4.1 - Explaining an election since 1997

Votes, seats and the return of Labour government

The general election on 4 July 2024 ended fourteen years of Conservative-led government. Keir Starmer became prime minister after Labour won 411 of 650 seats on the Commons Library's party classification. The Conservatives won 121 seats, the Liberal Democrats 72, Reform UK five and the Green Party of England and Wales four. These are election results, not a prediction or the present composition of Parliament. Commons Library results page, checked 14 October 2026.

Labour's UK vote share was 33.7%, far below the proportion of seats it won. Reform's 14.3% brought five seats, while the Liberal Democrats' 12.2% brought 72. Those different conversions show why the location of votes mattered. Labour could win many constituencies where opposition was divided, while Liberal Democrat support and campaign effort could be concentrated in competitive seats. A national vote total does not show which candidate was first in each constituency.

Labour's large majority created legislative capacity, but the party had to reconcile its promises with public expectations and financial constraints. The Conservative defeat made rebuilding its electoral coalition and leadership major questions for the party. Liberal Democrat gains strengthened a third-party presence; Reform's vote share showed substantial support despite limited seats. These consequences should be distinguished from later developments, which are not part of the election result itself.

Incumbent performance and the demand for change

Voters assessed an incumbent associated with the pandemic, rising prices, pressure on public services and several changes of prime minister. Governing competency concerns whether a party is seen as capable of managing such problems. Being blamed for a problem and having caused all of it are different claims: external shocks may constrain any government, but voters can still judge its response inadequate.

The cost of living and the NHS were prominent public concerns before the election. The ONS survey for 5–30 June 2024 found both cited as important issues by 87% of respondents, with the economy at 66%. Respondents could name several issues. These are public-priority estimates, not the percentage who voted Labour because of the NHS. They establish a plausible context for competing party appeals.

An argument about anti-incumbent voting therefore needs a chain: voters cared about services or household finances; some associated unsatisfactory performance with the governing party; an alternative appeared preferable or a protest vote became attractive. Labour, the Liberal Democrats and Reform could benefit through different routes. The loss of Conservative support cannot simply be treated as an equal gain for Labour.

The national result supports that distinction: Labour's vote share rose only modestly from 2019 while Conservative support fell sharply and other parties gained. Labour also received fewer votes in absolute terms than in 2019 because fewer people voted overall. Its landslide was therefore not evidence of an enormous uniform rise in enthusiasm for Labour. A balanced explanation combines rejection of the incumbent, the opposition's credibility, participation and the electoral system.

Competing offers and campaign strategies

A manifesto states what a party proposes to do in government. Labour's 2024 manifesto presented change with economic stability, spending rules and a promise not to raise income-tax, National Insurance or VAT rates. It offered measures on NHS waiting lists and clean energy. The strategy could reassure voters who wanted a different government but feared disruptive economic change. Its constraints could also disappoint voters who wanted a larger expansion of public spending.

The Conservatives' 2024 manifesto offered further tax reductions and protection for pensioners, alongside national service. Such proposals could appeal to particular groups, but an incumbent had to convince voters that fresh promises were credible after its existing record. A manifesto verifies the offer; it does not prove how many voters accepted it.

The Liberal Democrats placed health and care at the centre of their programme. Ed Davey's campaign appearances and discussion of his caring experience helped make that message visible; the party's manifesto-launch speech is partisan evidence of its intended presentation. Constituency targeting and local canvassing mattered because winning a seat required a local first place, not simply increasing the national total. Reform gave dissatisfied voters another alternative to the Conservatives, making a simple two-party account incomplete.

Leadership connected these offers to assessments of credibility. Starmer presented Labour as ready for government; Rishi Sunak defended the government's record and argued for continuity; Davey connected leadership presentation with care. Media appearances, short online messages and face-to-face campaigning gave parties several channels to reach voters. The analytical question is how communication changed salience or credibility, not whether an eye-catching moment happened. A campaign event's visibility alone does not establish a voting effect.

If you sit in 2027: Issue 3 does not list leadership separately in this case-study bullet. Issue 4 explicitly includes it for assessments from June 2028. The connections with policies, campaigns and context remain useful for both cohorts; any task relying specifically on the addition must be visibly labelled for 2028 onwards.

Age, gender, class and ethnicity

National post-election studies estimate how groups voted because the secret ballot does not record each voter's demographic identity. Ipsos interviewed 17,394 GB adults on 5–8 July through its address-recruited KnowledgePanel and weighted the data to population and regional election results. The table selects estimates from its 2024 profile.

Age strongly separated the two main parties, but other parties complicate that pattern: younger voters could choose the Greens, while Reform attracted some older voters. Education, housing, generational experience and policy interests may help explain age associations. An age table alone does not separate all those influences.

Labour's overall shares among men and women were similar. Ipsos estimated Reform support at 17% among men and 13% among women. That shows a gender difference for one party, not a uniform division between all men and all women. Likewise, its class breakdown showed relatively small Labour–Conservative differences between social grades, while Reform performed particularly strongly among C2 voters. Class dealignment weakens the connection between class identification and political choice. These occupational patterns are consistent with a weakening traditional alignment, but do not alone measure subjective identity or show that economic position has ceased to influence politics.

Ethnic minority voters were more Labour than white voters in the aggregate, but Labour's estimated minority share fell from 2019. Ipsos cautions that separate ethnic-group bases are small. Experiences, religion, income, age and policy disagreements vary within broad ethnic labels, so a single average must not become a stereotype. National subgroup estimates are preferable to assuming that everyone in a diverse constituency voted the same way.

Changing attachments, territory and participation

Ipsos estimated that, among surveyed 2019 Conservative voters who voted in 2024, 53% stayed Conservative, 24% chose Reform, 12% Labour and 7% the Liberal Democrats. Labour retained 71% of its corresponding 2019 voter group. This is evidence of movement between votes, not a direct measure of party identity. A voter can temporarily abandon a party while still feeling attached to it. Partisan dealignment concerns the weakening of durable identification; a repeated panel measuring identity as well as votes is better suited to testing it than a single result.

Territory mattered. Labour recovered many Scottish seats as the SNP fell from 48 to nine UK seats, while the Liberal Democrats increased their presence in areas with competitive Conservative contests. Regional industrial histories, housing, nationalism, local candidates and campaign targeting help explain why a national trend produces varied local outcomes. Commons results. Northern Ireland has a different party system and is excluded from the GB demographic surveys used here.

On the Commons Library's valid-votes-as-a-share-of-registration measure, turnout was 59.7%, below 2019. The South West had 64.9% and Yorkshire and the Humber 55.7%. These are territorial turnout results from its turnout analysis. The Electoral Commission's approximately 60% figure counts participation differently; choose and state a measure consistently.

Ipsos's demographic turnout estimates instead use all resident adults as the denominator: it estimated 37% for 18–24s and 43% for ethnic minority adults. These cannot be directly compared with the official registered-electorate percentage or with its own pre-2015 demographic estimates. Reporting the denominator is part of interpreting a political statistic, not a minor technical detail.

Disillusion can make people doubt that parties will solve problems; apathy can make them uninterested in participating. Low turnout is consistent with these explanations but may also reflect barriers or a predictable expected outcome. An angry voter may participate to punish a government. If you sit in 2027: use “no confidence” when explaining disillusion; the earlier glossary's “no confident” is corrected in Issue 4 and is not wording to memorise.