The exit poll has developed primarily as a media tool - it consists of an opinion poll compiled by conducting interviews with voters 'exiting' from the polling station. Since clearly they are inherently unreliable, not being an actual record of how people voted, it is important that they be distinguished from 'mef03b'.
The exit poll is a species of opinion poll and the media would be well advised to exercise all the same caution with them. See 'mee05a'. The additional problem with exit polls is that they appear to give an indication of how people have actually voted, as opposed to how they intend to vote, so it would seem clear that any restrictions that applied to opinion poll reporting while voting was continuing would apply with even greater force to exit polls.
The usual practice is to use exit polls as an early taster in results programmes, before actual results are available. Morning newspapers, whose deadlines are often before any significant results are available will also make use of them. The issue can be more complicated in countries spread across several time zones, where early exit poll results might have an effect on later voting patterns. In the United States most media voluntarily refrain from reporting exit polls or early results from the East Coast.
Part of the problem with exit polls, of course, is whether people will honestly report how they have voted. This is likely to be a particular issue in transitional democracies or in situations where there has been widespread intimidation. Voter education should have stressed that each person's vote was secret. In the Zimbabwe elections of 2000, a South African organization, the Helen Suzman Foundation, designed a much more complex exit poll format to take account of the danger that people would not honestly state how they voted. But one of the results of this complexity was that it was not released until later. Its main purpose, in the end, was to provide some evidence for how intimidation might have affected the result. Unfortunately, since the exit poll findings were not available until after the actual result was known, it did not receive the media attention that it deserved. With the media, speed is everything.
