What would good election results for Labour in May 2021 look like?

Luke Akehurst
© Twitter/@Keir_Starmer

Every year for the last decade I’ve written an article for LabourList setting out some benchmarks and historical comparisons for Labour for the May elections.

This year is a huge and complex set of elections because the May 2020 set were delayed and wrapped in with the May 2021 set. This makes it particularly hard to compare like with like, as we can’t disaggregate them and compare when they were last contested in the 2016 and 2017 cycles of council seats, which were quite different from each other in that 2016 was merely mediocre (Labour lost a net 18 councillors) and 2017 was truly catastrophic (Labour lost a net 382 councillors).

The pandemic may have an as-yet unquantifiable impact on turnout, and we won’t get the results overnight as social distant count arrangements mean that they will dribble out over the Friday, Saturday, Sunday, Monday. Don’t jump to conclusions about Labour’s performance based on early results on the Friday that may not be representative. One important vote, the Airdrie and Shotts parliamentary by-election, isn’t even happening until a week later on May 13th.

Where Labour will be fighting elections

The elections being fought on May 6th are as follows:

  • A parliamentary by-election in Hartlepool
  • Every seat in the Scottish parliament
  • Every seat in the Welsh Senedd
  • The London mayor and Assembly
  • 39 police and crime commissioners in England and Wales
  • 7 combined authority mayors
  • 5 single-authority mayors
  • The following four categories, which add up to 4,866 council seats up for election, about 25% of the approx. 20,000 councillors in the UK:
    • 21 county councils (every seat, so 1,632 councillors)
    • 35 metropolitan borough councils (one third of the seats in most of them but every seat in Doncaster, Rotherham and Salford, so 907 councillors)
    • 28 unitary councils (every seat in 13 councils, one third of the seats in 15 councils, so 1,241 councillors)
    • 60 district councils (every seat in 7 councils, half the seats in 6 councils, one third of the seats in 47 councils, so I make that 1,086 councillors)
  • Over 300 council by-elections in seats and councils that would not otherwise be up for election

The task for Labour in Hartlepool is straightforward – hold a marginal seat in what was described as the “Red Wall” in the last general election. Any kind of hold will be impressive in a constituency with a high level of support for Brexit and a history of anti-Labour protest voting.

In Scotland, Labour is defending 23 MSP seats won in 2016, only three of which were won at constituency level, the rest being regional list proportional top-up seats. Although Anas Sarwar has scored high personal ratings since taking over as leader, it will be difficult to translate this into even holding the same number of seats, as Alex Salmond’s new Alba party could eat into Labour’s regional list seats due to a quirk in the voting system (the SNP very rarely get top-up list seats as they perform disproportionately well in the constituencies, but a second nationalist party acting as a proxy and winning the second, regional votes from SNP supporters could qualify for some, at Labour’s expense).

As well as looking at whether Labour’s seat number goes up or down, we should look at whether its vote shares improve on the 22.6% (constituency) and 19.1% (regional list) that Kezia Dugdale got in 2016, and whether it can overtake the Tories and move into second place on either seats or vote share.

In Wales, Labour is defending 29 seats out of 60 in the Senedd and currently holds power with the support of the sole Lib Dem. Many of the current Labour-held constituencies have been lost to the Tories at Westminster level in the 2019 general election, and the few opinion polls taken this year show wildly differing potential results, so this is a highly unpredictable election but one where damage limitation may be the order of the day.

The mixture of constituency and regional top-up lists means that for all parties, losses of constituency seats can be offset by compensating list gains. In 2016, Labour got 34.7% of the vote in the constituencies and 31.5% in the top-up regional lost seats.

London should be fairly straightforward for Labour as Sadiq Khan has a huge poll lead and may even improve on the 44.2% first round vote he got in 2016. In the London Assembly, Labour is defending 12 of 25 seats under a mixed constituency and top-up list system, which is near the outer limits of what a proportional system will give the winning party. Havering and Redbridge is the most marginal target of the constituency seats, with a Tory majority of just 1,438, whilst Merton and Wandsworth is the trickiest defensive marginal for Labour, with a majority of 4,301.

The police and crime commissioner elections have never really caught the public’s imagination and have been afflicted by low turnout. In 2016, Labour won 15 out of then 40 PCCs. West Yorkshire’s PCC role is being abolished and merged with the new combined authority mayor, so Labour is actually defending 14 PCC seats now.

The most marginal defences for Labour are Cheshire and Derbyshire. The most marginal targets for Labour are Bedfordshire, Cambridgeshire, Devon and Cornwall, Norfolk, Northamptonshire, Staffordshire and Warwickshire, all with Tory majorities of under 10,000 under the supplementary vote system, where voters can express a first and second preference. It’s difficult to see which of these are realistic prospects for gains, as in some of them Labour got a very low share of first preference votes in 2016.

The combined authority mayors include high-profile Labour incumbents Andy Burnham and Steve Rotherham. The competitive fights are the West Midlands, where former Labour cabinet minister Liam Byrne is taking on incumbent Tory Andy Street, West of England where the Labour challenger is former MP Dan Norris, and Tees Valley, where seat losses in the general election suggest Labour’s Jessie Joe Jacobs has an uphill battle. There is also the newly created West Yorkshire mayor.

The single-authority mayors are mainly notable for two big cities: Bristol where Labour mayor Marvin Rees is one of the UK’s most senior Black politicians, and Liverpool, where anything could happen given the recent scandals and government investigation, but Joanne Anderson could become Liverpool’s first Black and first female mayor.

In the council elections, it is reasonable to expect a continuation of the pattern of the last decade of the distribution of Labour’s votes changing due to Brexit, boosting Labour in urban areas and university towns, and reducing its support in small towns and former mining areas in the Midlands and North. Unfortunately for Labour, the London boroughs where this pattern is helpful are not electing councillors this year (other than scattered by-elections), and only one third of seats in metropolitan areas are being contested, whereas every seat in the shire counties is being fought, so Labour will be fighting a primarily defensive battle on difficult electoral terrain, looking to minimise net losses of councillors and councils rather than make net gains.

How to assess Labour’s electoral success

There are at least four ways of measuring Labour’s national performance in council elections: national projected vote share (which the BBC calculates for the whole country including areas not voting this year), raw number of councillors, number of councillors gained or lost, and number of councils controlled.

Looking first at Labour’s national vote share, the estimated figures the BBC uses are as follows for the last decade:

  • 2010: 29% (general election result)
  • 2011: 37%
  • 2012: 39%
  • 2013: 29%
  • 2014: 31%
  • 2015: 30% (general election result)
  • 2016: 33%
  • 2017: 27%
  • 2018: 36%
  • 2019: 28%

Raw number of councillors is the national (Great Britain) total figure for how many Labour councillors there are, including all the thousands of councillors not up for election this year. This figure will have changed at the margins since 2019 due to by-elections and factors such as the abolition of Labour-controlled Corby council.

  • 2010: 4,831
  • 2011: 5,691
  • 2012: 6,559
  • 2013: 6,850
  • 2014: 7,098
  • 2015: 6,895
  • 2016: 6,859
  • 2017: 6,297
  • 2018: 6,468
  • 2019: 6,323

Earlier cycles than this tell us that over the lifetime of this parliament we need to build up to over 8,000 councillors if we are going to win a general election.

Next, we have the number of gains or losses. For comparison, here are the years since Margaret Thatcher came to power when Labour has made net gains (in the other 18 years not listed, we lost seats):

  • 1980: +601 Labour councillors
  • 1981: +988
  • 1983: +8
  • 1984: +88
  • 1986: +13
  • 1988: +76
  • 1989: +35
  • 1990: +284
  • 1991: +584
  • 1993: +111
  • 1994: +44
  • 1995: +1,204
  • 1996: +468
  • 2010: +372
  • 2011: +860
  • 2012: +847
  • 2013: +288
  • 2014: +256
  • 2018: +79

Lastly, control of councils. The number of councils Labour has controlled across England, Scotland and Wales has been as follows:

  • 2002: 136 (the last year in which we controlled more than the Tories)
  • 2003: 103
  • 2004: 94
  • 2005: 92
  • 2006: 75
  • 2007: 58
  • 2008: 46
  • 2009: 37
  • 2010: 54
  • 2011: 81
  • 2012: 114
  • 2013: 117
  • 2014: 120
  • 2015: 114
  • 2016: 114
  • 2017: 107
  • 2018: 105
  • 2019: 99

The location of councils and seats that change hands is important, too: we need to hold onto councils, or hold or gain seats, in areas similar to parliamentary seats we need to gain, to get an overall majority in a general election.

After the elections, I’ll write a follow-up to see how we did against these benchmarks.

More from LabourList


We provide our content free, but providing daily Labour news, comment and analysis costs money. Small monthly donations from readers like you keep us going. To those already donating: thank you.

If you can afford it, can you join our supporters giving £10 a month?

And if you’re not already reading the best daily round-up of Labour news, analysis and comment…