BBC HOMEPAGE | NEWS | WORLD SERVICE | SPORT | MY BBC help
news vote 2001search vote 2001
 You are in: Vote2001: Opinion Polls
VOTE2001 
Main Issues 
Features 
Crucial Seats 
Key People 
Parties 
Results &  Constituencies 
Candidates 
Opinion Polls 
Online 1000 
Virtual Vote 
Talking Point 
Forum 
AudioVideo 
Programmes 
Voting System 
Local Elections 
Nations 

N Ireland 
Scotland 
Wales 

BBC News

BBC Sport

BBC Weather
Monday, 18 June, 2001, 09:45 GMT 10:45 UK
ICM takes polling honours

David Cowling

Some 29 voting intention polls were published throughout the 2001 campaign.

Two of them (both by Rasmussen) registered the actual Conservative share of 33% at some stage in the campaign; only two registered the 19% share achieved by the Lib Dems; and every single poll put Labour above its 42% share.

The table below records the range of voting intention figures for the five polling companies throughout the campaign:

Lab range Con range LD range
Gallup 47-48 30-32 13-18
ICM 43-48 28-34 14-19
Mori 45-55 26-31 11-18
NOP 47-49 30-32 13-16
Rasmussen 44-46 32-33 13-17
Result 42 33 19

The fluctuations were less marked this time compared with 1997.

Whereas in 1997, four out of five polling companies put Labour at 50% or more at some stage in the campaign (compared with the party's eventual 44% share of the vote) this time only one out of five companies did so.

Of Mori's eight published polls, six put Labour on 50% or more.

Mori changed its methodology towards the end of the campaign and as a result cut Labour's share by an extraordinary 7% and increased the Lib Dem share by 5%.

The company claimed that this simply repeated what itdid in its final poll of the 1997 campaign, but the difference then was utterly negligible. Gallup also changed its methodology during the 2001 campaign, but the difference that resulted was only 1% in the Conservative share.

Robert Worcester of Mori published some highly sceptical remarks about the methodology of Rasmussen, but the latter emerged with a. better record than his own company.

The final test

However, the traditional test for the pollsters is their final published poll. So, what was the record for 2001?

The relevant polls are set out in the table as well as (in brackets) how far they diverged from the real result:

Con
%
Lab
%
LD
%
Lab lead
%
Gallup 30 (-3) 47 (+5) 18 (-1) 17 (+7)
ICM 32 (-1) 43 (+1) 19 (0) 11 (+1)
Mori 30 (-3) 45 (+3) 18 (-1) 15 (+5)
NOP* 30 (-3) 47 (+5) 16 (-3) 17 (+7)
Rasmussen 33 (0) 44 (+2) 16 (-3) 11 (+1)
Result 33 42 19 9
* NOP polled for the Sunday Times and therefore its final survey was sampled between 31 May-1 June.

ICM's final poll for the Guardian came the closest to the final result. It gave the Dem share exactly, understated the Tories by 1% and overstated Labour by just 1%.

Exit polls

Two exit polls were conducted throughout 7 June: by NOP for the BBC and Mori for ITN. At 10pm the two predictions were:

Actual BBC +/- ITN +/-
Lab seats 413 408 -5 417 +4
Con seats 166 177 +11 154 -12
LD seats 52 44 -8 58 +6
Lab majority 167 157 -10 175 +8

Both polling companies therefore provided their respective clients with remarkably accurate forecasts.

 A/V CONSOLE
BBC RADIO NEWS
BBC ONE TV NEWS

Internet links:


The BBC is not responsible for the content of external internet sites
©BBC