In May 2022, local elections will take place throughout Scotland and Wales, as well as in many parts of England, including the election of all local councillors in Greater London.
In Scotland, these will be the fourth set of local elections held using the Single Transferable vote (STV). STV puts power in voters’ hands and allows them to make nuanced choices, not just between parties but often between candidates of the same party. It is a straightforward system for voters to use, ranking candidates in order of preference – 1,2,3 etc. Electors can give their first preference to their favourite candidate, while also being able to influence the election’s outcome via their later preferences.
STV in Scotland sees councillors elected in multi-member wards of typically between three or four seats. These larger wards are designed to ensure that the diversity of opinion in an area is at least relatively well represented by those candidates who are elected and by extension secure a reasonably proportional outcome across the council area as a whole. It is highly unusual, almost impossible, for candidates from the same party to secure all the seats in one of these STV multi-member wards, unless it is a genuine reflection of overwhelming support for one party above all others.
This year’s local elections in Wales will take place under First Past the Post (FPTP). However, following the Senedd’s passing of the Local Government and Elections (Wales) Act 2021, Welsh local authorities will soon have the power to switch to using STV for their local elections.
In England, there has been no progress towards the adoption of STV for local elections and this year’s elections will again take place under FPTP.
London’s councils and the problems of FPTP
The election of all councillors across Greater London’s 32 boroughs take place in one go, every four years. As such, they provide some prime examples of the inadequacies of FPTP when it comes to local democracy in England, particularly as almost all wards in London are represented by more than one councillor.
In 2018, London saw elections across 632 council wards of various sizes:
- Four (<1%) single member wards
- Fifty-five (9%) two member wards
- Five hundred and seventy-three (91%) three member wards
Of these 628 multi-member wards, electing either two or three councillors, 565 saw one party win all of the seats up for grabs. This means that 90% of all the multi-member wards in London were represented by only one party after the last set of local elections in 2018. This widespread domination by one party at ward level, inevitably translates into some highly disproportional results at council level.
This contrasts starkly with the situation in Scotland, where no single party has ever won all the seats in an STV multi-member ward. Independent candidates have sometimes won all available seats but never a single political party. These much more representative ward-level outcomes meant that after the last set of Scottish local elections in 2017, no one party had a majority on any council in the country, though Independent candidates made up a majority of councillors in three authorities.
The rest of this paper focuses on three examples from the 2018 London local elections, highlighting disproportional outcomes at ward and council level, and contrasts these with what occurred in Scotland’s capital, Edinburgh, at the 2017 Scottish local elections.
Case Studies
| Party |
Votes |
Vote % |
Seats |
Seat % |
| Labour |
26295 |
24.5 |
8 |
13.3 |
| Conservative |
47777 |
44.6 |
50 |
83.3 |
| Liberal Democrats |
15896 |
14.8 |
0 |
0.0 |
| Green Party |
11460 |
10.7 |
0 |
0.0 |
| Independents |
2013 |
1.9 |
2 |
3.3 |
| UKIP |
3065 |
2.9 |
0 |
0.0 |
| Others |
642 |
0.6 |
0 |
0.0 |
| TOTAL |
107148 |
100.0 |
60 |
100.0 |
The above data highlight the disproportional nature of the 2018 local election outcome in Bromley. Despite a vote share below 50%[i], the Conservatives received over four-fifths of council seats, leaving all other parties badly under-represented.
The Bromley ward level data, below, shows how every ward in the borough saw candidates from only one party elected. This left thousands of supporters of other parties in each ward unrepresented. Across the borough, over half (51.4%) of votes went to parties that had no representatives elected in that ward.
Ward level results
| WARD |
COUNCILLORS ELECTED |
VOTES FOR BEST PLACED CANDIDATE OF NON-REPRESENTED PARTIES |
VOTES FOR BEST PLACED CANDIDATE OF NON-REPRESENTED PARTIES (%) |
| Bickley |
⬤⬤⬤ CON x3
|
2774 |
48.6 |
| Biggin Hill |
⬤⬤ IND x2 |
2072 |
63.8 |
| Bromley Common & Keston |
⬤⬤⬤ CON x3 |
2153 |
43.6 |
| Bromley Town |
⬤⬤⬤ CON x3 |
3759 |
63.8 |
| Chelsfield & Pratts Bottom |
⬤⬤⬤ CON x3 |
2345 |
44.5 |
| Chislehurst |
⬤⬤⬤ CON x3 |
2150 |
41.0 |
| Clock House |
⬤⬤⬤ LAB x3 |
4343 |
72.0 |
| Copers Cope |
⬤⬤⬤ CON x3 |
3893 |
63.0 |
| Cray Valley East |
⬤⬤⬤ CON x3 |
2399 |
59.7 |
| Cray Valley West |
⬤⬤⬤ CON x3 |
3104 |
67.9 |
| Crystal Palace |
⬤⬤ LAB x2 |
1505 |
43.1 |
| Darwin |
⬤ CON x1 |
448 |
27.2 |
| Farnborough & Crofton |
⬤⬤⬤ CON x3 |
2485 |
43.1 |
| Hayes & Coney Hall |
⬤⬤⬤ CON x3 |
2565 |
45.3 |
| Kelsey & Eden Park |
⬤⬤⬤ CON x3 |
3205 |
55.6 |
| Mottingham & Chislehurst North |
⬤⬤ CON x2 |
1380 |
55.4 |
| Orpington |
⬤⬤⬤ CON x3 |
2130 |
42.7 |
| Penge & Cator |
⬤⬤⬤ LAB x3 |
2606 |
44.4 |
| Petts Wood & Knoll |
⬤⬤⬤ CON x3 |
2233 |
39.5 |
| Plaistow & Sundridge |
⬤⬤⬤ CON x3 |
3521 |
63.9 |
| Shortlands |
⬤⬤ CON x2 |
1423 |
40.6 |
| West Wickham |
⬤⬤⬤ CON x3 |
2634 |
46.2 |
| OVERALL |
|
55127 |
51.4 |
| Party |
Votes |
Vote % |
Seats |
Seat % |
| Labour |
37835 |
51.5 |
51 |
85 |
| Conservative |
20270 |
27.6 |
9 |
15 |
| Liberal Democrats |
7208 |
9.8 |
0 |
0 |
| Green Party |
6344 |
8.6 |
0 |
0 |
| Independents |
1288 |
1.8 |
0 |
0 |
| Others |
455 |
0.6 |
0 |
0 |
| TOTAL |
73400 |
100 |
60 |
100 |
In Hounslow, the winning party, Labour, received a similar seat share (85%) as the Conservatives did in Bromley, this time with just over half the vote share.
The ward level data is also almost a mirror image of Bromley. Again, every multi-member ward is represented by only one party, with 46.8% of votes going to parties that did not succeed in having a representative elected in a particular ward.
| WARD |
COUNCILLORS ELECTED |
VOTES FOR BEST PLACED CANDIDATE OF NON-REPRESENTED PARTIES |
VOTES FOR BEST PLACED CANDIDATE OF NON-REPRESENTED PARTIES (%) |
| Bedfont |
⬤⬤⬤ LAB x3 |
1730 |
51.1 |
| Brentford |
⬤⬤⬤ LAB x3 |
1982 |
46.1 |
| Chiswick Homefields |
⬤⬤⬤ CON x3 |
2327 |
57.3 |
| Chiswick Riverside |
⬤⬤⬤ CON x3 |
2463 |
59.1 |
| Cranford |
⬤⬤⬤ LAB x3 |
1105 |
34.3 |
| Feltham North |
⬤⬤⬤ LAB x3 |
1872 |
56.1 |
| Feltham West |
⬤⬤⬤ LAB x3 |
1935 |
51.0 |
| Hanworth Park |
⬤⬤⬤ LAB x3 |
2177 |
62.3 |
| Hanworth |
⬤⬤⬤ LAB x3 |
1547 |
50.4 |
| Heston Central |
⬤⬤⬤ LAB x3 |
1077 |
34.5 |
| Heston East |
⬤⬤⬤ LAB x3 |
1046 |
34.9 |
| Heston West |
⬤⬤⬤ LAB x3 |
874 |
26.0 |
| Hounslow Central |
⬤⬤⬤ LAB x3 |
1629 |
40.4 |
| Hounslow Heath |
⬤⬤⬤ LAB x3 |
1414 |
36.8 |
| Hounslow South |
⬤⬤⬤ LAB x3 |
1936 |
47.1 |
| Hounslow West |
⬤⬤⬤ LAB x3 |
1275 |
36.6 |
| Isleworth |
⬤⬤⬤ LAB x3 |
1863 |
49.3 |
| Osterley & Spring Grove |
⬤⬤⬤ LAB x3 |
1998 |
47.6 |
| Syon |
⬤⬤⬤ LAB x3 |
1962 |
50.0 |
| Turnham Green |
⬤⬤⬤ CON x3 |
2111 |
56.7 |
| OVERALL |
|
34323 |
46.8 |
| Party |
Votes |
Vote % |
Seats |
Seats % |
| Labour |
6949 |
11.7 |
0 |
0.0 |
| Conservative |
17478 |
29.5 |
9 |
18.8 |
| Liberal Democrats |
28989 |
48.9 |
39 |
81.3 |
| Green Party |
2641 |
4.5 |
0 |
0.0 |
| Independent |
2863 |
4.8 |
0 |
0.0 |
| UKIP |
267 |
0.5 |
0 |
0.0 |
| Others |
82 |
0.1 |
0 |
0.0 |
| TOTAL |
59269 |
100.0 |
48 |
|
Kingston upon Thames is another example of one party getting over four-fifths of the seats, based on around half of the votes. In this borough it is the Liberal Democrats that benefit.
At ward level, although there are two wards with split representation, the vast majority of these multi-member wards are again represented by one party, meaning that over forty percent of votes went to parties who did not secure representation in a ward.
Ward level results
| WARD |
COUNCILLORS ELECTED |
VOTES FOR BEST PLACED CANDIDATE OF NON-REPRESENTED PARTIES |
VOTES FOR BEST PLACED CANDIDATE OF NON-REPRESENTED PARTIES (%) |
| Alexandra |
⬤⬤⬤ LIB DEM x3 |
1640 |
45.0 |
| Berrylands |
⬤⬤⬤ LIB DEM x3 |
1767 |
46.2 |
| Beverley |
⬤⬤⬤ LIB DEM x3 |
2139 |
51.9 |
| Canbury |
⬤⬤⬤ LIB DEM x3 |
2847 |
56.1 |
| Chessington North & Hook |
⬤⬤⬤ LIB DEM x3 |
1554 |
47.6 |
| Chessington South |
⬤⬤⬤ LIB DEM x3 |
1719 |
47.0 |
| Coombe Hill |
⬤⬤⬤ CON x3 |
1686 |
53.7 |
| Coombe Vale |
⬤⬤⬤ LIB DEM x2; CON x1 |
809 |
18.6 |
| Grove |
⬤⬤⬤ LIB DEM x3 |
1546 |
43.5 |
| Norbiton |
⬤⬤⬤ LIB DEM x3 |
1523 |
44.4 |
| Old Malden |
⬤⬤⬤ CON x3 |
1912 |
57.9 |
| St James |
⬤⬤⬤ LIB DEM x3 |
1926 |
52.8 |
| St Mark’s |
⬤⬤⬤ LIB DEM x3 |
1382 |
41.7 |
| Surbiton Hill |
⬤⬤⬤ LIB DEM x3 |
1504 |
39.7 |
| Tolworth & Hook Rise |
⬤⬤⬤ LIB DEM x3 |
1319 |
38.8 |
| Tudor |
⬤⬤⬤ CON x2; LIB DEM x1 |
694 |
18.5 |
| OVERALL |
|
25967 |
43.8 |
First preference votes
| Party |
Votes |
Vote % |
Seats |
Seat % |
| SNP |
49798 |
27.0 |
19 |
30.2 |
| Conservative |
51212 |
27.7 |
18 |
28.6 |
| Labour |
33916 |
18.4 |
12 |
19.0 |
| Green Party |
22907 |
12.4 |
8 |
12.7 |
| Liberal Democrats |
25154 |
13.6 |
6 |
9.5 |
| Independents |
1239 |
0.7 |
0 |
0.0 |
| Others |
401 |
0.2 |
0 |
0.0 |
| TOTAL |
184627 |
100.0 |
63 |
100.0 |
The data for Edinburgh, Scotland’s capital city, are in stark contrast to the three London borough examples. These data are from the last Scottish council elections in 2017 and the chart above shows a highly proportional result, with first preference vote shares closely matching seat share. As described above, STV gives voters the opportunity to select multiple preferences, so first preference vote share is not the only consideration when looking at the ‘fairness’ of an election outcome under STV. This at least partly explains why parties can sometimes get slightly more seats than another party while having slightly fewer first preference votes, as happened twice in Edinburgh in 2017. However, the fact that first preference vote share and seat share match so closely indicates a far more representative outcome for voters than we saw in London.
The Edinburgh ward level data again contrasts with the London examples. In no ward were all of the seats taken by one party, in fact in some wards four different parties saw candidates elected. The overall percentage of first preference votes for parties with no representation in a ward was 15.7%, much lower than the percentage of votes for the ‘top candidate’ of parties who failed to get elected in particular wards, in the three London boroughs.
| WARD |
COUNCILLORS ELECTED |
FIRST PREFERENCE VOTES FOR CANDIDATES OF NON-REPRESENTED PARTIES |
FIRST PREFERENCE VOTES FOR CANDIDATES OF NON-REPRESENTED PARTIES (%) |
| Almond |
⬤⬤⬤⬤ SNP; CON; LD x2 |
1384 |
9.7 |
| Pentlands Hill |
⬤⬤⬤⬤ SNP; CON x2; LAB |
1049 |
9.3 |
| Drum Brae/Gyle |
⬤⬤⬤ SNP; CON; LD |
1673 |
17.7 |
| Forth |
⬤⬤⬤⬤ SNP x2; CON; LAB; |
2036 |
19.6 |
| Inverleith |
⬤⬤⬤⬤ SNP; CON x2; LD |
3736 |
27.5 |
| Corstorphine/Murrayfield |
⬤⬤⬤ SNP; CON; LD |
1634 |
14.3 |
| Sighthill/Gorgie |
⬤⬤⬤⬤ SNP x2; CON; LAB |
1623 |
19.1 |
| Colinton/Fairmilehead |
⬤⬤⬤ CON x2; LAB |
3374 |
29.7 |
| Fountainbridge/Craiglockhart |
⬤⬤⬤ SNP; CON; GREEN |
1789 |
19.6 |
| Morningside |
⬤⬤⬤⬤ CON; LAB; GREEN; LD |
2427 |
17.9 |
| City Centre |
⬤⬤⬤⬤ SNP; CON; LAB; GREEN |
754 |
8.5 |
| Leith Walk |
⬤⬤⬤⬤ SNP x2; LAB; GREEN |
2487 |
23.4 |
| Leith |
⬤⬤⬤ SNP; LAB; GREEN |
1616 |
21.2 |
| Craigentinny/Duddingston |
⬤⬤⬤⬤ SNP; CON; LAB; GREEN |
448 |
4.2 |
| Southside/Newington |
⬤⬤⬤⬤ SNP; CON; LAB; GREEN |
1410 |
12.1 |
| Liberton/Gilmerton |
⬤⬤⬤⬤ SNP x2; CON; LAB |
1254 |
10.7 |
| Portobello/Graigmillar |
⬤⬤⬤⬤ SNP; CON; LAB; GREEN |
258 |
2.5 |
| OVERALL |
|
28952 |
15.7 |
[i] For multi-member London ward and overall council results, we have calculated vote share by using the vote of each party’s best-placed candidate. This is the approach taken by local election experts Colin Rallings and Michael Thrasher, directors of the Election Centre, a major resource for local election data in the UK.