Explanations

How are the overviews on OpenRCV produced, what do the terms and metrics mean, and where are the limits of the data? Here are the key points.

What is this?

OpenRCV makes the roll-call votes of the European Parliament searchable and traceable, with the voting record of every individual MEP. All information comes from official sources, almost all of it straight from the Parliament’s official Open Data Portal. No unofficial third-party sources, no estimates and no AI are used to prepare the data. Only two clearly labelled features use a language model: “Ask” and the summary of a document, which can be expanded in the vote view below “Procedure details”.

Explained in detail

Individual questions on pages of their own, with sources and current figures:

• Motion of censure on the European Commission
• Roll-call votes in the European Parliament
• Finding European Parliament vote results
• Vote corrections in the European Parliament
• European Parliament political groups and parties

What is a roll-call vote?

Parliament votes in several ways: by show of hands, electronically (totals only), by secret ballot (for example in elections) or by roll call (RCV). Only a roll-call vote records how each individual MEP voted.

OpenRCV only analyses these roll-call votes. The other votes have no individual records; they appear in the agenda, marked “not by roll call”, and as a number, for example in the statistics under “At a glance”.

Where does the data come from?

The main source is the European Parliament’s EP Open Data Portal: votes and how each MEP voted, vote corrections, MEPs, documents, procedures, agendas and speeches. Titles, committees and results are provided there officially and in many languages, so no machine translation is needed. The names of the topics (EuroVoc) come from the Publications Office of the EU. The portrait photos are the Parliament’s official photos; OpenRCV shows reduced copies of the originals from the Parliament’s website. If there is no copy yet, for example for a new MEP, OpenRCV takes the same official portrait from Wikimedia Commons (matched via Wikidata) or loads the photo directly from the Parliament’s website.

The data is available under CC BY 4.0. The prepared dataset can be downloaded freely (Data).

The “Ask” feature

With “Ask” you can look up votes, MEPs, political groups and countries in your own words. The answers are written by the language model Claude from Anthropic. It works only with the data of OpenRCV, that is the official voting data of the Parliament, and links to the votes and profiles it relies on.

A language model can be wrong or oversimplify connections. The linked sources and the figures on the pages themselves are authoritative. Apart from these answers and the labelled summaries of documents, everything on OpenRCV (results, metrics and graphics) is produced without AI.

OpenRCV does not store questions or answers; the history stays only in your browser until you close the tab. Where the questions are sent to be answered is explained in the privacy policy.

Plenary: agenda and proceedings

The Plenary section shows, for every plenary week since July 2024, what was on the floor: debates and votes day by day, with times and results. Past sitting days follow Parliament’s minutes with the actual times, upcoming days the draft agenda. The draft often changes until the plenary week itself, so OpenRCV fetches it again regularly.

The debate and the vote on the same matter are often days apart. OpenRCV links them when they fall in the same week and concern the same document (such as a report A10-…) or the same procedure, otherwise when their titles are very similar. Pointing at or tapping a linked item highlights its counterpart. Votes without a roll call are listed on the agenda but have no individual votes.

Procedural items such as the opening, the approval of the minutes or explanations of vote are hidden at first and can be shown. Document references such as A10-0265/2026 link to the full text on europarl.europa.eu, procedure codes such as 2026/2925(RSP) to the procedure file in Parliament’s Legislative Observatory.

Times. The time shown is the start of each roll-call vote according to Parliament’s decision data, in local time (Brussels and Strasbourg). During summer time these data give the time one hour too late; OpenRCV corrects this for each sitting day and checks it against the proceedings in the minutes. Votes by show of hands, electronic votes without a list of names and secret votes have no time. A voting session ends when no roll-call vote follows for more than 15 minutes. “On average” is the mean of the gaps of up to five minutes between two roll-call votes; “for half of them” refers to the median of the same gaps.

Time since the report. The days from the date of the committee report for plenary (A document) to the final roll-call vote. Files without a committee report, decisions without a roll call and ongoing procedures are not included; the Commission proposal and the work of the Council are not in the data. For legislation with an agreement with the Council, the time includes the negotiations. A report with two final votes appears as two dots. The grouping by type of procedure is the same as for the lead committees. For legislation, an agreement with the Council means that the main vote is on a provisional agreement or a joint text, and every third reading; all other readings, including the second, are shown under “No agreement”. If a title does not yet carry the procedure marks, for example in the latest week, the type of the procedure decides.

Speeches and speaking time. Counted are spoken contributions to plenary debates since July 2024, with their length according to the verbatim report. Since 23 October 2026, contributions from the chair (vice-presidents presiding over the sitting), written statements and duplicate entries no longer count; profiles show the contributions from the chair separately. Anyone who did not belong to a political group on the day, for example as a member of the Commission, is not counted. For the President, every entry counts as a contribution from the chair, because her chairing cannot be separated from her own speeches in the data; she therefore has no speeches and no rank. Speaking time by group counts by the group on the day of the speech and without rapporteurs; the reference value is the speaking time the agenda allocates to the groups, on sitting days with a speaking-time plan. Catch-the-eye speeches, blue cards and replies are part of the group’s time, so only shares are shown. Part of the speaking time is divided equally among all groups and the rest by size, so small groups receive more than their share of seats.

Reports by group. Counted are reports (A references) with a final roll-call vote since July 2024, each report once; with two final votes, the last one counts. Co-rapporteurs count proportionally, by their group on the day of the vote; anyone who no longer held a seat on that day is left out. The reference value is the share of seats on those days. Committees allocate reports through a points system based on group size; the number is not a measure of performance. Reports without a final roll-call vote, such as most immunity cases, are not included.

Main vote and sub-votes

A file often involves several roll-call votes: on individual amendments, paragraphs (§) and recitals, plus the final vote on the text as a whole. OpenRCV identifies the final vote (the “main vote”) from the official reference text and highlights it. The detail view shows its result first; the other votes on the file can be selected there.

If a file has no roll-call final vote (for instance because the final vote was taken by show of hands and only an amendment was voted by roll call), no vote is marked as the main vote; the list of votes then says “no final vote”.

Vote positions

• For / Against / Abstention: the vote cast.
• Excused: the MEP was officially excused on that sitting day.
• Did not vote: in office but did not take part in this vote (without an official excuse).

Vote corrections

After a vote, MEPs may put on record that they had intended to vote differently (for example after pressing the wrong button). OpenRCV shows such corrections transparently: the original vote stays visible and the intended one is marked.

Important: the official result is unaffected by corrections; they do not change the official vote counts. Metrics and statistics therefore use the official vote; the “Own positions” page can be switched to the vote after correction.

The metrics explained

• Participation: share of votes during the mandate in which the MEP cast a vote (For, Against or Abstention). Excused absence counts as not taking part; the profile also gives the figure without it.
• Cohesion: how uniformly a political group or another bloc votes: the share of votes cast that fall on the most frequent position (For, Against or Abstention), averaged over the votes.
• On the majority side: for an MEP, the share of their own For and Against votes in adopted or rejected votes that matches the result: For when adopted, Against when rejected. Abstentions and votes not cast are not counted. For a group, the share of adopted or rejected votes in which its majority voted that way; a majority of abstentions is counted, but never as a match.
• Against own group (dissent rate): share of votes in which an MEP votes differently from the majority of their own group. The MEP’s own vote is not counted towards the group majority. What counts is the group on the day of the vote; time without a group is not counted, as there is no group line. This figure implies no judgement: it may reflect a matter of conscience, the line of a national party or an arrangement within the group.
• Agreement / alliances: how often two groups take the same majority position. In the profile, likewise, how often an MEP votes like the majority of a group; when comparing two MEPs, how often both voted the same way in the final votes in which both took part.

Majorities: who carries a result

Below the result, the vote detail shows which political groups carried it and which were decisive by the numbers. In the list of votes, the “Majority” filter finds every final vote in which the majorities of particular groups voted for or against, across all part-sessions if you like. The statistics show under “Majorities” how often each pattern occurred. All three use the same rules.

• Majority of a group: the position (for, against or abstention) that occurs most often among the votes cast by its members. The group on the day of the vote counts. In the event of a tie, the group has no majority. The group majority is not the whole group: individual members may have voted differently, and those who did not vote are not counted. Non-attached Members do not form a group and are not listed, but their votes count in the result.
• Carried by: the groups whose majority voted the way the result went, i.e. for an adopted text and against a rejected one. If a vote fails to reach a higher majority even though there were more votes for than against, the detail instead names the groups whose majorities voted against.
• Decisive by the numbers: the groups without whose votes for and against the result would have been different, as if their members had not voted or had abstained. This is a calculation, not a statement about agreements. Large groups are naturally decisive more often than small ones.
• Required majority: Usually a majority of the votes cast is enough (more for than against, abstentions do not count, a tie means rejection). Some votes need a majority of Parliament’s component Members, for example objections to delegated acts: more votes for than half the Members in office on the day of the vote. A few also need two thirds of the votes cast, for example a motion of censure. OpenRCV takes the required majority from the official information on the vote; since October 2025 Parliament has stated it for every vote, before that only where it differed from a majority of the votes cast. If the official result cannot be recalculated with the rule (for instance for amendments to the draft budget), no “decisive” information is shown.

The “Majority” filter applies to final votes; the statistics apply to final votes or to all roll-call votes, depending on the dataset chosen. They show the six largest groups so that the patterns stay manageable; the filter offers all eight groups. The detail calculates for every split vote you select; split votes are taken to need a majority of the votes cast when the item itself does, otherwise “decisive” is left out. The information describes how people voted. It does not judge, and it does not name coalitions.

Line-ups, contested votes and reference values

Under “Majorities”, the statistics also show which groups carried a result together, as a line-up, for the selected plenary week or the parliamentary term and over time by half-year. The week in review counts them over all roll-call votes of the week, and the vote detail names the line-up of the selected vote. Line-ups are definitions by OpenRCV, not negotiated coalitions. They record which of the groups EPP, S&D, Renew, ECR and PfE were, by majority, on the side of the official result: adopted means for, rejected means against.

• EPP, S&D and Renew without ECR and PfE: the majorities of EPP, S&D and Renew were on the side of the result, those of ECR and PfE were not.
• EPP, S&D and Renew with ECR or PfE: EPP, S&D and Renew, plus ECR, PfE or both.
• EPP, ECR and PfE without S&D: EPP, ECR and PfE were on the side of the result, S&D was not. Renew may be among them; the statistics state how often.
• Other line-ups: all other cases, for instance when the majority of the EPP was not on the side of the result.

“Without” means: not on the side of the result by majority. The group’s majority voted differently or abstained, or the group had no majority because of a tie. As above, the majority of a group is the most frequent position among the votes cast by its members on the day of the vote; a group that mostly abstained did not carry the result. Greens/EFA, The Left, ESN and non-attached Members do not change the classification. If a vote failed a raised majority although more Members voted for than against, against counts as the side of the result. For these votes the vote detail names no line-up; they are still counted, also in the week in review. Votes without an official result are listed as “no result” outside the line-ups.

Over time: half-years by the date of the vote; the current half-year is provisional and has fewer cases. Values based on fewer than 10 votes are marked “few cases”. Many votes on one item can shape a half-year, so for all roll-call votes the table also shows the share weighted per item: each item counts once, and its individual votes share that weight. Examples: the three most recent votes per line-up, at most one per item.

Contested: between 30 and 70% of those voting voted in favour, and at least a fifth voted against. Those voting are all who voted for, against or abstained; the official figures apply. The “Contested only” switch counts only these votes, and wherever it changes figures, a line above them says how many there are. Contested does not mean rejected: most contested final votes are adopted, whereas about half of all contested roll-call votes are rejected. There are far fewer of them: below 10 contested votes the figures are marked “few cases”, and for people and parties with fewer than 20 contested votes counted, “too few cases” replaces the percentage. Contested refers to Parliament as a whole; a single group is called split when no position reaches 60% of its votes.

Reference value: how often someone who voted for in every vote would be on the side of the result (if at least half were adopted), otherwise someone who always voted against. It is calculated over the same set as the share next to it; on a profile, over the person’s own votes for or against in adopted or rejected votes. The reference value is not a judgement; it shows what share would result from the distribution of outcomes alone. When comparing two Members, the median of all pairs of current Members with at least 20 final votes in common is shown alongside, overall and per pair of groups.

Ties: the rules stay as they are for each figure. Line-ups and majority patterns treat a group with a tie as having no majority; in the statistics, “Political groups” and “Who votes with whom” take the first position in the order for, against, abstention. “Contested only” calculates each figure by its own rule, filtered and unfiltered alike.

The week in review explained

The week in review sums up a plenary week in a few figures. It appears in the plenary view above the list of votes, in the graphic for sharing and in the weekly feed, always with the same figures. OpenRCV calculates it from the official data according to fixed rules when it processes the data.

• Decisions: all votes in plenary according to the official lists of decisions, whether by roll call, by show of hands, electronic or secret, with every split vote and amendment counted separately; lapsed and withdrawn items are not counted. If the list of decisions is missing for a sitting day with votes, the review only gives the roll-call votes.
• Final votes: adopted, rejected and without result according to the official result.
• Smallest margin: the final vote where the fewest votes in favour would have had to be missing or added to change the result. It is measured against the required majority (see Majorities): for a majority of the votes cast, by the gap between votes for and against; for a majority of component Members, by the gap to the threshold (half the Members in office on the day of the vote, rounded down, plus one). Votes under another rule, such as two thirds, and votes whose official result cannot be recalculated this way are left out; at least two comparable final votes are needed. If two margins are equal, the earlier vote counts. If the margin is at most 5% of the Members in office (about 36 votes), it is called the narrowest result.
• Groups split: a group is split if at least 10 of its Members voted and no position (for, against, abstention) reaches 60% of its votes, as in the statistics. The review names the final vote with the most groups split; if there is a tie, the one with the largest share of Members who voted differently from their group's largest position, then the earlier one. The group on the day of the vote counts; non-attached Members do not form a group. If it is the same vote as the smallest margin, it is shown only once.
• Line-ups: which of the groups EPP, S&D, Renew, ECR and PfE were, by majority, on the side of the result, counted over all roll-call votes of the week by the same rules as in the statistics (see Line-ups); a definition, not a negotiated coalition. The narrow bar appears only from 20 roll-call votes.
• Participation: the votes cast (for, against, abstention) in all roll-call votes of the week, divided by the Members in office on the day of each vote. Members who did not vote are not counted, even if they were present.
• Provisional: while the plenary week is still running, the review is provisional and states when the data was last updated; it changes with every new voting session.

The review describes how Members voted. A small margin or a split group is not a judgement.

Amendments: who tables them, what is adopted

More than half of all roll-call votes concern amendments: proposals to change, delete or add to a text at a specific point. In the vote detail, every amendment shows who tabled it, and the list of votes on an item can be limited to the adopted ones. In the dataset “All roll-call votes”, the statistics show under “Amendments” how many amendments each political group brought to a roll-call vote, how many of them were adopted and how the majorities of the other groups voted on them. Below is a searchable list of all these amendments.

• What counts: roll-call votes that Parliament records as an amendment, with an amendment number, an amendment document or “Am” in the label. Not included are the main vote on an item (even where a provisional agreement is formally voted as a committee amendment), separate and split votes on the original text, and entire motions for resolutions tabled by a group.
• Tabled by: as stated by Parliament for the vote. One or more political groups, a group of Members (at least one twentieth of the component Members), the committee responsible or the rapporteur. Amendments from several groups, tabled jointly or in identical wording, have their own rows in the statistics. For some amendments, mainly to the budget, the data name no author; the statistics state how many.
• Adopted: the official result of the vote. The adoption rate is not a measure of success. Many amendments are tabled to make a position visible, without any prospect of a majority, while others have been agreed beforehand. Whoever requests a roll-call vote also decides which amendments are voted on by roll call.
• Majority of a group: as in the section Majorities, the most frequent position among the votes cast by its members on the day of the vote. In the event of a tie, the group has no majority.

Limitations: the data do not contain the wording of the amendments, only the point in the text and the number. The list links to the document of the item on Parliament’s website, to which the point and number refer; Parliament publishes the amendments themselves in separate sitting documents. Amendments decided by show of hands or by electronic vote without a list of names are missing, as are amendments withdrawn or fallen before the vote. The figures describe how people voted. They do not assess anything.

Mandate periods and MEPs joining mid-term

Participation and dissent rates only refer to votes within the respective mandate. MEPs who join during the parliamentary term receive no values for votes before their mandate began. This keeps the comparison fair.

How current is it?

The site checks its data against the portal automatically several times a day (new votes every ten minutes from 12:00 to 20:00, Monday to Thursday; vote corrections every two hours from 08:00 to 20:00, daily) and is rebuilt when there is new data; the source line of every page shows how current the data is (“Data as of”). New plenary weeks appear as soon as the data is available in the portal. Vote corrections are sometimes submitted days later; they are included automatically. Speeches are fetched again once a week, on Saturdays.

Subject areas: what the votes are about

Parliament indexes its documents with keywords from EuroVoc, the EU’s multilingual thesaurus (Publications Office, op.europa.eu/eurovoc). Each keyword belongs to a microthesaurus, and each microthesaurus to one of 21 subject areas (EuroVoc domains). OpenRCV uses this fixed structure and gives the subject areas simplified names, such as “Environment and climate” for ENVIRONMENT or “EU institutions, law, budget and enlargement” for EUROPEAN UNION.

• Assignment: a vote counts in every subject area that one of its document’s keywords belongs to, using all of the document’s keywords that OpenRCV holds (a few are still missing), not only the first eight shown with a file. A vote is therefore often in several subject areas, and the subject area figures add up to more than the number of votes. Some final votes do not (yet) carry keywords; for new ones they usually arrive a few days after the vote. They count in no subject area.
• Also: “Countries and regions” and “International organisations” name places and organisations, not a subject. They are shown apart from the other subject areas.
• Groups by subject area: the share of final votes in which a group’s majority voted in favour (the largest of the three blocks yes, no and abstention, group on the day of the vote), counted over the votes in which the group took part; a tie counts only in the denominator. Non-attached members have no common line and no value. Below 10 votes the page says “too few”. The reference value is the same share over all final votes with a subject area.
• Objections to Commission acts (Rules 114 and 115 of the Rules of Procedure) reverse the meaning: in favour means against the act. They are therefore left out of the values and are not counted as adopted files, but have their own view “Objections only”. Many objections concern authorisations of genetically modified plants and therefore count in several subject areas at once. Motions of censure and proposals to reject the Council position also stay outside the values; they are only counted.
• Cohesion: the average share of the largest block (yes, no or abstention) within the group, over the same votes as the values next to it. It is the same measure as in the statistics and the blocs.
• On a profile, “With own group by subject area” shows how often a person voted like the majority of their group at the time in final votes, from 10 comparable votes per subject area. As this is only about same or different, objections are included.
• In the statistics, the subject area view counts final votes, or in the “all roll-call votes” dataset every roll-call vote; split votes inherit the subject areas of their file.

The voting map

The voting map places all MEPs on two axes so that those who vote alike sit close together. It always uses all roll-call votes of the whole term; the choice of final or all votes, a week and “contested only” do not change it. It is recalculated together with the profiles.

• Calculation: every vote counts as a number, in favour 1, against minus 1, abstention 0; each vote is measured from the average of the votes cast, and a missing vote counts as that average, so it has no effect. A principal component analysis finds the direction in which the votes differ most (axis 1) and the next one independent of it (axis 2). Each axis shows the share of the differences it explains.
• Position: a person’s position is based only on their own votes cast, so that short mandates do not drift to the middle. It is given as the distance from the average of everyone placed, in standard deviations. A position requires at least 300 roll-call votes cast; up to 999 votes it counts as uncertain and is shown as a hollow dot. Anyone with fewer votes has no position; the President usually does not vote.
• Axes: the axes have no fixed meaning. They are described only by which groups’ majorities stood on which side in the votes that shape an axis most. The orientation of axis 1 follows the seating order in the chamber, and axis 2 puts the largest group on the positive side (at the top in the Map view); both are choices of presentation, not a statement about a political direction.
• Examples for each axis are final votes with a large weight on it in which the majority of every group named for that side voted in favour. Objections to acts, motions of censure and motions to reject are left out there, because in favour means against in those votes.
• Limits: groups request roll-call votes deliberately, and many of the votes that shape an axis are amendments by individual groups. The map therefore shows the broad pattern of voting behaviour, not positions on every question. For comparing individual people, “Most similar MEPs” remains authoritative (agreement in final votes); distances on the map reflect it only roughly.

Limits & notes

• The 10th parliamentary term since July 2024 is covered; earlier votes are missing.
• The metrics are based on the final votes per file by default. The statistics and the profiles can be switched to all roll-call votes, including amendments and split votes; the vote details always show them. The list of MEPs, the comparison and the Groups and countries page count final votes; there, only “Most divided votes” also gives the figure across all roll-call votes. The voting map always uses all of them.
• Almost everywhere, the group, party and country on the day of the vote apply, including MEPs who have left: anyone who switches counts with the old group until the switch and with the new one afterwards. This applies to the statistics, profiles, vote details and the Party view; in the Group view also to cohesion by half-year, in the Member state view to participation, and in both to all sections below the internal vote distribution, such as “Most divided votes”.
• By contrast, the member count, cohesion, majority side and internal vote distribution in the Group and Member state views, at the top and in the overview, use the current MEPs and their current group, across all their final votes of the parliamentary term. The statistics use the same cohesion for the delegations under “Distribution”. The voting map in the statistics, profile and comparison arranges the dots by current group. Cohesion and majority side therefore differ slightly from the figures for the groups in the statistics.
• OpenRCV is an independent project. The official result of the European Parliament always prevails; for details see Legal notice and notes.

Glossary

Technical terms are underlined with dots where they first appear on a page. Tapping, clicking or pressing Enter shows a short explanation with a link to this list; if you know the terms, you can turn the marking off here, in the “More” sheet or at the bottom of the sidebar.

Abstention, did not vote, excused

An abstention is a vote cast, neither for nor against. “Did not vote” means in office but not taking part in this vote, without an official excuse; “excused” means officially excused on that sitting day.

More: Vote positions

Allocated speaking time

Speaking time that the agenda allocates to the political groups for each debate. Part of it is divided equally among all groups, the rest by size.

More: Plenary: agenda and proceedings

Amendment

An amendment proposes to change, delete or add to a text at a specific point. It can be tabled by political groups, a group of at least one twentieth of the component Members, the committee responsible or the rapporteur.

More: Amendments: who tables them, what is adopted

Carried by, decisive by the numbers

A result was carried by the groups whose majority voted the way it went; groups are decisive by the numbers if, without their votes for and against, the result would have been different. This is a calculation, not a statement about agreements; large groups are naturally decisive more often than small ones.

More: Majorities: who carries a result

Cohesion

Cohesion shows how uniformly a political group or another bloc votes: the share of votes cast that fall on the most frequent position (For, Against or Abstention), averaged over the votes. 100 per cent means everyone who voted voted the same way.

More: The metrics explained

Contested

OpenRCV calls a vote contested when between 30 and 70% of those voting voted in favour and at least a fifth voted against (abstentions count as voting). This concerns Parliament as a whole; a single group is called split when no position reaches 60% of its votes.

More: Majorities: who carries a result

Debate

A debate is the discussion in plenary on an item on the agenda, often with contributions from the Commission and the Council. Speaking time is allocated to the political groups in advance; the chair gives the floor.

Dissent from the group

A dissenting vote is a vote against the majority of the MEP’s own group in the same vote. The MEP’s own vote is not counted towards the group majority, and the group on the day of the vote is what counts.

More: The metrics explained

EuroVoc

EuroVoc is the EU’s multilingual thesaurus, which Parliament uses to index its documents. OpenRCV uses these official keywords (also called topics on the pages) and their grouping into 21 subject areas; a vote can have several keywords.

More: Where does the data come from?

Final vote

The final vote is the vote on the text of a file as a whole, after the votes on amendments and individual parts. OpenRCV identifies it from the official reference text; the result at the top of a detail page refers to it.

More: Main vote and sub-votes

Group majority

A group’s majority is the position (For, Against or Abstention) that occurs most often among the votes cast by its members. Those who did not vote are not counted; in the event of a tie, the group has no majority.

More: Majorities: who carries a result

Keyword

A keyword is a term from the EuroVoc thesaurus that Parliament uses to index a document, such as food safety. The pages also call keywords topics; at most eight are shown with a file.

More: Subject areas: what the votes are about

Lead committee

The committee that prepares a report for plenary. Resolutions and objections often reach plenary without a lead committee.

More: Plenary: agenda and proceedings

Line-up

A line-up records which of the groups EPP, S&D, Renew, ECR and PfE were, by majority, on the side of the official result; OpenRCV distinguishes four, such as “EPP, ECR and PfE without S&D”. It is a definition by OpenRCV, not a negotiated coalition.

More: Majorities: who carries a result

Majority of component Members

Some votes need more votes in favour than half of the Members in office on the day of the vote, for example objections to delegated acts. Unlike with a majority of the votes cast, abstentions and missing votes then have the same effect as votes against.

More: Majorities: who carries a result

Majority of the votes cast

Usually a majority of the votes cast is enough: more votes for than against, abstentions do not count, and a tie means rejection. Where the Treaties or the Rules of Procedure require it, more is needed, for example a majority of Parliament’s component Members, and for a motion of censure also two thirds of the votes cast.

More: Majorities: who carries a result

Non-attached

Non-attached Members do not belong to any political group, the alliances of Members of similar political leaning from several member states. Their votes count in the result like all others; where OpenRCV shows them together, they are not a political group with a common line.

More: Majorities: who carries a result · Video: The Parliament

Not participating (Council)

A member state does not participate in a Council vote when the act does not apply to it, for example Denmark and Ireland for parts of justice and home affairs, or states outside an enhanced cooperation. It then does not count towards any majority.

Objection under Rule 114 or 115

With an objection under the Rules of Procedure, Parliament opposes a delegated act (Rule 114) or an implementing act (Rule 115) of the Commission; voting for the objection therefore means voting against the act. An objection under Rule 114 needs a majority of Parliament’s component Members, under Rule 115 either that majority or a majority of the votes cast, depending on the act; OpenRCV takes the required majority from the official information.

More: Majorities: who carries a result

On the majority side

Being on the majority side means voting the way the plenary decides: For on an adopted text, Against on a rejected one. For an MEP the figure is the share of their own For and Against votes that match the result (abstentions are not counted); for a group, the share of votes in which its majority voted that way.

More: The metrics explained

Participation

Participation is the share of possible votes that were actually cast, for, against or as an abstention; only votes during the term of office count. Where a second figure is shown, it leaves out the days on which the MEP was officially excused.

More: The metrics explained

Procedure type and reading

The procedure type determines how Parliament and the Council take part: in the ordinary legislative procedure both decide as equals, in up to three readings; in the consent procedure Parliament only approves or rejects, and in the consultation procedure it gives an opinion and the Council decides. Most laws are now adopted at first reading, after Parliament and the Council have agreed a compromise in trilogue.

Video: How an EU law is made

Qualified majority

In the Council of the EU, an act is adopted by qualified majority if at least 55% of the member states vote in favour and they represent at least 65% of the population; if not all take part, only the participating states count. Abstentions count like votes against, and a blocking minority needs at least four member states.

Rapporteur

The rapporteur is a member of the committee responsible, sometimes together with others: they draft the report, negotiate changes to it and present the result in plenary. Motions for resolutions have no rapporteur; for them, the page shows who tabled them.

Resolution

With a resolution, Parliament states a position, for example on current events or the situation in a country; it has no legal force. Motions for resolutions are tabled by political groups, committees or individual Members; a joint resolution is based on a motion by several groups.

Roll-call vote

In a roll-call vote, the way each individual MEP voted is recorded. Votes by show of hands or electronic votes with totals only have no individual records, so OpenRCV covers roll-call votes only.

More: What is a roll-call vote? · Video: The Parliament

Shadow rapporteur

Members who follow a report for their group when another group holds the rapporteurship.

Speech

A speech is a contribution to a debate in plenary as the EP records it in its open data: who, on which item, when and for how long. Contributions from the chair and written statements are marked in the list of speakers and do not count in the speech figures.

Split vote

Split votes are the roll-call votes on a file before the final vote, for example on individual amendments, paragraphs (§) or recitals. They determine which text is put to the vote at the end; the final vote decides on the file as a whole.

More: Main vote and sub-votes

Subject area

A subject area is one of the 21 domains of the EuroVoc thesaurus, such as environment and climate or transport. A vote counts in every subject area that one of its document’s keywords belongs to, often in several.

More: Subject areas: what the votes are about

Time since the report

Days from the date of the committee report for plenary to the final roll-call vote. For legislation with an agreement with the Council, it includes the negotiations.

More: Plenary: agenda and proceedings

Verbatim report

The verbatim report of proceedings reproduces the debates in plenary word for word, each speech in the language in which it was given. OpenRCV shows this wording unchanged and does not translate it.

Vote correction

With a vote correction, MEPs put on record after a vote that they had intended to vote differently, for example after pressing the wrong button. The official result does not change; OpenRCV shows the original vote and marks the intended one.

More: Vote corrections

Voting map

The voting map places every MEP with enough votes on two axes so that those who vote alike sit close together. It uses all roll-call votes of the term; the axes are described only by which groups they separate.

More: The voting map

Written statement

Members can submit a written statement on a debate instead of speaking (Rule 178 of the Rules of Procedure). It appears in the verbatim report but was not spoken, so it has no time and no speaking time.

Source: European Parliament open data (CC BY 4.0), prepared by OpenRCV.