Age appears a bigger factor in immigration debate than class

Before we talk about social grade or class and the immigration debate, we must talk about age.

It is a political truth universally acknowledged that the issue of immigration in divides Britain primarily along class lines.

Pundits and politicians from David Goodhart to Nigel Farage will tell you the same story.

While the metropolitan liberal elite that dominate all political parties are in favour of open borders, salt of the earth working folk worry about their wages and the country they once recognised.

The charge particularly wounds the Labour Party. After all, it was founded to represent the interests of the working class in parliament.

The charge is that it has been captured by middle class professional politicians, who want to impose liberal values on working class communities which are in fact essentially small-c conservative. The immigration debate is offered as exhibit A.

So, new data, which would seem to contradict this story from the Ipsos-Mori monthly issues tracker, caught my eye.

Some context: Various polling houses carry put ‘issue trackers’ that ask voters what they think the most important issues facing the country are. Ipsos-Mori is the most prestigious. It goes back more than 30 years allowing for very long-term comparisions. And it is completely “unprompted”.

Most polling houses give a list of issues for voters to choose from. Ipsos-Mori gives no list, ensuring that participants are not led in any way.

In their March 2014 tracker anaylysis, Ipsos-Mori gave this chart, showing the two answers given most by participants when asked: “What do you see as the most important issue facing Britain today?” and a follow up question:  “What do you see as the other important issues facing Britain today?”.

By taking all responses from the January, February and March surveys, they could break the sample up into six segments divided by class – or more exactly, social grade – and age, each segment being statistically significant. In marketing terms, ‘social grade’ is usually taken to measure class. Think of class as a general concept but grade as a technical definition, like the difference between ‘speed’ and ‘miles per hour.’ Crudely put, ABC1 refer to ‘middle class’ social grades and C2DE refer to ‘working class’ social grades. This is the chart that Ipsos-Mori released.

Age immigrationj

Ipsos-Mori released the chart to show how Britain divides along age and social grade lines on immigration, and it does: Older and C2DE voters are more likely to identify immigration as an important issue facing Britain than younger and ABC1 ones.

But what struck me was how much age appeared to explain immigration’s issue salience more than social grade. After all, immigration was the leading issue identified by 55+ year olds from both sets of social grades

Technically more “middle-class” ABC1 voters did so (48 per cent) than “working class” C2DE voters (46 per cent), but this was within the margin of error. Meanwhile, immigration was not in the leading two issues identified by the C2DE 18-34 year old segment or the ABC1 18-34 year old segment.

So I asked Ipsos-Mori if they could provide the data for all six segments on which proportion of each group identified immigration as an important issue facing Britain. The data looks like this (Note that the issues mentioned are immigration/race/race relations in that Ipsos-Mori group these issues together):

Age immigration 2j

So in general the older groups, and C2DE groups are more likely to identify immigration as an important issue. But what’s interesting is that age appears to explain more than social grade.

A simple test is this: For each segment, does the segment with which it shares the most similar result share its age profile or its social grade profile. So for example,  young ABC1 voters (23 per cent), has a closer result to young C2DE voters (29 per cent) than mid-aged ABC1 voters (32 per cent) or older middle class voters (48 per cent).

Overall, four segments have the most in common with segments with which they share an age profile (both 18-34 year old groups and both 55+ year old groups). One segment is closer to a segment of a similar social grade than a similar age (C2DE 35-54 year-olds (41 per cent), have a closer result to C2DE 55+ year olds (46 per cent) than ABC1 35-54 year-olds(32 per cent).) For ABC1 35-54 year olds, the result is tie (At 32 per cent, they are equidistant between C2DE 35-54 year olds (41 per cent) and ABC1 younger voters (23 per cent)).

It would appear from this that age, is more important than class in explaining immigration salience.

This is tangentially supported by the most the groups most likely to be concerned by immigration being the over 55s, the groups most likely to be out of the workforce and therefore not threatened by ‘undercutting’ of wages by migrants.

This, however, it could be argued, is an inelegant test. There are three age categories, but only two social grade categories. So while any segment can be compared with all other voters that do not belong to the same social grade but do belong to the same age range, they cannot be compared to all voters that shared the same social grade but not the same age range.

Thankfully, Ipsos-Mori provide the weightings for the different segments. So for each segment, we can see what proportion of all voters of the same age range but a different social grade identify immigration as an important issue, and what proportion of all voters of the same social grade but a different age arrange identify immigration as an important issue.

We can see whether segments tend to have a closer result to voters of a similar age range, or a similar social grade. The raw data and workings are here.

In the chart below the red bars show the result for each segment. The blue block to the left shows the result for all other voters of the same social grade, the green block to the right shows the result for all other voters of the same age range.

Age immigration 3j

For four out of the six segments (The 18-34 year old segments and the 55+ segments) have more in common with voters of a similar age than with those with social grade, by ranges of between 9 and 18 points. For two out of the six segments (both the 35 to 54 year old segments), have more in common with voters of the same social grade than  age, but by a smaller magnitude ( between 3 and 6). So we can say, drawing on this data, that age is likely to be more important than social grade in explaining the salience of immigration.

For comparison’s sake, we can repeat the exercise by looking at what proportion of each segment said that the economy was the biggest issue:

Age graph 4j

On this measure, five of the six segments are closer to the rest of their social grade than their age cohort, with ABC1 18-34 tied between their social grade and their age group. Here social grade segments are more likely to identify the economy as an important issue, with little pattern along age lines. This is a much clearer class division than we saw with immigration.

To see exactly how much both age and social grade explain both immigration and economy salience, we would need Ipsos-Mori to release the individual level data for their issues survey.

We can say there is some relationship between class and immigration salience. But this data suggests, that before we talk about social grade or class and the immigration debate, we must first talk about age.

Like this article? Sign up to Left Foot Forward's weekday email for the latest progressive news and comment - and support campaigning journalism by making a donation today. 

3 Responses to “Age appears a bigger factor in immigration debate than class”

  1. Gavin

    Indeed, age may well be the real explanation behind any differences in attitudes between classes. It partially depends on how class is measured, of course.

    If we use the class measure created by Mike Savage and Fiona Devine ( then the average age of the traditional working class is 66*, and if older people are more against immigration, then it is inevitable that measuring for working class will get the same.

    *Because the NRS class measure (i.e. the outdated ABC1 C2DE version) is based on ‘manual’ work, many of the jobs that young people do (call centres, retail) may well end up coded as C1. Certainly ‘students living away from home’ are coded C1 by Mintel, so this will skew the ABC1 sample to the young.

  2. Daniel Elton

    Thanks for that – will read when I get a chance

  3. interested

    Mass immigration forces wages down and keeps housing prices high. All the parties are in favour of this policy so it will continue indefinitely. That is it.

Leave a Reply