At 10pm on Thursday 8 June, British voters faced yet another political shock. Despite an overwhelming 21 percent lead (YouGov), the Conservatives failed to win enough seats to form a majority government.

As the count unfolded, the reality dawned that Theresa May did not receive the landslide that she had gambled on. Instead her party had to enter a confidence and supply arrangement with the Northern Irish Democratic Unionist Party (DUP). With Brexit negotiations kicking off on June 19, and rumours circulating that May is facing a mutiny from her own MPs, the ‘strong and stable’ government she promised now seems almost impossible.

One popular rough and ready way pundits and commentators used estimate the eventual Tory vote share in a given constituency was to take the 2015 Conservative vote share, add the 2015 UKIP vote share, and voila – Conservative majority! But, in reality, this did not prove to be the case.

How did it all go wrong?

There are a number of complex factors at play, but perhaps one of the more surprising outcomes has been the movement of former UKIP voters across to the Labour Party (as shown in a graphic from the Financial Times, here).

So how is it that a party commonly perceived as right-wing leaked votes to a party which is portrayed as the most left wing Labour Party in a generation? The answer lies in the economic views of UKIP members. Work by Bale, Webb and Poletti shows that strong party supporters of UKIP locate their party at 7.28 on a 0-10 left-right spectrum, whilst Labour voters placed their own party at 3.44. This gap of 3.84 suggests the two parties would be uncomfortable bedfellows.

However, subjective rankings might not tell the whole story. Bale, Webb and Poletti’s ‘objective measure’ place Labour at 1.60, and UKIP at 2.62, on the left-right spectrum – a gap of just 1.02. Bale, Webb and Poletti explain this by arguing that on economic issues UKIP and Labour are not so different – UKIP’s voters “are very possibly ‘welfare chauvinists’ with a populist (as opposed to class-based) antipathy toward big business.”

This is supported by Wheatley, who plotted the ideological location of party supporters based on two scales – an economic left-right scale, and a social conservatism open-closed scale. Interestingly, whilst in both 2015 and 2017 the median UKIP voter was a 0.2 on the closed-open scale (more closed than open), they were 0.4 on the economic left-right scale (economically left of centre). The median Labour voter, however, was around 0.7 on the open-closed scale in both elections (more open than closed), but on economics moved slightly from about 0.25 in 2015 to 0.2 in 2017 (economically left-wing).

So, we see a situation whereby economically the median UKIP and Labour voters are not very far away at all, and there is considerable overlap in the area covered by both parties in Wheatley’s graphs on that axis – the real gap is on the closed-open axis.

One of the key issues in this election was, obviously, Brexit. With both Corbyn and McDonnell committing themselves to a form of Brexit which does not include freedom of movement, they were able to effectively satisfy a segment of UKIP voters. This neutralised the closed-open gap between the two parties on the big issue of the day, and allowed for UKIP voters who had ideologically similar economic views to Labour to move over – and in many cases move back – to Corbyn’s Labour.

The argument laid out above is not to suggest that Labour wooed more UKIP votes than the Conservatives did – this was patently not the case. Instead, it is to show how some UKIP voters could, ideologically, migrate to Corbyn’s Labour Party and – perhaps – have denied Theresa May the majority she was so sure of winning.

This blog first appeared on POP Politics AUS, and can be found here: https://poppoliticsaus.wordpress.com/2017/06/12/strange-bedfellows-the-ideological-migration-of-ukip-voters-to-labour-in-the-2017-general-election/