Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Hoverable Dropdown

Saturday, 23 April 2016

London Assembly Projection, April 2016

So: YouGov have finally released a poll of London Assembly voting intentions, which I was beginning to think they'd never do.


A Universal Swing projection of the Assembly result based on the YouGov poll.

I warned a few weeks ago that Labour should expect a slight decline in their London Assembly results, as their 2012 score was a high point. Instead, they've blown past it. In 2012, Labour won 41% of the List vote; they're now polling at 45%. The Tories have dropped 3 points and the Greens 2 points, while both UKIP and the Lib Dems have improved solidly on their 2012 result.

I used a 'universal swing' seat calculator that I've constructed, and put the constituency and regional poll results through it.

If this poll was repeated on May 5th, Labour would win 12 seats on the London Assembly (no change), the Conservatives 8 (-1), UKIP 2 (+2), the Lib Dems 2 (no change) and the Greens just 1 - their Mayoral candidate Sian Berry (-1).

The breakdown

The 'no change' for Labour is a consequence of the proportional "Additional Member" system used in London. In the "first-past-the-post" section of the ballot, Labour would pick up 2 Assembly seats from the Conservatives (Croydon & Sutton and Havering & Redbridge) and would come within 0.2% of a third (Merton & Wandsworth).

The YouGov poll shows the Constituency voting intention as:

LABOUR 46% (+4%)
CONSERVATIVE 30% (-3%)
UKIP 9% (+5%)
LIBERAL DEMOCRAT 9% (-)
GREEN PARTY 6% (-3%)
OTHERS 0% (-3%)

In total, the constituency seat breakdown would be:

LABOUR 10 seats (+2)
CONSERVATIVE  4 seats (-2)

Under the D'hondt proportional allocation system, the remaining seats would be allocated to ensure that parties' seats reflect their overall share of the Party List (regional) vote. The List vote is as follows in the YouGov poll:

LABOUR 45% (+4%)
CONSERVATIVE 29% (-3%)
UKIP 9% (+4%)
LIBERAL DEMOCRAT 8% (+1%)
GREEN PARTY 7% (-2%)
OTHERS 2% (-5%)

The Tories would thus grab 4 List seats, Labour 2, UKIP 2, the Lib Dems 2 and the Greens 1.

Monday, 11 April 2016

The polls aren't wrong: Sadiq Khan will be London's next Mayor

With the London Mayoral election nearly upon us, opinion polls for the contest are finally beginning to emerge in force. And each one of them tells the same story: Sadiq Khan, the Labour candidate, is in the lead.

He wasn't always. Back when the race for the Mayoral nominations began, Khan and the Conservative candidate, Zac Goldsmith, were much closer - at one point Goldsmith had an 8-point lead. But as the race began in earnest, Khan opened up a large lead that hasn't gone away. Currently, a simple average of polls in the last month (7th March-7th April, when the last poll was conducted) puts Khan on 54% to Goldsmith's 46%.

Average poll results by month for the final round in the Mayoral election.

The most recent surveys show him with a lead of between 6 and 10 points over the Conservative candidate, Zac Goldsmith. While these could shift, there's less than four weeks left for him to close the gap.

But could he? Polls change. And could the polls change that dramatically between now and May 5th? Let's take a look at how the polls have changed in previous Mayoral elections.

2012

From 7th March-7th April 2012, Boris Johnson's average poll figures were 53.5% to Ken Livingstone's 46.5%. On election day, the actual result (with changes from March-April) were:

Johnson (Con) - 51.5% (-2%)
Livingstone (Lab) - 48.5% (+2%)

So in 2012, between where we are now and election day that year, Labour actually gained support, rather than losing it. And that was with a campaign that got quite bitter at times, and with a Conservative candidate who was unusually popular.

2008

From 7th March-7th April 2008, Boris Johnson's average poll figures were 54.2% to Ken Livingstone's 45.8%. On election day, the actual result (with changes from March-April) were:

Johnson (Con) - 53.2% (-1%)
Livingstone (Lab) - 46.8% (+1%)

So once again, back in '08, Labour gained on election day relative to the polls at this point in the race, though not as much as they would go on to do in 2012.

What this means for 2016

The average shift for each party, in the final round, in 2008-12 was a gain of 1.5% for Labour and a drop of 1.5% for the Tories. This would suggest that, rather than swinging to Goldsmith, the final result on May 5th should be the following:

Khan (Lab) - 55.9%
Goldsmith (Con) - 45.7%

While it may not be as decisive as that - a 56-44 victory would be the largest Mayoral victory margin since Ken Livingstone was elected as an independent back in 2000 - I think the nature of polling in London, and the margins that Khan is managing in the polls, suggest he is on track to be the next Mayor of London, and Zac Goldsmith will be going back to Richmond Park to be a backbencher for the next four years.

Friday, 8 April 2016

3 Reasons why May's elections would be bad for any Labour leader

As the May elections creep up upon us, the obituaries for Jeremy Corbyn's leadership are already being written (again). Rallings and Thrasher project that Labour may lose around 150 council seats across England, with those losses being split somewhat evenly between the Tories, Lib Dems and UKIP. In this blog I'd like to dwell on some statistical data to emphasise why these elections would have always been bad.

1. Corbyn's first election is being compared to Miliband's high point

The council seats up for election in 2016 were last elected in 2012. At that time, the country was two years into the Coalition government, austerity was setting in, we had just had the "Omnishambles" budget and Labour were riding high in the polls. Progressive Lib Dem voters, Labour's newly-energised base and general anti-government voters were all behind Labour in 2012, and Labour's eviction from government in 2010 had not been the total rout some had expected. Many of its ex-voters, fresh from backing the Lib Dems in 2010, had returned to the party. One survey by YouGov on 1st April 2012 gave Labour a 10-point lead over the Tories.

On election day, the Tories' unpopularity hurt them. They lost 328 council seats, with the Lib Dems losing 190. Labour collected 1,188 councillors - an increase of 534. In the London Assembly elections, their lead over the Tories was 10 per cent, and their estimated share of the national vote was 38 per cent to the Tories' 31 per cent. It was a good night for Miliband.

But this isn't 2012. Corbyn has only been in office for 9 months, the Tories have only just started to make the public angry again, it's less than a year since they won their first majority government since 1992 and - yes - Labour hasn't yet settled its internal struggles. (And let's face it, such struggles were still raging to a degree 9 months after Miliband's own election).

Labour was always going to lose seats in this election, whoever won the leadership election. There's a reason it lost in 2015, and 9 months simply isn't enough time to regain the trust of British voters. If Corbyn manages to lose only a handful of seats (maybe 10, 20 or 30) it will be an absolutely brilliant night. If he actually gains seats, I would consider that a stupendous result.

But let's not kid ourselves into believing that Labour in 2016 faces the same circumstances it did in 2012. I doubt that Ed Miliband could have pulled off gains in the circumstances of 2016.

2. UKIP are a factor that simply did not exist in 2012.

One factor that seems to have been under-acknowledged by the media is the presence of UKIP. Over the last few years, since their surge began, UKIP have been steadily picking up protest votes that used to go to the main party of opposition. In 2013, they won 147 councillors; in 2014, they got 166; and in 2015, they grabbed 202.

Yet in May 2012, the party was still a fringe organisation and grabbed just 21 seats, a gain of just 5. Britain today has a thoroughly different party system than it did in 2012, with a stronger Green Party, a dominant SNP, a surging UKIP and an anti-austerity Labour Party. Comparing Labour's results in these circumstances, to a 2012 result where UKIP essentially didn't even exist on a national level, is bizarre. Labour can no longer rely on the automatic support of everyone who doesn't like the incumbent government; because of this, its results were always going to be lower than in 2012, when it essentially swept everyone who disliked the Tories. That's no longer the case, and the numbers simply don't stand up to comparison.

3. Labour lost badly in 2015, and it was never going to do well regardless of the leader.

We have to be honest here: in 2015, Labour was walloped. After 5 years of an austerity government, it actually lost seats to the Tories in England and Wales and got nearly wiped out in the previous stronghold of Scotland. There have been many analyses of why that happened, but suffice to say, 12 months (3 of them without a leader) simply isn't enough time to recover to a 2012 level of support. British voters aren't hugely positive towards Labour right now, and any leader would have faced that challenge.

Would another leader have risen to it better? Possibly. But as things stand now, it's unreasonable to point to changes in councillor numbers from 2012-16 and claim this to be a disaster for Jeremy Corbyn. Sure, it would be bad, but is it surprising, given the context? Is it something that is solely his fault, and solely within his control?

No. Not really.