PREDICTING THE OUTCOME

Press, Volume XCI, Issue 27643, 27 April 1955, Page 13

PREDICTING THE OUTCOME

“Back A Landslide Each. Way” (Special Correspondent N.Z.P.A.) (Rec. 9 p.m.) LONDON, April 26. Those who want to predict the outcome of the British General Election must study both the detailed effect of the recent redistribution of Parliamentary seats and more generally the arithmetic of the British electoral system, says Mr David Butler, official Fellow of Nuffield College, Oxford, and author of “The British General Election of 1951,” writing in the “Daily Telegraph,” He says that of the 625 constituencies at present in existence, 215 will have their boundaries altered, although in 43 cases the alteration is negligible. In many of the remaining 172 constituencies, the changes are small and party prospects are not affected.

He says it is difficult to answer the auestion what the net effect of all wse changes is likely to be because information is hard to come by and because the consequences of the redistribution will vary with the result of the election. “It seems likely that if vcnnle were to vote in the next e'ect *'•■» just as they did in the last, the '~~njcrvatives would increase their majority by seven. This would give the Conservatives a majority over Labour of 33. If the Conservatives are given the benefit of every doubt, their majority is likely to be 45, and if Labour is given the benefit of every doubt, the Conservatives’ majority is likely to be 13.” Errors Widen

Mr Butler thinks the effect of the new boundaries will be limited and adds: “But that does little to reduce the uncertainty which anyone should feel in predicting the outcome. Even the most skilled of public opinion pollsters claims no more than accuracy to within 1 or 2 per cent, and the party machines have no secret methods which are any more reliable. An error of only 1 per cent, in estimating how the public vote will divide can lead to an error of 40 seats in estimating a Parliamentary majority. “This can be very simply shown. The British electorate behaves very uniformly in Parliamentary elections, both in the numbers going to the poll and in the turnover between the narties. A change in one constituency tends to be reflected very closely in all others. *

“One can assume that if nationally there is a 1 per cent, swing to one party, it will gain on the average as many seats as its opponents hold by that margin. “One reads in the phrases like

‘Conservatives confidently expect a majority of about 40 seats’ or Xabour circles still anticipate victory—but by only 30 seats.’ If in the last days of the 1951 campaign, only two people in every 100 had switched from Conservative to Labour, Labour would have had a majority of 40. And if two people in every 100 had switched from I -abour to Conservative, the Conservatives would have had a majority of 94. "So it may be as well to hedge your hets. Back a landslide each way, he adds.