摘要：The change of regime in Serbia in October came as a surprise to most people both inside and outside the country. Although surveys in the months before the election were fairly consistent in suggesting that the opposition would win at least a plurality, this was by no means universally accepted as a prediction of the future. In the first place, voter surveys of this type are notoriously unreliable when they are carried out in environments where answers are heavily consequential and might not be given freely, like Milosevic's Serbia. In the second place, experience from previous elections persuasively suggested that there was no reason to expect that the results would be consistent with the way people voted. In the third place, there was no confidence that Milosevic would recognize the results of an election he did not win. So in the two months leading up to the election, it was the consensus in reports by several international governments and monitoring agencies that no great changes ought to be expected from the elections which took place on 24 September.
