Spot the difference: early analysis of the relationship between Brexit vote and anti-Trump petition

Many people have argued that the victories of Brexit and Trump are linked, because of a "wave of populism." The petition to prevent Donald Trump from making a state visit to the UK offers an opportunity to test this.

Here's a very preliminary analysis of the relationship between UK votes at the EU referendum and the number of signatories to the anti-Trump petition. Just as a caveat, my research does not usually look at UK politics, so apologies in advance if I'm overlooking some important factors!

As the EU referendum did not use parliamentary constituencies as the data-gathering units, I've depended on the inferred data created by Chris Hanretty at UEA.

The petition data are collected at the constituency level, so we can pull these directly from the parliament website in JSON format. I captured the data at the moment the number of signatories tipped over a million.

Plotting the two data sets as maps gives us the images in Figure 1.

Figure 1: Constituency-based comparison of EU referendum vote and anti-Trump petition signatories. Apologies for missing much of Scotland due to data aggregation issues.

As we can see, there is clearly a relationship.

In England and Wales, the correlation is 0.87; a quick linear model found the relationship between the two variables to be highly significant. A correlation of 0.88 is achieved in Scotland, although this is based on an incomplete data set for Scotland (apologies for that).

Interestingly, this does not hold in Northern Ireland: there is no relationship between EU referendum vote and anti-Trump petition signatories.

So there we go. This is just a first hack, and as political scientists know, significance can quickly disappear when you introduce other variables. But at least at the moment, it looks like there is a relationship between the EU referendum vote and the number of anti-Trump signatories.

Steve doesn't do Facebook but he does do Twitter