Brexit: Why Nobody Can Predict How the U.K. Will Vote

This article originally appeared in The Politics Counter. Read the original article.

Anybody who says they know for sure how Britain will vote next week is fool or a liar, and possibly both. Since last weekend, we have had ten polls conducted by eight different companies. Five of the polls have been online, the other five conducted by telephone. As a result, we can be certain about some things, but less certain about others.

What we know for certain

There has been a marked shift since the start of June from "remain" to "leave". Every polling company finds more support for Brexit than they did in mid-to-late May. Movements in individual polls might reflect sampling fluctuations; but when every polling company, whatever their methods, finds the same direction of movement, this can no longer be reasonably explained by chance.

There are significant geographical differences. If the result is close, then London, Scotland, Northern Ireland and probably Wales will vote "remain", while the rest of England will vote "leave". Whatever the overall outcome, the political consequences of these divisions are likely to be huge.

There is a big generation gap and a big education gap. People over 65 are much more likely to support Brexit than people under 30; and those with university degrees are far keener of staying in the EU than those who left school at 15 or 16.

On the other hand, there is NO evidence of a significant gender gap. More women are undecided than men, so might well make up a majority of the late-deciders who could well make the difference next Thursday. However, among those who take sides, men and women currently hold similar views. This is unlike Scotland in 2014, when men voted on balance for independence, while women plumped decisively for keeping the U.K. together.

What we don't know for sure

While the polls agree about the direction of movement, they don't agree about the scale of the move or the current level of support for "leave" and "remain." At one of the scale, since mid-May. And in the past six days we have seen figures ranging from a two-point "remain" lead (Opinium) to a ten-point "leave" lead (ORB). I believe the race is currently neck-and-neck, but this is more a best guess than a firm estimate.

There are still differences between online and telephone surveys. Overall, these differences are less stark overall than they were; but even when they report similar figures overall, they disagree about the way different groups of voters line up. Online polls tend to report big "leave" leads of 20 points or more among people over 65 and those who voted Conservative in last year's general election. (There is, of course, a big overlap between these two groups.) Telephone polls tend to show a closer contest among Tories and the elderly. These differences provide another reason to be cautious about interpreting the polling data.

We have no real idea how high turnout will be next week. This could affect the result—but we can't be sure how. Let's suppose one or two million people make a late decision to go to their local polling station rather than stay at home. And let's assume they belong to groups who are least likely to vote in general elections. If they are predominantly young people (including those who have registered online in the past few weeks), then this is good news for "remain". But if they are mainly people who have the fewest educational qualifications, struggle to make ends meet, and live in parts of Britain where general election turnouts are way below the national average, then this will boost the chances of Brexit. Unless, that is, Labour manages to persuade its working-class supporters that it is in their self-interest to side with the overwhelming majority of Labour MPs and vote "remain."

The battle to offer the safer option

What, then, will happen in the final week? I argued in a previous blog that past referendums have usually produced a shift to the status quo, because people who decide late have opted for the safety of things continuing as they are, rather than take the risk of voting for change.

I stand by the basic judgement that late deciders will prefer safety to risk. However, I am no longer sure that this will benefit "remain". In past fortnight the Vote Leave campaign has successfully raised the saliency of immigration, while the Stronger in Europe campaign has so far failed to keep the economic risks of Brexit at the forefront of voters' minds.

This raises the prospect that the late deciders, who generally follow the news least, and have the weakest feelings about the EU per se, will take sides next week on the basis of "controlling our borders" rather than the dangers that Brexit poses for jobs, prosperity and living standards.

We should remember that public attitudes to immigration are prompted largely by the rise in insecurity and inequality over the past decade. Secure, well-paid jobs for millions of people are harder to come by, public services are being squeezed, most people in their twenties cannot afford to become home-owners, and so on. We can debate the reasons—technology, the 2008 banking crisis, demographic pressures on public spending, etc.

To many voters, however, globalisation and immigration are major culprits, making life easier for the well-off and harder for everyone else. If Brexit is seen as the route to keeping the rest of the world at bay, and giving Britons better-paid jobs, improved public services and a generally more secure future, then a final-week shift to safety could take late-deciding voters away from the status quo and towards Brexit.

On balance I still expect a shift to the status quo, and a narrow victory for "remain". But we are dealing in probabilities and not certainties. We have a blurred picture of the present, and can only guess the future. It is perfectly possible that the trends of the past fortnight will continue, and we shall vote to leave the EU.

The big question for the final week is: can Stronger In Europe make its economic warnings of Brexit stick? If they can, they will win; if they can't, they will lose.

Peter Kellner is a political commentator and the former president of YouGov. He tweets at @PeterKellner1.