169 Responses to “Canterbury”

1. Frederic Stansfield

   We need to remember that the Liberal Democrat party has moved to the right alongside the Coalition, and many of its leading people actually seen to find the economic liberalism this involves rather congenial. However, Liberal Democrat voters, particularly in a seat like this, are largely looking for an alternative to the Tories. Opinion polls are consistently showing the results for current voting intentions.

   It is to be seen how far the LIbDems will hold their tactical advantage now that they have taken second place. They did quite well at this in the local elections, where Labour (who are present less well organised that the LibDems here) to my mind clearly had poor results by comparison with those nationwide.

   The 1997 result showed that the comparative Labour and LibDem results here can switch rapidly because of their perceived strengths. One reason of course is the large churn of electors, and large number of first-time voters, because of the very large student vote.

   The Liberal Democrats will have got a considerable number of votes here on their student fees promise. This makes them vulnerable, particularly if Labour get their act together to have a sensible policy on this area (at present they certainly have not done so nationally, and are being far too cautious) and use it well in their campaign.

   Pete Whitehead’s calculations about the 2001 election are helpful and interesting. However, we should remember that in the 2001 election the Canterbury seat was very heavily targeted in the “New Statesman”‘s tactical voting campaign, which was taken up in the national newspapers. The reason was that Canterbury was the seat chosen to balance Folkestone and Hythe, where the then Conservative leader Michael Howard had only a small majority over the LIbDems. This campaign might well have shifted some LibDem votes to Labour in Faversham.

   There is of course another counter-factual in that Labour replaced a hopeless candidate by Emily Thornberry before the 2001 election. If they had had a longer campaign it might have helped them, although having said that Labour had one of their best results in the country here in 2001, even if it was not enough for them to win.

   On Pete’s figures it does look as though Brazier would have scraped hone in 2001 if the seat had then included Faversham. But it would not have been a one hundred per cent certainly.Labour had an experienced agent in 2001 and if the Conservative majority had been down to around 500 to 750, which is what I am now inclined to think likely, he would have been pressing for a recount.

   The LibDems are marketing their 2010 candidateGuy Voisey as their Parliamentary Spokesman here, clearly looking for the long-term. From their point of view, Labour have serious issues to get their act together here and their regional and national agents should be very concerned about this seat if they are not already.

   This seat is far from the safe haven for a redistributed Conservative that has been mooted. although Hugh Robertson who is currently MP for Faversham also has personal connections with Canterbury which would help get the Tories votes. But Julian Brazier is the probable candidate anyway.

   October 11th, 2011 at 11:33 am
2. H.Hemmelig

   “However, we should remember that in the 2001 election the Canterbury seat was very heavily targeted in the “New Statesman”‘s tactical voting campaign, which was taken up in the national newspapers. The reason was that Canterbury was the seat chosen to balance Folkestone and Hythe, where the then Conservative leader Michael Howard had only a small majority over the LIbDems.”

   Erm….Michael Howard did not become Tory leader until 2003.

   And nobody took any notice of those stupid tactical vote campaigns.

   October 11th, 2011 at 12:21 pm
3. Pete Whitehead

   Frederic it seems like you may require a few too many counter-factuals to get your desired result. I take the point of course about people in Faversham possibly voting differently in a different electoral context. This is always a caveat with notional results. However the LDs were fairly weak in the Faversham & Mid Kent seat where Labour had also emerged as clear challenegrs in 1997/2001. Quite simply there were not enough LD voters in Faversham to have made the difference, even if they had all switched to Labour (which of course would not have happened). It is possible in the best case scenario (for you) that the Conservative majority may have been below 1,000 – perhaps 500 to 750 as you say. That is not the kind of majority which would normally result in a recount, but let us accept that your experienced agent called for one. Unless the votes had been counted incorrectly in the first place a recount would merely have confirmed the Conservative win, not reversed it. I am not aware of any suggestions that there were counting irregularities in either Canterbury ot Faversham & Mid Kent. You may of course know differently

   October 11th, 2011 at 12:40 pm
4. Barnaby Marder

   Actually I’d contend that the LDs never had a chance of unseating Michael Howard. Even in 1997 they didn’t get his majority that small, and basically the local LDs there talked a good fight. They did for some time control the nearly coterminous Shepway council, but I don’t think the Tories were ever really going to lose Folkestone & Hythe.

   October 11th, 2011 at 2:48 pm
5. Frederic Stansfield

   I hope I am not being hopeful in a partisan way, although I must admit, Pete that some of us would like political results in this part of the world that would keep people on their toes.

   Canterbury elections were notorious in the nineteenth century, for corruption rather than counting irregularities. But I have not heard anything suggested more recently about Westminster election counts.

   You are right Pete, that a majority of 750 would be unlikely to get overturned on recount, and we are agreeing that in 2001 the Conservatives would very probably have got home narrowly on the new boundaries.

   H. Hemmelig. Thank you for correcting me about Michael Howard’s position in 2001, although the point remains that he was particularly targetted by the centre-left. And you need look no further than the result for this seat to see that people did take notice. The Conservative majority over Labour was nearly halved in an election where in many seats the Conservative majority slightly increased.

   And whilst I know that people’s reports of their past voting intentions are often poor, I myself was wavering between LibDem and Labour in this seat in 2001 and eventually voted Labour. It was the last time I voted for one of the three major parties in a General Election.

   Coming back to the issue of recounts, have we ever discussed what is the largest majority on a first count to be overturned on a recount?

   In addition, has anybody analysed first count and recount figures to estimate how accurate elections counts actually are? Frankly, I don’t believe that, for instance, if a candidate has been declared as getting 10,000 votes we can be sure that there were actually 10,000 votes for that candidate, as opposed to 9,999 or 10,001.

   October 12th, 2011 at 1:57 pm
6. Ben Foley

   I think to some extent practice varies from area to area, but in the counts I have been in, if the numbers don’t add up absolutely (number of papers issued = total number of votes + spoilt and unmarked etc), then efforts are made to trace the votes: but if the number unaccounted for is in the low tens at most, and the majority a lot bigger than that, then the issue is not pursued to the extent of counting every paper again.

   So, sometimes, when the numbers do add up perfectly, 10,000 does mean exactly 10,000, but quite often it could mean anywhere in the range 9,980-10,0020.

   As for numbers on recounts: I am pretty sure that revealing numbers that weren’t the finally declared number would technically be a breach of the secrecy required, so I doubt that systematic analysis of recount vs initial count would be possible.

   On top of that, bundling of votes is an issue: most recounts are at bundle level, so numbers will be more likely to move by 50 votes, or not at all, than move by a handful. So any statistical analysis would be very sensitive to whether there is the odd misplaced bundle identified in recounts.

   October 12th, 2011 at 3:19 pm
7. Jason Harcourt

   Indeed, having been involved in counting myself that’s how I have seen it done, first of all in the verification of votes, that the amount counted out of a ballot box should tally with the amount counted into it at the Polling Station – usually re-counted two or three times before “accepted” if the amounts don’t tally.

   Same again then with the total votes counted – should match the total verified. Which caused problems and a far later night than necessary when an efficient colleague and I ended up having to recount the votes miscounted by the imbeciles we had to work with…!

   October 12th, 2011 at 4:06 pm
8. Ben Foley

   Yes, Jason, that matches my experience from the other side of the table.
   I think though, there is insufficient clarity about precisely when and to what extent numbers that don’t match are ‘accepted’ (as you put it), and when and to what extent recounts take place generally.
   To me the need for clear pre-determined criteria was shown in Florida in 2000. Fortunately our counts have (thus far, at least) been well-enough staffed that there isn’t time for the courts to intervene while re-counting is going on.
   I would like there to be pre-determined criteria for recounts along the lines of “(first) full recount if majority looks like it will be less than 2% (plus 3 times the total of any discrepancies in numbers of ballot paper that have been ‘accepted’)”, so that recounts take place even without agents having to call for them.

   October 12th, 2011 at 9:37 pm
9. Frederic Stansfield

   Thanks for the comments about counting.

   We should not equate the demonstrable accuracy of checking the number of ballots counted against those issued with that of apportioning votes between candidates. Firstly, the verification is a simple count of pieces of paper, whereas the actual count entails discriminating between differently marked papers. And secondly, am I right in thinking that the ballots are verified by box? – it is easier to get a count of a thousand or two papers right than one of twenty thousand or so.

   Jason makes an important point about bundles. I actually remember being at a local election count (not in Canterbury) where a few Tory papers had been placed in front of a Labour bundle. The result was that the Conservative candidate was about to be elected, but after the Labour agent detected the difference the Labour candidate was elected instead with a majority of about 50. The Conservatives failed to ask for a recount!

   Jason’s point that there often has to be several checks at the verification stage before the total is accepted actually suggests that at the counting stage there are errors which are not found.

   I have only once been a Party scrutineer at a General Election count, a considerable time ago. My recollection is that we did correct votes about to be miscounted. I remember in party because I was there for the Labour Party and Labour were narrowly beating the LibDems into third place. So Tory votes miscounted as LibDem needed to be pointed out, but LibDem votes counted as Tory let go.

   If scrutineers do correct votes, the strong likelihood (cf. Signal Detection Theory) is that is that some indorrectly counted votes are going through. And again, if the number of votes counted does not match the number verified, which can be checked, this would again indicated that there are votes incorrectly allocated to candidates, which can only be checked by recounting.

   I am inclined to agree with Jason about the likely magnitude of errors. It is a matter to be established, but might guess is that the Standard Error for statisitics of General Election voting might be about 10.

   Why should the results of counts, other than the final one, be kept secret? My recollection is that they have come out in the past, I vaguely think for Peterborough when there were seven recounts. The first couple of counts were way out becasue some papers had been left in a ballot box, After that I think thy were varying by one or two each time.

   Presumably we should know if a recount has changed the result because the party requesting the recount will get their money back and it will show in the accounts.

   I am inclined to think that the bundles should be rechecked, by recounting them and by rifling through them to checked that there has not been some gross mistake such as putting a vote for one party at the front of a bundle for another, if the majority is less than just over 600. If the bundles are right, I cannot see any point in having a full recount of individual papers if the putative majority is over about 200 at most, and probably 100.

   If we don’t have actual data, we could presumably get relevant information by running dummy counts as an experiment. One for the political scientists to collaborate with experimental psychologists (or cognitive scientists).

   I think most of us suspect that there are unnecessarily many recounts, except in that parties want them for propaganda purposes. If some money were spent on establishing the accuracy of counts (instead of pretending they are perfect, which is ridiculous), this could make savings by reducing the number of recounts.

   Reduction of recounts would probably be best done by informal agreement between the major parties, with the knowledge of the returning officers. Formal legislation would have to take account of issues like the size of bundles. It could also cause inflexibility if some particular reason arises at the time to think that a count may be inaccurate – cf. the Florida “chads” issue.

   Finally, how many recounts have taken place in how many seats at recent General Elections? And how many results do we think to have been changed as a result of recounts (if you are replying, why not gossip about named seats!)? In response to the latter question, I suspect not many if any.

   October 15th, 2011 at 10:35 am
10. Frederic Stansfield

    I haven’t heard or seen anything of the Money Reform Party since the General Election, so far as I can recollect. Any news of them?

    February 7th, 2012 at 5:14 pm
11. swanarcadian

    Never heard of them, frankly.
    They don’t even appear to have a Wikipedia article yet (it just re-directs to Bromley & Chislehurst by-election, 2006)

    February 7th, 2012 at 6:59 pm
12. Frederic Stansfield

    Money Reform Party:-
    http://www.moneyreformparty.crg.uk

    February 8th, 2012 at 11:35 am
13. Frederic Stansfield

    Sorry, typo

    http://www.moneyreformparty.org.uk

    February 8th, 2012 at 11:36 am
14. Barnaby Marder

    The Conservatives comfortably held Blean Forest in a by-election on Thursday. Labour stood a candidate this time, having not done so last year, and got a creditable second with 24% of the vote, slightly more than half the Tory score.

    September 22nd, 2012 at 9:45 am
15. THE RESULTS

    Have to be honest I was very surprised to see Julian Brazier’s majority fall as sharply as it did in 2010. The Anti-Conservative vote in Canterbury is well-documented as we all know, and it appeared that a lot of Labour votes from 2005 went straight to the Liberal Democrats.

    September 22nd, 2012 at 4:26 pm
16. Joe

    Its a naturally divided opposition, even on a fairly low vote share the tories are still very safe here.

    September 22nd, 2012 at 5:50 pm
17. Chris K

    I don’t know Canterbury itself particularly well, but I used to go Whitstable a lot as my grandmother lived there.

    Aside from an area near the centre of the town (near the harbour) Whitstable is a fairly affluent place and I would imagine it to be safely Tory, and the same with Chestfield, Swalecliffe which are effectively part of the town. Seasalter is a little more run down but very small.

    I don’t know the latest population but would imagine Whitstable to have grown since the last census, and there is certainly some significant new build on the edge of the town.

    Given this, as long as Whitstable remains in the Canterbury constituency I would imagine it would be difficult for Labour/LD’s to win the constituency, however anti-Tory Canterbury may or may not be.

    September 22nd, 2012 at 6:24 pm
18. Barnaby Marder

    I suspect, from what I’ve seen of Whitstable, that it would have been fairly close between Lab & Con in 2001, but would usually be a fairly good town for the Tories.

    September 22nd, 2012 at 10:17 pm
19. Harry Porter

    I presume Whitstable has been part of the City of Canterbury district since 1974? Has it always been in this constituency?

    February 27th, 2013 at 8:34 pm