Democracy Denied: The 2019 Local and European Election Audit

Author:
Ian Simpson, Research Officer

Posted on the 22nd August 2019

Foreword

By Dr Jess Garland

Jessica Garland

If two party politics isn’t dead, it has certainly taken a big hit over the first half of 2019. The English local elections and European Parliament election, both held in May, saw big gains for those other than the ‘traditional’ two parties – Conservative and Labour.

Voters appear keen to shop around at the ballot box. However, breaking out of the constraints of the winner-takes-all First Past The Post (FPTP) voting system – one that is infamous for distorting results – can be difficult.

Our analysis of the English local elections, held under FPTP, reveals some startling injustices, which affect all parties negatively in one area or another. However, the real losers are voters who do not see their choices properly reflected in their local council chambers.

For example, in nearly half of all English local councils a single party was able to secure more than half of the councillors up for election, while winning fewer than half of votes cast across the local authority area. The most extreme example was Havant Borough Council, where the Conservatives won every single councillor up for election with only a 44% share of the votes cast, leaving the choices of a majority of voters unrepresented.

Other areas saw parties winning the most council seats, of those up for election, when they had not won the most votes in the area. One of the most striking examples was Basildon Borough Council, where Labour won less than a quarter of the vote – almost half the Conservatives’ vote tally – yet elected more councillors.

Even worse for democracy are the scourge of uncontested and under-contested seats where a party is guaranteed a seat or seats, due to a lack of candidates being put forward in their ward. We uncovered hundreds of uncontested and under-contested seats and wards in councils across England affecting over 800,000 potential voters. Democracy wasn’t just lacking in these wards – it was effectively cancelled.

The somewhat unexpected European Parliament election was therefore a breath of fresh air for many voters, in some respects. While the D’Hondt Closed Party List voting system isn’t perfect, it is a proportional system, which gives voters a better chance of their vote counting than FPTP does.

It is the Scottish local elections, however, that provide a real contrast with the English local election results. The proportional, multi-member Single Transferable Vote (STV) system, used in Scottish local elections since 2007, brings a host of democratic benefits, from greater competition and choice to much fairer and more proportional outcomes. We bring these to light by analysing the results of the most recent Scottish local elections, held in 2017.

As talk grows of a General Election, these discussions are far from academic. Voting systems have a profound effect on both representation and outcomes – and indeed how people express themselves. Polling for this report found that more than one in five people plan to vote ‘tactically’ in the next general election, not voting for who they want, but voting against someone else. This is not how democracy should be working. The problems caused by FPTP at local level in England apply across the UK at general elections.

This is the first major analysis of this year’s local elections – and we’re grateful to Democracy Club (English local election results provided by Democracy Club and checked by ERS staff against results published on local authority websites https://democracyclub.org.uk) for providing the full set of results and also to the many ERS supporters who donated to fund this report. All three sets of elections analysed here offer some fascinating insights into the state of politics in Britain – and how voting systems affect representation.

Introduction

By Ian Simpson

A state of turbulence defines British politics in the summer of 2019, the Government’s working majority (including the 10 Democratic Unionist MPs) cut to one, following the result of the Brecon and Radnorshire by-election on 1 August. Political discussion, meanwhile, is dominated by a potential autumnal clash between government and parliament over whether a ‘no-deal’ Brexit is acceptable.

While politics itself has felt unpredictable, the British electorate appears to have become more fragmented and volatile than ever before. A number of developments over spring and summer 2019 have provided evidence of this heightened volatility.

The English local elections, on 2 May, saw big net gains in councillors for the Liberal Democrats and Green Party, as well as a rise in the number of independents elected. Both the Conservatives and Labour suffered net losses of councillors. Less than a month later the newly formed Brexit Party topped the poll at the European Parliament election, pushing Labour into third place and the Conservatives into fifth. This was the first time since the 1918 general election that neither the Conservatives nor Labour finished in the top two in a UK-wide election. In addition, the parliamentary by-election in Peterborough, on 6 June, saw Labour’s candidate receive the lowest winning vote share in a parliamentary by-election since 1945 (Apart from a parliamentary by-election for the now defunct English Universities constituency) just 30.9% of the vote.

A string of general election voter intention poll records were also broken throughout May and June 2019. This period saw the combined Conservative and Labour vote share fall below 50% for the first time in polling history. Although opinion polls during the first half of August showed something of an increase in support for the two main parties, their combined vote share remained around 10–15 percentage points below the lowest seen at any general election for over a century.

This volatility and voter fragmentation is an important development in itself but is perhaps all the more remarkable given that just two years ago, the 2017 UK general election was viewed by some as a reversion to ‘two-party politics’, with the Conservatives and Labour garnering a combined 84.4% of votes cast in Great Britain. This was the highest Conservative and Labour combined vote share in a general election for nearly half a century, with the 1970 general election being the last time a higher total than this (90.0%) was recorded. However, it is important to note that the headline result of the 2017 general election masked the fact that around one in five people were estimated to have voted ‘tactically’, i.e. not for their first-choice party but for a party they perceived to be in a better position to beat a party that they disliked. The vagaries and iniquities of the FPTP electoral system helped paint an unrealistic picture of the extent to which ‘two-party politics‘ had been restored.

Prior to the 2017 general election, British electoral politics had seen a steady decline in the concentration of votes for the two main parties. It was highest in the immediate post-World War II period, reaching a high point of a combined 97.2% vote share at the 1951 general election. By the second decade of the 21st century, those voting this way had fallen from nearly all voters, to around two thirds, 66.6% voting this way at the 2010 general election and 68.9% at the 2015 general election. Processes of class and partisan dealignment are viewed as key drivers of this trend (Webb, Paul (2000). The Modern British Party System. London. SAGE Publications.), which the 2017 general election very much bucked.

It now appears at least possible that the long-term trend of a fracturing electorate is being re-asserted. As Sir John Curtice, Professor of Politics at the University of Strathclyde puts it: ‘There is little doubt that Britain’s traditional two-party system is facing its biggest challenge yet in the wake of the Brexit impasse’.

Our findings also highlight the unsuitability of FPTP for Westminster elections. Many people do not feel able to opt for their first choice under Westminster’s one-person-takes-all voting system. BMG polling for the ERS reveals that – similarly to 2017 – in a snap general election over a fifth (22%) of voters say they plan to opt for a party or candidate to keep out someone they dislike more.

Given the volatility and renewed fragmentation of the electorate and the possibility of a general election being fought in the near future, it is a good moment to explore how the use of different voting systems translates votes into representation.

Two large-scale sets of elections have taken place in Britain this year: English local elections, held in nearly 250 local authorities; and the European Parliament (EP) election, held throughout the whole of the UK. The use of both a majoritarian voting system (First Past The Post for the English local elections) and a proportional voting system (Closed List PR for the EP election) provides a useful point of comparison for this report. In addition, we also explore the most recent set of Scottish local elections, held in May 2017. These elections allow us to contrast the outcomes of the voting system used for Scottish local elections, Single Transferable Vote (STV) (the ERS’s preferred voting system), with both FPTP and another type of proportional system (Closed List PR).

For all three of these elections there will be a description of the voting system used, a synopsis of the overall results and an exploration of how each election fared in relation to a number of democratic outcomes.

English Local Elections 2019

Background

Thursday 2 May 2019 saw the biggest set of English local elections for four years. Elections took place in 248 local authorities in England (168 non-metropolitan districts; 47 unitary authorities; 33 metropolitan boroughs).

In 101 local authorities, one-third of councillors were up for election. In these authorities, one successful candidate was elected per ward, via the First Past The Post (FPTP) system. Each voter had one vote and the candidate with the highest number of votes in each ward was elected, irrespective of whether their vote share was 30% or 70%. All other candidates in the ward were unsuccessful, whether they came second or seventh, meaning the ballots of voters who supported these candidates were ignored.

In the other 147 local authorities where elections took place, all councillors were up for election. In these authorities, either one, two or three councillors were elected per ward. Councillors were again elected using the FPTP system with voters having one, two or three votes, depending on the number of councillors to be elected in the ward. Where voters had two or three votes, their votes were non-preferential, i.e. voters were instructed to mark their ballot paper with either two or three ‘X’s, rather than use numbers to indicate their preferred candidates. Where two or three candidates were elected, the two or three candidates with the highest number of votes in the ward were elected, with all candidates below the cut-off line being unsuccessful.

Results

Overall, the elections saw big gains for the Liberal Democrats, Green Party and independents, while the Conservatives suffered big losses and Labour saw a small reduction in their number of councillors. Ukip had a smattering of councillors elected, while neither the Brexit Party nor Change UK stood candidates in these elections.

 

Table 1: Councillors elected and council control by party, English local elections 2019

Party Councillors elected Change Councils controlled Change
Conservative 3564 -1330 93 -44
Labour 2021 -84 60 -6
Liberal Democrat 1352 +706 18 +10
Independent 1044 +604 2 +2
Green 265 +194 0
Residents’ Association 119 +49 2 +1
Ukip 31 -145 0
Other 14 +6 0
No Overall Control 73 +37

Some commentators saw the results through the prism of the ongoing Brexit debate, arguing that the Conservatives and Labour were punished over a failure to break the Brexit deadlock. However, with these being local elections, it is important not to underestimate the myriad local issues that will have played a role in many of the contests up and down the country.

Whatever the reasons, the 2019 local elections resulted in substantial changes to the political landscape of English local government. The Conservatives control fewer councils than they have done at any point for the last fifteen years. The Liberal Democrats more than doubled the number of councils they control, to the extent that they run almost as many councils as they did before the collapse in their support, precipitated by the formation of the national Conservative–Liberal Democrat coalition government in 2010. The Green Party elected more councillors than at any time for over twenty years, while over 1,000 independent candidates were elected, more than double the number that were elected four years earlier.

These elections were the most volatile set of English local elections, in terms of the number of council seats changing hands, for nearly a quarter of a century. The last time more seats changed hands was in 1995, when the Conservatives suffered a loss of over 2,000 council seats, mostly at the hands of a Labour Party riding high under its recently elected new leader, Tony Blair. In 2019, the fragmented nature of the political landscape is reflected in the fact that while the Conservatives again suffered a big net loss of seats, Labour also suffered a net loss, with the gains being shared among other parties, independents and local groups.

Uncontested and under-contested seats

A key feature of representative democracies is that free and fair elections take place, where citizens are able to choose who represents them at a particular level of government. However for some people in England, the right to participate in deciding who should represent them at local authority level does not exist, or is limited. This is due to uncontested or under-contested seats which have been in evidence in English local government for decades and continued to blight people’s democratic options in this year’s local elections.

Uncontested seats occur where the number of candidates standing for election in a ward matches the number of councillors to be elected in that ward. In other words, there are only enough candidates for the seats available. These candidates become councillors without any votes having been cast, meaning that there is no democratic decision for the electors of such wards to make. Our analysis finds that in this year’s local elections 150 councillors were ‘elected’ in this way, across 47 local authorities, representing nearly one-in-five councils where elections took place. Over 250,000 potential voters were denied any say in the selection of their local councillor(s) this year (Where there was evidence that there have been boundary changes in a council since 2015, we have used figures from the most recent Local Government Boundary Commission for England review for that local authority).

In addition to wards where no voting took place, there were a significant number of ‘under-contested’ wards. Under-contested wards are multi-member wards (where more than one councillor is elected) where a party was guaranteed to win at least one seat because there were not enough candidates to contest every seat in that ward. For example, in Melton Dorian ward on Melton Borough Council in the East Midlands, three councillors were to be elected and there were only four candidates (three Conservatives and one Green). This meant that the Conservatives were guaranteed to elect at least two councillors in this ward.

As well as the 150 councillors who won without a single vote being cast in uncontested seats, there were an additional 154 council seats guaranteed for one party or another before a vote had been cast, in under-contested wards that spanned 54 local authorities. Around another 580,000 potential voters had their choice of local councillor limited in this way.

Table 2: Local authorities with high instance of Uncontested Seats and Guaranteed Seats, English local elections 2019

Authority Region Previous Current Uncontested/Guaranteed seats Seats available Uncontested/Guaranteed seats (%)
Fenland East of England Con Con 15 30 50.0
Rutland East Midlands Con Con 12 27 44.4
Melton East Midlands Con Con 11 28 39.3
South Holland East Midlands Con Con 13 36 36.1
North Kesteven East Midlands Con NOC 10 43 23.3
Oadby & Wigston East Midlands LD LD 6 26 23.1
Lichfield West Midlands Con Con 10 47 21.3
Wychavon West Midlands Con Con 9 44 20.5
Great Yarmouth East of England Con Con 8 39 20.5
Sevenoaks South East Con Con 11 54 20.4
West Suffolk East of England New New 13 64 20.3
East Lindsey East Midlands Con Con 9 55 16.4
South Staffordshire West Midlands Con Con 8 49 16.3
Copeland North West Lab Lab 5 33 15.2
Breckland East of England Con Con 7 49 14.3
Selby Yorkshire & Humber Con Con 4 28 14.3
Bolsover East Midlands Lab NOC 5 37 13.5
Ashford South East Con Con 6 46 13.0
West Devon South West Con Con 4 31 12.9
West Lindsey East Midlands Con Con 4 36 11.1
Epping Forest East of England Con Con 2 18 11.1
Staffordshire Moorlands West Midlands Con NOC 6 56 10.7
Hambleton Yorkshire & Humber Con Con 3 28 10.7
Malvern Hills West Midlands Con NOC 4 38 10.5
Allerdale North West NOC NOC 5 49 10.2
Mid Sussex South East Con Con 5 50 10.0
Ryedale Yorkshire & Humber NOC NOC 3 30 10.0

Even in wards where there were no uncontested or guaranteed seats, there were not always a wide range of parties for voters to choose from. There were 817 wards, in 141 local authorities (close to half of councils where elections took place), where voters had a choice between only two parties, or between a candidate representing one party and an independent candidate.

With voters showing a renewed desire to support a wide range of parties (see Introduction), it is likely that many of the approximately 2.65 million voters who live in wards where only a binary choice was on offer, were left frustrated by their options.

These examples of voters being denied a say in deciding their local representatives, or facing a limited choice of options, are linked to the voting system. Under FPTP the barriers to success can be very high. There are many areas of England where representatives of one party or another may feel that it is almost impossible for them to gain representation under the FPTP system, even in areas where they might have fairly significant levels of support. It is therefore understandable, if disappointing, that parties decide not to stand candidates in all areas, instead focusing limited resources on the places where they have the best chances of being able to jump the FPTP hurdle and achieve representation on local councils.

On the same day as the English local elections, elections also took place for Northern Irish local councils. In contrast to the English local elections, no voters in Northern Ireland were denied a say due to living in an uncontested electoral area. Northern Ireland has long used the STV voting system for local elections and the section of this report that focuses on the 2017 Scottish local elections, which also used STV, will explore why using this system reduces the chances of uncontested seats occurring. This report focuses on the Scottish local elections, rather than the Northern Irish local elections because the party systems in England and Scotland are much more comparable (the Great Britain-based parties either do not participate in elections in Northern Ireland or are very minor players).

Proportionality of results

In the following section we analyse the political outcomes of the 2019 local elections in more detail. The 150 councillors ‘elected’ in uncontested seats, without voting taking place, are excluded from the analysis throughout the rest of this chapter.

Analysis of election results in each English local authority reveals that in nearly half of councils where elections took place, one party was able to win a majority of seats up for election with fewer than half the votes cast across the council area. This outcome occurred in 115 local authorities, representing 46% of councils where elections took place this year. In the most extreme case, the Conservatives won every single seat up for election on Havant Borough Council while winning only 44% of votes across the borough.

It is surely something of a democratic anomaly that in so many places a single party was able to win over half the available council seats (and in the vast majority of cases go on to run the local authority), when over half the votes in that area went to candidates not representing that party.

This is an anomaly that hurts all parties and their supporters in one area or another. The Conservatives, Labour and the Liberal Democrats all lose out or benefit in different places (see Table 3) – risking complacency through lack of scrutiny in some areas, and artificially eradicating their pockets of support in others.

In each of the areas it is voters that lose out, with councils often being run on a one-party basis, when this simply does not reflect the level of support for each party at a local level. In many cases, even where a party is able to achieve over 50% support at the ballot box, they are still vastly over-represented in terms of seats. For example, Labour won every single seat up for election on Sandwell Metropolitan Borough Council, in the West Midlands. While the party was able to win a healthy 59% of votes across the council area, this still means that over two-fifths of votes went unrepresented in the council chamber. As a 2015 study for the ERS found, these types of results are not conducive to healthy local government scrutiny.

Table 3: Top 10 over-represented parties, English local elections 2019

Authority Party over-represented Council control Votes for party (%) Seats for party (%) Gap (%)
Havant Con Con 43.9 100.0 56.1
Redditch Con Con 40.5 90.0 49.5
City of Lincoln Lab Lab 44.5 90.9 46.4
Tameside Lab Lab 46.2 89.5 43.2
Sandwell Lab Lab 58.5 100.0 41.5
Eastleigh LD LD 52.1 92.3 40.2
Wigan Lab Lab 41.4 80.0 38.6
Manchester Lab Lab 58.5 97.0 38.4
Tamworth Con Con 42.4 80.0 37.6
Salford Lab Lab 41.7 78.9 37.3

It is again worth noting the contrast with this year’s local elections in Northern Ireland, where no party got an unfair majority of seats on a minority of votes. The part that the ranked-choice system of STV plays in ensuring fairer electoral outcomes at local elections will be explored in the chapter on the 2017 Scottish local elections.

Another way of measuring the proportionality of electoral outcomes is via the Loosemore-Hanby Deviation from Proportionality (DV) Index. The DV Index is calculated by adding up the difference between each party’s vote share and their seat share in each electoral area and dividing by two, giving a ‘total deviation’ score. The higher the score, the more disproportionate the result.

In the case of local elections, the key DV scores to be calculated are those for each individual local authority, as these represent 248 separate elections. These scores reflect the cumulative difference between vote share and seat share for all parties across each local authority (Independents and local groups such as Residents’ Associations were grouped together for the purposes of creating DV scores. Other political parties, who received fewer than 5% of votes across the country, were also grouped together)The mean DV score across the 248 local councils that held elections was 19.6. This score is very much in line with DV scores for recent general elections, also held under FPTP, with the mean DV score across the last six general elections being 20.3At the very disproportionate end of the scale, there were a number of councils with extremely high DV scores (see Table 4), with five councils having scores above 40.0 and one council (Havant) having a score above 50.0. These scores will be placed into greater context when they are compared with DV scores for the EU Parliament election and the 2017 Scottish local elections, in later chapters of this report.

Table 4: Top 30 most disproportionate results, English local elections 2019

Authority Region Previous Current DV Score
Havant South East Con Con 56.1
Redditch West Midlands Con Con 49.5
City of Lincoln East Midlands Lab Lab 46.4
Tameside North West Lab Lab 43.2
Eastleigh South East LD LD 42.0
Sandwell West Midlands Lab Lab 41.5
Wigan North West Lab Lab 38.6
Manchester North West Lab Lab 38.4
Tamworth West Midlands Con Con 37.6
Salford North West Lab Lab 37.3
South Tyneside North East Lab Lab 36.1
North Tyneside North East Lab Lab 36.0
Oldham North West Lab Lab 35.7
Chorley North West Lab Lab 34.1
Bracknell Forest South East Con Con 33.7
Leicester East Midlands Lab Lab 33.6
Broxbourne East of England Con Con 33.4
East Hertfordshire East of England Con Con 33.4
Gateshead North East Lab Lab 32.8
East Suffolk East of England Con Con 32.7
Portsmouth South East NOC NOC 32.5
Coventry West Midlands Lab Lab 32.1
Epping Forest East of England Con Con 31.8
Newcastle-upon-Tyne North East Lab Lab 30.9
Stevenage East of England Lab Lab 30.9
Exeter South West Lab Lab 30.7
Rochdale North West Lab Lab 30.6
Central Bedfordshire East of England Con Con 30.0
Wealden South East Con Con 29.8
South Lakeland North West LD LD 29.8

N.B. DV score calculations exclude the 150 councillors ‘elected’ in uncontested seats, where voting did not take place. Also, DV scores are based on elections that took place in May 2019, not the overall political make-up of councils. In some local authorities, all councillors were up for election but in others only one-third were.

Another way of highlighting the disproportionality of the results are the number of votes, on average, each party requires to get a councillor elected. This data also shows how all parties suffer under FPTP, in one region or another. The details for each region of England where elections took place and for England as a whole are provided in the Appendix to this report.

A glance at some of the results highlights some major discrepancies. For example, in the North East it took 2,124 Conservative votes to elect a Conservative councillor and only 1,327 Labour votes. In the South East it took 2,357 Labour votes to elect a councillor but only 1,325 Conservative votes. While in the South West the Liberal Democrats benefited, needing only 1,401 votes to gain a councillor, whereas Labour needed 3,132 votes. Across the country as a whole, the Green Party fared worse than the other parties. They needed over 4,000 votes to elect a single councillor, whereas the Conservatives, Labour and Liberal Democrats required under 2,000.

Wrong winners

Another democratic anomaly in evidence at the English local elections were the instances of ‘wrong winners’. These are instances where the party that got the most votes in a local authority did not get the most councillors elected.

This is a localised example of something that has happened on a national scale, in UK general elections, under FPTP. In 1951, Labour won the most votes across the United Kingdom (48.8%) but the Conservatives won a majority of seats. In February 1974, the Conservatives won most votes (37.8%) but Labour won the most seats.

In the 2019 English local elections, there were 17 local authorities where the party getting the most votes did not get the most councillors up for election (see Table 5). Again, it was not one particular party that benefited, with multiple parties benefitting and losing out across the country.

Table 5: Councils where parties winning most votes did not win most councillors, English local elections 2019

Authority Party with most votes Votes (%) Seats (%) Party with most seats Votes (%) Seats (%)
Arun Con 42.4 38.9 LD 30.2 40.7
Basildon Con 43.1 35.7 Lab 24.4 42.9
Bedford Con 31.3 27.5 LD 30.9 37.5
Blackpool Con 44.4 35.7 Lab 43.9 54.8
Brighton & Hove Green 36.4 35.2 Lab 34.6 37.0
Chelmsford Con 40.6 36.8 LD 39.9 54.4
Chester West & Chester Con 39.2 40.0 Lab 36.7 50.0
Cotswold Con 42.5 41.2 LD 41.8 52.9
Gravesham Con 41.1 40.9 Lab 40.3 54.5
Kingston-upon-Hull LD 43.0 47.4 Lab 40.1 52.6
Lewes Green 27.9 22.0 Con 27.1 46.3
Portsmouth LD 27.8 40.0 Con 26.3 46.7
South Oxfordshire Con 34.9 27.8 LD 25.0 33.3
South Ribble Con 45.0 43.8 Lab 34.7 45.8
Southend-on-Sea Con 27.0 23.5 Lab 21.5 29.4
Stockton-on-Tees Con 33.2 25.0 Lab 32.6 42.9
Stoke-on-Trent Con 34.9 34.1 Lab 32.5 36.4

 

European Parliament Election 2019 (Great Britain)

Background

The 2019 European Parliament (EP) election was not due to take place in the UK. The original date for the UK’s exit from the European Union (EU) was 29 March 2019, two years after the UK triggered the Article 50 process of leaving the EU. After extending the leaving date, the UK was required to hold elections yet even following the extension of the exit deadline, it was not until 7 May 2019 that the UK Government officially confirmed that elections to the EP would take place in the UK. The election took place just over two weeks later, on 23 May, alongside elections in all other EU countries during the same week.

Prior to 1999, the UK (apart from Northern Ireland) used FPTP for EP elections. However, from 1999 onwards it was specified that all EU countries should use a proportional representation (PR) voting system, with the specific system to be decided by individual member countries. Despite Northern Ireland having used STV for all previous EP elections, the decision was made that the rest of the UK would use a different proportional voting system, the Closed Party List system. Under this system, Britain is split up into 11 relatively big regions (South West; South East; London; Eastern; Wales; East Midlands; West Midlands; North West; Yorkshire & Humber; North East; Scotland). These areas each elect a certain number of MEPs, ranging from three in the North East to 10 in the South East, depending on the size of the population in the region.

The specific form of Closed Party List PR used is called the D’Hondt voting system. Each voter has one vote, with which to mark an ‘X’ next to one party or one independent candidate. MEPs are then allocated within each region via the following formula: in the first round of counting, the party with the most votes wins a seat for the candidate at the top of its list. In the second round, the winning party’s vote is divided by two and whichever party comes out on top in the re-ordered result wins a seat for their top candidate. The process repeats itself, with the original vote of the winning party in each round being divided by one, plus their running total of MEPs, until all the seats for the region have been taken.

While being a proportional representation system, the Closed Party List system has some notable differences from STV. Firstly, a voter only has one vote, rather than being able to rank their preferences, as they can under STV. Secondly, under the Closed Party List system, a voter cannot indicate any preference between candidates from the same party, as the order in which they are elected has been decided before the election by the parties themselves, which is why the system is called a ‘closed list’. Under STV, voters are usually able to indicate preferences between candidates from the same party as well as between candidates from different parties (something which is elaborated on in the section on the Scottish local elections). Both of these differences mean voters have less power under the Closed Party List system than they do under STV.

A third consequence of the Closed Party List system is that it can cause uncertainty among voters and make them question the most efficient way to use their vote. In the build-up to the European Parliament election there was much discussion, particularly among pro-EU inclined voters, about how to use their one vote in a way that would maximise the number of ‘Remain’-inclined MEPs in a particular region. Such tactical voting can be difficult to work out under this system. This debate about tactical voting would have been far less relevant under STV, where voters have the reassurance of knowing that they can influence the election outcome with an unlimited number of preferential votes – not having to place all of their eggs (or ‘X’s) in a single-vote basket.   

Results

The outcome of the elections saw two parties with clear positions on the question of Brexit come first and second in terms of votes and MEPs. The Brexit Party, with a policy of leaving the EU via a ‘clean-break Brexit’ gained nearly a third of votes and just over two-fifths of MEPs, while the Liberal Democrats, who advocated a policy of ‘stopping Brexit’ came second, with just over one-fifth of votes and MEPs. Labour and the Conservatives, who had adopted more nuanced approaches to Brexit and could perhaps be seen as divided on the issue, were third and fifth respectively, in terms of both votes and the number of MEPs.

Table 6: European Parliament 2019 election results (Great Britain)

Party Votes Votes (%) Vote change since 2014 (%) Seats Seat share (%) Seat change since 2014
Brexit Party 5,248,533 31.6 +31.6 29 41.4 +29
Lib Dem 3,367,284 20.3 +13.4 16 22.9 +15
Labour 2,347,255 14.1 -11.3 10 14.3 -10
Green 2,023,380 12.1 +4.2 7 10.0 +4
Conservative 1,512,147 9.1 -14.8 4 5.7 -15
SNP 594,553 3.6 +1.1 3 4.3 +1
Plaid Cymru 163,928 1.0 +0.3 1 1.4
Change UK 571,846 3.4 +3.4 0 0.0
Ukip 554,463 3.3 -24.2 0 0.0 -24
Others 264,780 1.5 -3.7 0 0.0
Total 16,648,169 100.0 70

Uncontested seats

Perhaps unsurprisingly, there were no uncontested seats in the EP election in Britain. Voters in each region were presented with a wide range of parties and independent candidates competing for their votes, with a mean average of 10 parties or independent candidates standing per region. The proportional nature of the voting system clearly encouraged parties to think they had a reasonable chance of achieving representation, even where it was very unlikely they would top the poll in a particular region.

Proportionality of results

As MEPs are elected to represent the nation at the European Parliament, the key DV score is the one for Britain as a whole, which was 13.7, lower than the mean English local authority DV score of 19.6 recorded for this year’s English local elections. This reflects the fact that the Closed List PR system produced a more proportional result than FPTP did at the English local authority level. However, it is worth noting that the overall EP DV score masked some high DV scores in particular regions.

A downside of the system used for EP elections is that some regions elect only a small number of MEPs and these regions tend to have much higher DV scores, reflecting less proportional results. For example, the North East has the lowest number of MEPs (three) and the highest (most unproportional) DV score (41.9). The South East has the highest number of MEPs (10) and the lowest (most proportional) DV score (10.9). The differences in outcome are highlighted by the fact that Labour were able to gain an MEP in the South East with 7% of the votes in the region, while the Liberal Democrats were unable to gain an MEP in the North East, despite getting 17% of the votes there.

Table 7: European Parliament election 2019 (regional DV scores)

Region MEPs DV Score
North East 3 41.9
Wales 4 32.7
East Midlands 5 30.6
Scotland 6 21.8
South West 6 23.5
Yorkshire & Humber  6 18.7
East of England 7 16.6
West Midlands 7 13.1
London 8 18.6
North West 8 17.2
South East 10 10.9
Total (Great Britain) 70 13.7

Scottish Local Elections 2017

Background

The last set of Scottish local elections were held on 4 May 2017. This was the third time that the STV system was used for local elections in Scotland and the first time that 16- and 17-year olds were able to vote in a local election. Thirty two councils and 1,227 council seats were up for election.

The first Scottish local election held using STV, in 2007, saw massive changes in how elections and local democracy worked in Scotland. Voter choice more than doubled, uncontested seats were eliminated, and the councils where single parties had ruled with virtually no opposition for decades, became a thing of the past. The 2017 local elections largely saw a consolidation of these democratic improvements. Voter choice remained high, increasing on 2012 levels and no council in Scotland saw a single political party take majority control without the support of the majority of voters.

What is Single Transferable Vote (STV)?

The Single Transferable Vote (STV) is a form of proportional representation which uses preferential voting in multi-member constituencies.

Voters are invited to place the candidates on the ballot paper in order of preference by placing a ‘1’ against the name of the candidate they prefer most, a ‘2’ against the candidate who is their second preference, etc. Voters are free to choose how many candidates they rank.

If a voter’s preferred candidate has no chance of being elected or has enough votes already, that person’s vote is transferred to another candidate in accordance with their instructions, minimising the number of wasted votes.

STV ensures that very few votes are ignored, unlike other systems, especially First Past the Post, where many votes make no contribution to the result.

STV typically produces results that see power shared fairly across different parties, creates greater competition for seats (as candidates have a greater chance of success) and gives voters greater choice between both parties and individual candidates.

Candidates and Seats

STV encourages parties to stand candidates in more areas, as the chances of achieving representation are higher. There are no ‘no-go’ areas as there are under FPTP because candidates do not need to top the poll in order to win a seat.

The number of candidates standing in 2017 increased slightly on 2012 levels. There were 2,572 candidates compared to 2,496 in 2012. These 2,572 candidates contested seats in 354 multi-member wards giving voters on average a choice of 7.3 candidates. This is over double the choice compared to 2003, when the elections were held under FPTP and there was an average choice of 3.4 candidates. In 2007 there were 2,599 candidates contesting 353 multi-member wards (7.4 candidates on average).

As a rule, it is now unusual to see uncontested seats in Scottish local elections. In 2017 there were just three uncontested wards where the number of candidates standing was the same as the number of seats available. The uncontested wards were on the islands of Orkney and Shetland and on the Kintyre peninsula. This was the first time that any wards had been uncontested since the introduction of STV for Scottish local elections.

Preferences

In 2012 there was an increase in the proportion of voters expressing more than one preference.35 That year 86% of ballot papers contained a second as well as a first preference. In 2017, once again 86% of voters gave more than one preference. Sixty-one percent of voters cast three preferences and 29% used four preferences. These figures are slightly higher than in the previous election in 2012. In other words, the vast majority of voters are using the power that STV hands them.

Table 8: Power sharing in Scottish councils, as at June 2017 (one month after local elections)

Con Lab LD SNP Ind
Majority control 0 0 0 0 3
Minority administration 0 6 0 7 0
Coalition/power-sharing 7 7 5 8 12

In June 2017, one month after the local elections, the Conservatives were sharing power in seven councils, with Labour also sharing power in seven councils and governing as a minority in six. The SNP were governing as a minority in seven councils and sharing power in eight, while the Liberal Democrats were sharing power in five councils. Independent councillors were running three councils and sharing power in 12.

These results better reflect the diversity of voters’ views compared to the English local elections. No party was able to gain total control of a council, while getting fewer than half the votes, whereas this occurred in over two-fifths of councils in England. In Scotland, parties are required to share power and work together, in a way that reflects how local voters have voted, rather than dominating council chambers on a minority of votes.

The data provides further evidence that Scottish local election voters have a much higher chance of seeing their votes count, than their English counterparts do. In Scotland, in 2017, three-quarters of voters (75%) saw their first preference candidate elected. This is very similar to the 2007 Scottish local elections (74%) and 2012 Scottish local elections (77%), both also held under STV. In contrast, only 55% of votes in the 2019 English local elections, held under FPTP, went to winning candidates with all other votes being wasted. This figure is similar to the 52% of voters who saw the candidate they voted for elected in the last FPTP Scottish local elections, held in 2003. The benefits of STV stretch even beyond this comparison, however, as Scottish voters are also able to have an influence on the outcome with their subsequent preferential votes, meaning that even if their first choice candidate is not elected, they still have a chance of contributing to the election of their second, third or subsequent choice candidate.

Proportionality of results

Having produced DV scores for the 2017 Scottish local elections, using first preference votes to compare party vote shares with seat shares, it is clear that STV enabled a more proportional outcome than occurred in either the English local elections, using FPTP or the European Parliament election, using the D’Hondt method of Closed List PR.

The mean DV score for individual local authorities across Scotland was 9.7,  half the equivalent score for the English local elections (19.6). This score provides clear evidence that STV produces much fairer and more proportional outcomes than FPTP.

The table below lists the Scottish local authorities with the highest DV scores (or most disproportionate outcomes). When compared to the English councils with the highest DV scores, in Table 4, it is clear that the Scottish councils have far more proportional outcomes  (the highest DV score in Scotland is lower than the mean local authority DV score across all English councils). In sum, the councillors elected in Scotland much more accurately reflect what voters want.

Table 9: Top 10 most disproportionate results, Scottish local elections 2017

Authority Previous Current DV Score
Dundee SNP  NOC 17.8
East Dunbartonshire NOC NOC 17.2
Midlothian NOC NOC 16.4
North Lanarkshire NOC NOC 13.1
Dumfries & Galloway NOC NOC 12.8
Fife NOC NOC 12.8
East Lothian NOC NOC 12.3
Aberdeen NOC NOC 12.2
South Lanarkshire Lab NOC 11.8
Glasgow Lab NOC 11.2

Conclusion

Conclusion

Voters in England suffer clear disadvantage in comparison to their Scottish counterparts when it comes to their participation in local government elections. FPTP causes many voters to be disenfranchised through uncontested or under-contested seats, while many more see their votes wasted under a system that does not properly reflect how people have voted.

The outcomes of the elections analysed in this report provides clear evidence that it is time for England (and Wales, where councils may soon have the option of adopting STV) to join Scotland and Northern Ireland in using STV for local elections.

On a range of democratic outcomes, STV performs better than FPTP for local elections. The scourge of uncontested seats has been virtually eliminated in Scotland (and Northern Ireland); one-party dominated councils have been ended in Scotland too, meaning parties have to work together to get things done locally; local representation much more closely matches how people actually voted; the link to a local electoral area is retained; and voters are given maximum choice (they can give as many preferences as they want and are often able to choose who they prefer between candidates representing the same party, as well as choosing between parties).

The renewed fragmentation of the British electorate, as explored in the Introduction to this report, makes a move towards a proportional system for UK general elections even more compelling too. Even though general election opinion polls in summer 2019  have shown something of an increase in the combined Conservative and Labour vote share, none have come close to matching even the lowest combined Conservative and Labour general election vote share for over a century (66.6%, across Britain, at the 2010 general election).

It seems likely that the next general election, to be held under FPTP, will see ‘random’ results on a scale not seen before – with voters denied the opportunity to express themselves in a way that a better voting system would make possible. By using winner-takes-all FPTP, Westminster is an anomaly among central government legislatures throughout Europe. It is increasingly becoming an anomaly within the UK, too, with the Scottish, Welsh and Northern Irish devolved parliaments using varying forms of mixed or proportional voting systems, along with local authorities in Scotland and Northern Ireland. The vote-wasting machine of FPTP should be replaced at long last – with a system that reflects how voters want to be represented today.

Appendix

Votes, councillors elected and votes required to elect a councillor, by region and overall, English local elections 2019

‘Ind’ refers to Independents and local groups such as Residents’ Associations.

‘Oth’ refers to other political parties that received fewer than 5% of votes across the country.

North East Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 125314 59 21.6 17.6 2124
Green 36132 4 6.2 1.2 9033
Lab 209455 159 36.1 47.3 1317
LD 64306 36 11.1 10.7 1786
Ind 93431 70 16.1 20.8 1335
Oth 50794 8 8.8 2.4 6349
Overall 579432 336 100 100 1725

 

North West Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 599679 341 29.0 30.0 1759
Green 161160 25 7.8 2.2 6446
Lab 802524 537 38.8 47.3 1494
LD 239618 101 11.6 8.9 2372
Ind 174341 124 8.4 10.9 1406
Oth 89681 8 4.3 0.7 11210
Overall 2067003 1136 100 100 1820

 

Yorkshire & Humber Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 392617 196 28.9 37.0 2003
Green 135348 15 10.0 2.8 9023
Lab 432023 166 31.8 31.3 2603
LD 206923 71 15.2 13.4 2914
Ind 106100 67 7.8 12.6 1584
Oth 87261 15 6.4 2.8 5817
Overall 1360272 530 100.0 100.0 2567

 

East Midlands Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 599388 494 34.5 40.2 1213
Green 69636 12 4.0 1.0 5803
Lab 616216 401 35.4 32.6 1537
LD 229671 133 13.2 10.8 1727
Ind 175515 187 10.1 15.2 939
Oth 48097 3 2.8 0.2 16032
Overall 1738523 1230 100.0 100.0 1413

 

West Midlands Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 456511 379 38.4 47.4 1205
Green 102443 37 8.6 4.6 2769
Lab 337765 208 28.4 26.0 1624
LD 110098 59 9.3 7.4 1866
Ind 117589 106 9.9 13.3 1109
Oth 63999 10 5.4 1.3 6400
Overall 1188405 799 100.0 100 1487

 

East of England Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 773273 639 37.6 47.7 1210
Green 176982 48 8.6 3.6 3687
Lab 442695 213 21.5 15.9 2078
LD 402523 251 19.6 18.7 1604
Ind 209046 180 10.2 13.4 1161
Oth 51350 9 2.5 0.7 5706
Overall 2055869 1340 100.0 100.0 1534

 

South East Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 1237468 934 38.6 49.8 1325
Green 298328 81 9.3 4.3 3683
Lab 598646 254 18.7 13.6 2357
LD 653162 369 20.4 19.7 1770
Ind 315462 232 9.8 12.4 1360
Oth 104353 4 3.3 0.2 26088
Overall 3207419 1874 100.0 100.0 1712

 

South West Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 581543 380 34.9 37.5 1530
Green 117082 43 7.0 4.2 2723
Lab 238062 76 14.3 7.5 3132
LD 458031 327 27.5 32.2 1401
Ind 217388 187 13.1 18.4 1163
Oth 52022 1 3.1 0.1 52022
Overall 1664128 1014 100.0 100.0 1641

 

England (overall) Total votes Councillors elected % votes % seats Votes required to elect councillor
Con 4765793 3422 34.4 41.4 1393
Green 1097111 265 7.9 3.2 4140
Lab 3677386 2014 26.5 24.4 1826
LD 2364332 1347 17.1 16.3 1755
Ind 1408872 1153 10.2 14.0 1222
Oth 547557 58 4.0 0.7 9441
Overall 13861051 8259 100.0 100.0 1678

Read more publications...