Reforming the mayoral voting system: do ministers’ arguments stack up?

This week, the Elections Bill was amended to replace the Supplementary Vote (SV) system used for electing mayors and police and crime commissioners (PCCs) in England and Wales with First Past the Post (FPTP). But, as Alan Renwick and Alejandro Castillo-Powell argue, the arguments put forth by ministers are not as convincing as they might at first appear.

Ministers plan to replace the Supplementary Vote (SV) system used for electing mayors and police and crime commissioners (PCCs) in England and Wales with First Past the Post (FPTP). The most detailed explanation for the change given so far appeared in a press release last month, which gave five arguments for the switch: (1) SV increases the number of spoilt ballots; (2) it allows ‘loser’ candidates to win; (3) FPTP improves accountability by ‘making it easier for voters to express a clear choice’; (4) FPTP ‘is the world’s most widely used electoral system’; and (5) SV is ‘an anomaly’ and ‘out of step with other elections in England’. In the written statement that first trailed the proposals, Home Secretary Priti Patel also said (6) that the change ‘reflects that transferable voting systems were rejected by the British people in the 2011 nationwide referendum’. Some of these arguments were repeated, though in less detail, when the matter was considered at the Committee Stage of the Commons bill’s scrutiny earlier this week.

This post assesses the government’s claims in turn. Some have merit, but important counterarguments are ignored. Voting systems should be treated with care: it is all too easy for those in power to manipulate them to their own advantage. Ministers have not adequately made the case that the change will be good for democracy.

1. Does Supplementary Vote lead to more spoilt ballots

Elections using SV in the UK typically see higher numbers of spoilt ballots than do those using FPTP. Electoral Commission data shows that 0.8% of ballots cast in local council elections in May – under FPTP – were rejected. That compares to over 2% in most elections under SV. In PCC elections, they stood at 2.9% this year, down from 3.4% in 2016. In London mayoral elections, they have ranged between 1.8% in 2012 and 4.3% this year. They have generally been around 2% in other combined authority mayoral elections, peaking at 2.2% in 2018.

That SV elections show higher rates of rejected ballots does not mean that SV itself is necessarily the culprit. The jump in such ballots in this year’s London mayoral election points to another factor: ballot paper design. The Electoral Commission notes the use in that contest of a new, untested design, split over two columns because of the large number of candidates, which voters described as ‘being confusing/complex’. Poor design similarly led to more spoilt ballots in the 2007 Scottish local and parliamentary elections. Another factor may be deliberate spoiling of ballot papers: the Electoral Commission noted anecdotal evidence of this in the 2012 PCC elections.

So SV elections do see more spoilt ballots than FPTP elections, but improved ballot paper design – and clearer guidance for voters – might ameliorate the problem.

2. Does Supplementary Vote allow loser candidates to win?

The government’s second argument is that under SV ‘loser’ candidates can win on second preferences. But that argument is circular – such candidates are ‘losers’ only under the rules of FPTP. Take this year’s North Wales PCC election. On first preferences, the Conservative candidate won 32% of the votes, the Labour candidate 29%, and the Plaid Cymru candidate 28%. With such numbers, it is quite possible that most voters preferred either the Labour or the Plaid candidate over the Conservative – the latter was not the self-evident ‘winner’. In fact, under SV, many Plaid voters expressed a second preference, and two thirds of them chose the Labour candidate, giving him victory. It is not obvious why that was wrong.

To count only first preferences is to give them a meaning that they do not possess. A first preference is just a ranking among the available candidates. Among a different field of candidates, a first preference might have become a second, or a second a first. Voters whose first preference is for one candidate may nevertheless have very clear preferences among others too. To suggest that lower preferences are ‘worthless’ – as the government’s press release did – does not make sense.

The choice between SV and FPTP does not actually affect the result very often. Our analysis suggests that the allocation of second preferences has affected the result in 8% of SV elections since their introduction in 2000. In 2021, there were three: the PCC elections in North Wales (as above) and Dyfed–Powys, and the mayoral election in Cambridgeshire and Peterborough.

3. Does Supplementary Vote harm accountability?

Ministers in their press release last month said that FPTP would improve accountability by ‘making it easier for voters to express a clear choice’, and because it ‘allows voters to kick out the politicians who don’t deliver’.

SV and FPTP both allow voters to remove incumbents they dislike. The differences relate to whether voters can express a clear choice. That takes us back to the two points that have already been explored. First, SV may slightly increase the proportion of spoilt ballots, perhaps by a percentage point. If so, it inhibits some voters from expressing a clear choice. Second, however, FPTP forces many voters to engage in complex tactical voting calculations to decide how to vote. To return to the North Wales example, many voters had a preference for either Labour or Plaid over the Conservative. Under FPTP, they would have had to work out which had the greater chance of success. It is not obvious how forcing voters into such difficult calculations empowers them to ‘express a clear choice’.

So there are two considerations at play here. Unless voters who accidentally spoil their ballots have very different preferences from those who do not, the spoilt ballot consideration can affect the result only in the very tightest of races. But considerations around tactical voting can have much bigger effects. That suggests that accountability is clearer under SV.

4. Is First Past the Post ‘the world’s most widely used electoral system’?

There is no definitive catalogue of subnational electoral systems. At national level, however, we can use International IDEA’s Electoral System Design Database. For national legislative elections, FPTP is the second most common system, used in 28% of countries, behind List Proportional Representation, on 39%.

But the UK uses SV to elect individual executive officeholders, not collective bodies. So the relevant comparison at the national level is with presidential elections. Here, 22 countries – 19% of those with presidential elections – use FPTP. Nowhere uses precisely the SV system. But SV is very closely related to other systems allowing voters to express multiple preferences, including the Two-Round System (TRS) and the Alternative Vote (AV). TRS is used in 85 countries: 75% of those with presidential elections. AV is used in one (Ireland).

Thus, the great majority of countries with elections to executive offices eschew FPTP in favour of a system allowing second preferences to be counted.

5. Is Supplementary Vote an ‘anomaly’ in the UK?

The government’s press release described the use of SV for mayoral and PCC elections as an ‘anomaly’. That is true in the sense that these are the only public elections in England using this system. On the other hand, these are also the only public elections to executive offices. As we elaborate below, it is perfectly defensible to advocate FPTP for elections to collective bodies such as parliaments and councils, but SV for executive offices such as mayors and PCCs.

Indeed, all the main political parties use preferential voting systems to choose their leaders. Labour and the Liberal Democrats use AV. The Conservatives use an Exhaustive Ballot, whittling the candidates down to two before a final run-off. Why such voting systems would be right for these elections but not for public elections is unclear.

6. Does the change reflect voters’ rejection of preferential systems in 2011?

The government’s final argument is that the reform ‘reflects that transferable voting systems were rejected by the British people in the 2011 nationwide referendum’. Voters did indeed vote against introducing AV for elections to the House of Commons, by an overwhelming 68% to 32%.

As already mentioned, however, there may be good reasons for favouring different voting systems to elect different offices. Putting aside the many misleading arguments that were thrown around in 2011, the best argument for FPTP against preferential systems such as AV or SV is that the latter can sometimes give even greater over-representation to the largest party. In 1997, for example, Lib Dem second preferences would probably have added yet more to Labour’s landslide victory. But this argument applies exclusively to elections for collective bodies; it is irrelevant to elections for single executive offices.

For this and other reasons, a simple extrapolation from the 2011 referendum results to public preferences on the present issue is not sound. Indeed, current ministers’ predecessors understood that when they introduced the SV voting system for PCC elections in the months after that referendum.


SV elections in the UK are associated with slightly higher rates of rejected ballots than are FPTP elections. But no other plausible argument for switching to FPTP has been given. Indeed, the case in terms of clear accountability runs the other way. In this circumstance, it would be better to seek improvements to the operation of SV, rather than abandon it.

Furthermore, unilateral adoption of electoral reform by one party is always problematic: the risk is that the party will fix the rules to suit its own interests. Some form of independent review – such as the citizens’ assembly posited by a proposed amendment to the Elections Bill that was defeated at Committee Stage – should always be held first. That should be followed by thorough parliamentary scrutiny – which has also been curtailed in this case, by the late introduction of the proposal through an amendment.

Electoral systems are fundamental parts of democracy and should be treated with great care. Making changes based on flawed arguments can only do harm.

About the authors

Alan Renwick is Professor of Democratic Politics at UCL and Deputy Director of the Constitution Unit.

Alejandro Castillo-Powell is a Research Volunteer at the Unit.