims that regardless of regime type some states over time (as measured by time series analysis, etc.) tend to have more trouble and conflict with one another than with other regime types.
A focal point of the theory is also based upon the perceived ability of federal states to handle diverse interests (or heterogenous group interests) and conflicts. For example, Elazar (1994a), in his Federalism and the Way to Peace, posits that “federalist solutions” to conflicts have until now been a rare focus of international relations. Elazar critically speaks of a missed opportunity in noting that the George H. Bush administration, in the days leading up to and through the end of the 1991 Gulf War, never seriously considered a federal solution in dealing with Iraq and Saddam Hussein . Elazar specifically claimed that 1991 Iraq--with regionally congregated groups of minorities of Shiites in the east, Kurdish peoples in the north, and its Sunnis in the west--was definitely a prime candidate for a federal peace treaty. In turn, particularly concerning Israeli and Palestinian territorial divisions of power in the near future, Elazar claimed that it is possible that similar federal or confederal solutions eventually would open the way for a more sustained peace in all of the Middle East.
Finally, Maoz and Russett (1993) have concurred that federalism likely has a strong negative effect on the propensity of such a nation to wage international war. They state: "Federalism is probably not as severe as a constraint on foreign policy as on domestic policy, but even on foreign policy it somewhat restricts the ability to mobilize economic and political resources rapidly in the event of a serious international dispute. It also provides an institutionalized base from which regional political leaders can challenge government policy (Maoz & Russett: 629)."
In contrast, Zinnes and Merrit (1991) have hypothesized that the differences in regime types might have little or nothing to do with foreign policy processes and negotiations, as foreign policy processes are usually dominated by elites who drive the process in approximately the same manner around the globe regardless of regime type. Zinnes and Merrit (1991: 230) also find that another variable, namely economic structure, of societies is likely of greater importance. This implies that regardless of regime type, "economically modernized societies have too much to lose to gamble on any serious war." Similarly, Gartzke (1998) has indicated that far too much, which has been ascribed to democratic peace, is actually the result of commonly perceived national interests. For this reason, a political-economic control variable is developed and applied in this paper to help test that dimension of federal and democratic peace theories.
One of the proposed models for the research is as follows: No War^=If [(fed)*(polity)*(demostate))] is positive versus a positive [(fed)*(polity)*(demostate)]
In line with replicating a part of Lake’s research, three hypotheses were proposed. Initially, they focus on (dyadic) year relationships of actors whereby at a minimum a pair of actors face off in conflict. The dependent dichotomous variable for war is based on having a level 6 score in the fatality category of the MID data set.
Of the hypotheses to be tested one of them is clearly normative, another one is clearly structural, while the others are mixed.