Twenty isn’t plenty
The Conservatives’ vote share of 20% in historical context
I have a clear recollection of sitting in a presentation around the turn of the century. Sir Robert Worcester told the assembled MORI audience that the two main political parties in Britain had core levels of support of around 30% of the voting public. This was “tested to destruction” by Labour in 1983 and 1987, and by the Conservatives in 1997.
He was talking about election results - “real” voting - not voting intention polls of the kind that made the headlines last week. Ipsos’ latest measure, collected in late February, published in March, and based on those intending to vote, recorded the Conservative’s share at a breath-takingly new low of 20%.
This was significant because no pollster has a longer time-series than Ipsos, formerly Ipsos MORI, before that, MORI. Sir Robert started it in 1976 and it is without equal in terms of its longevity (see here).
The Conservatives have been close to this low before. They were the choice of 22% under John Major in December 1994 and May 1995, 23% in July 1994, in July and August 1997, shortly after Labour’s landslide win, and 23 per cent in December 2022.
However, a share between 20% and 29% is not that unusual. Among 719 polls since 1976, Ipsos recorded a vote share for the party in the 20s on 151 occasions. That’s an incredible (and fitting) 20% of the time, but there’s an important caveat. The time-series isn’t confined to Ipsos’ standard monthly polls.
As elections approach, polls are in greater demand, more are conducted. This has changed over time. For example, while Ipsos conducted 11 polls in 2017 and 10 in 2019 — the last two election years — this compared to 12 in just four months before the 1997 election.
This possibly reflects the last two elections being ‘snap’ elections and a change in the media as well as the polling landscape (there are many more wannabe pollsters). One effect is that periods such as 1992–1997, when the Conservatives were limping along, are over-represented in the time-series.
The 151 times the Tories have been in the 20s compares to Labour’s much smaller 42, but this is unsurprising too given blue has been the colour of the Government for two-thirds of the time-series. Even in good times, Governments can expect to suffer mid-term discontent.
It will be of no comfort for the Tories that Labour still holds the unenviable record of the lowest poll rating in the series; 18% in May 2009. Nor that Labour bounced back to 25% a month later and the party was back in the 30s by the end of the year. Or that vote shares are subject to statistical margins of error and that outliers can occur 5 times of 100.
There have been four occasions when both parties concurrently had a share in the 20s — twice in 1981, twice in 2019 — and several times when they have both been in the 30s. This, and low vote shares for both parties at the same time, has usually occurred because of two factors. First, a strong challenge from ‘third’ parties — most obviously, the Liberal/SDP Alliance during the 1980s, the Liberal Democrats and UKIP more recently. A second factor is relatively unpopular leaders simultaneously at the helm of both the main parties.
The key topic of interest is what history tells us might happen next. On each occasion that one or other party got a sub-30% share, they went on to lose at the subsequent election.
The nearest we have to an exception was February 2017. Jeremy Corbyn’s Labour was on 29% then and trailed Theresa May’s Conservatives by 11 percentage points. Labour, or perhaps more accurately the Conservatives (hamstrung by Brexit and a terrible election campaign), reduced this gap to 2 points by election day just a few months later.
The boot is on the other foot now. It’s a governing party not an opposition trying to build from a position nine percentage points behind where Corbyn’s Labour opposition party were. That looks like an exceptionally tall order especially for as long as Keir Starmer avoids the sort of gaffes Theresa May’s team delivered.
It’s not uncommon to spot roadside traffic calming signs in Britain suggesting that twenty is plenty. In polling terms, twenty is nowhere near good enough.
