Archive of Voting Records
The Voting Archive contains the historical roll-call votes from 1814–1940 and data from the Norwegian Parliament’s electronic voting system, starting in the 1979/80 session. It includes all non-unanimous votes in a continuous time series.
The data includes individual voting records, party codes, and matter references.
About the Voting Archive
The Voting Archive covers two periods: the historical roll-call votes from 1814–1940, and the votes from the electronic voting system from the 1979/80 session up to the most recent completed parliamentary session. The historical period 1814–1940 is described in more detail further down this page.
The archive from 1979 onward includes only votes cast using the electronic system. Unanimous decisions and some proposals, often from smaller parties, were handled without formal voting and are therefore not included. This can give the impression of more political agreement than actually existed.
The number of recorded votes increased significantly in the 1990s due to changes in voting procedures and more proposals. The budget reform introduced in 1997 led to fewer votes per session.
The variable id_vot links the voting tables together by uniquely identifying each vote. It is based on the Storting's internal numbering system but adapted to account for variations over the years. Additionally, id_vot is designed to be sorting-friendly, ensuring the correct chronological order of votes. Note that the date part of id_vot is not necessarily correct – use id_vot as an identifier, and read the date from the date column:
-
Sessions 0–89 (1814–1940):
An eight-digit number consisting of the year, chamber and a running vote number within the year and chamber.
Example: 19383006 is the id_vot for the 6th vote in the Lagting (chamber 3) in 1938. -
Sessions 124–154 (Autumn 1979 – Spring 2010):
Voting numbers restarted from 1 for each matter, and matters restarted from 1 for each agenda—typically one per day. id_vot was composed of date – agenda – matter – vote number.
Example: 2007-05-10-077-01-003. -
Session 155 (Autumn 2010 – Spring 2011):
Each matter received a unique running number. id_vot was simplified to: date – matter – vote number.
Example: 2011-05-10-50324-001. -
From Session 156 (Autumn 2011) onward:
Each vote received its own unique running number. id_vot was further simplified to: date – vote number.
Example: 2016-05-10-07213.
Background
Data from 1979–1989 was manually registered by NSD using printouts and parliamentary records.
From 1989/90 onward, data is transferred directly from the Parliament's systems. NSD supplements this with matter information and subject codes.
Except for the period 1981–1989, the data includes how each representative voted.
Data Quality
The data quality is generally high. Some votes, particularly from smaller parties, are missing because they were not registered through the electronic system.
This can affect analyses of party distance or agreement.
Use
Voting data is often used to measure political distance between parties.
This is done by comparing how often two parties vote the same, resulting in an agreement index ranging from 0% to 100%.
These results can be analyzed by session, topic, committee, or matter type.
Examples
| Vote | Party A (%) | Party B (%) | Disagreement Index | Agreement Index |
|---|---|---|---|---|
| 1 | 100 | 100 | 0 | 100 |
| 2 | 0 | 100 | 100 | 0 |
| 3 | 100 | 0 | 100 | 0 |
| 4 | 60 | 40 | 20 | 80 |
Representative Level
The representative-level tables show how each individual representative voted in each vote. They can be linked with biographical and matter data.
Party Level
The party-level tables present aggregated voting results at the party level, showing the number and percentage of votes for or against proposals.
Sources
- Parliament's electronic voting system
- Stortingstidende (the official report)
- Parliamentary Records, Part 9 (register volume)
The Voting Archive 1814–1940
The historical part of the archive contains around 7,500 divided votes from the Storting's plenary sessions, Odelsting, and Lagting in the period 1814–1940. All votes were conducted by roll call, making it possible to see exactly how each representative voted. The archive offers unique insights into Norwegian political history, party alignments, and parliamentary decision-making in important and often controversial matters.
The source material for the personal votes 1814–1940 does not show whether the representative voted for or against the proposal, but whether the person voted with the majority or with the minority. The votes are being reviewed from 1940 and backwards in time to determine whether each vote was for or against, but this is an extensive task that is far from complete. For the years where the vote is still only known as majority/minority (at the time of writing 1814–1917), the personal and party votes are held in separate tables (see vot_person_eldre_flertall_mindretall and vot_parti_eldre_flertall_mindretall in Variable Description).
Content
The historical part includes all divided votes with roll call from 1814–1940, that is, matters where representatives disagreed, and each vote was recorded by name.
Historical calculations show that unanimous votes were often far more common than divided ones. For example, in 1893 there were 427 unanimous votes compared to 231 divided votes; in 1896 the numbers were 451 versus 187. Between 1892 and 1897, roll call votes accounted for about 40% of all divided votes. Based on these figures, the archive covers approximately 20% of all votes during that period.
The use of roll call was not random. According to historian Ottar Dahl (1972), it was applied to matters considered particularly important or controversial, based on political judgement by the Storting's president or individual representatives. Divided votes without roll calls often had larger majorities, while roll calls tended to be used in matters with sharp political divisions. After the mid-1920s, roll call became less common, likely due to changes in parliamentary procedures and working methods.
Background
The voting data was collected at the Department of History at the University of Oslo between 1968 and 1972, under the direction of historian Ottar Dahl. The material was initially recorded electronically on punch cards and stored in two separate datasets: a representative register with the names of all those who voted, and the voting itself, that is, how each representative voted in each matter.
The matter descriptions originally existed only in written form. When the material was transferred to NSD (now part of Sikt – the Norwegian Agency for Shared Services in Education and Research), the information was digitised and made machine-readable.
For the period 1857–1940, NSD added:
- Subject codes from the Storting's thesaurus (also used for modern voting data)
- Identity numbers from the Politician Archive, enabling analyses of voting behaviour by profession, education, party, and geographical affiliation
Data Quality
The quality of the data varies between periods. There is no full documentation of the original registration process, and no systematic checks have been made against original sources such as Stortingstidende. Nevertheless, the material provides valuable insight. Evidence suggests that different individuals worked on the material at different times, which may explain quality variations. Further analysis is available in John-Erik Ågotnes: Roll Call Votes (NSD Report No. 1, Bergen 1989).
Sources
- Main source: Stortingstidende
- Other sources: Punch card data and a printed index compiled by Ottar Dahl, University of Oslo
Variable Description
Main table, giving an overview of the votes. So far we have registered votes for the period 1814–1940 and from 1979 up to the most recent completed parliamentary session.
Identifies the individual vote. Based on the Storting's own numbering, but adjusted for the fact that this numbering has varied over the years. In addition, id_vot is sorting-friendly, so that the sort order gives the correct order of the votes. The many different formats reflect the origin of the voting data.
Note that the date part of id_vot is not necessarily correct. Treat id_vot as an identifier, and read the date from the date column.
- Sessions 0–89 (1814–1940): An eight-digit number consisting of the year, chamber and a running vote number within the year and chamber. Example: '19383006' is the id_vot for the 6th vote in the Lagting (chamber 3) in 1938.
- Sessions 124–154 (1979–2010): The numbering of votes started at 1 for each matter voted on, and the numbering of matters started at 1 for each agenda, usually one agenda per day. The different chambers (plenary, Odelsting and Lagting) have separate series of agenda numbers. For these years id_vot is composed of date – agenda – matter – vote number. Example: '2007-05-10-077-01-003'.
- Session 155 (2010–2011): Each matter received a unique running matter number from the Storting. This simplified id_vot to date – matter – vote number. Example: '2011-05-10-50324-001'.
- From session 156 (2011–2012) onwards: Each vote received its own unique running vote number from the Storting, and id_vot was simplified to date – vote number. Example: '2016-05-10-07213'.
The Storting's running identifier for the vote. NULL for older votes, before session 156 (2011–2012).
Date of the vote. The votes from 1814–1940 lack a date, and the value is NULL, but the year can be read from id_vot. Be aware, however, of sessions that span two calendar years.
Hour, minute and second of the vote. Kept in a separate variable since sittings in the Storting can extend past midnight. The time of the vote will then technically be the next day, but is still registered on the date when the sitting started. If, for example, the date is 10 May and the time is 01:23, it means in reality that the vote took place on 11 May, since the date changed when the clock ticked past midnight. With date and time in the same variable, this link would be problematic.
Numbering of the period when the Storting sits together, currently from autumn to spring. See i_stortingssesjoner. Example: Session 166 indicates H2021–V2022.
Number referring to the agenda. Usually there is one agenda per day. The numbering starts at 1 for a new session.
Which stage of consideration this vote concerns. For legislative matters there is consideration in two sittings in plenary, in some cases three sittings.
Code indicating whether the vote took place in plenary, the Odelsting or the Lagting. See the table i_vot_sal for codes.
Vote number from 1 and upwards within the relevant matter.
If the vote is alternative, that is, two proposals set against each other, in modern times two votes are held, For or Against each of the proposals. votnr_alt then holds the votnr of the other vote, where the counter-proposal is voted on. Earlier, the alternatives could be set against each other in a single vote, in which case votnr_alt equals votnr. It is -1 if the vote is not alternative, NULL if data is missing.
What type of vote (the committee's proposal, private, alternative, etc.) it concerns; NULL for missing data. The table i_vot_vottyp contains codes, which are currently:
- 1 Vote on the recommendation's proposal
- 2 Vote on a personal proposal
- 3 Alternative vote between the recommendation and a personal proposal
- 4 Alternative vote between two personal proposals
What the vote is about, in plain text.
Number of votes for if known. NULL if unknown, or in the case of a unanimous decision. In a direct alternative vote, as was common in earlier times, ant_for is the number of votes in the alternative that gained a majority.
Number of votes against if known. NULL if unknown, or in the case of a unanimous decision.
Number not present if known. NULL if unknown, or in the case of a unanimous decision.
Is 1 if the proposal was carried, 0 otherwise.
How the result of the vote came about, in plain text as defined by the Storting:
- enstemmig_vedtatt – Carried unanimously
- ikke_spesifisert – Presumably always a vote using the electronic system
- manuell – Manual vote
- vedtatt_med_president_dobbeltstemme – Tie, carried with the President's casting vote
- forkastet_med_president_dobbeltstemme – Tie, rejected with the President's casting vote
- vedtatt_mot_N_stemmer – Carried against the number of votes given in ant_mot
- forkastet_mot_N_stemmer – Rejected against the number of votes given in ant_for
Is 1 if the vote relates to the matter as a whole, 0 if it relates to one or more voting proposals. (The Storting's data service calls this variable Fri_votering.)
Is 1 if the vote was actually carried out, 0 otherwise, such as when the outcome is clear in advance.
Who was President of the Storting during the session. See the table i_politikere for names and other personal data. See the table sto_representant for party.
The volume of Stortingstidende (the official report) in which this vote was published.
Page number in Stortingstidende (the official report) where the vote is described.
Any remark.
The votes' connection to matters.
Identifies the individual vote. Mandatory. See the description under vot_votering for details about the different formats of id_vot.
Running matter number, index in sto_sak. NOT NULL
Here we find what each individual representative voted in each vote. In December 2025 the personal voting data was updated with data from data.stortinget.no. The update covered votes from autumn 2011 to spring 2025. This data replaced the old data for that period, that is, from autumn 2011 onwards. Earlier personal votes were taken from the original table, now renamed vot_personvotering_gml.
Identifies the individual vote. Mandatory and unique. See the description under vot_votering for details about the different formats of id_vot.
The break in format, which currently lies between session 154 and 155, may shift backwards in time, since we now have unique matter numbers all the way back to 1986.
Identifies the representative who voted. See the table i_politikere for names and other personal data. See the table sto_representant for party. As the representative is unknown in some votes, only identified by seat number, party is also introduced here as a variable.
Where in the chamber the representative was seated.
The representative's party at the vote. To ensure safe updating, the party should really be found by looking up the table sto_representant, but personal identification is often missing in the old voting tables, especially session 155, and then the inherited variable parti is the closest we get.
Indicates the vote cast. The table i_vot_votering contains codes, which are currently:
- 1 For
- 2 Against
- 3 Absent
NOT NULL
Votes in the Storting broken down by party. Based on the table vot_person.
Identifies the individual vote. Mandatory and unique. See the description under vot_votering for details about the different formats of id_vot.
Numeric party code defined in the helper table i_parti.
Indicates the vote cast. The table i_vot_votering contains codes, which are currently:
- 1 For
- 2 Against
- 3 Absent
NOT NULL
Number of votes with this value of the variable votering.
The source material for personal votes 1814–1940 does not show whether the representative voted for or against the proposal, but whether the person voted with the majority or with the minority. The votes are being reviewed from 1940 and backwards in time to determine whether each vote was for or against, but this is an extensive task that is far from complete. This table covers the years in which the variable votering still means 1 = the majority and 2 = the minority, at the time of writing the years 1814–1917.
Identifies the individual vote. Mandatory and unique. See the description under vot_votering for details about the different formats of id_vot.
Identifies the representative who voted. See the table i_politikere for names and other personal data. See the table sto_representant for party. As the representative is unknown in some votes, only identified by seat number, party is also introduced here as a variable.
Where in the chamber the representative was seated.
The representative's party at the vote. To ensure safe updating, the party should really be found by looking up the table sto_representant, but personal identification is often missing in the old voting tables, and then the inherited variable parti is the closest we get.
Indicates the vote cast:
- 1 Voted with the majority
- 2 Voted with the minority
- 3 Absent
NOT NULL
Votes in the Storting broken down by party. Based on the table vot_person_eldre_flertall_mindretall.
The source material for personal votes 1814–1940 does not show whether the representative voted for or against the proposal, but whether the person voted with the majority or with the minority. The votes are being reviewed from 1940 and backwards in time to determine whether each vote was for or against, but this is an extensive task that is far from complete. This table covers the years in which the variable votering still means 1 = the majority and 2 = the minority, at the time of writing the years 1814–1917.
Identifies the individual vote. Mandatory and unique. See the description under vot_votering for details about the different formats of id_vot.
Numeric party code defined in the helper table i_parti.
Indicates the vote cast. The table i_vot_votering contains codes, which are currently:
- 1 Voted with the majority
- 2 Voted with the minority
- 3 Absent
NOT NULL
Number of votes with this value of the variable votering.