This has become the key determinant of our election resultsby Paula Surridge / December 18, 2019 / Leave a comment
“Education, education, education,” the mantra of the first Blair government, and one which heralded social change on a scale that has yet to be appreciated. The target of 50 per cent of young adults entering higher education, only recently met, is reshaping the social and the political landscape.
Unlike the social changes of the Thatcher era such as Right to Buy, it was not clear at the time it was announced that getting more young people into higher education would be electorally advantageous for Labour. In the previous election, the Labour Party had been much more successful winning the votes of those without degrees than those who held them. However, one of the most important changes over the period since then, and particularly evident in the electoral maps of 2017 and 2019, is the relationship between the educational profile of areas and how they voted, notably also a key feature of the EU referendum vote. While the detailed analysis of these patterns will not be fully available for some time, some glimpses of its silhouette are possible using the aggregate data on how different types of place voted in 2019 and the longer-term trends since 2010 (the first election held under the current constituency boundaries).
The key to understanding why education is important is political values. Politics has long been thought of as a battle between the left and the right, but this captures only part of the story. There is another dimension of social values, which is not correlated with economic positions, but which includes issues increasingly salient in our politics such as criminal justice, immigration, security and foreign aid. This “other” dimension is much more strongly related to referendum vote than the traditional left-right values are. On these “social” issues a significant portion of voters who are otherwise on the “left” in terms of economics are in more conservative positions. The core demographic driver on these social issues is education: crudely, those with degrees are less socially conservative than those without. It will be some time before we are able to look at how the values of individuals mattered for voting in 2019. But the patterns in terms of education for constituencies, and particularly how these have changed over time, are revealing.
Figures quoted throughout are simple bi-variate correlations for seats in England, measuring the extent to which two things vary at the same time. The higher this is in absolute terms the stronger is the connection between them, while negative values indicate that as one rises the other falls. Positive numbers indicate both factors vary in the same way.
Looking only at seats in England (the Scottish and Welsh electoral choices are now too different to take all parts of the UK together), we see significant changes in the relationship between the proportion of the electorate with level 4 qualifications (broadly speaking degree level or higher) and electoral outcomes. In 2010, the correlation between the Conservative vote share and this measure was a small to moderate positive one (0.23): there were on average more Conservative voters in places where there were more people with degrees. By the 2017 election this correlation was no longer evident: there was no relationship between the proportion with level 4 qualifications and the Conservative vote. But in 2019 this correlation turned negative (-0.20). In other words, in 2019 there were fewer Conservative voters in places where there were more people with degrees.
We can also see an electoral pay-off of the expansion of higher education beginning to happen for Labour. Even by 2010 it was not clear that the rising numbers of graduates would be to Labour’s advantage; there was a moderate negative correlation for Labour (-0.30) such that places with more graduates had fewer Labour voters. This correlation has gradually reduced over the elections since 2010 and in 2019 was very close to zero (-0.09).
Two other elements are critical to this story of realignment around an education cleavage—the Liberal Democrats and turnout. The Liberal Democrat vote has long been stronger among those with degrees than those without. But in 2010 they rose to be top of what was broadly a three-way split among these voters. Their more recent collapse in part explains the dynamics in patterns of education and voting. However, the new Liberal Democrat phoenix rising from the ashes of electoral defeats in 2015 and 2017 represents an even more socially liberal, and highly educated, part of the electorate.
In 2010, the Liberal Democrat vote share was already more highly correlated with the proportion of the electorate with degrees than either the Conservative or Labour share (0.43). This correlation further strengthened between both 2015 and 2017 and between 2017 and 2019, as the party made a conscious play for the most liberal (and remain leaning) groups of voters. In 2019 this relationship between Lib Dem vote share and the proportion of the electorate with degrees is one of the strongest correlations between demographic characteristics and party support (0.577), stronger even than the correlation between Labour vote share and the proportion of people aged 18-24 in a constituency.
In the coming weeks many pieces will be written looking at the fortunes of the Conservatives and Labour in 2019 and beyond, but in trying to understand any class based realignment—the “Blue collar Conservative” vote—ignoring the rebirth of the Liberal Democrats among the well-educated middle class will mean only a partial picture is painted. The Labour Party seeks these voters based on their socially liberal values, while the Conservatives seek to appeal to them on economics and the fortunes of each may rest on whether the Liberal Democrats can further strengthen their vote share in these groups.
Finally, it is important not to overlook the impact of education on turnout. Notwithstanding a small decline in turnout between 2017 and 2019 (the jury will be out for some time on how much this represented a lack of enthusiasm for the choices or the time of year), elections since a low point in 2001 had been showing gradually increasing turnout. While in 2017 this became talked about as the infamous “youthquake,” even in that election the change in turnout was more closely related to education levels than age per se. Education levels have long been recognised at the individual level as a strong predictor of engagement with politics, and turnout in particular. While this effect has also strengthened since 2010, there was no change in this between 2017 (0.63) and 2019 (0.63); in both elections, places with fewer graduates had lower turnout.
In the few days that have passed since the election, the discussion has focussed heavily on whether Brexit or Corbyn were to blame for Labour’s catastrophic results. But the truth is more complex than this binary suggests. Rather both the troubles caused to Labour by the Brexit vote and Corbyn’s leadership can be understood as consequences of changes in the character of the Labour Party, which have moved it away from what has been thought of as its traditional voters. Political competition is now fought over more than just economics—and while it is easy to point to Corbyn’s left-wing economic manifesto as a source of weakness, it was often the non-economic aspects which were least favoured by the public and particularly by voters in the Labour heartlands. As Labour starts to try and rebuild its electoral coalition, understanding changes to the social and political landscape set in motion by the last Labour government, 20 years ago, will be central to finding its way back.