Held on 1997-05-01. The result below is the official outcome; underneath it, where we hold the polls, is how close each pollster’s final poll came.
How the final polls compared
Each pollster’s last poll in the 30 days before this election, and how far it was from the result. Only parties the pollster reported are counted.
| Pollster |
Final poll |
Days before |
Parties scored |
Difference |
| ICM |
1997-04-29 |
2 |
3 |
1.3 percentage points |
| Party |
Poll |
Result |
Difference |
| Conservative Party |
33.0% |
31.0% |
+2.0 percentage points |
| Labour Party |
43.0% |
44.0% |
-1.0 percentage points |
| Liberal Democrats |
18.0% |
17.0% |
+1.0 percentage points |
|
| Gallup |
1997-04-29 |
2 |
3 |
1.7 percentage points |
| Party |
Poll |
Result |
Difference |
| Conservative Party |
33.0% |
31.0% |
+2.0 percentage points |
| Labour Party |
46.0% |
44.0% |
+2.0 percentage points |
| Liberal Democrats |
16.0% |
17.0% |
-1.0 percentage points |
|
| Harris |
1997-04-28 |
3 |
3 |
2.0 percentage points |
| Party |
Poll |
Result |
Difference |
| Labour Party |
48.0% |
44.0% |
+4.0 percentage points |
| Liberal Democrats |
15.0% |
17.0% |
-2.0 percentage points |
| Conservative Party |
31.0% |
31.0% |
0.0 percentage points |
|
| Ipsos MORI |
1997-04-30 |
1 |
3 |
2.0 percentage points |
| Party |
Poll |
Result |
Difference |
| Conservative Party |
29.0% |
31.0% |
-2.0 percentage points |
| Labour Party |
46.0% |
44.0% |
+2.0 percentage points |
| Liberal Democrats |
19.0% |
17.0% |
+2.0 percentage points |
|
| NOP |
1997-04-29 |
2 |
3 |
4.0 percentage points |
| Party |
Poll |
Result |
Difference |
| Labour Party |
50.0% |
44.0% |
+6.0 percentage points |
| Conservative Party |
28.0% |
31.0% |
-3.0 percentage points |
| Liberal Democrats |
14.0% |
17.0% |
-3.0 percentage points |
|
How this is measured: for each past election we take each pollster’s last poll in the 30 days before polling day and compare it, party by party, with the result. The figure is the mean absolute difference across the parties that appear in both. Only parties the pollster actually reported are counted. A late shift in opinion is not the same thing as a polling error.