Data / Compare presidential elections
Shifting Ground.
Two rounds. Your view.
Choose parties and election years. Each round has its own adaptive scale, exact figures and CSV export. New years are discovered from the shared archive.
What do shares of registered voters mean?
The usual share divides a candidate’s votes by valid candidate votes. Registered-voter mode divides that same count by everyone registered, including abstainers and people casting a vote blanc or an invalid ballot.
Example: 30 out of 60 valid votes is 50%. If 100 people were registered, those same 30 votes are 30% of the electorate. The count and legal outcome are unchanged.
Vote blanc: participating without choosing a candidate
A vote blanc is an empty envelope or an envelope containing a blank ballot. It differs from abstention, because the voter participates, and from a vote nul, an invalid ballot.
It can express “I came to vote, but I do not choose any of these candidates.” It is sometimes used as a protest, although the total cannot reveal every voter’s motive.
Votes blancs count toward turnout, not valid candidate votes. Separate counting from invalid ballots began in 2014, so older elections show a combined total. They do not automatically cause another election.
Official explanation ↗The initial view uses the three most recent elections. Add older years to extend the comparison; the automatic scale grows if their selected scores need more room. Only checked years are plotted, with horizontal positions based on calendar year.
Shares of valid candidate votes. Identical colors identify the same labelled party lineage. Missing candidacies remain gaps. UMP/LR includes predecessors; Mélenchon’s 2012 point belongs to the Front de gauche coalition, not LFI itself.
Automatic party scales use the median leading score across the archive, add 5% headroom, and round upward to a five-point ceiling. They then expand when the selected values require it. This robust typical score prevents one exceptional result setting every default scale.
First round
Party / lineage candidate share
Zero baseline. Automatic scale. Typical leading score plus breathing room gives a 35% baseline ceiling. The scale expands to include the selected elections and parties. Each panel has its own labelled scale; slopes between panels are not directly comparable.
| Year | Renaissance (En Marche! → LREM) | Rassemblement national (formerly FN) | Parti socialiste / SFIO predecessor | Les Républicains / UMP / RPR / UDR |
|---|---|---|---|---|
| 2012 | Not available / no candidacy | 17.90% | 28.63% | 27.18% |
| 2017 | 24.01% | 21.30% | 6.36% | 20.01% |
| 2022 | 27.85% | 23.15% | 1.75% | 4.78% |
Second round · Runoff
Party / lineage candidate share
Zero baseline. Automatic scale. Typical leading score plus breathing room gives a 60% baseline ceiling. The scale expands to include the selected elections and parties. Each panel has its own labelled scale; slopes between panels are not directly comparable.
| Year | Renaissance (En Marche! → LREM) | Rassemblement national (formerly FN) | Parti socialiste / SFIO predecessor | Les Républicains / UMP / RPR / UDR |
|---|---|---|---|---|
| 2012 | Not available / no candidacy | Not available / no candidacy | 51.64% | 48.36% |
| 2017 | 66.10% | 33.90% | Not available / no candidacy | Not available / no candidacy |
| 2022 | 58.55% | 41.45% | Not available / no candidacy | Not available / no candidacy |
Renaissance (En Marche! → LREM) · names and continuity
En Marche! in 2017 became La République en marche, renamed Renaissance in 2022. The 2022 election preceded that rename. Party history ↗
Rassemblement national (formerly FN) · names and continuity
Front national was renamed Rassemblement national in 2018. These are the same party across a change of name. Party history ↗
Parti socialiste / SFIO predecessor · names and continuity
PS succeeded SFIO in 1969. Mitterrand’s 1974 candidacy had broader Union of the Left support. Candidate totals cannot be separated into the votes of each supporting party. The 1965 Mitterrand entry is a CIR-led candidacy backed by SFIO and PCF, not a result for a PS organization that did not yet exist. Party history ↗
Les Républicains / UMP / RPR / UDR · names and continuity
UMP was renamed Les Républicains in 2015. UMP was formed in 2002 from RPR and other organizations: earlier RPR and UDR entries are predecessor-lineage comparisons, not an unchanged party. The earlier Gaullist series also includes UNR–UDT in 1965 and UNR support in 1958. These are labelled predecessor organizations, not LR under its present name. Rival RPR members Chirac and Balladur are listed separately in 1995, never added together. Party history ↗
Sources & definitions
The direct-election archive now begins in 1965. The 1958 electoral-college election is separate and excluded from popular-vote comparisons. Counts for pre-2002 pages follow their linked historical proclamations and transcriptions; percentages are recalculated from those counts.
For 1988, the first-round data use the final corrected totals annexed to the 11 May proclamation, rather than the earlier 27 April declaration. Blank and invalid ballots are combined throughout these historical returns.
In 1995, Chirac and Balladur were rival RPR members. Both appear separately on the results page and in lineage histories. A single-line party comparison uses the highest-vote candidate in that labelled lineage, not the sum of both candidacies. This is not a combined party-vote total.
National totals, France entière, from the linked national returns. Pre-2002 pages identify their Constitutional Council proclamations or historical transcriptions. Candidate share = votes ÷ valid candidate votes; turnout = ballots cast ÷ registered voters; vote blanc / invalid percentages = those ballots ÷ ballots cast. Registrations can differ between rounds. Rounded percentages may not sum to 100%.
Before 2014 the published totals combine votes blancs and invalid ballots. Separate blank and invalid series therefore have gaps for 2002, 2007 and 2012. The combined series is comparable across all five elections. Missing party candidacies are gaps, not zero votes.
Colors are stable identifiers, not official political classifications. Rename continuity is distinguished from mergers, predecessors, coalitions and shared endorsements. The old lookbacks reach 1969 and use secondary historical compilations, linked per year; the direct-election pages use their linked national returns and historical transcriptions.
- 2022: National election return ↗
- 2017: National election return ↗
- 2012: National election return ↗
- 2007: National election return ↗
- 2002: National election return ↗
- 1995: Constitutional Council returns, transcribed by Université de Perpignan MJP ↗ · Historical result table ↗Percentages are recalculated from the cited counts, not copied from rounded or truncated percentages in the transcription.
- 1988: Constitutional Council final proclamation, 11 May 1988 ↗ · Historical result table ↗Uses the corrected first-round totals annexed to the final proclamation, not the earlier 27 April declaration. The corrected valid-vote total is 30,436,734; the earlier declaration used 30,406,038. Candidate counts and turnout here belong to the corrected version.
- 1981: Constitutional Council returns, transcribed by Université de Perpignan MJP ↗ · Historical result table ↗Percentages are recalculated from the cited counts, not copied from rounded or truncated percentages in the transcription.
- 1974: Constitutional Council returns, transcribed by Université de Perpignan MJP ↗ · Historical result table ↗Percentages are recalculated from the cited counts, not copied from rounded or truncated percentages in the transcription.
- 1969: Constitutional Council returns, transcribed by Université de Perpignan MJP ↗ · Historical result table ↗Percentages are recalculated from the cited counts, not copied from rounded or truncated percentages in the transcription.
- 1965: Constitutional Council returns, transcribed by Université de Perpignan MJP ↗ · Historical result table ↗Percentages are recalculated from the cited counts, not copied from rounded or truncated percentages in the transcription.
This section presents numerical results and descriptive facts, not critical readings. Comparisons do not identify individual voter transfers or establish the causes of a change.
Teaching and learning
Intended audience: Undergraduate students in French studies, history, political science and related fields; independent adult learners.
Suggested uses: Guided data interpretation, election comparison and evidence-based discussion.
After using this resource, learners should be able to:
- Interpret the election results, turnout figures and ballot statistics presented in this resource.
- Identify the election, round, geographic scope and denominator before comparing figures.
- Distinguish descriptive patterns in the data from claims about individual voting motives.
Use the objectives that correspond to the material available on this page. Collection pages provide entry points to individual resources; unfinished or unavailable material is not a completed learning activity.
Original materials © Dr. Amine Zidouh / French Politics, licensed under CC BY-NC 4.0 unless otherwise indicated. When sharing or reproducing covered material, credit Dr. Amine Zidouh and French Politics, link to the source and license, and identify changes. Images, protest signs, artworks and other third-party materials are excluded. See Rights, Credits & Educational Reuse for scope, attribution and the as-is disclaimer.