Brexit and the opinion polls

Last week was a good one for the 'remain' camp in the UK's EU referendum campaign. The Treasury published an analysis of the economics of leaving the UK which concluded that a Brexit would involve significant net costs for the UK. Later in the week US President Barack Obama warned that outside the EU the UK would be "at the back of the queue" for a trade deal with the US.

But where does public opinion stand?

No opinion polls had, as of Sunday evening, been released which covered the period following President Obama's comments on the risks of Brexit. But even before the President's comments the lead for the remain camp had widened.

Taking an average of the six major polls published up to 19th April, adjusted for the removal of the 'don't knows', the remain vote stood at 54% and the 'leaves' at 46%. This is the widest lead for remain since 23rd February.

This is a snapshot of public opinion. But taking a longer perspective on the opinion polls what lessons emerge?

First, polls carried out by different pollsters frequently give different results. Thus ComRes and ICM both conducted polls between the 8th and 10th April. ComRes gave remain a seven percentage point lead; ICM put leave three percentage points ahead. This difference reflects a wider conundrum. On-line polls tend to show a stronger showing for the leaves than telephone polls. Unfortunately we don't know which approach is right, something that introduces a further uncertainty into gauging public opinion.

Second, public opinion is changeable. Last summer, a poll by IPSOS Mori, one of the most longstanding pollsters on the EU issue, showed that 66% of UK voters support EU membership, the highest reading in more than 35 years. Since then there has been a reduction in the lead for remain, probably partly in response to the migration crisis. Our calculations show that, on average, remain had a three percentage point lead over leave in April's polls so far, down from 14 percentage points in June of last year. More recently, in the space of just over four weeks, between 10th January and 14th February, the lead for the remain camp using a smoothed, six-poll average, went from 10 percentage points to zero.

Third, UK public opinion often shifts in tandem with the relative economic fortunes of the UK and of its major EU partners. In 1975, at the time of the last UK referendum on membership of what was then the European Economic Community, the UK was the "sick man of Europe", wracked by high inflation and low growth. To a troubled UK Germany offered a model of prosperity and stability. In 1975 the electorate voted by 67% to 33% to stay in, a level of support which, as far as we can tell, has never been repeated. UK public support for the EU also surged in the early 1990s as the UK fell into recession and Germany boomed in the wake of reunification. Conversely the euro crisis of 2011-13 saw UK public opinion turn cooler on membership of the EU.

Fourth, support for the EU is lowest among less affluent voters. Currently support for remaining in the EU is running at just 32% among skilled and unskilled manual labourers and the unemployed. More educated, affluent voters tend to be strongly pro-EU.

Fifth, young voters, those aged 18 to 24, tend to support EU membership, with April's polls showing support running at around 60%. Older voters, those aged 60 and above, are more sceptical, with support for remaining currently at 32%.

Sixth, unsurprisingly, views on Europe vary by political affiliation. A majority of Labour and Liberal Democrat supporters favour remaining in the EU. Only 37% of Conservative supporters polled in April would vote to remain in the EU. Puzzlingly, an average of 6% of UKIP supporters polled so far this month say they would vote to remain in the EU. Voters in Scotland and London tend to show higher levels of support for the EU than other regions. The former is in marked contrast to the referendum in 1975 when Scotland was resolutely Eurosceptic and the Scottish National Party campaigned to leave the EEC.

What is clear, and last year's General Election result demonstrated this, is that political polling is not a precise science. BBC's analysis of data from 92 opinion polls carried out in the run up to the election showed 56% had predicted a Labour lead rather than a Conservative majority.

There are, of course, other measures that can be used to assess the likely outcome of the referendum.

As of Friday, the odds offered by bookmaker Paddy Power implied a 33% chance of Brexit. It seems probable that betting markets are focussed on the significant number of don't know voters, averaging at 15% in April's polls so far, and expect a repeat of the 2014 Scottish independence referendum, where the don't knows tended to go with the status quo.

In its April issue of economic forecasts, Consensus Economics polled economists on the likely outcome of the EU referendum. Both UK and euro area economists assign the same probability to Brexit at 41%, a rather higher figure than betting markets.

Turnout will be key. Older people are more likely to vote than the young. In both the Scottish referendum and the General Election, older voters strongly favoured the 'no' campaign and the Conservatives respectively. Given younger voters are more likely to vote for remaining in the EU, a low turnout will favour the leave camp while a high turnout should favour the remain camp. But, as post-summer and recent moves in polls suggest, public opinion is volatile and susceptible to outside events.