In a recent Note we investigated the incidence of tactical voting in the British election of 1987. A major hypothesis was that in situations where it made sense (i.e., in which voters preferred a party that was a long way from contention, so that a vote for that party was likely to be wasted), tactical voting would be much more frequent than had hitherto been assumed. We discovered that this was indeed the case; in what we called objectively tactical situations, tactical voting was as high as 25 per cent or more in the general case, rising to more than 50 per cent among highly educated weak partisans who had strong negative feelings about the winning party. In the process, however, we discovered what appeared to be an anomaly in need of further exploration.
