An academic's look at some insights from past bye-elections
What do bye-elections tell us about Irish politics? With contests due in Dublin Central and Galway West on 22 May, a century of results offers a sense of campaign strategies and voting behaviour.
Since 1923, Irish voters have gone to the polls in 136 bye-elections, filling 138 Dáil seats, with two more vacancies to be decided in May.
The first of these contests took place in a State emerging from civil war, within a party system that was far removed from the one we see today. While some patterns have shifted with time, others have remained remarkably stable. This article examines six aspects of bye-elections that highlight both continuity and change in Irish politics.
1. Deaths caused most bye-elections historically
Bye-elections in Ireland have been triggered by a variety of causes. The death of a sitting TD has been by far the most common, accounting for 93 bye-elections, or two-thirds of all cases. This was particularly pronounced in earlier decades: in the 1950s alone, 20 bye-elections followed the death of a TD.
Resignations make up the next largest share (21%), covering departures for personal reasons, political controversies, or appointments as EU Commissioner, ambassador, or to the judiciary.
Eleven bye-elections were triggered by TDs elected to another office: seven resigned to take up European Parliament seats (all since 2010), and four were elected President: Seán T O'Kelly in 1945, Éamon de Valera in 1959, Erskine Childers in 1973, and Catherine Connolly in 2025.
The remaining six bye-elections arose from three candidates being elected simultaneously in two constituencies during the 1920s, and three disqualifications due to imprisonment (Henry Coyle, 1924), an alleged assault on a member of An Garda Síochána (Seán McGuinness, 1926), and bankruptcy (James Larkin, 1928).
We need your consent to load this flourish contentWe use flourish to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences
2) The typical bye-election now attracts 10-15 candidates
Except for one uncontested bye-election in 1933, every contest has featured at least two candidates. From the 1920s to the 1960s, no bye-election attracted more than six candidates, with averages ranging from 2.4 in the 1930s to 3.5 in the 1950s and 1960s.
Candidate numbers rose sharply from the 1980s, as the dominance of Fianna Fáil and Fine Gael weakened. As the party system fragmented, more candidates entered the field.
In the 1980s and 2000s, around ten candidates typically competed in a bye-election. The Dublin North bye-election in 1998 drew the highest number on record, with 19 candidates, followed by the Dublin Bay South contest in 2021 with 15. This increase in competitiveness reflects a broader shift also seen in general elections, where more candidates now compete for seats in Dáil Éireann.
We need your consent to load this flourish contentWe use flourish to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences
3) The winner of a bye-election typically receives around 30% of first-preference votes
Since the 1990s, the average first-preference vote share of the subsequent winner has been 29.7%. A candidate who secures around three in ten first preferences is well placed to win.
In 81% of bye-elections since 1923, the candidate leading on first preferences went on to take the seat. In the remaining 19%, a candidate who trailed on the first count comes through to win. This pattern has remained broadly stable, despite the rise in candidate numbers.
Political scientist Stephen Quinlan finds that two factors roughly double a candidate's chances of winning: being a relative of the outgoing TD, or having previously held a Dáil seat. Both advantages reflect name recognition, which remains a decisive factor in Irish elections.
Bye-election victories have also served as a route to party leadership. As Michael Gallagher, professor emeritus at Trinity College Dublin, notes, eight of the 138 TDs elected through bye-elections later became party leaders: Seán Lemass and Brian Cowen (Fianna Fáil), Enda Kenny (Fine Gael), William Norton, Brendan Corish, and Ivana Bacik (Labour), Seán MacBride (Clann na Poblachta), and Des O'Malley (Progressive Democrats).
4) Government parties have won just three of the last 30 bye-elections
Bye-elections have long tended to favour opposition parties, but the scale of the shift is stark. In the 1920s, government candidates won 22 of 27 contests. That dominance faded over time and has now effectively reversed: since 1994, 90% of seats went to opposition candidates or Independents.
The well-documented tendency for governments around the world to lose support between general elections is what political scientists call the 'midterm election effect'. Bye-elections intensify this dynamic in two ways. First, research shows that government supporters are less likely to turn out, shifting the electorate on polling day towards opposition voters. Second, voters often use bye-elections to signal dissatisfaction, knowing that even an opposition gain will rarely threaten the government's majority in the Dáil.
Turnout data reinforces the point. Stephen Quinlan compared turnout in each bye-election with that in the same constituency at the previous general election and found lower participation in 77% of cases, with government supporters particularly unlikely to mobilise. The gap has widened sharply over time: on average, turnout was four percentage points lower in the 1920s, rising to 10 points in the 1930s and 24 points in the 2010s. In all 30 bye-elections since 1994, turnout has fallen below the level recorded at the preceding general election in the same constituency.
We need your consent to load this flourish contentWe use flourish to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences
5) 121 unsuccessful bye-election candidates later became TDs
Bye-elections often serve as stepping stones for unsuccessful candidates rather than dead ends. Tracking the political careers of the 600 candidates who competed in the 138 bye-elections reveals that 121 unsuccessful candidates were later elected to Dáil Éireann. For 66 of them, the bye-election was their first attempt to win a seat.
The pattern cuts across parties. Fianna Fáil later returned 37 such candidates, followed by 26 each for Fine Gael and Labour, and nine for Sinn Féin. A further nine Independents and five Green Party candidates also used a bye-election as a springboard to the Dáil.
Should parties always field their strongest candidates? The evidence points that way. More than 120 unsuccessful candidates eventually became TDs, and for many the bye-election was a first step. Name recognition and a connection to the local constituency can pay off even in defeat.
Yet, incumbents in the same constituency may see things differently. A bye-election candidate from the same party who wins the seat, or simply performs well, becomes a potential rival at the next general election.
If bye-election experience boosts future prospects, sitting TDs have reason to be cautious. Whether TDs fully back their party's bye-election candidate, or quietly prefer a weaker result, remains an open question.
We need your consent to load this flourish contentWe use flourish to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences
6) A strong social media presence does not replace door-to-door campaigning
Detailed survey data on voting behaviour in bye-elections are limited. However, the 2024 general election exit poll asked voters which sources they relied on most when forming their views.
More than a third of voters aged 18 to 24 cited social media and messaging apps as their main source. Among those aged 25 to 34, one in four did so. The figures drop sharply with age: just 6% of those aged 50 to 64, and 2% of those over 65, identified social media as their primary influence.
The pattern reverses for 'traditional' media. Around 60% of voters over 65 pointed to television or radio as their main source of information. Newspapers, in print and online, account for roughly one in four voters across most age groups, although less among the over-65s.
These patterns shape how campaigns operate. Social media offers access to younger voters, but precise constituency targeting remains constrained by technical limits and regulation. The 2024 general election also showed that a large online following does not translate reliably into votes.
Direct contact remains decisive. Knocking on doors continues to be one of the most effective strategies in Irish elections, and online campaigning does not replace face-to-face engagement.
We need your consent to load this flourish contentWe use flourish to manage extra content that can set cookies on your device and collect data about your activity. Please review their details and accept them to load the content.Manage Preferences
Lessons for Dublin Central and Galway West
The upcoming bye-elections in Dublin Central and Galway West will test whether these historical patterns hold. If past trends continue, government parties face an uphill battle, successful candidates will need strong local profiles and door-to-door campaigns, and the results are likely to deviate from national opinion polls.
Voters get to decide when they go to the polls on 22 May.
Further information
Folke, O., Rickne, J., & Smith, D. M. (2021). Gender and dynastic political selection. Comparative Political Studies, 54(2), 339–371.
Gallagher, M. (1996). By-elections to Dáil Éireann 1923–96: The anomaly that conforms. Irish Political Studies, 11(1), 33–60.
Gallagher, M., O'Malley, E., & Reidy, T., Eds. (2025). How Ireland Voted 2024. The New Normal? Palgrave Macmillan.
Louwerse, T., & Müller, S. (2025). Irish Polling Indicator Datasets: Stable Version (1987–2024). Harvard Dataverse.
Müller, S., & Louwerse, T. (2020). The electoral cycle effect in parliamentary democracies. Political Science Research and Methods, 8(4), 795–802.
Smith, D.M., & Martin, S. (2017). Political dynasties and the selection of cabinet ministers. Legislative Studies Quarterly, 42(1), 31-165.
Quinlan, S. (2023). Special elections in alternative vote electoral systems: Exploring turnout and the vote in Irish by-elections 1923–2019. The British Journal of Politics and International Relations, 25(1), 156–177.
