James Kanagasooriam is an advisory board member of Onward and a strategy consultant at OC&C. Between 2014-2018, he was Head of Analytics for Populus, where he and his team utilised data science techniques to optimise political campaigning. Their efforts contributed to Ruth Davidson’s Scottish Conservatives’ successful 2017 Election.

It might not feel like it this week, but the Conservatives’ problem with younger voters is a bigger problem for them than Brexit. This morning, Onward publishes a big new report on the age gap in British politics – now the most important indicator of vote intention. The stark reality in the data is that the Conservative Party’s age curve is not only extreme – you now need to be 51 years old before you become more likely to vote Conservative than Labour – but worse than the age curve for Leave. This is often missed because support for Leave is higher amongst older voters than voting Conservative.

You don’t get any prizes for knowing that young Britain is skewed considerably towards Remain and Labour. One of the more distasteful arguments put about in the media about Leavers is that they are facing a natural demographic decline – just last week a young commentator on a BBC panel said that delivering on the “will of the people” now meant delivering on the “will of dead people”. But the Conservative age curve is much, much worse – and spells steady decline for the Party in the long-run unless something is done at pace.

The Conservative curve gets steeper the older you are – right up until very old age. The difference in voting intention between people in their 40s and 30s is smaller than the difference between people in their 50s and 40s. This differs from Leave, which arcs over middle age – as set out below. There are huge numbers of middle age voters who were willing to vote for Leave but are not currently willing to vote Conservative. That is a problem for the Tories.

Figure 1: Observing the problem – disaggregating the Leave issues from the Tory one

Why is this? It is because the Leave divide is primarily an educational gap, not an age gap. When you mathematically adjust for education, ethnicity, social and economic worldviews and housing tenure, the difference in Leave and mean voting behaviours by age is almost entirely explained. But when you do the same for the Conservatives, we find a significant chunk of the difference unexplained at the demographic top and bottom of the electorate as set out below.

If we look at the maths, we see something powerful at play. The Conservative vote is crumbling faster than the Leave vote over time. In effect, the Conservatives are about to lose a group of older voters who are more likely than they should be to vote Conservative – once we have factored in everything we know influences voting behaviour – which will be replaced by a group of voters who are much more likely than they should be, controlling for every conceivable factor, to vote Labour. In politics, cohorts matter and the next one could kill the idea of a Conservative majority.

Figure 2: The bow tie of doom: actual versus statistically inferred Conservative age curve

This flies in the face of most Conservative strategists’ thinking. Currently the Party is relying on the idea that if we a) build more houses b) build a vision of Britain that creates prosperous high paying jobs and is spread outside city centres c) provide people with the skills and retaining capabilities for a flexible and changing job market without expelling them from their hometowns, networks and families, then they will necessarily vote Tory. We will be able to create more Conservative voters through good government, as we have done in the past – or so the story goes.

Here’s the problem: people aren’t equations. They are social animals they take their cues from the people around them. They practice what I call “empathic voting”: their political actions are influenced by their generation and the times they live through. The political contract is not such that if Governments pull certain policy levers and create certain conditions voters respond in kind. Leaders matter, external moments matter. Brand matters. The liberal consensus of the late 1990s and 2000s didn’t come out of nowhere – it stemmed from a generation of leaders who grew up in the shadow of the Berlin Wall. They thought Fukuyama was right and they governed accordingly. Today no voter under the age of 40 has had a similar formative experience. First time voters at the last election were not even politically conscious when British capitalism was last deemed to be in good health. This is why – crudely – the older postie votes Tory and the younger investment banker votes Labour; each are voting – to a degree – in empathy with their own generational cohort.

The converging trends above are what I term the “bow tie of doom” – a possible cohort effect in different generations’ voting behaviour. A generation of younger people which – controlling for everything about them, including their views, where they live, their ethnicity, their income – are less Conservative than they should be, sitting alongside a cohort of older voters of whom many tick the Conservative box when they are not really Conservative in any meaningful, demographic or philosophical way. The Party might benefit from a political premium right now but it is unrooted in anything except time – and the clock is ticking.

There is a lack of longitudinal data to fully validate this cohort thesis, but the Onward data, a detailed snapshot, gives us a clear indication that we are not making conditions ripe for conservatism – and even if we were, it wouldn’t be enough. In fact, we are creating battalions of anti-Tory shock troops – disheartened by austerity, uninitiated in prosperity, alienated by the Conservative brand. If every generation has a centre of gravity in its voting patterns, today’s is far away from the current Conservative Party.

We need to learn the right lessons now. The solution is threefold. First, the Conservatives need to radically rethink what they stand for. It can’t be everything – we should pick five things and double down on them again and again until we are blue in the face. This is hard in Government – you are by default responsible for everything – but ruthless clarity and urgency of purpose are essential if we are to cut through. Mass movements spring from evocative narrow issues, not from broad manifestos.

Second we need to completely retool our polling and data operation. Right now, CCHQ’s operation is limited in both breadth and depth. They know a little about a few things: they need to know a lot more and bring in the best brains to drive it. Good businesses start with their customers and repeatedly refine the proposition until they get product-market fit. They don’t start with the product and then try and find an appropriate market.

Third, we need to build a policy platform underpinned by data and tailored to our audience. Onward’s work identifies the precise policy areas that three million young considerers would be swayed by – and many are Conservative strengths. If we can’t build a compelling proposition around low personal taxes, responsible capitalism, protecting the environment, reforming public services and being proud to be British, we deserve to lose.

If we spend the next two years arguing about Brexit we will lose the next election, no matter where we end up. If we are to ever have a majority again, we need to reverse the age gap.