Norwich North by-election poll results
The results are in for the Norwich North by-election. As you can see, I have illustrated it with a pie chart rather than a first-past-the-post-legitimizing bar chart.
There are four sets of hard numbers we can use to discuss this:
- The current July 2009 Parliamentary election poll
- The previous 2005 Parliamentary election poll
- The aggregate of June 2009 local election votes for wards in the constituency (link)
- The 26 June ICM telephone poll (link)
Here they are lined up in a nice handy-to-read table.
| Poll | Conservative | Labour | Libdem | Green | UKIP | Other | Non-voting |
|---|---|---|---|---|---|---|---|
| 2005 Parliament | 15638 | 21097 | 7616 | 1252 | 1122 | 308 | 29944 |
| 2009 Local | 10756 | 4708 | 4626 | 4290 | 2106 | 228 | 49798 |
| ICM (scaled) | 11688 | 10313 | 5157 | 4813 | 2406 | 40518 | |
| 2009 Parliament | 13591 | 6243 | 4803 | 3350 | 4068 | 2322 | 40518 |
I have taken the ICM poll’s headline predictions from table 5 of their report and multiplied the percentages by the actual turnout.
The figures show a collapse in the Labour vote by a factor of more than three (with the LibDems declining as well) between 2005 and 2009. But Labour also has the greatest proportionate increase (ignoring UKIP) between its Local election performance the previous month and the by-election.
Turnout by the supporters of a party is one of the crucial factors in any election. Table 2 of the ICM poll report gives a really detailed breakdown of turn-out predictions. In the poll, 40% reported that they were certain to vote, and 4% reported that they were almost certain (score of 9 out of 10) to vote. The actual turnout was 45.9%.
One BBC reporter tried to explain the results by merely subtracted the two numbers, and writing:
Some 14,800 people who voted Labour in 2005 did not vote on Thursday. There was no Crewe and Nantwich-style flood of Labour voters to the Tories.
By actually reading the ICM poll I get a different story, because according to it 49% of Labour voters said they were certain to vote, 45% of Lib Dem voters were certain to vote, and only 37% of Conservative voters were certain to vote.
In other words, Labour voters were turning out in greater proportions than supporters of the other parties.
We need to look at its numbers for the 2005 voting statements, because the former MP Ian Gibson was especially popular. But here we find that 51% of those who voted Labour in the 2005 General election were certain to vote in 2009, and 51% of those who voted Conservative in 2005 were certain to vote.
Clearly there is some vote switching going on, so we look at tables 3 and 4. Table 3 has the raw figures, including a whopping 34% “Don’t know”s and “Refused”s, while table 4 scales these non-responses out of the numbers.
Of those who gave an answer to the pollsters (Table 4), 91% of those who voted Conservative in 2005 intended to vote Conservative in 2009, with 6% going to the Green party. But only 54% of those who voted Labour in 2005 intended to vote Labour in 2009, with 17% going to Conservatives, 11% to LibDem, and 14% going to Greens. Also of note: 21% of 2005 LibDem voters intended to defect to Green and 12% to UKIP.
So the story appears to show that half the Labour votes were fleeing to other parties, but those who remained with Labour were still going to turn out.
Or, another possibility, Ian Gibson’s personality attracted votes from supporters of all other parties, and the 2009 results simply reflected a return of these votes to the parties to whom they belonged.
That, and a whole load of the “Don’t know”s turned out for the Conservatives.
The ICM poll in table 3 (the one which includes the “Don’t knows”) also reports that 40% of the 18-25 group said they intended to vote Conservative (support of the over 65s was 34%), and 39% of the Unemployed intended to vote Conservative. Also, 37% of students were Lib Dems, and none were Green.
This, to me, sounds like bonkers, and I could be reading the tables incorrectly. Nevertheless, if the Conservative party knew they were going after — and capturing — the young or unemployed vote, then a Male, 34-65, B, Other like myself is going to be pretty unfamiliar with the thrust of their campaigns.
The Guardian printed in its byelection post-mortem:
The campaign started after the Conservatives declared that they might have to cut public spending in most government departments by 10% after the general election, and Labour attacked [the Conservative candidate] aggressively on this issue. One Labour leaflet suggested that the Tories could close up to 10% of schools in the country, and another said the Tories were “threatening to do away with free TV licences and bus passes for the elderly”.
…
Tory strategists believe that Brown was using the byelection to road-test a “Tory cuts” campaign and that the result shows that this approach does not work.
The threat to free TV licenses and bus passes for wrinkly grannies is here. I can’t find the 10% of schools will close leaflet.
I wish I knew precisely how the Labour party supposedly worked out that attacks on the Conservative party for alleged public spending cuts weren’t working. Knowing about the flows and sources of information is critical to the public understanding of the process, as well as explaining some of the systematic failures whereby reasoned public opinion does not get translated into public policy.
On the plus side: aside from all the other disagreeable content in this final Labour leaflet, I notice that it is the only leaflet that published the correct ICM poll results for the constituency on its front page, rather than some other bogus figures intended to mislead the public into voting tactically in the wrong direction.
Roll on the next by-election, which I believe will be Glasgow North East.
Here, the real story is not the expenses scandal which precipitated the resignation of the sitting MP Michael Martin, but the fact that for the last decade that he has held the office of the Speaker of the House, the people of Glasgow East have been officially denied any democratic representation.
I wonder if anyone there has noticed.
