some random exit polls/election day polls for other ballots

[Defenestrated Fipplebox]

Sept 11, 2021 20:03:29 GMT nodealbrexiteer said:

Sept 11, 2021 19:50:04 GMT Defenestrated Fipplebox said:

Sorry bad question. 

Are they in the week  before each polling day or 5 days or fewer etc. 

Thanks

What johnloony said-they are from polling day itself,if you like the rolling results from the exit poll vote shares as polling day progressed.

Thanks. 

[nodealbrexiteer]

An excellent article on ITN's 1987 exit poll from our very Robert Waller:

amsr.contentdm.oclc.org/digital/collection/p21050coll2/id/8327/rec/1

Some points I picked out:

Total respondents 20172. Over 15500 voters at 100 polling stations in the seat projection poll(60 Con seats largely vulnerable to Lab and 40 Con seats largely vulnerable to the Alliance-Con leads up to 20%). Over 4500 voters in the analysis poll at 50 polling stations-134 sampling points in total as 16 polling stations common to both polls. 10pm prediction based on votes up to 8.45pm cut off.

Plus 720 postal voters filled in questionnaires. After 3rd result, increased Con majority at Basildon seat forecast was for Con overall maj of 94(up from 68 at the start-actual 102).

[Andrew_S]

Sept 23, 2021 19:54:49 GMT nodealbrexiteer said:

An excellent article on ITN's 1987 exit poll from our very Robert Waller:

amsr.contentdm.oclc.org/digital/collection/p21050coll2/id/8327/rec/1

Some points I picked out:

Total respondents 20172. Over 15500 voters at 100 polling stations in the seat projection poll(60 Con seats largely vulnerable to Lab and 40 Con seats largely vulnerable to the Alliance-Con leads up to 20%). Over 4500 voters in the analysis poll at 50 polling stations-134 sampling points in total as 16 polling stations common to both polls. 10pm prediction based on votes up to 8.45pm cut off.

Plus 720 postal voters filled in questionnaires. After 3rd result, increased Con majority at Basildon seat forecast was for Con overall maj of 94(up from 68 at the start-actual 102).

I don't know how you find documents like this but they're always interesting.

[nodealbrexiteer]

Bit on NOP 1989 Euro elections exit poll:

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/322/rec/4

Halfway down page Harold Wilson comments on BBC oct 74 poll...

amsr.contentdm.oclc.org/digital/collection/p21050coll2/id/4727/rec/38

[nodealbrexiteer]

NOP by election exit polls for the BBC: Kensington(result Con 42, Lab 38,SLD 11,SDp 5,Green 2.5,Other 2 and forecast 40/40/11/5/2.5/2)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1113/rec/1

Vale of Glamorgan(forecast Lab 51, Con 36, Plaid 4,Democrats/SLD 3,SDp 2,greens2, Other 2, result 49/36/4/4/2/2/3)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1165/rec/2

Glasgow Govan(forecast SNP 51,Lab 38, Con 6,SLD 2,Other 2 and result 49/37/7/4/3)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1182/rec/3

[nodealbrexiteer]

Pre-Election and Exit Polls in the 1992 Campaign by Martin Collins-can't find his report on the flawed 1987 BBC Gallup poll:

amsr.contentdm.oclc.org/digital/collection/PapersOffprnt/id/2608/rec/22

[The Bishop]

Sept 24, 2021 10:30:04 GMT nodealbrexiteer said:

NOP by election exit polls for the BBC: Kensington(result Con 42, Lab 38,SLD 11,SDp 5,Green 2.5,Other 2 and forecast 40/40/11/5/2.5/2)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1113/rec/1

Vale of Glamorgan(forecast Lab 51, Con 36, Plaid 4,Democrats/SLD 3,SDp 2,greens2, Other 2, result 49/36/4/4/2/2/3)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1165/rec/2

Glasgow Govan(forecast SNP 51,Lab 38, Con 6,SLD 2,Other 2 and result 49/37/7/4/3)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1182/rec/3

Byelection exit polls were a regular thing until the 1992 GE polling fiasco, after which they seem to have fallen off the radar completely.

[nodealbrexiteer]

Sept 24, 2021 11:51:12 GMT The Bishop said:

Sept 24, 2021 10:30:04 GMT nodealbrexiteer said:

NOP by election exit polls for the BBC: Kensington(result Con 42, Lab 38,SLD 11,SDp 5,Green 2.5,Other 2 and forecast 40/40/11/5/2.5/2)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1113/rec/1

Vale of Glamorgan(forecast Lab 51, Con 36, Plaid 4,Democrats/SLD 3,SDp 2,greens2, Other 2, result 49/36/4/4/2/2/3)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1165/rec/2

Glasgow Govan(forecast SNP 51,Lab 38, Con 6,SLD 2,Other 2 and result 49/37/7/4/3)

amsr.contentdm.oclc.org/digital/collection/NOP_Reports/id/1182/rec/3

Byelection exit polls were a regular thing until the 1992 GE polling fiasco, after which they seem to have fallen off the radar completely.

Which whilst it makes sense meant that the BBC and ITN trying to recover from the debacle of the 1992 exit polls didn't give themselves anything to experiment with to improve their techniques. Neither did an exit poll for the 1994 Euros though MORI's exit poll for the Electoral reform society in London alone in the 1994 Euros helped them when doing ITN's 1997 poll.

[hullenedge]

Sept 24, 2021 11:56:40 GMT nodealbrexiteer said:

Sept 24, 2021 11:51:12 GMT The Bishop said:

Byelection exit polls were a regular thing until the 1992 GE polling fiasco, after which they seem to have fallen off the radar completely.

Which whilst it makes sense meant that the BBC and ITN trying to recover from the debacle of the 1992 exit polls didn't give themselves anything to experiment with to improve their techniques. Neither did an exit poll for the 1994 Euros though MORI's exit poll for the Electoral reform society in London alone in the 1994 Euros helped them when doing ITN's 1997 poll.

The BBC was badly wounded by the 1992 'fiasco' following on from the 1987 'fiasco' and some concern about Vincent Hanna and his sampling techniques for his by-election polls. Very grateful that we now have the integrity of the academics for the exit poll.

[nodealbrexiteer]

More on that 1994 MORI Euros London poll:

amsr.contentdm.oclc.org/digital/collection/MORI-BPO/id/1159/rec/5

[Davıd Boothroyd]

The last byelection exit polls were in Kincardine and Deeside and in Langbaurgh in November 1991. The Langbaurgh one was spectacularly bad - showed Labour running away with it, when actually it was quite narrow.

[nodealbrexiteer]

Jun 29, 2021 21:47:23 GMT nodealbrexiteer said:

I today received some disappointing news from the BBC regarding a freedom of information request on the early missing exit poll data i alluded to on other threads and this one-the data if it exists is not something they are obliged to give out so won't give it out-something to do with exemptions under the Act for artistic and journalistic matters-I may yet query why as it's hardly state secrets!

ico.org.uk/media/action-weve-taken/decision-notices/2021/4018433/ic-121565-q7f2.pdf

The ICO decision notice

[nodealbrexiteer]

in.booksc.eu/book/50957897/f19dd1

full version of Payne/brown 1975 article on 1974 elections forecasting, just click download(with discussion at end of paper which wasn't on other version i posted before).

[nodealbrexiteer]

Forecasting on British election night 1997

like above needs no log in just download

jp.booksc.eu/book/8972481/c679aa

Forecasting the British Election to the European Parliament

ur.booksc.org/book/25832109/c4dc52

[nodealbrexiteer]

Sadly some not so good news from the BBC archives re a request I put in in June and chased up by my MP re the early missing prior forecasts, thanks to hullenedge and my friend GC for their help and support:

Firstly, can I apologise for the length of time it has taken to respond to your request. Our Reading Room was closed between March 2020 and early June 2021 due to the COVID situation, and we built up a large backlog of research enquiries during this time. We are working through the backlog, but inevitably it is taking us much longer than usual to answer enquiries.



We have now been able to conduct a full search of the archive and there are no file titles or index entries that relate specifically to the Payne / Brown seat forecasts. Additionally, we have searched more generally for files about the General Elections in February and October 1974 and in 1979 and identified approximately 50 files that have been retained for the archive. There were also in excess of 100 files meeting this criteria that were not retained for the archive.



Usually for a research task of this size we would invite you to view the material yourself in our Reading Room, but as you have explained that you’re not able to visit and additionally we have a long waiting list for appointments due to the COVID-related backlog, we have reviewed the 50 surviving files in detail on your behalf. Whilst there are a few incidental references to Clive Payne and Philip Brown and their work, I can confirm that we don’t hold their seat forecast data that you have requested. Had we found their data, we’d have been happy to find a way of sharing it with you. Given that you’ve already been in touch with the academics directly, it’s difficult to know what else to suggest.



I’m sorry we’re not able to help on this occasion, and that it has taken such a long time to provide an answer.



Yours Sincerely,

James Codd



_______________________

James Codd
Archives Collection Manager (BBC Written Archives Centre)

Archives Technology & Services

BBC Technology Group
BBC Written Archives Centre, Peppard Road, Caversham Park, Reading RG4 8TZ. 020 8008 5681 / 07725 903896

[nodealbrexiteer]

From Brown/Payne 1975 article in Journal of the Royal Statistical Society re the BBC October 1974 exit poll boob:

There was one notable departure from previous practice in that we were provided
with a nationwide post-poll as well as two individual constituency polls. The nationwide poll sampled one polling station in each 155 constituencies. There are good
reasons to suspect the value of such a poll as the voting history of these polling
stations is not known (see, for example, Smith, 1969). The use of the nationwide
poll resulted in a very poor initial non-computer prediction given out at the beginning
of the television coverage and was to some extent responsible for the initial pro-Labour
bias in the computer prediction.

[nodealbrexiteer]

email to me from Professor Michael Thrasher from 2013:

Sky did not have an exit poll in 1992. The on-air forecast that gave the Conservatives a majority of over 20 seats was made only after we received some results – I don’t recall the exact number. I do not know the exact numbers in terms of House of Commons forecast I’m afraid.

The 1997 forecast was made after the first result came in from Sunderland. Again, we did not use an exit poll.

The 2001 forecast was made in the same way but I don’t recall the exact figures. I don’t recall a phone-in poll but I would not have used those data to do a seat forecast in any case.

The first time that Sky News was involved with an exit poll was in 2010.

[nodealbrexiteer]

From Britain at the Polls 1974(on the 2 channels exit polls)

The BBC-TV forecasting operation came badly unstuck because
of the way the television program, without consultation or approval
by its polling organization, presented its post poll information. Eager
to be first with news, the BBC, within minutes of the ballot boxes
being closed, forecast a Labour lead of 180 seats over the Conservatives, and a 10.5 percentage-point Labour lead in the popular vote.
The BBC's pundits proceeded to analyze these figures as if they were
real results, even though the projection was only based upon information obtained from persons voting before 8 :15 p.m. The BBC-London
team used the Harris post poll in Scotland to project an SNP gain of
approximately 15 seats. Approximately an hour later, the BBC
presented a revised forecast, based upon interviews completed up to
10 :00 p.m. The final report to BBC-TV from the nationwide Harris
post poll was Labour, 42.0 percent, Conservatives 34.2 percent,
Liberals 18.1 percent, and nationalists and others, 5.7 percent, a
7.8 percent Labour lead in Great Britain. The error in estimating the
gap was 4.3 percentage points, and there was a 1.4 percentage point
average error in estimating each party's share of the vote. Both
figures were slightly more accurate than the final Harris poll in the
Express, but not more accurate than some newspaper polls (see
Tables 10-1 and 10-2). The BBC-TV forecast of seats based upon
the final poll figures gave Labour a 100-seat lead over the Conservatives, 340 to 240. The forecast gave the Liberals 15 seats and the
nationalist and other parties 40 seats. The Guardian's critic commented:
Later figures (after 10 o'clock) reduced the predictions
somewhat but were still wide of the mark-and viewers
then saw the BBC men make the crude mistake that the pollsters themselves now avoid: trying to turn estimated shares
of the vote into seats.

ITN, by contrast, was much more cautious in the presentation
of the findings supplied to it by ORC. Instead of conducting a
nationwide post poll, ORC was asked to concentrate polling efforts
on key categories of constituencies in England and Scotland. ITN's
forecast before the first result was primarily based upon an ORC reported swing of 2.3 percentage points to Labour in marginal seats,
an estimate within 0.1 of a percentage point of the final swing in the
marginals. Concentrating resources in a small number of constituencies gave the ITN poll a larger number of interviews in the places
where swing could make seats change hands. The translation of popular preferences into seats was treated as a projection from survey
data, and not given special ITN program emphasis. Labour was
estimated to have a 54-seat majority over the Conservatives, 325 to
271 (actual result: 43), and a 15-seat overall majority (actual result: 3)
with the Liberals winning 11 seats, and nationalists and others, 28.
The level of error in the Harris post poll for BBC is less surprising when one considers the difficulties of organizing day-long
interviewing of voters on a massive scale. Both Harris and ORC used
extremely clustered samples. The BBC forecast was based upon
interviews taken in eighty-five different constituencies in England,
but the interviews covered only one polling station in each constituency. Since voting records are not published at this level, it is
difficult for a survey organization to know whether the place chosen
to represent the constituency is in fact representative in its voting,
or whether it returns a lopsided six-to-one majority for one party.
Refusal rates among prospective respondents are also higher in this
situation than in doorstep interviewing, because people are in a hurry
to get away, once having voted.

[nodealbrexiteer]

May 27, 2021 22:01:38 GMT nodealbrexiteer said:

1992:BBC-after 2 results-11.18pm from Con short by 3 to Lab short by 18(presumably 308/293 assuming at this stage margin of error was plus/minus 15 seats still on the 2 main parties-the confidence interval would narrow later of course)-central forecast Con short by 18,after Basildon 309/292/22/28(range of forecast Con short by 2 to Lab short by 19)-4 results approx 11.45pm,approaching 00:30 Con short by 14,after 45 results Con short by 12(range Con maj of 7 to Lab short by 23),00:42 approx Con short by 11(range Con maj of 9 to Lab short by 25),after 201 results at 01:24 Con short by 4(range Con short by 17 to Con maj of 19 so plus or minus 13 seats now),01:49(332 results) Con maj of 5,about five to two-Con maj of 1!(362 results),after 418 results a little after five past two Con maj of 3,same maj forecast after 437 results,after 523 results 02:40ish Con maj of 9(Hung parliament still possible),after 532 results Con maj of 13 then moments later Con 331, Lab 274,Lib Dem 20,Others 26-Con maj of 11,then moments later back to Con maj of 13!(range Con maj of to Con maj of 23 plus/minus 5 seats),just after five to thre share forecast for Great Britain of 42/36/18/4,just after Con maj of 11 on board behind Dimbleby,after 595 results Con maj of 15-333/274/20/24,03:32 after 597 results Con maj of 17,after 611 results back to Con maj of 15,after 612 results back to Con maj 17,early in morning programme national shares given as 43/36/18/3,Con maj still 17,Con maj of 19 after 627 results,(after 629 COn reach 326 needed for win),after 634 1pm 10th April 1992 still maj of 19,national share 43/35/18/4,after 640 result Con maj of 21(all 3 main parties on their final totals by now),(seats under PR 279/229/119),

(youtube copy gone so this information is from Getty images clips and I will add more should I find more in the future)ITN-somewhere between 1am and 2am Con 324,Lab 277, Lib Dem 24, Others 26,also 329/272/24/26,then between 2am and 3am Con 333,Lab 272,Lib Dem 21,Others 25.

I also once saw on the old TV Ark a Sky News forecast which overestimated Con on 342(33 majority) and Lab on 268 ish but it's not on the new Tv Ark

More from Getty images re ITN 1992-between 11pm and midnight forecast changed to Con 310,Lab 289,Lib Dem 25,Others 27

[nodealbrexiteer]

An interesting snippet highlighting the different uses of the BBC and ITN 2001 exit polls-it's from a John Curtice/David Firth article on the 2005 exit poll:

This was not the first time that such an approach had been adopted. ITV’s exit polls had long been conducted as far as possible in the same polling districts as at the previous election. The BBC’s history of exit polls is a shorter one, but in 2001 it too adopted a strategy of revisiting the same polling districts as it had done in 1997. Thus when the two organisations decided in 2004 to lay down their rivalry and commission a joint exit poll, largely on the grounds of cost, it was relatively easy for them to agree that the 2005 poll should be based on those polling districts that had been covered by one or other of them in 2001. There had, however, been some differences between the approach used by ITV in 2001 in selecting polling districts and that used by the BBC. While it had been disproportionately conducted in marginal constituencies, and in particular those marginal between Labour and the Conservatives, the 2001 BBC poll, conducted at 90 polling stations, had covered all kinds of constituencies. In contrast the parallel ITV poll, conducted in 100 locations, had only visited polling stations in constituencies that had previously been marginal. Either they were constituencies that had been marginal between Labour and the Conservatives (all of which were constituencies held by Labour in 2001 with majorities over the Conservatives of 25% or less) or they were ones 4 that were closely contested by the Liberal Democrats and the Conservatives (a mixture of seats held relatively narrowly by either party). In excluding any other kind of marginal, this approach reflected the fact that these had hitherto been the two main political ‘battlegrounds’ at British general elections. The two organisations adopted their different approaches because their polls were designed to achieve somewhat different objectives. ITV’s poll was only intended to produce a seat projection, not forecast shares of the vote, and thus it could focus exclusively on the two types of constituency where seats were most likely to change hands. In contrast, not only was the BBC’s poll intended to produce a forecast vote share, but it was also used, alongside the results from those seats that declared first, in order to produce an updated forecast; see, for example, Brown et al. (1999). As most of the constituencies that declare early are safe (Labour) constituencies (e.g., Payne, 2003, p.202) it is only possible to make an early check on the likely accuracy of an exit poll if it contains at least some polling stations located in safe constituencies