3.1 - Closed Proportional List System

Party-ranked closed lists

A closed list asks voters to choose between party lists rather than rearrange the candidates within one list. Parties announce their candidates in a ranked order before the election. If a party wins three seats, its first three eligible candidates are elected; the voter cannot move its fourth candidate above its first.

“Closed” describes candidate order. “Proportional” describes the aim of translating party support into seats. These are separate properties. A list system can give several parties representation while leaving considerable power over personnel with party selectors. Independent candidates can stand; a vote for an independent supports that individual rather than a multi-candidate party list.

For example, imagine Party Willow's list is Maya, Noor, Owen and Pia. Winning two seats elects Maya and Noor, not whichever two names voters personally like best. List ranking can help a party organise a balanced team, but the composition depends on its nomination decisions rather than a voter's preference between those candidates.

From the May 2026 election, the Senedd uses closed lists in 16 constituencies with six seats each, making 96 members. Voters have one vote for a party or independent. This replaces the former AMS combination of constituency and regional votes. The Electoral Commission's 2026 Senedd guide supplies the rule and date; it does not demonstrate how every voter judged the change.

If you sit in 2027: Pearson Issue 3 lists Supplementary Vote (SV) in 3.1 and its glossary. Issue 4 replaces it with CPLS, first assessed in June 2028. The two SV parts below provide your required operation and evaluation. CPLS is the canonical course teaching for 2028 onwards. Use clearly series-labelled practice that matches your exam year when a question requires only one issue's system. Pearson amendment notice.

Allocating seats from party votes

A proportional allocation divides whole seats among competitors; it does not normally multiply each vote percentage by the number of seats and round every party independently. That could allocate too many or too few seats. Different list systems use different rules. The Senedd uses d'Hondt, which repeatedly awards the next seat to the highest quotient.

For a list allocation without an AMS constituency tier, divide each party's vote total by one plus the number of list seats it has already won in the constituency. All parties start with zero seats. Recalculate after each award. This differs from Scottish AMS, where constituency seats are already included in the divisor. Electoral Commission seat-allocation guidance, updated 26 March 2026.

In an invented six-seat constituency, A receives 43,000 votes, B 31,000, C 19,000 and D 7,000. Each has at least six candidates available, and none of the winning quotients ties. The table shows the quotients immediately before each award.

A wins three seats, B two, C one and D none. The first three candidates on A's list are elected, along with the first two on B's and the first on C's. A's 43% of votes becomes 50% of seats; D's 7% earns none. Six indivisible seats cannot exactly reproduce every percentage. A larger number of seats can allow a finer-grained result, while d'Hondt can favour larger lists relative to some alternative proportional formulas.

This is a formative illustration of allocation and its political consequences, not an additional official calculation family. In actual Senedd counting, an independent drops out after winning their one seat, and a party with no unelected listed candidate left cannot receive further seats. Electoral Commission count guidance.

Party fairness and candidate control

Closed lists can improve the representation of party support while weakening a different kind of voter control. Several parties can win seats in an area, including parties unlikely to finish first in a single-member constituency. Fewer votes may fail to contribute to representation. However, proportionality is approximate, and smaller parties may still struggle in constituencies electing only a few members, as D does in the worked example.

The ballot can be simple: one choice rather than a long candidate ranking. Parties can offer an identifiable programme and use list selection to widen the range of people elected. Yet the voter must accept the ordering. A candidate highly placed on a successful list can depend more directly on party selectors than on voters' judgement of that individual. The system does not prevent accountability: voters can reject the party, and elected members still answer for their actions. It changes where control over personnel lies.

Multi-member representation offers several people to contact, but shared responsibility may make accountability less direct. More proportional party representation can also make a single-party seat majority less likely. A coalition may represent a wider range of voters but require negotiations that alter promises. A coalition government means parties share government, normally through a policy and ministerial agreement. A party merely voting to support another party's minority government is not necessarily a coalition partner.

Consider this original teaching source: “A legislature should reflect the different parties voters support. Equally, citizens should be able to judge individual representatives, and a government should be answerable for its decisions.” These are three democratic criteria, not proof that any one system fulfils all three. The worked allocation supports a claim about party representation; fixed ordering raises the candidate-choice concern. Neither tells us how well a future government will govern. A fair argument identifies the criterion, explains the relevant mechanism, and weighs its limits.

If you sit in 2027: Supplementary Vote operation

Supplementary Vote elects one office holder using a first preference and a possible second preference. It is a majoritarian system, designed to widen support for the winner, rather than a proportional method for distributing several legislative seats. A voter marks one first choice and may mark a different second choice.

If a candidate receives more than 50% of valid first preferences, that candidate wins. Otherwise the two leading candidates remain and the others are eliminated, subject to the election's tie rules. Examine the ballots whose first preference was for an eliminated candidate. Their second preference transfers only if it names one of the two continuing candidates. Add those valid transfers to the finalists' first-preference totals; the finalist with the larger final total wins. The second preferences of voters whose first choice already reached the final two are not added again.

In an invented election, A has 40 first preferences, B 35 and C 25, from 100 valid ballots. Nobody exceeds 50. C is eliminated. Of C's ballots, five second preferences go to A, fifteen to B, and five have no valid second preference for either finalist. A finishes with 45 and B with 50; B wins. Ninety-five ballots contribute to the final comparison. B has more than half of that final total, but not more than half of the original 100. Exhausted or unusable second preferences explain the distinction.

SV is not STV. It elects one winner, records at most two preferences and eliminates everyone outside the top two in one step. There is no multi-member quota or transfer of an elected candidate's surplus. It is also not Alternative Vote, which can continue eliminating the lowest candidate and transferring fuller rankings over several rounds.

In the 2021 London mayoral election, Sadiq Khan's first-preference total was 1,013,721 and Shaun Bailey's 893,051. After the valid second preferences were added, their totals were 1,206,034 and 977,601 respectively. These figures demonstrate the two stages, rather than a direct election by first preferences alone. London Assembly election review, January 2022, Appendix 2.

If you sit in 2027: SV merits, limits and changes

SV lets a voter support a preferred candidate and express a fallback, which can broaden the winning candidate's support compared with a single plurality contest. A voter whose first choice is unlikely to win need not immediately abandon that candidate. A contender may seek second preferences from other parties' supporters, encouraging appeal beyond a narrow first-choice base. For a single executive office, this can strengthen the political claim to legitimacy without pretending to represent every party proportionally.

The fallback is limited. A voter must anticipate which candidates will reach the final two if they want their second preference to remain usable. A second choice for a third-placed candidate cannot help once that candidate is eliminated. In the worked count, five ballots drop out of the final comparison; the winner therefore does not have a majority of all original votes. The system is simple compared with a full preferential count but more complex than choosing one candidate, and mistakes or unmarked second choices can reduce transfers.

An original teaching source makes the dispute clear: “A mayor should gain support beyond a narrow first-choice lead. Two preferences can encourage that. But asking voters to predict the top two restricts genuine choice, and a majority of the surviving ballots is not the same as a majority of everyone who voted.” The first argument values breadth of consent; the second qualifies both choice and the mandate claim. The correct response is to explain these mechanisms, not assume that either preference system guarantees universal approval.

Dated legal update, checked 14 October 2026: the Elections Act 2022 replaced SV for mayoral and police-and-crime-commissioner elections with FPTP, used from May 2023 to May 2026. The English Devolution and Community Empowerment Act 2026 reintroduced SV, with implementation through election conduct rules. The Commons Library's 7 October 2026 briefing records this reversal; the 21 May 2026 ministerial statement distinguishes the Act from the steps for implementation. Consequently, Issue 3's glossary reference to SV ending in 2022 is historical, not a permanent description of current law. Its 2027 examination requirement remains SV.

The contrast with CPLS concerns electoral purpose as well as preferences. SV chooses one executive winner with potentially wider consent. CPLS distributes several legislative seats among party lists. An argument that one is “more democratic” needs to state whether it values a mandate for one office holder, proportional party representation or control over individual candidates. It cannot apply the same seat-sharing test to a single-office election without explaining the different purpose.