General Election Forecast based on Uniform Swing by Region

On May 15, YouGov produced a General Election opinion poll broken down by the eleven Regions of Great Britain. It seems impressive with 14395 GB Adults, plus 1040 London Adults, 1017 Scottish Adults and 1018 Welsh Adults. However, with fieldwork on 24 April to 05 May, it might be a little out of date. By combining this with the General Election Results by Constituency (available for the British Election Study) I have been able to produce a crude forecast for the General Election on June 8th.

The starting point in the General Election Results of May 2015, shown in Figure 1. Since then Con has gained 1 seat from Lab (Copeland), and lost a seat to LD (Richmond). The sole UKIP MP, Douglas Carswell, left the Party in March to become an Independent. These are ignored.

Note that the 18 Northern Ireland constituencies are not included. The 331 Con seats are against 326 for a majority in the House of Commons.

I made the following assumptions.

  1. Within each constituency, for each Party I have assumed the change in the vote is the difference between the regional share of the vote in 2015 and the opinion poll share from YouGov.
  2. If the constituency vote share in 2015 was less than the regional drop in vote share between 2015 and current opinion poll, then the vote is nil.
  3. A party may have a predicted vote despite not having a candidate. There are two instances where this is possible. First is that UKIP are not standing candidates in every constituency. Second is that the Progressive Alliance of Lib Dems, Greens, Labour and the SNP are standing down candidates to maximize the impact of the anti-Tory vote. Guido Fawkes’ summary of 16th May is here.

This simple model produces the forecast in Figure 2.

Implied Conservative majority is 128, up from 12 in the previous Parliament. The Lib-Dems also increase there number of seats, whilst SNP lose 9. UKIP’s “gain” is in Buckingham, the seat of the Speaker. This is due to a flaw in the crude model.

The Party gains by region are in Figure 3

Of note is that Labour do not gain a single seat, as YouGov estimate that their popularity has dropped in all but two regions. In the South East and the South West Labours’ presence is quite low. The SNP in Scotland lose seats to both the Conservatives and the Lib-Dems, but the loss of 9 seats is would still mean they have 47 of 59 seats.

These switches in seats are shown in detail in Figure 4.

My very crude forecast can be compared to the current forecast by Martin Baxtor at Electoral Calculus of Con 409, Lab 167, LD 7 & SNP 46. The Conservative majority is 168, 40 more than my own.

Another comparison is the mid-point of the spread betting at Sporting Index. This Con 399, Lab 159, LD 15.5, SNP 45.5. The implied Conservative majority is 148, bang in the middle of my own and Martin Baxters’.

In subsequent posts I intend to

  • Clear up the obvious errors.
  • Refining the forecast for Scotland based on the local election data of May 5th.
  • Look at the forecast for Wales, where I believe YouGov might be out of line with popular opinion.
  • Update in relation to more recent, but National, polling. For instance the recent strengthening of the Labour poll share and the fall in the UKIP share.

Kevin Marshall

 

Implications of the Forecast Local Council Elections Results in England, Wales and Scotland for GE 2017

Summary

I bring together Local Election forecasts for England, Wales and Scotland made over the past few weeks, before the General Election announcement a week ago. In the three countries I forecast that the Labour Party would lose 700 seats, and the Conservatives to gain over 500 seats. The most dramatic changes I forecast are in Scotland, where the Conservatives should end up with more seats than Labour, firmly cementing their place as the second party of Scotland. The SNP I forecast to have a net gain of 100 seats, despite likely losing seats to the Conservatives.
The impact of the snap General Election is likely to reinforce the dramatic exchange of seats that I forecast, with the additional impacts of (a) confirming the collapse in UKIP support; (b) highlighting the re-emergence of the Liberal Democrats as the major alternative to the Conservatives in much of England; and (c) providing a signal that the peak SNP dominance in Scotland has passed.

 

In the last few weeks I have made some results forecasts for the forthcoming local elections in England (here and here), Wales and Scotland occurring on May 4th. I was forecasting some big changes in numbers of seats. Since then a snap General Election has been called for June 8th. This may affect the forecasts, although given I have used mostly GE opinion polls, maybe not quite so much as local election forecasts based more upon recent local election by-election results. Rather than try to reforecast based upon widely fluctuating opinion polls (such as in last Sunday’s papers ), I will try to evaluate the impact of my forecasts being correct in the context of the narrative for the parties for the last few weeks of the General Election campaign.

In England, only a small proportion of councils are up for re-election. The 35 councils are a mixture of shire counties and unitary authorities. They are predominantly in areas with Conservative members of parliament, although there is also the Labour stronghold of Durham, along with Lancashire and Derbyshire where Labour managed to regain council majorities in 2013. In terms of councils involved, this is a very low number. There were 124 English councils with elections in 2016 and 279 in 2015. Entirely absent are any council elections in the major English cities.

In Wales and Scotland all council seats are up for re-election. Although with similar numbers of council seats in the two countries as the English councils, these are far more significant politically.

Figure 1 summarizes my forecasts of seat changes for each country both against the last elections in 2012 and 2013 and the previous round of elections in 2007 and 2008.

I believe that a major influence on the UK local election results is the state of national opinion. This is in general direction of opinion and not the percentage share of the vote. For instance, in 2007 and 2008 the Labour Government was trailing the Conservatives by quite a long way in the opinion polls – possibly as much as 15 points. By 2012 and 2013 the situation was reversed, with Labour in opposition being around 8 points ahead of the Conservatives. Proportionately, Liberal Democrats as the junior partner in a coalition government, suffered greater reversals than the Conservatives.  This is strongly seen in the English council results, with the exception of the UKIP factor. The 2013 council elections demonstrated a game-changing http://www.bbc.co.uk/news/uk-politics-22382098 rise of UKIP outside of elections to the European Parliament.  In Wales, the National Opinion influence is less marked, but it is still there. Scotland is the least influenced by National Opinion trends, as there are strong factors unique Scottish factors, such as the replacement of the Labour Party with the SNP as the major left of centre party.

Before looking at the likely consequences for the General Election, I have shown the approximate forecast council seat numbers in figure 2.

Political Implications of the English Council Results

With hindsight my UKIP seat estimate is probably too high. But even if they retain 40 to 50 seats it will still be a major reversal on 2012. It will be very hard to convince voters in England – especially in the pro-Leave areas of the country – that they are a credible alternative political party, whether on Brexit or populist issues such as immigration.

The Liberal Democrats fell into third place behind Labour in 2013. I will be very surprised if that they do not regain second place. Expect to see a lot made of being in most of Britain the only credible alternative to Conservative pro-Brexit juggernaut.

The Labour Party will excuse the results as being in Tory areas, claiming that there is a different picture in the rest of England, particularly in London. They will try change the conversation to winning the mayoral elections in Manchester and Merseyside.

 

Political Implications of the Welsh Council Results

Plaid Cymru might only make small gains, and it will be in a limited number of areas. They will use this to persuade voters of being an anti-Brexit and pro-Welsh voice. By softening the Welsh language aspects, they might seek to extend their reach beyond the West Coast.

The Liberal Democrats will look at actual council successes to drive home their Pro-Remain message, particularly in the capital, where they will be looking to regain Cardiff Central.

For the Conservatives, coming in third place behind Plaid Cymru, despite large gains, may give confidence to other parties. If they come second (and/or regain control of the two councils lost in 2012) then this will be a landmark achievement. A YouGov opinion poll for Cardiff University and ITV on 24/04/17 of voting intentions for the Welsh local elections gave Labour 28%, Conservatives 26% and Plaid Cymru 19%, against my assumptions of 29%, 16% and 17%. The main inaccuracy of the poll I believe is that it calculates Independents plus minor parties at 12%, half the level of 2012. I assume a 1% gain. Given such varied support across the councils, along with first asking General Election opinions, it is easy to understand how this discrepancy might arise. However, it will be the actual results that will decide who has the greater accuracy.

 

Political Implications of the Scottish Council Results

It is in Scotland that the results will reverberate most strongly if my forecasts are correct. There are a couple of points to remember about the peculiar Scottish context.

First is that Scotland is embarking on the third set of council elections using the single transferable vote system (STV), with council areas divided into supersized wards of three or four seats. It means that if a party selects too many candidates in a ward, they may end up with less Councillors elected as candidates of the same party compete against each other for votes. They parties therefore try optimization of candidates based on forecasts. As I found, the Labour Party, instead of optimization based on a dramatic forecast fall in their vote, chose to largely match candidates with existing Councillors. They are therefore likely to lose proportionately more seats as a consequence of this quite rational decision.

Second, is the growth in the SNP means that the change in the vote from council election to council election does not strongly reflect the swings in the UK-wide opinion polls over the same period. For instance, although the Labour share of the vote in 2012 was higher than in 2007, the rise was much smaller than the SNP rise in vote share. However, I still forecast SNP vote share to increase, resulting in a net gain in Councillors. But these net gains will include very large gains in Labour councils, such as Glasgow, alongside loses elsewhere, especially to the Conservatives. There will be a strong message for those who oppose independence that the SNP juggernaut might have peaked and could signal the turning of the tide on the Independence issue.

Lib Dems may not have a significant showing, as in many areas where they are strong in the Westminster elections, are where there is a strong representation of Independents. Indeed, the councils of Orkney and Shetland, where the Lib Dems have their only Scottish MP, are all Independents. Whether they will win over some votes as the party of double-Remainers, remains to be seen. An extensive Scottish Yougov poll  published in January, but based on sampling in late 2016, showed that the double-Remainers were 28% of voters. That includes 21% who voted in one or none of the two referendums.

From my forecast the Conservatives will easily replace Labour as the second political party represented on Scotland’s councils. Due to Labour having too many candidates, they could pick up seats in unexpected areas. Conversely, if the Conservative surge continues, then Ruth Davidson may end up apologizing for having too few candidates, as vote share in some wards might easily elect the available candidates.

 

Implications for the UK General Election

A loss of 700 seats across three countries would be a huge message that the Labour Party under Jeremy Corbyn’s leadership was heading for an even worse result than in 1983. The benchmark in England and Wales would be the woeful results of 2007 and 2008, which were 530 seats lower than the last set of elections, compared to my forecast of 455 seats lost.

The Conservatives will want to downplay strong gains in England and Wales in case complacency reduces turnout on June 8th. But they will want to emphasize their confirmed position as the opposition in Scotland.

The Liberal Democrats will announce that they are back, and concentrate on gains made in particular areas where they are clearly the main opposition to the Conservatives Hard Brexit. They will emphasize any gains they might make in the council areas where they had Westminster MPs until 2015.

UKIP may lose more than the rounded 100 seats I forecast. It will show that, unlike in 2015, they are no longer the protest vote of the disaffected. Given that their main cause has been achieved, and supporting the Conservatives is the surest way to ensure Brexit is enacted in full, this may signal the end of the UKIP as a national force on June 8th.

The biggest losers could end up being the SNP, despite their likely winning control of a number of councils and destroying the Scottish Labour Party. Their demonisation of the Conservative and Unionist Party as the subjugators of the Scots will not hold as much sway when the spread of their Councillors is much broader than in living memory and the Conservative voter is more likely to be your own neighbour.

Kevin Marshall

Is the Rawlings and Thrasher English Elections Forecast a bit timid?

At the weekend I posted my forecast for the English council elections to take place on May 4th, using a comparison with opinion polls both now and 2013, along with the forecast made by Harry Hayfield at Political Betting.

On Sunday, The Times published a forecast from Colin Rawlings and Michael Thrasher, both Professors of Politics at the University of Plymouth.

Again they use local by-election results a a basis for their forecast, and come up with similar results to Hayfield for the Lib-Dems (+9%) and UKIP (-12%). The biggest difference is with Labour, predicting no change, as against -4% for Hayfield. More relevant is the forecast for seats, where they predict Conservative +50, Labour -50, Lib Dem +100, UKIP -100

Despite these chaps having considerable experience of election data, having written a number of post-election reports for the Electoral Commission, I believe that their forecasts might be somewhat out. This I have split into share of the vote and seat predictions.

Rawlings & Thrasher share of the vote prediction

I have summarized some data in figure 1.

Rawlings & Thrasher base their forecast on the Notional National Change, so I have shown the difference between those figures and the General Election Opinion Polls. There are reasons why the local council elections are different from the opinion polls, aside from the fact that opinion polls may not fully reflect actual decisions.

In 2013 UKIP benefitted in the local elections as people used their vote as a protest against the Coalition Government and/or in support of an EU Referendum.  That no longer exists as Brexit is underway; whilst UKIP is in disarray and has fund-raising issues. In this I would agree with Rawlings and Thrasher in a huge drop in support for UKIP, swinging from outperforming the opinion polls, to underperforming.

With the Liberal Democrats, I would agree that they should not only outperform the opinion polls, but do so by a greater margin than in 2013, when being in Coalition Government damaged their brand as being the alternative to the Conservatives in the non-socialist areas. This year they may also benefit as being the Party of those most opposed to Brexit. But the actual share of the vote was 20%. For the notional share of the National Vote to rise nine points would indicate a larger rise to over 30% of the actual vote. It would effectively make the arch-Remainers effectively the biggest beneficiaries of the fall in support for UKIP. That is possible, but would imply very large underlying switches in effective party support. Also, this large net increase in Lib-Dem vote share would be despite no net movement in the general election opinion polls in the last five years. Even if local by-elections results indicate such a swing, when voting for full councils I do not believe that this will replicate this result.

The forecast for Labour to have no change from 2013 is perhaps the most out of line with reality. Just four years ago Labour seemed likely to be in a position to be win the 2015 election. Now they are 13 points lower in the opinion polls. Most of the councils concerned are in the Conservative heartlands, with Labour traditionally trailing in third place. They thus have less far to fall than nationally, but barring a miracle in the next four weeks, they will lose vote share.

By difference, this leaves the Conservatives making much larger gains that 5% of the Notional National Vote.

Rawlings and Thrasher seat change predictions

This I think is the most timid part of the seat changes to repeat they are

Conservative +50, Labour -50, Lib Dem +100, UKIP -100

Prof John Curtice on the Sunday Politics thought, based upon a 12 point swing in the opinion polls, Labour could suffer a much bigger loss of seats than the 50 forecast by Rallings and Thrasher. (hattip Guido Fawkes)

But to get an idea of the likely level of seat changes, it is worth looking at the changes in the similar (but not quite identical in terms of seats and councils) council elections in 2009 and 2013. These, from Wikipedia (here and here) are shown in Figure 2.

Please note that the numbers do not quite stack up due to (a) one more council in 2013 (b) boundary changes (c) by-elections. But broad comparisons can still be made.

UKIP gained 139 seats in 2013. The swing from 2009 to 2013 was probably greater than the counter swing they will likely suffer next month, so the loss of 100 seats is a reasonable estimate.  Similarly the Lib Dem estimate of 100 seats gain seems about right, despite the likely net change in fortunes maybe be less at a National level not likely nearly countering the loss in popularity between 2009 and 2013.

When it comes to Labour, they are nationally in a worse position than in 2009. So their loss of seats could be greater than then suffered in 2009. A reduction of 300 seats, or nearly 60% of the defended seats, seems a reasonable forecast, with a 200 seat loss or less being a relatively good night for Labour with their current unpopularity.

The Conservatives should easily be the biggest gainers. Most of the councils are in Conservative Heartlands. Others, like Lancashire and Derbyshire, are where they have managed majorities in the recent past. Yet in 2013 the Conservative share of the vote was below that achieved in the UK as a whole in the General Election just two years later. Two years after the General Election, the party is in a considerably stronger position, so should gain considerably. A gain of 300 seats is my forecast.

Final points

An aspect to consider in local elections is the impact of turnout. In their report on the 2013 English and Anglesey Council Elections, Rawlings and Thrasher noted that turnout was 31%, compared with over 39% in 2009.

In the similar 2009 report they note that this turnout was higher than in the previous three years.

Could the relatively high turnout be due to the confidence of Conservative voters? In 2009 the Conservative vote was 44.5%, 10% higher than in 2013. Labour had a 12.7% share (over 8% lower) and Lib-Dems were at 24.9% (5% higher). Then the Conservatives were a year away from going back into Government. In 2013, they looked like reverting to opposition. In 2017, there is a new sense of optimism among Conservative supporters, not seen since the end of the Falklands War – or at least with their Brexit supporters. These are concentrated in the middle-aged and older people, among whom turnout is usually higher than in the population at large. On the other hand, the Labour-supporting Remainers might stay at home, being doubly-demoralised. The change in seats could be even bigger than in either 2009 or 2013.

These are of course my forecasts compared with those of others who have more experience in these matters. What is crucial is how analyzing the difference between forecasts and actual results can deepen our collective understanding of what is happening in an interesting period in British history.

Kevin Marshall

 

Local Elections Forecast May 2017

At Political Betting Harry Hayfield has produced a forecast for the County Council elections in May 2017, just six weeks away. I have long followed this blog, though I neither bet nor follow the politics of the blog propriotor Mike Smithson (Lib-Dem & Pro-Remain on Brexit). However, it is the very differences that provide an alternative perspective. Harry Hayfield’s basic assumptions County Council elections in England are:-

In the local by-elections, areas that voted REMAIN saw the following change:
Con -9%, Lab -4%, Lib Dem +12%, UKIP -12%, Greens 0%, Ind -10%, Others +23%
Where as areas that voted to LEAVE saw the following change:
Con +4%, Lab -4%, Lib Dem +7%, UKIP -9%, Greens +1%, Ind +4%, Others -3%

With any forecast there is a lot of work involved in converting these into seats. Any baseline, when compared to the actual results will help develop an understanding of both of National opinion in turbulent times and also how it impacts at the local politics. What I would note is that the Counties are mainly Conservative territory; has a large Independent / Other representation (though this has been declining for decades); and is predominantly Remain voting, but not extremely more than the UK average.

But that said Harry Hayfield’s forecasts seem to be at odds with the opinion polls. Taking the UK Polling Reports opinion poll figures for (a) 28/04/13 to 10/05/13 and (b) 05/02/17 to 19/02/17 (the latest available) the change in voting intention is about

Con +11%, Lab -13%, Lib Dem +0%, UKIP -2%, Greens +2%.

Since then the Conservatives and Lib Dems at improved slightly at the expense of UKIP and Labour. The latest YouGov figures, for 25-27 March,  are

Con 43%, Lab 25%, Lib Dem 11%, UKIP 10%

And the change on 2013 are about

Con +13%, Lab -14%, Lib Dem +1%, UKIP -4%

The actual share of the vote – skewed by being counties – was

Conservative 34.3%, Labour 21.1%, UKIP 19.9%, Lib Dem 13.8% and Green 3.5%

therefore – Other 7.4%

In terms of the County Councils, perhaps 85% were for Leave in the EU Referendum, on the basis that maybe three-quarters of Parliamentary Constituencies in in England and Wales voted for Leave, and most of the Remain vote was in the Metropolitan areas. So why should the Conservatives only improve by 4% in most of the Council areas on 2013 when they are 13% up in the opinion polls nationally?

Harry Hayfield’s assumptions diverge from the opinion polls from being based upon recent by-elections nationally. That is upon a maybe half a dozen by-elections a week, These one-off events usually have low turnouts, and are highly impacted by protest votes.  The biggest reason for protest votes at the present time is the Brexit issue. It will have an impact on the local elections, with the Remain protest vote being stronger than the Leave support vote. But this will be much less than in by-elections due to (a) higher turnout due to area-wide elections (b) higher turnout on average as the counties tend to have higher turnout than the Metropolitan areas (c) the protest vote is concentrated in the cities, especially London. The latter is evidenced by the two recent very large petitions to change the referendum rules after the EU referendum and to cancel the State Visit of President Trump.

That said, using by-election results is a as good as any considering the political landscape has changed dramatically in the past four years. By using clear assumptions on forecasts it is possible to understand where they go wrong. But that understanding will be only on the empirical differences between forecast and actual. What I will now do is to give a forecast based on reasons (maybe my unfounded opinions) with some notes on the parties. This might muddy the waters a bit, but here goes.

For UKIP 2013 elections were a “game-changer“, to use Nigel Farage’s term. UKIP then got nearly double the share in the County vote as in the National opinion polls. A party campaigning on a single national issue made huge gains on votes at local government level. The 20% of vote share was a precursor to the 26.6% in the European Parliament elections achieved in 2014.  This impact is now gone, and the grassroots support needed to field and promote thousands of candidates may have diminished due to internal strife. Local by-elections may not fully reflect this impact, as they can concentrate their much reduced resources. UKIP’s share may go below the National Opinion polls. On these grounds I will go for a -13% change on 2013, giving UKIP 7% of the vote.

The Liberal-Democrats, are usually the main opposition to the Conservatives in the Counties and also usually perform much better than in General Elections. In 2013 they were in Coalition with the Conservatives, which severely damaged their standing as a local opposition. The 13.8% of the vote was 3-4% ahead of the opinion polls. This margin might not be much more than in previous 20 years, but in those years the Lib-Dems were doing much better nationally. Any bounce I believe will be mostly due to being the respectable face of hard-line opposition to Brexit. This could be bet positive, but put some folks off voting for them on local issues. I would guess at an increase of 5% giving the Lib-Dems 19% of the vote.

The Labour Party cannot fall nearly as far in the English Counties, where support is much lower than the national average. At the end of April and beginning of May 2013 they were polling 38-39% nationally, but achieved 21.1% of the vote. Will they still fall? In 2013 they managed to overtake the Lib-Dems with 538 to 352 seats, as against 247 to 476 in 2009, so the 21.1% was actually an improvement. Nationally Labour are doing worse (or at least no better) than in 2009. What is more, they have lost a lot of their working class base, which would have formed some of the core support in the Counties. I would estimate a 5% fall.

The Independents were once a major contingent in local politics, but only achieved 7.4% of the vote in 2013. How will this go? I would hope for an increase as people vote again for local representatives based on local issues. With the decline of UKIP, and maybe the Lib-Dems tarnished in the eyes of some for going very anti-Brexit, I think they may gain ground to around 10%. The Greens maybe the same.

By difference, the Conservatives will have a 11% increase in the vote share to 45% of the vote share. This is fairly similar to the change in national opinion polls over the last four years (from 30 to 41%). Indeed, it might be a low estimate, as in 2013 the Coalition Government was suffering from normal mid-term blues. At present, the Government is in the unusual position of being significantly more popular than when elected.

So, in conclusion, Harry Hayfield’s estimate was as follows

In the local by-elections, areas that voted REMAIN saw the following change:
Con -9%, Lab -4%, Lib Dem +12%, UKIP -12%, Greens 0%, Ind -10%, Others +23%
Where as areas that voted to LEAVE saw the following change:
Con +4%, Lab -4%, Lib Dem +7%, UKIP -9%, Greens +1%, Ind +4%, Others -3%

The LEAVE areas are at least 5 times more than the REMAIN areas. My estimate, for all English areas with a complete council elections in 2013 & 2017 is as follows.

Con +10%, Lab -5%, Lib Dem +5%, UKIP -12%, Ind/others +2%.

It is the actual results that will matter, and possible lessons to be learnt during a time of massive change in British politics.

Kevin Marshall

UKIP did not lose the Tories the Election

The notion that UKIP lost the election for the Conservatives is erroneous.

This claim originated by Richard North on EU Referendum, and repeated by Conservative Home (with figures), John Redwood and Cranmer.

UKIP cater for a niche of voters who would otherwise (mostly) vote Conservative. However, a mainstream party cannot cater for all tastes. If the Tories became more euro-sceptic to squeeze the UKIP vote, they would most probably have lost more votes to the Lib-Dems and Labour. Any main-stream political party must be a broad church. The problem with our current political opinion is that we had two left-of-centre parties that got over 50% of the vote, a mainstream right-of-centre party that got 36% of the vote and UKIP that got 3%.

The conclusion for the Conservatives is not to try to appeal to a very broad church by merging many different opinions. Rather, they must capture a vision that people can empathise with, as did New Labour and Thatcherism. The time to introduce this was not with the launch of the manifesto, but two or three years before an election. Further that vision should also be an implicit attack on the alternatives.

A positive vision to vote for; and the opposing failures to vote against.