British Election Study: Post Election 2019

[clyde1998]

I've been having a look through the panel data for the post election wave of the British Election Study (Wave 19; 2014-2023), which is publicly available. So I thought I'd post some of the interesting pieces of data I can find in it.

Most electoral analysis look at age groups, as oppose to single year age, so I thought I'd graph the change in party's shares by people's age as a continuous. As there's an obvious margin of error for each year, the trend line is a five-year moving average.

(Just as a side note, I did graph Plaid Cymru and UKIP separately in the graph. Looking at it now, it would probably be best to have grouped the SNP and Plaid Cymru and included UKIP with others. Also everything at age 85 or older are grouped together.).

At 46, you clearly become more likely to vote Conservative than Labour. Those who were 46 at the general election were born in 1973 - the year Britain entered the European Community. There's also a significant drop in the Labour vote share at age 40, those born in 1979 - the year Margret Thatcher came to power. The age of these changes may just be co-incidences, but I thought it was very interesting nonetheless.

I'll look at doing the same graph for Scotland and Wales separately, so the SNP and Plaid Cymru figures get some real meaning. The data is detailed enough to tell me which constituency the person responding voted in, so I may be able to do this for seats where the Brexit Party had a candidate too.

[The Bishop]

This clearly shows that the Tories depend on over-60s for their majority - below that its roughly speaking a draw.

[J.G.Harston]

Apr 25, 2020 6:16:18 GMT clyde1998 said:

At 46, you clearly become more likely to vote Conservative than Labour. Those who were 46 at the general election were born in 1973 - the year Britain entered the European Community. There's also a significant drop in the Labour vote share at age 40, those born in 1979 - the year Margret Thatcher came to power. The age of these changes may just be co-incidences, but I thought it was very interesting nonetheless.

It's an interesting distinction, it would be useful to see some graphs with the horizontal axis labelled by year of birth rather than age. It would investigate, whether it was old people, for example, "to blame" for Brexit, or people born before a certain year. Was the Lab/Con crossover 46 a decade ago? Two decades ago? If it is moving, how fast is it moving compared to the passage of time?

It's similar to when people say "you don't look like you're in your 70s", but the commentator is mentally comparing with their grandmother 50 years ago, somebody born in the 1890s, not somebody born in the 1940s.

[clyde1998]

Apr 25, 2020 9:17:45 GMT The Bishop said:

This clearly shows that the Tories depend on over-60s for their majority - below that its roughly speaking a draw.

The data implies Labour had a plurality of votes up to the age 65 and an outright majority of votes (50%+) up to the age of 39.

38% of the Conservative vote came from retirees (those who said they've retired), compared to 15% of the Labour vote. Mind, the Conservatives also had a plurality of votes from those who are in work - being Con 39%, Lab 36% across GB.

---

Among retired voters (26.8% of voters) - the Leave/Yes votes are percentages within this sample:

	England	Scotland	Wales
Conservative	65.2%	37.5%	48.1%
Labour	17.8%	13.5%	29.5%
Lib Dem	10.5%	12.1%	5.4%
SNP/PC	-	35.0%	8.1%
Green	2.0%	0.3%	0.1%
Brexit	2.7%	0.9%	7.7%
Other	1.8%	0.8%	1.1%
Leave	66.0%	45.1%	60.8%
Yes	-	35.9%	-

Retirees' vote by Leave/Remain vote:

	England	Scotland	Wales
Conservative	83.9%/30.5%	64.0%/19.6%	65.9%/21.6%
Labour	7.3%/37.4%	8.0%/17.3%	13.7%/54.4%
Lib Dem	2.1%/26.4%	5.5%/16.8%	1.2%/11.7%
SNP/PC	-	19.3%/45.2%	4.3%/12.0%
Green	1.1%/3.5%	0.2%/0.3%	0.0%/0.3%
Brexit	3.8%/0.1%	1.8%/0.3%	13.3%/0.0%
Other	1.7%/2.0%	1.1%/0.4%	1.7%/0.0%

In a partial answer to J.G.Harston 's point, it appears that the 2016 referendum vote is a major factor into why there was a very large age gap as retirees who voted remain only had around a one in three chance of voting Conservative in England (lower in Scotland and Wales). On the flip side, full-time students in England who voted leave had over a 50% chance of voting Conservative compared to 19% among all students (the samples for Scotland and Wales are too small for meaningful figures), albeit your chance of being a leave voter decreases as you get younger. I'll have a look at some of the older datasets to see what the historical voting patterns were in terms of age.

[clyde1998]

One of the questions within the survey asks respondents to rate each party on a scale of 0-10 as to how much they like them; this was asked about the Conservatives, Labour, the Lib Dems, the SNP (across GB), Plaid Cymru (in Wales), Brexit and the Greens. This can be taken as a rough guide as to the respondents' vote in a ranked preference situation.

In Scotland, each seat either has the SNP first or second - meaning that each seat has a pro-independence party and a pro-union party as one of the first two parties. Some suggestions of tactical voting to reduce the number of SNP seats have come up. Just how effective would that be?

For the following figures, I assume:

• Everyone who voted for one of the two parties involved would continue voting for that party.
• Where the party they voted for isn't involved, their vote is given to the relevant party with the highest 'like' score (where tied, half their vote is given to each)
• If a respondent gives both parties a score of 0, I assume they wouldn't vote.

 
Across Scotland, the SNP has a significant lead compared to each unionist party:

All seats	Opp	SNP	WNV
Conservative	30.8%	56.9%	12.3%
Labour	33.0%	50.0%	17.0%
Lib Dem	32.2%	53.4%	14.4%

However, it's fair to assume that each party would do better in areas where they are already the main challenger to the SNP. Reducing the contests between each party to where their contest takes place (seats where each party is the main opponent to the SNP), you get:

1st unionist	Opp	SNP	WNV
Conservative	36.0%	52.5%	11.4%
Labour	36.5%	50.0%	13.5%
Lib Dem	43.1%	46.6%	10.2%

Notably, the Lib Dems come closest to overtaking the SNP, while the Conservatives are furthest away.

---

Just as a side note, the mean score for each party was:

All Scottish seats

• SNP - 4.69
• Greens - 4.58
• Labour - 3.70
• Lib Dem - 3.28
• Conservatives - 2.58
• Brexit - 1.54

Conservative largest unionist party

• SNP - 4.42
• Greens - 4.42
• Labour - 3.39
• Lib Dem - 3.21
• Conservatives - 3.06
• Brexit - 1.81

Labour largest unionist party

• SNP - 5.05
• Greens - 4.83
• Labour - 4.10
• Lib Dem - 3.21
• Conservatives - 1.95
• Brexit - 1.20

Lib Dems largest unionist party

• Greens - 4.68
• SNP - 4.36
• Lib Dem - 3.87
• Labour - 3.47
• Conservatives - 3.04
• Brexit - 1.73

[jamie]

Interestingly short thread on values and 2019 voting. Data point that immediately stood out to me was the Tories doing better with the culturally conservative left than the liberal right. More generally, it’s notable that cultural views are more predictive of voting Conservative than Labour (Labour underperform among the centrist and particularly right wing cultural liberals who are more amenable to the Lib Dems).