Overstated Opinion Polls can undermine election results
The UK lesson could impact New Zealand
Dr Muriel Newman
Whangarei, Sunday, June 18, 2017
The country that gave the world Monty Pythons Flying Circus, Mr Bean, and The Goon Show, has just held a General Election.
In true comic form, standing alongside the British Prime Minister Theresa May as results were read out for her Maidenhead constituency, were Lord Buckethead, Howling “Laud” Hope, and Elmo.
Unsurprisingly, Mrs May won with 37,718 votes. Lord Buckethead, who described himself as an “intergalactic space lord” and stood on a platform of strong but “not entirely stable leadership,” won 249 votes.
Howling “Laud” Hope of the Monster Raving Loony Party gained 119 votes.
And Elmo, who clearly does not have a wide circle of friends and family to draw upon for electoral support, won three votes!
However, there was no trace of humour in the overall result for the Conservative Party. Coming from a commanding 24-point lead in the polls six weeks before Election Day – with an expectation of winning almost 100 extra seats – the Conservatives lost their majority in Parliament after what has been described as the worst election campaign in living memory.
With New Zealand’s General Election less than 100 days away, our politicians will no doubt be dissecting the UK’s shock result for clues as to why the Tories did so poorly, while the UK Labour Party did so well – despite having a leader who’d been described as ‘unelectable.’
The poor result has largely been attributed to the fact that the electorate was not convinced that an election – three years earlier than scheduled – was necessary.
When told of the snap election, Brenda from Bristol became an instant celebrity by capturing the mood of the nation: “You’re joking, not another one? Oh, for God’s sake, I can’t stand this. Why does she need to do it?”
Voters were sceptical about the Government’s claim that they needed a stronger mandate for Brexit negotiations; instead they saw political opportunism to take advantage of a weak and divided Labour Party. As a result, they punished the Tories at the Ballot Box for putting their Party interests ahead of the good of the country.
Instead of increasing their numbers in the 650 seat Parliament, the Conservatives fell from 331 seats to 318 – well short of the 326 seats needed for an outright majority. Against all predictions, Labour gained 30 seats, rising to 262. The Scottish National Party (SNP) – pushing for independence – fell 21 seats to 35. The pro-Europe Liberal Democrats Party (Lib Dems) increased by 4 seats to 12. The Northern Irish Democratic Unionist Party (DUP) – opposing independence – increased 2 seats to 10. And the UK Independence Party (UKIP), which successfully campaigned for Britain to leave the EU, lost its single seat and most of its support, going from 3,881,000 votes to 595,000.
Pact with DUP
With the election delivering a hung Parliament, Theresa May lost no time in negotiating a ‘confidence and supply’ agreement with the DUP, for a working majority of 2 seats in the new Parliament. This is a far cry from the Conservatives’ direct majority of 6 seats before the election, and – with the support of allies – an effective majority of 17.
Lord Richard Balfe of Dulwich
I asked this week’s NZCPR Guest Commentator Lord Richard Balfe of Dulwich, a Member of the British House of Lords, to share some observations about the UK election:
“I can say without hesitation that this was the worst election I have known in my sixty years of election watching. A seemingly insuperable lead of over 20% disappeared as blunder after blunder was made by our supposed team of brilliant strategists.
“Many conclusions can be drawn but let me just mention two.
“Negative campaigning just does not work. Constantly denigrating the Labour Leader, when his answer was ‘I do not do personal politics’, lost votes. Attacking every expenditure plan however reasonable just made the Conservatives look like people without a plan.
“The impact of social media is much more significant than was expected. But any Party that looked at how Trump and Sanders – and for that matter Macron in France – campaigned, should have seen this coming.
“So where do we go from here? The current Government seems to have patched up some sort of deal with the Ulster based Democratic Unionist Party. These are probably the most illiberal group of politicians in the House of Commons. It is hard to see this Government lasting its full term and already people are looking round for a new Conservative leader. Meanwhile Jeremy is closer to number 10 than many ostriches imagine.”
Lessons for New Zealand
So what lessons can we in New Zealand learn from the UK election?
Firstly, opinion polls can no longer be trusted as reliable predictors of election outcomes.
In retrospect, the British Prime Minister’s biggest mistake was to call an election on the basis of poll results – especially in light of the UK pollsters’ poor track record.
In the lead up to the 2015 General Election, while the polls predicted a hung Parliament, the Conservatives gained a majority for the first time since 1992. The resulting calls for state regulation of the polling industry, led to an independent inquiry by the British Polling Council and the Market Research Society.
Their investigation revealed that not only had poll samples been unbalanced – with Tory voters under-represented and Labour voters over-represented – but the difference in turnout rates between age groups had not been properly accounted for, with young voters overestimated.
Wrong about Brexit
UK pollsters were also wrong about the Brexit referendum, predicting the ‘remain’ vote would win by a comfortable margin, when the ‘leave’ vote won by 52 to 48%.
Last week’s election became the third in a trifecta of shock results, delivering a hung Parliament instead of the Conservative majority the polls had predicted.
A part of the problem was that UK pollsters miscalculated where the collapsing UKIP vote would go, by predicting it would mainly benefit the Conservatives, since they were taking the UK out of the EU. But that underestimated Labour’s promise to follow through with the EU exit, since, with both main parties pledging to honour Brexit, UKIP supporters were free to vote on other policies. As a result, their 3.2 million votes were split between the two main parties.
An over-reliance on polling also contributed to the Tory campaign disaster.
When Theresa May announced the election in April, she was polling at 54% in the preferred Prime Minister stakes – well ahead of Labour’s Jeremy Corbyn on 15%.
This led party strategists to design a presidential style campaign focussed on Theresa May. But in a country that cares more about issues than personalities, a campaign focusing on the leader – instead of the party – was a major blunder.
May, no campaigner
To make matters worse, Theresa May is no campaigner – she even refused to take part in televised leaders’ debates. When she did appear in public, she was described as wooden, repeating her ‘strong and stable leadership’ mantra like a cracked record.
The lesson that polls cannot be trusted is not, of course, unique to Britain.
Last year’s polls in the US, indicating Hillary Clinton would win the Presidency, were also wrong.
It is a lesson that needs to be taken to heart here as well, especially since inflated polling can lead to complacency – whereby parties think they can sleepwalk to victory.
An analysis of New Zealand’s election polling between 1999 and 2011 by Gavin White, the Research Director of UMR, found that our polling companies have a tendency to overstate the votes for National and the Greens, and to understate the vote for NZ First: “One way of looking at this is to take the average error for these four parties across the 19 final polls included in this dataset. That shows that the average error is: National 2.7 percent too high, Labour 0.7 percent too high, Greens 1 percent too high, and NZ First 1.5 percent too low.”
Applying these adjustments to the average of the two latest political polls – a Colmar Brunton poll published on June 7 and a Newshub Reid Research poll published on June 15 – would give National 45.5%, Labour 27.5%, the Green 9.7%, and NZ First 10.7% with the Maori Party on 0.8%, UnitedFuture on 0.1%, and ACT on 0.9%.
These results show that National will need NZ First to govern.
The second lesson from the UK election is that voters cannot be taken for granted.
To most British voters, the election felt unnecessary. They believed their only purpose was to give Theresa May a landslide she didn’t deserve – and they rebelled!
The importance of not taking electors for granted is a lesson that all political parties in New Zealand should have already learnt from the Northland by-election in 2015, when voters, in what had been a very safe seat for National, voted for Winston Peters.
No voter loyalty
A third – and related lesson – is that voter loyalty can no longer be relied on.
These days, voters are less likely to blindly follow their traditional voting patterns – if their preferred Party is not doing enough to address what they consider to be key issues. Instead, they are now far more prepared than they ever were, to shift their allegiance and their vote.
This may well be a feature of our September election.
In particular, there is a groundswell of discontent regarding National’s handling of Maori issues, and unless the Party addresses this in their Manifesto, there is every likelihood that their vote will decline further than they might expect.
Policy is significant
This brings us to the next lesson from the UK election – parties must produce well designed policy manifestos to win the support of voters.
In that regard, the Conservatives failed badly. Their manifesto, which was the antithesis of an inspiring vision for the future, was likened to “three spoonfuls of arsenic” – with callous cost-cutting policies that placed an enormous financial burden on vulnerable elderly people needing care. Forced to backtrack on the policy, Theresa May insisted that nothing had changed, reinforcing perceptions that she didn’t care and couldn’t be trusted.
In comparison, Labour’s manifesto was bright and appealing to voters with its slogan ‘for the many, not the few’. The policy prescription for a fairer Britain was aimed directly at Labour’s heartland voter base, with promises that resonated including nationalising the railways, introducing a living wage, and eliminating interest on student loans.
This leads to the fourth lesson – good campaigning counts.
While establishment candidate Theresa May hid away in 10 Downing Street, Jeremy Corbyn – like Donald Trump – positioned himself as an anti-establishment outsider. In doing so, he successfully tapped into a genuine vein of anger and fear amongst millions of voters, concerned about the quality of their healthcare, education and public services.
Corbyn’s old style campaigning drew massive crowds in up-beat rallies. Using the campaign style of Bernie Sanders in the US Democratic primaries, he surrounded himself with crowds of young fans, using social media to boost his surging popularity.
His messages resonated: Labour’s campaign had “changed the debate and given people hope. Hope that it doesn’t have to be like this; that inequality can be tackled; that austerity can be ended; that you can stand up to the elites and the cynics. This is the new centre ground.”
A final lesson from the UK is that how well politicians respond to the unexpected during an election campaign can make all the difference. While the Conservatives overestimated the importance of Brexit to voters, they underestimated growing concerns over security. The Manchester bombings and London Bridge attacks created a powerful concern that Theresa May had not done enough to combat terrorism. Jeremy Corbyn capitalised on this by pledging to do whatever it would take to make the UK safe.
The impact of this disastrous election campaign on the UK will be significant. Not only is Theresa May’s leadership now threatened, but Britain’s negotiating position with the European Union is inexorably weakened.
Dr Muriel Newman is Director of the New Zealand Centre for Political Research, a web-based free weekly Newsletter, NZCPR Weekly. The above article, which appeared in her weekly edition dated June 18, 2017 has been reproduced here with her permission.