Local Council By-election Review for 2018

[middleenglander]

There have been 271 by-elections for 272 seats during 2018 with 76 seats (28%) changing hands. The results can be summarised by:

Party	Candidates	Defended	Retained	Gained	Lost	Won	retention rate
Conservative	269	132	98	21	34	119	74%
Labour	245	85	72	11	13	83	85%
Liberal Democrat	212	25	17	24	8	41	68%
SNP	10	3	2	2	1	4	67%
Plaid Cymru	2	1	1			1	100%
Green	127	1		2	1	2	0%
UKIP	58	8			8		0%
Independent ^	73	10	2	11	8	13	20%
Localist ^^	22	7	4	4	3	8	57%
No Description	5			1		1
Other	18
Total	1,041	272	196	76	76	272	72%

   ^    73 Independent candidates contested 63 by-elections
  ^^   Active for Plymouth (1), Ashfield Independent (1), Ashford Independent (1), Epsom & Ewell Residents (2), Farnham Residents (1),
         Farnworth & Kearsley Residents (1), Harlow Alliance (2), Independent - Save Our Stretton (1), Lincolnshire Independent (4), Loughton Residents (1),
         Nottingham Independent (1), Our West Lancashire (1), Stratford First (1), Thurrock Independent (1), Yorkshire (3)
^^^  Blue Revolution (1), BNP (2), Bus Pass Elvis (2), Christian Peoples (1), Democrats & Veterans (2), English Democrat (1), For Britain (1), Libertarian (2),
         National Flood Prevention (1), Something New (1), Women's equality (4)

At the May annual elections there were 41 "double" elections where a by-election was run along with the annual election.  The conventions used above were:
* where there were 2 candidates for a party then 1 was for the annual election and 1 for the by-election, where only 1 that was for the annual election and
* the highest placed elected candidate won the annual election and the second placed the by-election.

Conservatives had candidates for 99% of the 272 seats, Labour 90%, Liberal Democrats 78%, Green 47% and UKIP 21%.
- nearly 49% of the vacancies were Conservative seats, 31% Labour, 9% Liberal Democrats, 4% Independents, 3% UKIP, 3% Localists and 2% others.

Conservatives gained 21 seats and lost 34, a net loss of 13
- gained 4 seats from Labour, 6 from Liberal Democrats, 5 from UKIP, 5 from Independents and 1 from Localist whilst
- losing 7 to Labour, 18 to Liberal Democrats, 6 to Independents, 1 each to Green, Localists and No Description.
Labour gained 11 seats and lost 13, a net loss of 2
- gained 7 seats from Conservatives, 1 from Liberal Democrat, 1 Green and 2 UKIP whilst
- losing 4 to Conservative, 2 to Liberal Democrats, 4 to Independents, 2 to Localists and 1 SNP
Liberal Democrats gained 24 seats and lost 8, a net gain of 16
- gained 18 seats from Conservatives, 2 from Labour and 1 each from Independent, Localist, SNP and UKIP but
- lost 6 to Conservatives, 1 to Labour and 1 to SNP
SNP gained 1 seat from both Labour and Liberal Democrat but lost 1 to Liberal Democrats, a net gain of 1
Greens gained 1 from Conservatives and 1 from an Independent whilst losing 1 to Labour, a net gain of 1
UKIP lost 5 seats to Conservatives, 2 to Labour and 1 to Liberal Democrats, a loss of 8
Independents gained 11 seats and lost 8, a net gain of 3 
- gained 6 seats from Conservatives, 4 from Labour and 1 from a Localist whilst
- losing 5 to Conservatives and 1 each to Liberal Democrat, Localist and Green
Localists gained 4 seats and lost 3, a net gain of 1 
- gained 1 seat from Conservative, 2 from Labour, and 1 from Independent but
- lost 1 seat each to Conservative, Liberal Democrat and Independent
No Description gained 1 seat from Conservative

[greenchristian]

Dec 24, 2018 0:36:23 GMT middleenglander said:

There have been 271 by-elections for 272 seats during 2018 with 76 seats (28%) changing hands.

Which was the one seat that had two by-elections, then?

[middleenglander]

Dec 24, 2018 8:47:00 GMT greenchristian said:

Dec 24, 2018 0:36:23 GMT middleenglander said:

There have been 271 by-elections for 272 seats during 2018 with 76 seats (28%) changing hands.

Which was the one seat that had two by-elections, then?

Lancaster, University & Scotforth Rural on 17 May. One Labour defence and 1 Green, although former Councillor was sitting as Labour.

[yellowperil]

Dec 24, 2018 9:35:47 GMT middleenglander said:

Work In Progress

So how did 2018 by-elections compare with recent years?

Year	England	England	Scotland	Scotland	Wales	Wales	GB total	GB total
	seats contested	changed	seats contested	changed	seats contested	changed	seats contested	changed
2018	258	26%	9	56%	5	60%	272	28%
2017	337	29%	8	38%	3	33%	348	29%
2016	279	29%	18	72%	23	30%	320	32%
2015	184	22%	38	29%	29	24%	251	24%
2014	239	31%	17	41%	15	27%	271	31%
2013	310	29%	14	43%	13	8%	337	29%
2012	232	25%	3	100%	8	63%	243	27%
2011	162	28%	18	44%	12	75%	192	32%
2010	261	30%	4	25%	16	38%	281	31%

Presumably the apparent higher volatility in Scotland is down the fact that in effect a by-election there is fought under a different electoral system compared to the main elections?

[john07]

The results in Scotland can be totally random and have nothing to do with prevailing opinion.

It’s all down to which councillors resign/die and who was the leading party for the division in question.

Best not to try and judge anything from the results.

In fact it is arguable that there should not be AV by-elections in STV divisions.

[middleenglander]

Dec 24, 2018 10:49:36 GMT john07 said:

The results in Scotland can be totally random and have nothing to do with prevailing opinion.

It’s all down to which councillors resign/die and who was the leading party for the division in question.

Best not to try and judge anything from the results.

In fact it is arguable that there should not be AV by-elections in STV divisions.

Over the 9 years the number of Scottish seats changing hands averaged 44%, probably due to the electoral system, whilst in Wales the average was 35%, a fair number of single member, small electorate seats which tend to change hands more readily.

[middleenglander]

So how did 2018 by-elections compare with recent years?

Year	England	England	Scotland	Scotland	Wales	Wales	GB total	GB total
	seats contested	changed	seats contested	changed	seats contested	changed	seats contested	changed
2018	258	26%	9	56%	5	60%	272	28%
2017	337	29%	8	38%	3	33%	348	29%
2016	279	29%	18	72%	23	30%	320	32%
2015	184	22%	38	29%	29	24%	251	24%
2014	239	31%	17	41%	15	27%	271	31%
2013	310	29%	14	43%	13	8%	337	29%
2012	232	25%	3	100%	8	63%	243	27%
2011	162	28%	18	44%	12	75%	192	32%
2010	261	30%	4	25%	16	38%	281	31%

The number of by-elections vary with the electoral cycle as well be being affected by former Councillors resigning after becoming MPs.  The number changing hands however remains fairly constant at around 30%, a figure that goes back beyond 2010.

Comparing just England in 2018 with 2010 and 2014, the same stage of the electoral cycle, we have:

	2018	2018	2018	2014	2014	2014	2010	2010	2010
	seats   contested	changed    hands	changed	seats    contested	changed    hands	changed	seats   contested	changed    hands	changed
January to April	71	28	39%	46	15	33%	41	18	44%
May election - single elections	15	4	27%	30	6	20%	38	10	26%
May election - double elections	41	5	12%	40	11	28%	45	10	22%
Rest of May to August	63	19	30%	44	17	39%	48	18	38%
September & October	36	7	19%	43	15	35%	58	13	22%
November & December	32	5	16%	36	9	25%	31	10	32%
Total	258	68	26%	239	73	31%	261	79	30%

The number of seats contested is similar across the three years as is the proportion changing hands apart from possibly the last quarter of 2018.  Furthermore the profile of elections occurring across the year is not dissimilar, with nearly half the occurring before or during the annual elections in May.

Finally, how do the net gains / losses, including for Scotland and Wales, compare with previous years?

	Conservative	Labour	Liberal         Democrats	Green	UKIP	Independent	Localist	SNP /   Plaid Cymru	Other
2018	-13	-2	+16	+1	-8	+3	+1	+1	+1
2017	-13	-1	+20	+6	-13	-5	+5	0	+1
2016	-32	-7	+28	+1	-3	+3	+4	+4	+2
2015	+16	-3	0	-2	-10	-2	-4	+4	+1
2014	-10	0	+6	+3	+10	-8	0	+1	-2
2013	-30	+21	-4	-2	+12	+5	+3	-6	+1
2012	-15	+15	+2	0	-2	-1	+1	0	0
2011	-17	+27	-7	+1	0	-5	0	+2	-1
2010	-24	+30	+4	0	0	-12	+1	+1	0

Thus 2018 was not significantly different to 2017.  Although the Labour net losses were only small, the party has not had net gains since 2013.

[johnloony]

Dec 24, 2018 8:47:00 GMT greenchristian said:

Dec 24, 2018 0:36:23 GMT middleenglander said:

There have been 271 by-elections for 272 seats during 2018 with 76 seats (28%) changing hands.

Which was the one seat that had two by-elections, then?

If there have been 271 by-elections for 272 seats, then it is logical to ask "Which was the one by-election that had two seats, then?", but not the other way round.

[Deleted]

Dec 24, 2018 10:49:36 GMT john07 said:

In fact it is arguable that there should not be AV by-elections in STV divisions.

It is arguable that we shouldn’t be lumbered with this half baked electoral system at all. 

[john07]

Dec 25, 2018 14:50:48 GMT Sandy said:

Dec 24, 2018 10:49:36 GMT john07 said:

In fact it is arguable that there should not be AV by-elections in STV divisions.

It is arguable that we shouldn’t be lumbered with this half baked electoral system at all. 

Indeed it would remove a large number of Conservative Councillors throughout Scotland.

[Deleted]

Dec 25, 2018 20:07:43 GMT john07 said:

Dec 25, 2018 14:50:48 GMT Sandy said:

It is arguable that we shouldn’t be lumbered with this half baked electoral system at all. 

Indeed it would remove a large number of Conservative Councillors throughout Scotland.

Following your pitiful drubbing and receiving the most votes in exactly one council at the last locals i really think it would worry your lot more.

[andrew111]

Dec 25, 2018 14:50:48 GMT Sandy said:

Dec 24, 2018 10:49:36 GMT john07 said:

In fact it is arguable that there should not be AV by-elections in STV divisions.

It is arguable that we shouldn’t be lumbered with this half baked electoral system at all. 

Getting rid of a fair voting system that gives representation to the widest range of people is exactly what I would expect from Tories (and sadly also Labour, most of the time)

On the subject of AV for by-elections, it is the best system available when only one person is elected (and would be better in England too!). The problem in analysing the results lies in comparison with which councillor is being replaced. A more sensible comparison if you want to see how different Parties are doing is with the votes cast for each Party in the last normal election. That applies under any voting system.

[andrew111]

So, looking long term do we conclude that in the coalition years the Lib Dems took seats from Tories and lost them to Labour? And that this source of gains for Labour dried up in 2015? Since when the Lib Dems have continued gaining from the Tories but stopped losing to Labour and indeed gaining a little?

[finsobruce]

Dec 26, 2018 12:34:12 GMT andrew111 said:

Dec 25, 2018 14:50:48 GMT Sandy said:

It is arguable that we shouldn’t be lumbered with this half baked electoral system at all. 

Getting rid of a fair voting system that gives representation to the widest range of people is exactly what I would expect from Tories (and sadly also Labour, most of the time)

Apart from when we actually introduced it in Scotland, of course. If the Tories want to be daft enough to get rid of a system that saved their bacon north of the border that's up to them. But I somehow don't think they will. 

I'd be very pleased if it was introduced for local elections in England as well. 

[Deleted]

Dec 26, 2018 12:34:12 GMT andrew111 said:

Dec 25, 2018 14:50:48 GMT Sandy said:

It is arguable that we shouldn’t be lumbered with this half baked electoral system at all. 

Getting rid of a fair voting system that gives representation to the widest range of people is exactly what I would expect from Tories (and sadly also Labour, most of the time)

On the subject of AV for by-elections, it is the best system available when only one person is elected (and would be better in England too!). The problem in analysing the results lies in comparison with which councillor is being replaced. A more sensible comparison if you want to see how different Parties are doing is with the votes cast for each Party in the last normal election. That applies under any voting system.

I’m sure I have clashed with you over this before so I will keep it brief. There is no system fairer than one vote for all electors, and the candidate with the most votes is elected. Very few people understand STV which is why there are always many spoilt ballots. Under FPTP there is a clear connection between the electorate and their representatives. 

[mattb]

Dec 26, 2018 17:11:11 GMT Sandy said:

There is no system fairer than one vote for all electors, and the candidate with the most votes is elected. Very few people understand STV which is why there are always many spoilt ballots. Under FPTP there is a clear connection between the electorate and their representatives. 

erm … STV (Single Transferrable Vote) is by definition "one vote for all electors". Do you have any evidence that there are "always many spoilt ballots"? I don't think the facts support that?

[Deleted]

Dec 26, 2018 18:23:11 GMT mattb said:

Dec 26, 2018 17:11:11 GMT Sandy said:

There is no system fairer than one vote for all electors, and the candidate with the most votes is elected. Very few people understand STV which is why there are always many spoilt ballots. Under FPTP there is a clear connection between the electorate and their representatives. 

erm … STV (Single Transferrable Vote) is by definition "one vote for all electors". Do you have any evidence that there are "always many spoilt ballots"? I don't think the facts support that?

I think you know exactly what I mean and are splitting hairs. 

And yes, the Electoral Commissions report on the 2017 local elections support that. Indeed, there was an increase in the number of spoilt ballots from 2012. 

[Deleted]

Dec 26, 2018 17:11:11 GMT Sandy said:

Dec 26, 2018 12:34:12 GMT andrew111 said:

I’m sure I have clashed with you over this before so I will keep it brief. There is no system fairer than one vote for all electors, and the candidate with the most votes is elected. Very few people understand STV which is why there are always many spoilt ballots. Under FPTP there is a clear connection between the electorate and their representatives. 

I feel this is dishonest. I get that some people prefer FPTP on ideological grounds (although perhaps unsurprisingly, I disagree), but I think your points are rather poor.


There is no system fairer than one vote for all electors

Fairly intuitive, but problematic for several reasons:

1. STV can definitely be considered as a 1 person 1 vote system. That's how it works and I know you're well aware of that fact. The fact that multiple preferences can be given doesn't change that. The same can be said of List PR, or Alternative Vote.

2. Not all plurality voting systems used in this country apply that principle anyway. Lots of council elections in England & Wales see voters getting to cast multiple votes (bloc voting).

3. Whilst intuitive, I don't think it's an all-important rule. Equal number of votes for each voter makes sense - e.g. a cumulative voting system where voters can cast a fixed max number of votes between multiple parties, or that voters can vote once in both rounds of a two-round system.

and the candidate with the most votes is elected.

Why is this a fair thing if that candidate does not get more than half of the votes? How can it legitimately be said that this person is the best single person to represent the views of their electorate? Even with a majority of votes - is having a single representative for an electorate the fairest option? I don't think so. Plus List PR broadly applies this principle anyway, again it's not exclusive to FPTP.

Very few people understand STV which is why there are always many spoilt ballots.

I don't think this is necessarily a bad point but I feel it's misguided. Firstly, I think most voters understand it to a sufficient extent. Nonetheless the fact that a sizable proportion of the electorate don't shouldn't be ignored. However I think this problem is partly because how novel the system is in Scotland (this isn't anywhere near as much as an issue in NI, Ireland, Australia, Malta et al.) and partly because of the confusion with the number of different systems used there. I also understand that while the number of spoilt ballots is higher than usual, the main issue is voters casting votes with just a single preference. When voters do this, the system degenerates to 'Single Non-Transferable Vote', which often leads to popular candidates getting a huge surplus of votes which then can't be transferred to similar candidates - which isn't good for anyone. Australia's Group Voting Ticket averts this problem - voters can vote specifically for individuals if they wish, or they can vote for party lists - when voters cast just a single preference then, the system degenerates to List PR. Finally, list systems are very, very simple to understand.

Under FPTP there is a clear connection between the electorate and their representatives.

I genuinely don't see how this isn't the case with STV, AV, MMP etc.; I recognise that it is an issue with MEPs which represent very large regions (with no 'constituency MEPs' beneath them).


Ultimately, FPTP reduces choice to whoever the Blues have put up or whoever the Reds have put up, forcing voters to choose between two choices which they may not really like at all. How is that fair? For example, my father is a lifelong Conservative voter, who voted to remain and felt we shouldn't have had the referendum in the first place, and supports the Withdrawal Agreement and the current Premiership. Yet his MP is an ERG-er who voted to oust the PM. What choice does that leave him?

One of the most common complaints about List PR is that it reduces voters' choices in individuals (despite open list systems existing). FPTP is simply a closed list system with a list of length 1. One of the main issues with STV is voters casting a single preference. In FPTP, doing so is mandatory.

[andrew111]

Dec 26, 2018 17:11:11 GMT Sandy said:

Dec 26, 2018 12:34:12 GMT andrew111 said:

Getting rid of a fair voting system that gives representation to the widest range of people is exactly what I would expect from Tories (and sadly also Labour, most of the time)

On the subject of AV for by-elections, it is the best system available when only one person is elected (and would be better in England too!). The problem in analysing the results lies in comparison with which councillor is being replaced. A more sensible comparison if you want to see how different Parties are doing is with the votes cast for each Party in the last normal election. That applies under any voting system.

I’m sure I have clashed with you over this before so I will keep it brief. There is no system fairer than one vote for all electors, and the candidate with the most votes is elected. Very few people understand STV which is why there are always many spoilt ballots. Under FPTP there is a clear connection between the electorate and their representatives. 

Others have replied in depth, so I will only make one comment: like many people of all parties my vote has never helped elect an MP, in over 40 years of trying. I even tried voting Labour in a marginal seat in 2015 and it still didn't work.

Therefore I have never felt any close connection with my MP. Under multimember STV for most of my life I would have had an MP closely aligned with my views. Right now that might not be true, but at least I would be able to prioritise Labour and Tory candidates who favour fair voting systems (and believe me I would write to the candidates and find out their views on this and some other issues I feel strongly about)

[lbarnes]

Dec 26, 2018 18:43:05 GMT Sandy said:

Dec 26, 2018 18:23:11 GMT mattb said:

erm … STV (Single Transferrable Vote) is by definition "one vote for all electors". Do you have any evidence that there are "always many spoilt ballots"? I don't think the facts support that?

I think you know exactly what I mean and are splitting hairs. 

And yes, the Electoral Commissions report on the 2017 local elections support that. Indeed, there was an increase in the number of spoilt ballots from 2012. 

Using the number of spoilt votes to determine how democratic a system is seems very niche. They constitute a very small proportion of all votes and there are many reasons for them. Not understanding the process is only one and is very difficult to evaluate.

Undoubtedly a big reason for there being more spoilt votes in Scotland in the local elections of 2017 was that there was simply a higher turnout there anyway.

The English elections of that year took place in areas where there is a higher than average turnout on the whole and the figure was around 35%. In Scotland it was 46%. That's about one third more and is a far better indicator of the demicracy of a system and how much faith the local electorate place in it.