Letter: Deceptive crumb of comfort for the Conservatives

Sir: Ivor Crewe's article 'Voters show unprecedented degree of dissatisfaction' (8 May) offers the Conservatives a crumb of comfort in what he calls the raw voting figures from last Thursday's county elections. Yet these are merely estimates of the national share of the vote, derived from a limited number of marginal parliamentary constituencies used by the election-night programmes to make parliamentary projections. They assume the sample is representative and that voting changes are uniform.

Detailed analysis of the votes cast shows that the assumptions are flawed and Professor Crewe's reading of the figures misleading. He concludes that his estimated Tory vote (31 per cent) is only fraction-

ally lower than in 1985. Yet in county after county, if he had added up and compared the full figures, he would have found a large drop. Thus, here in Kent and his own county of Essex, the Conservative share was four points lower than in 1985.

This indicates that the true reading of the Conservatives' national support from the shire county elections is 28 to 29 per cent, barely ahead of the Liberal Democrats (about 27 per cent) and easily the lowest level of support the party has ever had in a set of local elections.

The reason the actual figures are a lot worse for the Tories than the projections on election night, used by Professor Crewe, is that there was such substantial regional and tactical variation in the way people voted; in addition, the Conservative slump was greater in what were previously their stronger areas than in the more marginal seats used for the sample. This also means, of course, that any projection from the mainly Southern shire counties should be treated with considerable caution since London, Scotland and the mainly Northern metropolitan counties did not vote.

However, it also means that the parliamentary projections on the BBC and ITN were wide of the mark. Both showed the Conservatives winning almost 200 seats and the Liberal Democrats 50-something. Constituency by constituency calculation here in Kent shows that the Conservatives would have lost 12 of their 16 seats, with seven going to the Liberal Democrats and five to Labour. As Kent was far from the Conservatives' worst area (it was the only county council on the south coast where they remained the biggest party), this indicates that Conservative Central Office should be very grateful that the full scale of their votes lost (and their MPs in danger) was obscured by the way the results were reported.

Yours faithfully,

MICHAEL STEED

Canterbury,

Kent

8 May