Higher education participation rates don’t predict Labour votes

The weekend’s YouGov MRP proved grim reading for Conservatives in general and the Telegraph in particular – but realistically it wasn’t much of a surprise to anyone who has been paying attention. Looking at the new English, Welsh, and Scottish constituencies only the poll points to a 1997 style landslide for Labour. You’ll know I can’t resist plotting an MRP poll result, so do please get stuck in.

[Full screen]

You can view either the MRP results, or a projection of 2019 results onto the new constituencies. The chart on the right shows the MRP party percentages – you need to mouse over the constituency in question, something best achieved by using the region and constituency filters first. It’s a big slow chart because there are still somehow now UK-wide boundary files easily available, I’ve had to roll my own.

All this is fun for a certain strain of political nerd, but there is a higher education angle here too – YouGov use level of education as one of the variables in calculating the constituency totals from their c11,000 sample, but they don’t use any measure of parental aspiration.

As per the discourse, the kind of people that would encourage their children to go to university are supposedly anti-Conservative by instinct, and we are led by certain academic voices to believe that it is this type of people that will quash the UK’s populist revolution.

I’m using the OfS’ TUNDRA mapping of the likelihood of most 18 year olds in an area to go to university (scaling up from MSOA to constituency), and seeing if we can spot any patterns among the constituencies likely to change hands.

Reader. We cannot.

[Full screen]

Labour are projected to gain Ashfield (progression to HE rate of 27 per cent) and Harrow East (67 per cent), while the Conservatives will hold Finchley (65 per cent) and Havant (26 per cent). Labour will hold the seat with the highest progression rate in the country (Harrow West, 67 per cent) and the lowest (Bristol South, 23 per cent). There’s not really any way to distinguish these populations using the TUNDRA data, so whatever is going on it is not driven by higher education aspirations.