Can we use a by-election result to predict the future - Behind the Headlines
Following the sudden death of Labour's Hefin David (MS), a by-election took place in Caerphilly in October 2025. Prior to the by-election, the assumption across most media outlets was that Reform’s Llyr Powell would win the seat, breaking a 100-year Labour stretch in the constituency.
However, the media was proven wrong and Plaid Cymru’s Lindsay Whittle won the seat with 47% of the vote, a majority of 3,848.
Dr Jac Larner an expert on elections, political psychology, voting behaviour and national identity explores how this result can act as a predictor of things to come.
The Caerphilly by-election result—Plaid Cymru establishing a sizable majority over Reform UK, with Labour recording their worst performance since 1910—reveals patterns that will define May's Senedd election. While by-elections often invite overinterpretation and excessive focus on local factors, research suggests Caerphilly reflects deeper structural shifts in how Welsh voters are realigning ahead of what may prove the most consequential devolved election in the devolution era.
The starting point is unprecedented governmental unpopularity. Both the Welsh Government and UK Government now register their lowest satisfaction ratings in the entire history of devolution. Perceptions of both governments have moved in lockstep since 2020. Previously, the Welsh Government benefited from being relatively more popular than Conservative UK governments. Now Labour governs at both levels, and the Welsh Government has reached depths of unpopularity where 'relatively better' no longer suffices—a majority of Welsh voters now believe it is doing a bad job. It has become difficult, and probably futile, to differentiate whether the Welsh or UK Government inflicts greater damage. Both are profoundly unpopular and appear inseparable in voters' minds.
Against this backdrop, we are witnessing unprecedented voter mobility. Tracking flows since before the 2024 general election reveals that switching happens predominantly within political blocs, not between them. Progressive-identifying Labour voters overwhelmingly are moving to Plaid Cymru with some minor movement to the Greens and the Liberal Democrats. Simultaneously, rightward migration from Conservatives to Reform strengthens Reform's electoral position while hollowing out the Conservatives, Reform's only viable coalition partner. This generates a strategic paradox: Reform may emerge as the largest party yet struggle to govern in a system explicitly designed to frustrate single-party majorities. This pattern of within-bloc switching, rather than across-bloc realignment, suggests voters are seeking alternatives that better represent their values while maintaining their broader ideological orientation.
This matters enormously for May 2026. Labour faces erosion from the progressive left while the Conservatives face fragmentation on the right. The traditional major party squeeze on smaller parties has reversed—voters now actively seek alternatives to what they perceive as a failed Labour-Conservative duopoly at both levels of government.