<P> According to one study, (Ray 2003) which uses a rather restrictive definition of democracy and war, there were no wars between jointly democratic couples of states in the period from 1816 to 1992 . Assuming a purely random distribution of wars between states, regardless of their democratic character, the predicted number of conflicts between democracies would be around ten . So, Ray argues that the evidence is statistically significant, but that it is still conceivable that, in the future, even a small number of inter-democratic wars would cancel out such evidence . </P> <P> Some authors criticize the definition of democracy by arguing that states continually reinterpret other states' regime types as a consequence of their own objective interests and motives, such as economic and security concerns (Rosato 2003). For example, one study (Oren 1995) reports that Germany was considered a democratic state by Western opinion leaders at the end of the 19th century; yet in the years preceding World War I, when its relations with the United States, France and Britain started deteriorating, Germany was gradually reinterpreted as an autocratic state, in absence of any actual regime change . Shimmin (Shimmin 1999) moves a similar criticism regarding the western perception of Milosevic's Serbia between 1989 and 1999 . Rummel (Rummel 1999) replies to this criticism by stating that, in general, studies on democratic peace do not focus on other countries' perceptions of democracy; and in the specific case of Serbia, by arguing that the limited credit accorded by western democracies to Milosevic in the early' 90s did not amount to a recognition of democracy, but only to the perception that possible alternative leaders could be even worse . </P> <P> Some democratic peace researchers have been criticized for post hoc reclassifying some specific conflicts as non-wars or political systems as non-democracies without checking and correcting the whole data set used similarly . Supporters and opponents of the democratic peace agree that this is bad use of statistics, even if a plausible case can be made for the correction (Bremer 1992), (Gleditsch 1995), (Gowa 1999). A military affairs columnist of the newspaper Asia Times has summarized the above criticism in a journalist's fashion describing the theory as subject to the no true Scotsman problem: exceptions are explained away as not being between "real" democracies or "real" wars . </P> <P> Some democratic peace researchers require that the executive result from a substantively contested election . This may be a restrictive definition: For example, the National Archives of the United States notes that "For all intents and purposes, George Washington was unopposed for election as President, both in 1789 and 1792". (Under the original provisions for the Electoral College, there was no distinction between votes for President and Vice-President: each elector was required to vote for two distinct candidates, with the runner - up to be Vice-President . Every elector cast one of his votes for Washington, John Adams received a majority of the other votes; there were several other candidates: so the election for Vice President was contested .) </P>

How does the united states and authoritarian regimes both claim to be democratic