1997 ITN/MORI General Election Exit Poll

MORI interviewed a total of 15,761 voters as they left polling stations on 1 May 1997. Interviews were conducted at one polling station in each of 100 Conservative-held marginal constituencies (those where the Conservative notional lead in 1992, using the Rallings & Thrasher figures to compensate for boundary changes, was less than 20% of the votes cast). Respondents were selected at random using a fixed interval, and asked to fill in a (secret) ballot paper to indicate how they had just voted. Refusals (15% of those approached) were replaced with similar voters outside the fixed interval. The exit poll was conducted for ITN, and the data available at 10 p.m. (based on those who had voted by 9.15) was used as the basis of their broadcast seats projection.

Results were telephoned back through the day and the data were processed by MORI On-Line Telephone Surveys. The poll measured the average change in the vote for each of the three main parties, and the cumulative results through the day at each phone-in cut-off time are shown below. (The poll was not intended to make any national measurement of the vote, only to allow predictions in the 100 constituencies that had been chosen as most useful in projecting the number of seats, so comparisons are made with the outcome in the constituencies polled rather than with the national result.)