With the advent of the public opinion poll, politicians began to have access to highly reliable information about the electorate's views on issues. Prior to this development, party leaders could only make educated guesses about public opinion. These guesses, however, were often incorrect, since they were based on unsystematic evidence. But armed with polls, parties should avoid such errors, approximating Downs' (1957) assumption of certainty. If so, rational parties should converge near the center of the distribution of public opinion. Or in other words, parties should no longer polarize on highly salient issues that confront the nation. This conclusion has important implications for the study of partisan realignments. The best work on the subject by scholars like Sundquist (1983) and Carmines and Stimson (1989) argue that one requirement for a realignment is that the parties must polarize on an issue of high salience to the public. Yet well-informed, rational parties should not engage in such behavior, suggesting that critical realignments may be things of the past. Note that partisan change still occurs--perhaps along the lines of Key's (1959) notion of secular realignment or Carmines' and Stimson's (1989) concept of "issue evolution."