Sunday, 30 March 2014

Just who IS voting for UKIP? (Part I)

There have been countless attempts to explain the growth in UKIP support before and since 2010. For some, like Paul Goodman at Conservative Home, the growth has come at the expense of the Conservatives for the most part. For others, like Rob Ford and James Bloodworth, Labour have reasons to be fearful of the rise in UKIP support. The truth probably lies somewhere in-between, which is perhaps why Paul, Rob and James are right to stand their ground.

This article brings together what I know about the demographic groups which have VOTED for UKIP between 2010 and 2013. For the purposes of this article I am going to discuss the demographic groups which have voted for UKIP in the 2010 general election. I have already analysed the 2012 and 2013 local elections and will introduce my findings in future posts.

Below I am presenting two maps showing where the party stood candidates and the spatial distribution of UKIP vote share at the 2010 general election.

The broad pattern is of the party doing well in eastern England, the south west and parts of the west Midlands. Of course this is all relative. The party only polled over 10% in one constituency (Buckingham).

Next I want to introduce the data I will use to describe the demographic composition of UKIP voters. The Mosaic data I use breaks down the population of the UK into 69 different segments. The charts below shows the likelihood of each of these groups identifying with either the Conservatives, Labour or the Liberal Democrats. At this stage I only want you to look at the patterns. I will go on to explain what A1, A2, etc means but for now take a moment to look over the charts. I have the proportion of each of these demographic groups for every neighbourhood, ward and parliamentary constituency in the United Kingdom.

So, what does this show? Those demographic groups with a score above the axis line (eg D14 for the Conservative party) are much more likely to identify with that party. I want you to identify the patterns yourself but let me pick out a couple of examples. First, the 'grey' voters (L52 to L55). These groups identify with the Conservatives and, to a lesser extent, the Liberal Democrats. I will describe them in more detail later but these voters are elderly and mostly comfortable. This group is much less likely to identify with the Labour party. Second, the 'G' groups (G26 to G34) are unlikely to identify with the Conservatives, and much more likely to identify with the Liberal Democrats or Labour. These voters are early-career graduates, young professionals or students.

These are just two examples. I will explain in more detail who these groups are, and which have voted for UKIP, in subsequent posts but I want to use this first post to introduce this data. Next look at the graphic below. This shows the turnout record for each of the 69 groups at the 2010 general election. Those above the line have good turnout; those below the line do not.

Again, it's for the reader to interpret these results as they see fit. However when combined with the party ID charts above we can begin to see a problem for Labour. Many of the groups which identify most strongly with Labour also have the poorest turnout records. For example the 'I' groups (I39 to I44) identify with Labour for the most part but didn't turn out in 2010. These groups are urban, on lower incomes and reside in and around town centres across the UK. Their communities are often ethnically diverse and, as I will go on to describe, there are also groups here which have been attracted to UKIP (I43 - Town centre populations with transient single populations, for example).

The next stage is to start to pick apart the voting records of each of these groups in the 2010 general election and subsequent local elections. I have been able to show the correlation between each of these demographic groups and vote share for all the major parties. This article is intended to show those voters which have been persuaded by UKIP. In future posts I will describe in more detail who these voters were in 2012 and 2013. For now though I will begin by showing the chart relating those demographic groups which voted for UKIP in the 2010 general election. Remember to put this graphic in context. In 2010 UKIP were polling at much lower numbers than they are today. For the most part they were portrayed as an off-shoot of the Conservative party. Take a look at the chart below.

There are some surprises here. For the most part the UKIP demographics seem to match those from the Conservative party ID chart above. The A,B,C,D groups identified more strongly with the Conservatives, and some of them voted UKIP in 2010. These groups are rural, wealthy, suburban or professional. I will describe them in more detail in future but they align with the Conservatives and UKIP. No real surprises there. However take a look at the E19 group. This group identifies somewhat with Labour but turned out to vote UKIP in 2010, albeit in smaller numbers. This demographic group is approaching retirement and reside in suburban semis in industrial towns, often in the north-west, Yorkshire and the West Midlands.

Now look at J46. This demographic group does not align strongly with any of the major political parties but has a very strong turnout record. This demographic group are blue collar workers in their early 30s with young families. Average household income is around £30,000 per annum, they struggle to make ends meet and likely have some household debt. They are employed in intermediate occupations and are geographically distributed across the country, with larger concentrations in the East, Midlands and North West of England. In 2010 this group came out to vote UKIP.

K51 voters are often indebted families living in low rise estates, most typically in the south-east. This group aligns with Labour but came out to vote UKIP in 2010. This group are among the poorest in society, struggling to make ends meet and relying often on benefits. Until recently this group, despite having few qualifications, has found it relatively easy to find work. However the economic downturn has hit them hard. Immigration is likely a deeply personal matter, given that they will often be in direct competition with low-skilled migrants for the few jobs available. They probably voted UKIP through sheer anger at the perceived failure of the Labour government to deal with this issue. As I will go on to show, they appear to be staying away from Labour.

I have carried out some research into the demographic groups which voted for UKIP in 2010 previously on my blog. Below is an interactive chart to show which demographic groups were positively (and negatively) correlated with UKIP vote share. Care needs to be taken when interpreting this table. In 2010 UKIP was a party which polled below 5% in many constituencies across the country. Still, they are an indication of the kinds of demographic groups UKIP appeals to.

Finally I want to present a spatial analysis of specific areas of the UK. I have used a balance measure to show which neighbourhoods of the UK are receptive to UKIP. These are created by summing the voting record of each of the 69 demographic groups. They provide an interesting snapshot of the neighbourhoods and regions of the country where UKIP, based on demographics and voting behaviour, can expect a favourable recpetion.

As the map below shows UKIP should expect a fairly good reception away from the towns and cities throughout the south-east, the east and west Midlands. This is relatively unsurprising given that many of these areas are represented in parliament by Conservative MPs. As I have shown above there are similarities in the types of demographic groups which are attracted to UKIP or the Conservatives (older, rural, comfortably-off, empty nesters, etc.). Many of these groups reside in the shires and semi-rural small towns of the south-east. In the towns and cities, however, a demographic boot stamps all over the likely reception UKIP will receive. There are simply too many of the demographic groups which are strongly unattracted to the party. That's not to say there aren't pockets of areas around the major towns and cities which could have good proportions of UKIP-friendly groups.

London is not great for UKIP for the most part. However there are pockets of neighbourhoods within the M25 which the party can expect to do well in. For instance Havering and Barking in north-east London is a potentially fertile area for UKIP, as the map shows. In fact the party has perfromed well in local elections in north Havering, confirming somewhat what this map shows.

The next map shows East Anglia, a region where UKIP has performed well in European elections. I want to use this map to conclude this post, but also as a conduit to my next post. Look carefully at the map. You will see a flaming red region in Cambridgeshire. According to my data this region offers a tremendous demographic opportunity for UKIP. There are just the right proportions of those demographic groups which the party attracts. In 2013 local elections were held in the area. The three electoral divisions highlighted in bold saw convincing victories for UKIP candidates, including one division (Ramsey) where the party secured 67% of the vote!

The next post will move forward to discuss the performance of UKIP in the 2012 and 2013 local elections. As I will show, not only has the party strengthened its hold on the 2010 cohort, it has also broadened its appeal. For now though, I will exit the stage.


  1. Just to say thanks for providing these charts and commentary. The graphics look great.

    1. Thanks Matt, much appreciated. I will be doing the same sort of things for the 2012 and 2013 local elections later this week.