Here to Stay: Two Decades of PR in Britain

Ian Simpson, Research Officer

Posted on the 31st July 2021

1. Foreword
  1. Foreword
7. Conclusion
  1. Conclusion



Across the UK, proportional representation is here to stay. Westminster needs to catch up.

As supporters and campaigners for electoral reform, it’s often easy to forget to celebrate the advances that have been made towards fairer votes. The damage done by winner-takes-all voting at Westminster overshadows so much of our politics.  

 While elections for the House of Commons remain warped, we are proud that for devolved elections in London, Wales, Scotland, and Northern Ireland, First Past the Post has been consigned to the dustbin of history.  

 Across the UK, elections with a proportional representation system are becoming the norm. Millions of votes have been cast across the UK using PR, with voters rightly expecting to secure fair representation, not the pale imitation of democracy that one-party-takes-all politics provides. 

 The benefits of this are clear. In devolved elections, voters’ voices are amplified through proportional representation – a stark contrast to the silencing effect of Westminster’s system. More than two decades in, the benefits are clear. 

But while elections for Westminster and English councils continue to lag behind (Wales has recently passed legislation letting councils switch to the Single Transferable Vote), they will keep failing voters:  reducing scrutiny, skewing resources, and silencing millions. It is a recipe for alienation, disengagement and division. 

 There is still some way to go, but our experiences in Scotland, Wales, Northern Ireland and London have shown the benefits of a fairer system: fostering cooperation and giving voters real power. 

 Now Westminster must follow suit. It cannot take two more decades for the Commons to catch up. 

 Rather than rolling back preferential voting – as UK ministers plan – we urge politicians to get with the times and back truly democratic elections at last.

Signed by over 650 ERS supporters in Scotland, Wales, Northern Ireland and London.



On 6 May 2021, the sixth set of elections took place for each of the devolved elected institutions that exist within the three nations of Great Britain – the Scottish Parliament, Senedd Cymru (Welsh Parliament) and the London Assembly. The first elections to these bodies took place in 1999 (Scottish Parliament and, as it was then known, Welsh Assembly) and 2000 (London Assembly). This means that systems of proportional representation (PR) have now been in use in Britain for over two decades. 

During this time, PR has become a fact of life for millions of voters. Indeed, the Scottish and Welsh governments have passed legislation allowing for proportional representation to be introduced for local government elections within those nations. The Single Transferable Vote (STV) system replaced First Past the Post (FPTP) for local government elections in Scotland, following legislation passed by the Scottish Parliament in the form of the Local Governance (Scotland) Act 2004.1Local Governance (Scotland) Act 2004.  In Wales, local authorities will have the option of switching from FPTP to STV, after their next set of elections in 2022, following the passage of the Local Government and Elections (Wales) Act 2021.2Local Government and Elections (Wales) Act 2021. 

 This report will explore the 2021 election results for each of these devolved institutions, as well as looking at how these compare with the results of the previous five sets of elections. We will review how the Additional Member System (AMS), a version of proportional representation that is used across the three institutions, produces much fairer outcomes for voters than the FPTP system that is still used for UK general elections.


For this report we spoke with voters and the following politicians:

Clive Lewis MP
Clive Lewis is the Labour MP for Norwich South (since 2015) and formerly served as a shadow minister.

Cllr Dave Dempsey
Cllr Dave Dempsey is a Scottish Conservative politician who serves as the Leader of the Opposition on Fife Council. He was elected in 2007 as part of Scotland’s first wave of PR elections for local government.

Baroness Jenny Randerson
Baroness Randerson is a Welsh Liberal Democrat member of the House of Lords. Prior to her peerage she was an Assembly Member for Cardiff Central from 1999 to 2011 when she served in the Welsh Labour-Lib Dem administration of the 2000–2003 Welsh Assembly government.

Cllr Caroline Russell AM
Caroline Russell is a Green Party politician and activist. Since May 2016, she has been a member of the London Assembly and serves as a councillor for Highbury East Ward on Islington Council.

History and Responsibilities of Devolved Bodies

History and Responsibilities

The Scottish Parliament was created in 1999, having been approved in a 1997 referendum. The Scottish Parliament is the most powerful of the three devolved bodies and has always worked on a ‘reserved powers’ basis, meaning that any areas of responsibility not specifically reserved to the UK parliament are devolved to the Scottish Parliament.3Sargeant, J. and Shuttleworth, K. (2021). Scottish parliament. Institute for Government, 21 May.

The Scottish Parliament has the power to pass legislation in most social policy areas, including health and social care, housing and education. It also has control over local government, agriculture and most aspects of transport, as well as the justice system. Most tax policy is reserved to the UK parliament but the Scottish Parliament has greater powers in this sphere than the other devolved bodies. It has powers over local property taxes, stamp duty, landfill tax and the power to set all rates and bands of income tax, other than the personal allowance.

“We can learn so much about the success of PR in Scotland, Wales and London. Giving all voters a voice and having politicians that work together to get things done is a far cry from the Punch and Judy politics of Westminster.”
Clive Lewis MP

Senedd Cymru was originally created as the National Assembly for Wales, in 1999. Its creation was also approved by a referendum in 1997.

The powers of the Senedd have increased over time. On its creation, the Assembly took over the functions of the Welsh Office of the UK government, with the power to make secondary legislation in these areas, meaning laws created under powers given by specific acts of the UK parliament. However, unlike the Scottish Parliament, it did not immediately have the power to pass primary legislation.

As part of the Government of Wales Act 2006, the Welsh Assembly was enabled to pass pieces of primary legislation, dependent upon the UK parliament giving it the power to do so on a case-by-case basis. In 2011 there was another referendum on extending the Assembly’s law-making powers, which was approved by voters. The Wales Act 2014 devolved some fiscal powers to the Assembly, including stamp duty, business rates, landfill tax and a portion of income tax. The Wales Act 2017 made a number of further changes, including moving to the ‘reserved powers’ model used in Scotland. At the same time, the Assembly was given power over elections, energy and transport. This Act also allowed the Assembly to change its name, which it duly did. From 6 May 2020, the institution became known as Senedd Cymru or Welsh Parliament.4Shuttleworth, K. (2021). Senedd Cymru (Welsh parliament). Institute for Government, 21 May.

“Electoral reform is not just about a different way of counting votes, it’s about a different way of doing politics – one that prioritises an equal and powerful voice for every voter, where parties work together with the best interests of their constituents in mind.”
Clive Lewis MP

In London, a 1998 referendum on the creation of the Greater London Authority, encompassing a Mayor of London and a London Assembly, was approved by voters.

Unlike the other two devolved institutions, which are responsible for whole nations, the London Assembly does not have law-making powers. The main function of the London Assembly is to hold the mayor to account. It is worth noting that the mayor is another position that is currently elected via a non-FPTP system, the Supplementary Vote (SV), which gives voters the option of expressing a second preference. Disappointingly, in March 2021, the Home Secretary announced the government’s intention to reduce voter choice by moving to FPTP for London mayoral elections, as well as Combined Authority mayoral and Police and Crime Commissioner elections.5Scolari, F. (2021). We can’t let the government force a broken system on important elections. ERS Blog, 19 March.

The London mayor has defined powers across a number of areas including transport, housing, economic development, planning, policing and fire and rescue. The mayor acts as the Police and Crime Commissioner for London, while another major function is to oversee Transport for London, having responsibility for its strategic direction and setting public transport fares. 

The London Assembly organises committees that investigate the policies and programmes of the mayor, as well as those of bodies under control of the mayor. The mayor is required to appear before the Assembly ten times per year to answer the questions of Assembly members. The Assembly has the power to amend or vote down the mayor’s annual budget. However, this requires a two-thirds majority and has never happened.6Hall, D. and Paun, A. (2021). Mayor of London and London Assembly. Institute for Government, 21 May.   

“For years I lived in England and never lived in a place with an MP I had voted for. It made me feel unrepresented and disconnected, I had the feeling that this made politics tribal and stifled informed debate.”
Jean Bell (voter in Scotland)

The Additional Member System – How it works in Scotland, Wales and London

The Additional Member System (AMS) is used for elections to all three devolved bodies and is also sometimes referred to as a Mixed Member Proportional (MMP) system.7 This system is a combination of FPTP single-member constituencies and regional, multi-member constituencies, where a form of Party List PR
is used.

Voters have two ballot papers. On the first is a list of candidates who are standing as constituency representatives. Voters mark a cross next to one candidate and whichever candidate gets the most votes within the constituency wins the one seat available. On the second ballot paper are a list of parties, with the names of the candidates representing each party on their list underneath. The voter puts a cross next to the party list they want to vote for.

Once the votes in the FPTP constituencies have been counted and representatives elected, the party list votes in each electoral region (areas that cover a number of FPTP constituencies) are then counted. List seats are distributed with the aim of ensuring that the overall share of seats for each party matches the share of party list votes received, as far as possible.

The party list representatives are allocated on the basis of a modified D’Hondt formula. In each electoral region, the number of votes each party receives on the list vote is divided by the number of FPTP seats that party won within that particular region, plus one. That means if Party B received 50,000 votes on the list vote in Region X, and had won two FPTP seats in that region, their initial list vote would be 16,667 (50,000 divided by three). The party with the highest modified list vote receives the first list seat in that region. Each time a party wins a list seat, their vote is divided again, so if Party B received a list seat, on top of their two FPTP constituency seats, their new list vote would be 12,500 (50,000 divided by four). This process continues until all list representatives have been elected in each region.

Although all three of the devolved bodies are elected by AMS, each version works slightly differently. The number and type of seats in each institution is outlined in Table 1. The party list seats for the Scottish Parliament are elected across eight electoral regions (seven MSPs for each region). The Senedd party list seats are elected across five electoral regions (four MSs for each region). For the London Assembly, the party list seats are allocated on a London-wide basis.

The party list seats make up only a third of the seats in the Senedd, around 10 percentage points lower than the proportion of party list seats in the Scottish Parliament and the London Assembly. This has implications for the proportionality of Senedd outcomes and is discussed later in the report.

Table 1: Number and type of seats in Scottish Parliament, Senedd Cymru, and London Assembly

Institution FPTP constituency seats FPTP constituency seats (%) Party list seats Party list seats (%) Total number of seats
Scottish Parliament 73 56.6 56 43.4 129
Senedd Cymru 40 66.7 20 33.3 60
London Assembly 14 56.0 11 44.0 25


Scottish Polling Station A Boards

Scottish Parliament


The Scottish Parliament’s electoral system was in the spotlight prior to the 2021 election, as there was much pre-election discussion of parties ‘gaming the system’ with the creation of several list-only parties such as Alba and All for Unity.8Smith, A. (2020). ‘Gaming’ Scottish proportional electoral system ‘is bad for democracy’. The Scotsman, 19 July. However, in terms of proportionality of outcome, the result was very much in line with the five previous Holyrood elections. As has always been the case, the AMS system produced a much fairer outcome for voters than the FPTP system produces for UK general elections.

Examining the results of the separate parts of the election, the FPTP constituency seats and the regional list seats, reveals how the proportional electoral system works to ensure a much more equitable outcome than would be the case under Westminster-style FPTP.

The data in Table 2 shows the kind of Holyrood chamber that would have resulted from an election held under Westminster-style FPTP. The Scottish National Party (SNP) won almost 85 percent of constituency seats and would have dominated the new parliament, despite receiving fewer than half the votes (47.7 percent) across the FPTP constituency contests. Parties receiving a combined 52.3 percent of votes would have been virtually wiped out under this winner-takes-all approach.

Table 2: Vote share and seat share in FPTP constituency seats, Scottish Parliament election, 2021

Party MSPs (%) Vote (%) Difference (%)
Scottish National Party 84.9 47.7 +37.2
Conservative 6.8 21.9 -15.1
Labour 2.7 21.6 -18.9
Liberal Democrat 5.5 6.9 -1.4
Green 0.0 1.3 -1.3
Others 0.0 0.6 -0.6
Total 100 100

This is the kind of result that is all too familiar at UK general elections, where every seat is allocated on the basis of FPTP. In every UK general election since 1997, the largest party in Scotland has won a majority of Scottish seats on a minority of votes. The same thing has happened only once at a Scottish Parliament election over the same period, in 2011 (see Table 3). Even on that occasion, the over-representation of the largest party was far smaller than that seen in any UK general election since 1997. 

On average, across the last seven UK general elections, the largest party has won 75 percent of Scottish seats on the basis of just over 43 percent of votes. Across the six Scottish Parliament elections during the same period, the average result for the largest party is a much fairer 45 percent of seats, having received 37 percent of regional votes. The regional vote share is the best data to use for this comparison, as an aim of the AMS system is for the overall proportion of seats to match the regional vote shares as closely as possible.

Table 3: Largest party over-representation at Scottish Parliament and UK general elections, 1997–2021

Scottish Parliament elections (AMS)

Year Largest party Total MSPs (%) Regional vote (%) Difference (%)
2021 SNP 49.6 40.3 9.3
2016 SNP 48.8 41.7 7.1
2011 SNP 53.5 44.0 9.5
2007 SNP 36.4 31.0 5.4
2003 Labour 38.8 29.3 9.2
1999 Labour 43.4 33.6 9.8
Mean 45.1 36.7 8.4

UK general elections – Scotland (FPTP)

Year Largest party Total MPs (%) Vote (%) Difference (%)
2019 SNP 81.4 45.0 36.4
2017 SNP 59.3 36.9 22.4
2015 SNP 94.9 50.0 44.9
2010 Labour 69.5 42.0 27.5
2005 Labour 67.8 38.9 28.9
2001 Labour 76.4 43.3 33.1
1997 Labour 77.8 45.6 32.2
Mean 75.3 43.1 32.2

As Table 4 shows, the allocation of regional list seats acts to make the overall outcome much more proportional. Having dominated the constituency seat contests, the SNP secured only two of the 56 regional list seats, while the Conservative, Labour and Green parties picked up seats in a way that meant their overall seat share closely matched their regional list vote share (see Table 5).

Taking a closer look at what happened to the SNP regional seats in South Scotland provides an example of the AMS system in action. The SNP gained two FPTP constituency seats in the South Scotland region, Ayr from the Conservatives and East Lothian from Labour, meaning they won six of the nine FPTP constituency seats in the region, compared to just four constituency seats in 2016.

However, at 37.6 percent, the SNP’s vote share in the South Scotland regional vote was almost identical to its vote share in 2016, 38.3 percent.Consequently, the two FPTP seats the SNP gained in the region were offset by the party losing two of the three South Scotland list seats they had won in 2016. The Conservatives and Labour, who also had similar South Scotland list vote shares to 2016, both picked up one extra South Scotland list seat to compensate for their constituency losses.

Table 4: Allocation of regional list seats, Scottish Parliament election, 2021

Party Central Glasgow Highlands & Islands Lothian Mid & Fife North East South West Total
Scottish National Party 0 0 1 0 0 0 1 0 2
Conservative 3 2 4 3 4 4 3 3 26
Labour 3 4 1 2 2 2 3 3 20
Liberal Democrat 0 0 0 0 0 0 0 0 0
Green 1 1 1 2 1 1 0 1 8
Total 7 7 7 7 7 7 7 7 56

Looking at the overall result of the election (see Table 5), the largest party, the SNP, won almost half (49.6%) of all MSPs. This share of seats is 9.3 points higher than the party’s vote share on the regional list vote (40.3%). This is typical of the ‘over-representation’ of the largest party under the Scottish version of AMS, where across the six Scottish parliamentary elections the largest party has been ‘over-represented’ by an average of 8.4 points (see Table 3). It is also typical to see parties in second, third and fourth position achieve overall seat shares that closely match their regional vote shares, with some smaller parties achieving no representation. In 2021, just over five percent of regional list votes went to parties that won no MSPs. However, of these parties that received no representation, the single biggest one, Alba, received just 1.7 percent of list votes.

“Using PR in Scotland removes the need for voters to vote for something they don’t want, to stop them getting something they want even less…”
Cllr Dave Dempsey

Table 5: Overall seat share versus regional vote share, Scottish Parliament election, 2021

Party Total MSPs (%) Regional vote (%) Difference (%)
Scottish National Party 49.6 40.3 +9.3
Conservative 24.0 23.5 +0.5
Labour 17.1 17.9 -0.8
Green 6.2 8.1 -1.9
Liberal Democrat 3.1 5.1 -2.0
Alba 0.0 1.7 -1.7
All For Unity 0.0 0.9 -0.9
Others 0.0 2.6 -2.6
Total 100 100

A common way to measure the overall proportionality of an election result is via a Deviation from Proportionality (DV) score. The DV score shows the extent to which an election result deviates from proportionality, i.e. from what it would look like if seats were proportional to votes gained by each party. It gives a percentage of seats in parliament which are ‘unearned’ in proportional terms.

There are various ways of measuring DV scores. We have used the Loosemore-Hanby index, which is calculated by adding up the difference between each party’s vote share and their seat share and dividing by two. This gives a ‘total deviation’ score – the higher the score, the more disproportionate the result.

As Table 6 indicates, AMS consistently ensures far more proportional results than we see for Scottish seats at UK general elections, where FPTP is used. Across the last seven UK general elections, the mean DV score for Scotland is 32.3, whereas for Scottish Parliament elections it is just 10.7.

Table 6: Deviation from proportionality scores, Scottish Parliament and UK general elections (Scotland), 1997–2021

Year Scottish Parliament elections UK general elections (Scotland)
2021 9.9
2019 36.4
2017 22.5
2016 8.3
2015 44.9
2011 11.9
2010 27.5
2007 13.0
2005 28.9
2003 10.7
2001 33.1
1999 10.6
1997 33.1
Mean 10.7 32.3

Voter experiences in Scotland

Voters appear to appreciate their increased representation in Holyrood compared to Westminster. In our preparations for this report, the ERS spoke to voters in Scotland about their experiences of using PR.  

Lothian resident Martin Muncer told the ERS: “From 20 years of experience, the Additional Member System (AMS) of proportional representation is far superior to the first-past-the post (FPTP) system used for the UK parliament, and is much more effective in giving a balance of the political opinion that is held across the region.”

Izzie Kirkpatrick had felt disenfranchised under the Westminster system saying: “I vote in Scotland, and I feel like my vote counts for more in our elections.  [In Westminster] the First Past the Post system seeks to perpetuate a two-party system which is no longer the case. It’s a massive disservice to the electorate… I’m very politically minded and the UK voting system leaves me feeling utterly disenfranchised.”

Others recognised the ‘satisfying simplicity’ of knowing seats fairly reflect share of the vote. “With two votes in Scottish Parliament polls I have confidence that my choices of an individual MSP and a [regional] party representative effectively reflect my political views without the high possibility of wasting my vote. This in contrast to the outdated Westminster system which regularly produces huge, ‘wasted’ party majorities in safe seats,” voter Alan Bell said. 

Rosemary Osman had assumed voting with PR would be confusing when she moved to Scotland, but when it came to the ballot box, found it ‘perfectly logical and easy to navigate’: “Crucially, though, I felt as though my vote was going to be counted fairly. During the other 41 years of my voting life – in England – although I always voted I rarely felt that my vote would make any difference to the outcome. In my opinion FPTP is undemocratic and does not properly represent all citizens.”


Welsh Polling Station

Senedd Cymru


Elections for the Welsh Parliament were also held in May 2021. 

If the Senedd election had been held using Westminster-style FPTP, the result, as in Scotland, would have been an overwhelming majority for the largest party, on the basis of fewer than half of votes cast (see Table 7). Labour won just over two-thirds (67.5 percent) of the FPTP constituency seats, from two-fifths (39.9 percent) of votes cast across these seats.

“We’ve got the constituency seats in Wales, under AMS. Of course, you have safe Labour constituency seats. But they’re not as comfortable as they are in England, because if you represent a valleys seat with a majority of 20,000, there is always someone else working on your patch [from the regional list seats].”
Baroness Randerson

Table 7: Vote share and seat share in FPTP constituency seats, Senedd election, 2021

MSs (%) Vote (%) Difference (%)
Labour 67.5 39.9 +27.6
Conservative 20.0 26.1 -6.1
Plaid Cymru 12.5 20.3 -7.8
Liberal Democrat 0.0 4.9 -4.9
Green 0.0 1.6 -1.6
Abolish Welsh Assembly 0.0 1.6 -1.6
Reform UK 0.0 1.6 -1.6
UKIP 0.0 0.8 -0.8
Others 0.0 3.2 -3.2
Total 100 100

Again, as with Scotland, the shape of the results seen across the Senedd constituency seats closely mirrors the shape of the results we tend to see in Wales at UK general elections, held under FPTP. At every UK general election since 1997, the largest party in Wales, always Labour, has won a majority of Welsh seats. Only once, in 1997, did Labour also receive a majority of votes in Wales (see Table 8).

The largest party across these last seven UK general elections has on average won just over 70 percent of seats, based on 44 percent of votes. In contrast, across the six Senedd elections to date, no party has won a majority of seats, though Labour has won exactly half the seats on three occasions, in 2003, 2011 and 2021. On average, across the six Senedd elections, the largest party has won 48.1 percent of seats, having received 34.4 percent of regional list votes, a more proportional outcome than is in evidence at UK general elections in Wales.

“Another thing that made it very significant was all the women – half of us were women. One of the things that any list system gives you, it’s the individual political parties feel obliged, when they stand back, to make their lists more balanced.”
Baroness Randerson

Table 8: Largest party over-representation at Senedd and UK general elections, 1997–2021

Senedd elections (AMS)

Year Largest party Total MSs (%) Regional vote (%) Difference (%)
2021 Labour 50.0 36.2 +13.8
2016 Labour 48.3 31.5 +16.8
2011 Labour 50.0 36.9 +13.1
2007 Labour 43.3 29.6 +13.7
2003 Labour 50.0 36.6 +13.4
1999 Labour 46.7 35.5 +11.2
Mean 48.1 34.4 +13.7

UK general elections – Wales (FPTP)

Year Largest party Total MPs (%) Vote (%) Difference (%)
2019 Labour 55.0 40.9 +14.1
2017 Labour 70.0 49.0 +21.0
2015 Labour 62.5 36.9 +25.6
2010 Labour 65.0 36.2 +28.8
2005 Labour 72.5 42.7 +29.8
2001 Labour 85.0 48.6 +36.4
1997 Labour 85.0 54.7 +30.3
Mean 70.7 44.1 +26.6

Table 9 shows how the allocation of the 20 Senedd regional list seats ensured a more proportional Senedd chamber than would have been the case if the election had been held under FPTP alone. As in Scotland, the party that dominated the constituency seat section of the election picked up few seats, with Labour winning only three out of the 20 available. The Conservatives and Plaid Cymru picked up eight regional list seats each, meaning that, just like in Scotland, the seat shares of the second and third placed parties closely matched their vote shares. 

The Liberal Democrats showed how it is possible for parties to achieve representation in the Senedd, without winning any FPTP constituency seats. The party lost their sole constituency seat, Brecon and Radnorshire, to the Conservatives. However, they won a seat in the region in which Brecon resides, Mid and West Wales, by securing nearly seven percent of votes across the region.

Table 9: Allocation of regional list seats, Senedd election, 2021

Mid & West North South East South Central South West Total
Labour 2 1 0 0 0 3
Conservative 0 2 2 2 2 8
Plaid Cymru 1 1 2 2 2 8
Liberal Democrat 1 0 0 0 0 1
Total 4 4 4 4 4 20

The overall Senedd result (see Table 10) shows that that the largest party, Labour, was ‘over-represented’ by 13.8 points, when comparing its overall seat share with its regional vote share. As in Scotland, this ‘over-representation’ of the largest party closely matched the average across the six devolved elections to date, which stands at 13.7 points (see Table 8). 

As in Scotland, some smaller parties did not receive representation in the Senedd. Taken together, parties who received 13.7 percent of regional list votes won zero seats between them. This is almost three times the vote share of parties who received no representation in the Scottish Parliament.

Indeed, the Green Party, who received no representation in the Senedd, secured slightly more regional list votes than the Liberal Democrats across Wales as a whole. However, whereas the Liberal Democrats were able to secure 6.8 percent of votes in the Mid and West Wales region, which was enough to elect a regional member, the highest the Greens managed in any particular region was 5.7 percent in South Wales Central, which was not enough to secure them an ‘additional’ member for that region.

Table 10: Overall seat share versus regional vote share, Senedd election, 2021

Total MSs (%) Regional vote (%) Difference (%)
Labour 50.0 36.2 +13.8
Conservative 26.7 25.1 +1.6
Plaid Cymru 21.7 20.7 +1.0
Liberal Democrat 1.7 4.3 -2.6
Green 0.0 4.4 -4.4
Abolish Welsh Assembly 0.0 3.7 -3.7
UKIP 0.0 1.6 -1.6
Reform UK 0.0 1.1 -1.1
Others 0.0 2.9 -2.9
Total 100 100

Table 11 provides another indicator of the greater fairness of the Senedd election results compared to UK general elections in Wales. The mean DV score for the six Senedd elections is 15.1, more than 10 points lower than the mean score for the last seven UK general elections across Wales, which is 26.9. However, it should be noted that the mean DV score for the six Senedd elections is over four points higher than the mean score for the six Scottish Parliament elections (10.7).

One of the reasons for this is the higher proportion of FPTP constituency members in the Senedd, compared to the Scottish Parliament. Whereas 56.6 percent of MSPs are FPTP constituency representatives, this rises to 66.7 percent of MSs in the Senedd. Consequently, allocation of the ‘additional’ members provides less scope for balancing out the results from the FPTP constituencies. It is also worth noting that the mean ‘over-representation’ of the largest party is more than five points higher across the six Senedd elections (13.7 points), than it is across the six Scottish Parliament elections (8.4 points)

Table 11: Deviation from proportionality scores, Senedd and UK general elections (Wales), 1997–2021

Year Senedd elections UK general elections (Wales)
2021 16.4
2019 14.2
2017 21.0
2016 16.9
2015 27.3
2011 14.7
2010 28.8
2007 17.7
2005 29.9
2003 13.8
2001 36.5
1999 11.2
1997 30.4
Mean 15.1 26.9

Potential improvements to the system

Major reform of the Senedd’s electoral system is on the agenda. Recent years have seen two reports calling for reform, one from the Expert Panel on Assembly Electoral Reform, published in November 2017,9Expert Panel on Assembly Electoral Reform (2017). A parliament that works for Wales: The report of the Expert Panel on Assembly Electoral Reform. and the other by the Committee on Senedd Electoral Reform, published in September 2020.10Welsh Parliament Committee on Senedd Electoral Reform (2020). Senedd reform: The next steps. Both reports recommend increasing the size of the Senedd to between 80–90 MSs, to acknowledge its increased powers since its creation, and so that the Senedd can properly hold the Welsh government to account. Alongside this proposed enlargement, both reports also recommend replacing the AMS system with STV.

Following the Senedd election, it appears possible that these proposed reforms will be taken forward. Welsh Labour’s manifesto indicated that ‘we will build on the work of the Senedd Committee on Electoral Reform…and develop proposals to improve the representation of the people of Wales in their Parliament’.11Welsh Labour (2021). Moving Wales Forward: Welsh Labour Manifesto 2021. Since the election, First Minister Mark Drakeford has indicated his backing for reform, stating ‘I think there is a significant tide across parties in the Senedd to grasp this difficult issue of making the Senedd fit for the responsibilities it now discharges’.12BBC News (2021). Mark Drakeford backs calls for bigger Welsh Parliament. BBC News, 19 May. Both Plaid Cymru and Liberal Democrat manifestos explicitly committed to increasing the size of the Senedd and introducing STV.13Plaid Cymru (2021). Vote for Wales: Senedd Election Manifesto 2021.; Welsh Liberal Democrats (2021). Put Recovery First: 2021 Manifesto.
However, at present it is unclear when and how these reforms might be taken forward. 

A major reform, such as increasing the size of the Senedd, provides a prime moment to consider how MSs are elected and we agree with the conclusions of the Expert Panel that STV is the best system available.14Blair, J. (2017). Wales needs a stronger Senedd that’s fit for the future. ERS Blog, 12 December. STV puts more power in the hands of the voters, giving them the opportunity to rank candidates in order of preference, meaning if their first choice is not elected, there is still the potential for their vote to influence the outcome. As has already been noted, 13.7 percent of list votes went to parties that have no Senedd representation. Under STV, people who voted for these parties would have had the opportunity to ensure their vote was not completely wasted, by indicating their second and subsequent voting preferences.

Voter experiences in Wales

Voters in Wales were positive about the increased representation in the Senedd compared to Westminster. 

First-time voter Madeline Dhesi told the ERS: “The May Senedd elections were my first-time voting. I was so grateful – after doing the vote count in the 2019 GE and seeing so many wasted votes – that at least half of my votes could come from what I actually wanted and not tactical voting. Whilst that vote counted it would feel more valued if there were more seats available on the regional list and I hope when the Senedd expands the new members are chosen via PR and not more constituency seats.”

Meg Underdown said: “I live in Wales and have voted in the Senedd elections. I definitely think that my views are better represented by a system of PR”.

The contrast between the Welsh experience and voting in England was noted by Rhianon Turrell as a much more positive one: “I voted for the first time in Wales this year having moved from the Midlands. It was a breath of fresh air. It was a simple process and I am now for the first time (in 50 years of voting) represented at the Welsh Assembly by people I voted for, having lived in constituencies all my life which never changed hands. The system produces a result which reflects the people’s wishes and produces a balanced outcome with much less posturing by politicians and representatives who just get on with the job. I just wish the same could apply to Westminster and we might achieve a full democracy.”

“I tell you what really annoys me. When you talk to people about proportional representation, it comes down in the end to, ‘oh, it’s complicated and people find it difficult’. That’s a load of rubbish. They use PR for all sorts of organisations. If they live in Scotland, Wales, or Northern Ireland, they do PR for their elections.” Baroness Randerson

London Polling Station

London Assembly


May 2021 also saw the sixth set of elections to the London Assembly.

As with the two other devolved institutions, if Assembly members were elected via FPTP alone, then the largest party, Labour, would have won over half the seats from a minority of votes cast (see Table 12). Indeed, across the six London Assembly elections, in 84 FPTP constituency contests, no party other than Labour or the Conservatives has won a FPTP constituency seat.  The AMS system has ensured that supporters of other parties have had their voices heard.

“PR would bring a diversity of voices into Westminster. When I stand up in the Islington Council chamber, I think I represent something between 17% and 19% of Islington residents – but I’m just 2% of the seats in the chamber. But I have voices of people with me. In the London Assembly – I stand with other Greens, and we have among that body a proportionate presence. I think the Greens add another perspective and the Assembly is the richer for having that diversity.”
Cllr Caroline Russell AM

Table 12: Vote share and seat share in FPTP constituency seats, London Assembly election, 2021

AMs (%) Vote (%) Difference (%)
Labour 64.3 41.7 +22.6
Conservative 35.7 32.0 +3.7
Green 0.0 13.0 -13.0
Liberal Democrat 0.0 10.3 -10.3
Others 0.0 3.1 -3.1
Total 100 100

Table 13 shows that across the last seven general elections, the largest party in London, always Labour, has won a majority of seats. On only one occasion, in 2017, did they win a majority of votes. In contrast, no party has ever won a majority of seats on the London Assembly. The average seat share for the largest party at Assembly elections is 41.7 percent, on the basis of 34.6 percent of regional list votes, whereas across the seven UK general elections since 1997, the average is 65.6 percent of seats, from 45.6 percent of votes.

Table 13: Largest party over-representation at London Assembly and UK general elections, 1997–2021

London Assembly elections (AMS)

Year Largest party Total AMs (%) Regional vote (%) Difference (%)
2021 Lab 44.0 38.1 +5.9
2016 Lab 48.0 40.3 +7.7
2012 Lab 48.0 41.1 +6.9
2008 Con 44.0 34.6 +9.4
2004 Con 36.0 28.5 +7.5
2000 Lab/Con 36.0/36.0 30.3/29.0 +5.7/+7.0
Mean 41.7 34.6 +7.2

UK general elections – London (FPTP)

Year Largest party Total MPs (%) Vote (%) Difference (%)
2019 Lab 67.1 48.1 +19.0
2017 Lab 67.1 54.5 +12.6
2015 Lab 61.6 43.7 +17.9
2010 Lab 52.1 36.6 +15.5
2005 Lab 60.3 39.1 +21.2
2001 Lab 74.3 47.4 +26.9
1997 Lab 77.0 49.5 +27.5
Mean 65.6 45.6 +20.1

In the 2021 Assembly election, the Green Party and Liberal Democrats were able to pick up regional list seats, as shown in Table 14, which meant that their overall share of seats closely matched their regional list vote share (see Table 15). However, a proliferation of smaller parties, representing 12.1 percent of all regional list votes cast, received no representation, largely due to the small size of the Assembly limiting more proportional outcomes.

“If the London Assembly was elected under FPTP you just wouldn’t have that diversity of ideas… With PR you do get a broader range of people.”
Cllr Caroline Russell AM

Table 14: Allocation of London-wide list seats, London Assembly election, 2021

London-wide list seats
Labour 2
Conservative 4
Green 3
Liberal Democrat 2
Total 11

Table 15: Overall seat share versus regional vote share, London Assembly election, 2021

Total AMs (%) Regional vote (%) Difference (%)
Labour 44.0 38.1 +5.9
Conservative 36.0 30.7 +5.3
Green 12.0 11.8 +0.2
Liberal Democrat 8.0 7.3 +0.7
Others 0.0 12.1 -12.1
Total 100 100

As with the other devolved institutions, a comparison of the London Assembly DV score with the DV scores for the relevant portion of UK general elections, shows how the PR system produces overall results that are far fairer than those produced by FPTP. Across the six London Assembly elections, the mean DV score is 12.7, whereas for the last seven UK general elections, in London, the mean DV score is almost twice this, at 20.9 (see Table 16).

Table 16: Deviation from proportionality scores, London Assembly and UK general elections (London), 1997–2021

 Year London Assembly elections UK general elections (London)
2021 12.1
2019 19.0
2017 12.6
2016 12.0
2015 20.1
2012 9.2
2010 19.3
2008 14.3
2005 21.2
2004 13.7
2001 27.0
2000 14.9
1997 27.5
Mean 12.7 20.9

Voter experiences in London

Voters in London spoke of the lack of pressure to vote tactically in the Assembly elections compared to other elections in the capital and the benefit of knowing their vote would be counted. 

Stephanie Jacques told the ERS: “I felt empowered to vote in the London elections. For once I didn’t feel torn between voting for who I want to vote for and who I have to tactically.”

Kate Hames spoke of the relief of knowing her vote would count under AMS: “I know that my London Assembly vote counts. FPTP is outdated and undemocratic, often leading to big majorities for governments who have less than 50% of the national vote. I also think it sustains adversarial politics, where we should be moving to more grown-up, collaborative decision-making.”

Peter Rodrigues added: “Voting in the London election made me more confident that my vote would count and not be a wasted vote.”

The benefit of having many views reflected on the Assembly was noted by long time voter Tom Brown: “I have voted in every London Mayoral and Assembly election since the office and system were established in 2000. I object very strongly to any proposal to replace the Preference system for Mayor with FPTP, which would be retrograde and anti-democratic in permitting a candidate with potentially only small minority support to win the election… If any change is made, it should be to a full preferential ranking of all candidates on the ballot paper, to ensure the result reflects the widest range of expression of voters’ preferences.” 

“With the London Assembly elections people say ‘it’s the only time I get to vote for a Green who can win’. That’s worrying – people know that the FPTP voting system makes it hard to get their candidate elected.”
Cllr Caroline Russell AM



After more than two decades of proportional representation being used for elections to devolved bodies in each nation of Great Britain, it is clearer than ever that FPTP for UK general elections is an outdated anomaly.

The data in this report reveals how PR produces much fairer outcomes for devolved elections in Scotland, Wales and London than FPTP in the equivalent nations and regions at Westminster level.

The testimony of PR supporters in each of these areas emphasises how voters can feel much more engaged in the democratic system when they feel their vote actually counts.

English voters are in danger of being left behind. Not only are the devolved parliamentary elections in Scotland and Wales run under PR, but both parliaments have legislated for the further expansion of PR systems into local elections in each country. In Scotland, there have already been three sets of local elections held under STV and Welsh councils will have the opportunity to move to STV for their elections, after the next set of Welsh local elections in 2022.

Now that European Parliament elections no longer take place, voters in England, outside of London, no longer have the opportunity of taking part in elections held using PR systems. English voters are stuck with the outdated and unfair FPTP system for both UK general elections and for their local elections. 

More than 20 years of experience have shown that voters in all parts of Britain can easily use and understand PR systems, and many greatly value the benefits that these systems bring. It is long overdue that all elections in the UK were run via proportional representation, with FPTP consigned to history.

Read more publications...