Electoral Systems: FPTP, AMS, STV and CPLS

First-past-the-post (FPTP) divides the country into single-member constituencies. Each voter marks one cross for one candidate, and the candidate with more votes than any rival takes the seat. Nothing else is required — not half the vote, not a minimum turnout, not the approval of anyone who voted for someone else.

That winning condition is called a plurality. It is why FPTP is described as a plurality system rather than a majoritarian one.

Worked example — a four-way marginal. Suppose a constituency casts 45,000 valid votes:

Candidate	Party	Votes	Share
A	Labour	15,300	34.0%
B	Reform UK	13,500	30.0%
C	Conservative	9,000	20.0%
D	Liberal Democrat	4,500	10.0%
E	Green	2,700	6.0%

Candidate A is elected on 34.0% of the vote. Two conclusions follow, and both are worth marks:

• 66.0% of voters chose somebody else. The seat's MP was opposed by two-thirds of those who turned out.
• The majority is 1,800 votes, or 4.0%. A uniform swing of just 2.0% from A to B would flip the seat, which makes this a marginal. Where a party's lead is large enough to survive ordinary swings, the constituency is a safe seat instead.

On the Electoral Reform Society's definition, a vote is wasted if it went to a losing candidate or exceeded what the winner needed. Here that is 29,700 losing votes plus A's surplus of 1,799 above the 13,501 needed to win — 31,499 votes, or 70.0% of everything cast.

Applied 650 times over, the plurality rule produces a national result that can look very little like the national vote. Professor Sir John Curtice described the 2024 general election as the most disproportional outcome in British electoral history.

Party	Votes (%)	Seats	Seats (%)	Difference
Labour	33.7	411	63.2	+29.5
Conservative	23.7	121	18.6	−5.1
Liberal Democrat	12.2	72	11.1	−1.1
Reform UK	14.3	5	0.8	−13.5
Green	6.4	4	0.6	−5.8
SNP	2.5	9	1.4	−1.1

Labour converted a third of the vote into nearly two-thirds of the seats and a Commons majority of 174, on a turnout of 59.7%. Reform UK took a larger share of the national vote than the Liberal Democrats — 14.3% against 12.2% — and finished with 5 MPs to their 72.

The mechanism is vote distribution, not size. The Liberal Democrats' support is concentrated in identifiable regions, so it converts into seats; Reform UK's was spread thinly across hundreds of constituencies, coming second in 98 of them and first in five. FPTP rewards geographically clustered support and punishes evenly spread support.

The single most useful sentence you can write about FPTP is that it counts seats, not votes. Every advantage and disadvantage of the system follows from that.

The Additional Member System (AMS) is a hybrid. It keeps FPTP constituencies and then bolts a corrective list tier on top. Scotland's 129-member Parliament has used it since 1999: 73 constituency MSPs elected by plurality, plus 56 regional MSPs returned from eight regions of seven.

Each voter gets two ballot papers:

1. A constituency vote for an individual candidate, decided by FPTP.
2. A regional vote for a party list.

The list seats are then allocated by the D'Hondt method, but with a twist that does the corrective work: a party's regional vote is divided by the number of seats it already holds in that region — constituency seats plus any list seats won in earlier rounds — plus one. A party that swept the constituencies starts with a large divisor and is pushed down the queue.

Worked example — a simplified region with 6 constituency seats and 4 list seats. Party P swept 5 of the constituencies, Q won 1, and R won none. The regional list votes were P 60,000, Q 45,000, R 32,000.

Round	P (÷ seats + 1)	Q	R	Seat awarded to
1	60,000 ÷ 6 = 10,000	45,000 ÷ 2 = 22,500	32,000 ÷ 1 = 32,000	R
2	60,000 ÷ 6 = 10,000	45,000 ÷ 2 = 22,500	32,000 ÷ 2 = 16,000	Q
3	60,000 ÷ 6 = 10,000	45,000 ÷ 3 = 15,000	32,000 ÷ 2 = 16,000	R
4	60,000 ÷ 6 = 10,000	45,000 ÷ 3 = 15,000	32,000 ÷ 3 = 10,667	Q

Final tally: P 5 seats, Q 3, R 2. On 43.8%, 32.8% and 23.4% of the regional vote, the parties end with 50%, 30% and 20% of the ten seats. P won no list seats at all — its constituency success had already used up its entitlement.

At the Scottish Parliament election of 7 May 2026 the six largest parties under AMS returned SNP 57, Labour 17, Reform UK 17, Green 15, Conservative 12, Liberal Democrat 10. The SNP, on 57, fell well short of the 65 seats needed for a majority in a 129-seat chamber.

STV is used for the Northern Ireland Assembly (90 members, 18 constituencies returning five each) and for Northern Irish and Scottish local council elections. Voters rank candidates 1, 2, 3… across and within parties, in any order they like.

A candidate is elected on reaching the Droop quota:

$\text{quota}=\lfloor \frac{\text{total valid votes}}{\text{seats}+1}\rfloor +1$

Worked example — a five-seat constituency with 30,000 valid votes.

$\text{quota}=\lfloor \frac{30,000}{6}\rfloor +1=5,000+1=5,001$

Candidate	Count 1	Murphy's surplus	Count 2
Murphy	7,200 — elected	quota retained	—
Wilson	4,600	+250	4,850
Adams	3,900	+1,200	5,100 — elected
Bell	3,600	+149	3,749
Kelly	3,400	+600	4,000
Doyle	3,300	—	3,300
Four others	4,000 between them	—	4,000

Murphy clears the quota on the first count with a surplus of $7,200-5,001=2,199$. Every one of Murphy's ballots is re-examined for its next available preference, and the 2,199 surplus votes are distributed in proportion to those preferences. That is enough to lift Adams over the line.

Once no continuing candidate holds a quota, the lowest-placed candidate is eliminated and their ballots transfer at full value to the next preference marked. Counting continues until five candidates hold quotas, or until only five remain in the race.

STV's distinctive feature is choice within a party. A voter can rank one Sinn Féin candidate first and another fourth, which turns a party-versus-party election into a candidate-versus-candidate one as well.

The Closed Proportional List System (CPLS) is the newest system on the specification, and Wales is the reason it is there.

On 7 May 2026 the Senedd was elected under a wholly new arrangement. It expanded from 60 to 96 members, elected from 16 constituencies returning six members each, formed by pairing Westminster seats. Each voter casts one vote, for a party, and seats within each constituency are allocated by D'Hondt. AMS, used in Wales from 1999, was dropped.

"Closed" is the operative word. The party sets the order of names on its list before polling day, and voters cannot alter it. If a party wins three seats in a constituency, the first three names on its list are returned. Independent candidates may also stand.

Worked example — a six-member constituency casting 60,000 votes. Divide each party's total by 1, 2, 3… and award the six seats to the six largest quotients.

Divisor	Party A (25,000)	Party B (17,000)	Party C (11,000)	Party D (7,000)
÷ 1	25,000 ①	17,000 ②	11,000 ④	7,000
÷ 2	12,500 ③	8,500 ⑤	5,500	3,500
÷ 3	8,333 ⑥	5,667	3,667	2,333
÷ 4	6,250	4,250	2,750	1,750

The six highest quotients are 25,000, 17,000, 12,500, 11,000, 8,500 and 8,333, so the seats split A 3, B 2, C 1, D 0. Party D, on 11.7% of the vote, wins nothing: with only six seats to share, a party needs roughly a seventh of the vote to be confident of one, which is the effective threshold created by the district size.

This is the comparison the specification names explicitly: FPTP against a different system used in a devolved legislature. As of September 2026, the sharpest available version of it is Westminster 2024 against the Senedd 2026.

Plaid Cymru won 43 of the 96 seats on 35.4% of the vote; Reform UK took 34 on 29.3%; Labour 9 on 11.1%; the Conservatives 7 on 10.7%; the Greens 2 on 6.7%; the Liberal Democrats 1 on 4.5%. Turnout was 51.6%, the highest at any Senedd election and the first above half. Labour left government in Wales for the first time since 1999.

Measure	Westminster 2024 (FPTP)	Senedd 2026 (CPLS)
Largest party's votes → seats	33.7% → 63.2%	35.4% → 44.8%
Biggest over-reward	Labour, +29.5 points	Plaid Cymru, +9.4 points
Biggest under-reward	Reform UK, −13.5 points	Green, −4.6 points
Outcome	Single-party majority of 174	No overall majority; Plaid Cymru six seats short of the 49 needed
Constituency link	One named MP per seat	Six members per constituency, no personal vote

The chart makes the headline point: under CPLS the two bars track each other for every party, while under FPTP they diverge violently. But CPLS in Wales is proportional rather than perfectly proportional, because six-member districts still leave small parties short — the Greens polled 6.7% and took 2.1% of the seats.

	FPTP	AMS	STV	CPLS
Proportionality	Poor	Good	Good	Good, limited by district size
Constituency link	Strong, one named member	Partial: constituency and list members	Shared between several members	Absent
Choice between candidates of one party	None	None	Full	None
Likely government	Single-party majority	Coalition or minority	Power-sharing coalition	Coalition or minority
Ballot complexity	Lowest	Two ballots to understand	Ranking, quotas, transfers	Lowest
Who controls candidate selection	Local party, then voters	Party for the list tier	Voters	Party alone

The recurring trade-off is fair representation against effective government. FPTP buys decisive single-party government at the cost of a distorted result. The three proportional systems buy an accurate result at the cost of coalition bargaining after polling day rather than before it. The Financial Times put the same point the other way round on 8 May 2026: FPTP forces parties to build their coalitions internally, before an election, which is why the UK has developed large broad-church parties.

A second trade-off runs alongside it. STV maximises voter choice but produces the most complicated count; CPLS produces a clean proportional result but hands candidate selection entirely to party managers. Neither is straightforwardly more democratic than the other, and saying which you think matters more is where AO3 marks are earned.

1. Writing about the Supplementary Vote

Issue 4 of the specification, published in May 2026, replaced SV with the Closed Proportional List System in Paper 1, 3.1. Textbooks and past papers written before that date still cover SV at length. Time spent on it is time not spent on assessable content.

2. Calling FPTP a majoritarian system

FPTP requires a plurality, not a majority. In the worked example above the winner took 34.0%. Getting this distinction right in the first line of an answer signals precision immediately.

3. Saying proportional systems are simply fairer

Proportionality is one criterion among several. A system is also judged on the strength of the constituency link, the stability of the government it produces, the clarity of accountability, and how much control voters have over which individuals are elected. CPLS scores highly on proportionality, has no constituency link at all, and gives voters no say over which individuals are elected.

4. Mixing up the two D'Hondt applications

Under AMS the divisor counts constituency seats already won in that region as well as any list seats won so far, plus one, which is what makes the top-up corrective. Under CPLS the divisor counts list seats already won, plus one, because there are no constituency seats to count. Same arithmetic, different starting point.

5. Using devolved examples without dating them

Wales changed system in 2026, so "the Senedd uses AMS" is now wrong. Scotland still uses AMS and Northern Ireland still uses STV. Attach a date — "at the Senedd election of 7 May 2026" — and the examiner can see the knowledge is current.

6. Describing a system without evaluating it

Paper 1 questions on this content area are worth 30 marks and carry AO3 credit for judgement. Explaining how STV counts votes is AO1. Saying whether its complexity is a price worth paying for the choice it gives voters, and why, is what the AO3 marks are for.