This month’s elections for mayors and police and crime commissioners were contested under a revised voting system. In a post published yesterday, Alan Renwick found that this change had a substantial impact on the results, to the benefit of the Conservatives. Here, he concludes that it also harmed democracy. 

Elections of mayors and police and crime commissioners (PCCs) were previously held under the Supplementary Vote (SV) system, where each voter could express first and second preferences. Now they take place using First Past the Post (FPTP), where there is a vote for a single candidate. The previous post in this series showed that this change produced a marked shift in the outcome of the elections held earlier this month, and that it did so entirely to the benefit of the Conservatives.  

That a change in the rules should favour those in power who instigated it is already cause for concern: democracy requires a level playing field. But ministers might defend the reform on the basis that the new system is superior on democratic grounds to its predecessor and that it was introduced fairly. Both of these claims therefore require interrogation. How do the two systems compare in terms of democratic quality? And was the process through which the change in voting system came about appropriate?  

Which voting system is more democratic? 

As I outlined in a blogpost published when the bill changing the voting system was before parliament in 2021, ministers argued that FPTP is the more democratic system: SV, they said, allows losing candidates – those coming second in terms of first preferences – to win. But this argument is circular: it works only if we have already accepted the FPTP definitions of ‘winner’ and ‘loser’.  

The underlying logic of ministers’ case was that second preferences should not count: that only first preferences should have value. Their press release in 2021 quoted Winston Churchill, who apparently said in 1931 that, under transferable vote systems such as SV, ‘the decision is to be determined by the most worthless votes given for the most worthless candidates’. The mere fact that something was said by Winston Churchill does not, however, make it right. In fact, no reasonable argument can support this view. Preferences are first or second (or whatever) only among a given set of candidates; with the addition or subtraction of a candidate, a first preference may become a second, a second a first, and so on. A Reform supporter’s preference for Conservative over Labour is no more or less real depending on whether they happen to have a Reform candidate running in their area. It is, of course, possible that a voter may not have strong preferences between some candidates. But that could just as easily mean that they do not have a strong first preference as that they lack a clear second preference.  

Another claim – familiar to anyone who remembers the referendum in 2011 on introducing SV’s close cousin the Alternative Vote (AV) system for Westminster elections – is that such systems give some voters a second vote by counting their second preferences. This claim is simply wrong. Under such systems, each vote is tallied once in each count – either for the same candidate if that candidate remains in the race, or for a different candidate if the first is eliminated. 

Pointing to the 2011 referendum, ministers also defended the reform when it was introduced on the basis that ‘transferable voting systems were rejected by the British people’ in that vote. AV was indeed then defeated for elections to the House of Commons by an overwhelming margin of 68% to 32%. But there are legitimate arguments for rejecting AV when electing a collective body such as the House of Commons while choosing it when electing single executive offices such as mayors or PCCs. In particular, in a parliamentary election, AV can exaggerate a landslide victory even further than does FPTP. Current ministers’ predecessors apparently understood that when they introduced the SV voting system for PCC elections in the months after the 2011 referendum.  

The most important consideration in assessing whether SV or FPTP is the more democratic system is simply the question of which leads to outcomes that better reflect voters’ preferences. Given all the evidence from surveys and election results, as explored in the first part of this post, it is overwhelmingly likely that, in multiple PCC areas that were won by the Conservatives in the recent elections under FPTP, most voters with a view in fact preferred the Labour (or in some cases Lib Dem) candidate over the Conservative candidate. The democratically better voting system would have allowed them to express that preference and it would have produced that outcome. In terms of which system yields the more democratic translation of valid votes into seats in each area, therefore, SV is unambiguously superior to FPTP. 

Spoilt ballots 

One further point concerning the democratic quality of the system does deserve more detailed consideration. Ministers argued in 2021 that SV is complex for voters, leading to more spoilt ballots. Is that true, and, if so, how far should it change our overall assessment? 

Spoilt ballots numbers have indeed been higher in elections using SV than under FPTP. In PCC elections, the proportion was 2.8% in 2012, 3.4% in 2016, and 2.7% in 2021. It reached an exceptional 4.3% in the 2021 London mayoral election – though it was only 1.9% in other mayoral contests on the same day. In local council elections using FPTP, by contrast, these proportions were 0.6% in 2016 and 0.8% in 2021. 

Until this month, it was hard to know how far these inflated spoilt ballot numbers were caused by SV. A role for other factors was certainly plausible. PCC elections receive minimal attention. When they are run concurrently with other elections, therefore, some voters arriving at the polling station to vote for a councillor or mayor may just cast their PCC ballot unmarked. In London in 2021, meanwhile, there were particular problems with a poorly designed ballot paper.  

Now, however, we can directly compare the results of mayoral and PCC elections under SV with those under FPTP. In this month’s elections, the proportion of spoilt ballots across the mayoral elections fell to 0.7%; in the PCC elections, it fell to 1.6%. The former is in line with other UK elections. The latter remains somewhat higher, suggesting that some voters indeed simply disregard these contests. Overall (and excluding the London 2021 anomaly), the number of spoilt ballots was just over a percentage point down on previous elections. Other factors may have been involved, but the switch from SV to FPTP is much the likeliest explanation for the drop. 

That reduction in spoilt ballots is a clear benefit from the change in voting systems, and it should not be dismissed as insignificant. At the same time, ministers cannot trumpet it as a gain without acknowledging that the introduction of voter ID requirements may have generated a larger drop in effective participation. Furthermore, if the ballot papers being used for SV were confusing some voters, the first step should have been to research better ballot paper designs, not to ditch a system that in other respects was superior. 

Most importantly, the gain from fewer spoilt ballots under SV needs to be balanced against the loss in the democratic quality of the elections already described. Unintended spoiling of ballots is unlikely to have changed the result in any of the elections discussed here, whereas the switch in counting rules changed the result in multiple cases. Further, the proportion of voters who, having backed a minor candidate, were unable to influence the final choice, at 26.6% in the PPC elections, far exceeded the proportion inadvertently casting an invalid ballot. Being denied a valid vote at all may seem more serious than being denied a second preference. But PCC elections are extremely low-information contests. Voters in Thames Valley (say) cannot be expected to work out for themselves whether the Labour or the Lib Dem candidate has a better chance of winning, and they will receive no help in the matter from campaigners or the media. FPTP requires many voters to make tactical calculations in order to express their true preferences; but the information required for those calculations is often unavailable. Why being denied an effective voice by this form of uncertainty should matter so much less than being denied it by confusion over the ballot paper is not obvious.  

In sum, SV is clearly more democratic than FPTP for single executive offices. All the major parties – including the Conservatives – seem to recognise that by using transferable vote systems (or closely related multi-round systems) to elect their leaders. The change to FPTP harmed democracy.  

How was the reform introduced? 

What, then, of the process by which the reform was introduced? The rules of the political game should never be fixed simply for the advantage of the player currently in power. The benefit that the Conservatives derived from the change, combined with the absence of good democratic arguments for the change, gives reason to doubt that this principle was upheld. But what of the process itself?  

In brief, it fell far short of any reasonable standard that we ought to accept. When it legislated to give the Scottish Parliament and the Senedd control over the voting systems by which these chambers are themselves elected (in, respectively, 2016 and 2017), Westminster stipulated that any reforms would require a two-thirds majority of members. Given these parliaments’ proportional systems at present, this meant that changes would always need cross-party support. Yet ministers made no attempt to seek any such agreement for the reforms to mayoral or PCC elections, which were opposed by all parties besides the Conservatives and the DUP.  

More than that, the change was made without even the normal level of scrutiny for legislation at Westminster. When the Elections Bill was introduced to parliament in July 2021, it included no mention of the proposal. The measure was proposed by ministers as an amendment, with no prior consultation on it having occurred. Formal evidence sessions with witnesses had by then already concluded; questioning of witnesses on the matter was prevented, and ministers refused to reopen evidence taking. The government simply used its Commons majority to force the measure through. The principle that changes to the electoral system should be treated with particular sensitivity was utterly ignored. 

What should happen now? 

The switch from SV to FPTP has therefore harmed democracy both through its effects and through the manner of its imposition by the ruling party. It should be reversed.  

Still, care is needed. Repeating the current government’s error by introducing a further change without proper scrutiny and attempts to build wide agreement would be wrong. Following the general election in July, any new government should therefore engage with a wide range of views before proceeding. It should ensure that full parliamentary scrutiny can take place. And it should seek support from as broad a set of parties as possible. Doing so would increase the legitimacy of the change and make it less likely that the rules could be switched again in the future. 

This is the second of a two-part series. Part one was published yesterday. As the general election approaches, this blog will continue to publish posts regularly. Sign up via the sidebar on the left side of the screen to be notified when new posts go live.

About the author

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

Featured image: Keir Starmer and Victoria Starmer voting (CC BY-NC-ND 2.0) by keir.starmer.mp.