The Committee Remarks Archive
After each election, MPs are assigned to parliamentary committees. These committees prepare recommendations on legislative matters. When there is disagreement, parties or individuals can include separate remarks. These remarks form the basis of the Committee Remarks Archive.
About the Committee Remarks Archive
The basic unit in the archive is a remark. All remarks in reports from Storting’s standing committees, submitted to Storting or Odelsting between 1945 and 2005, are included. For each remark, parties are coded as either participating or not. Remarks are identified by session, report number, report type, committee, and page in the document.
Since 1996, the archive has been expanded to include:
- The full text of the remark from the report, linked to the coded remark
- Information on participation at the individual representative level
About Committees, Reports, and Remarks
The committees of the Norwegian Parliament (Storting) prepare cases to be debated and decided by the Storting or Odelsting. Each committee’s work corresponds to the responsibilities of one or more ministries, and the process usually begins with a proposal or report from the government.
After each parliamentary election, representatives are assigned to committees in proportion to their party’s size. Each representative serves only one committee, which elects its own chair, deputy chair, and secretary.
A large part of the legislative process takes place in the standing committees. Here, cases are discussed, and information is gathered from ministries, organizations, and industry. Committee members also consult their parliamentary party groups, so the parties’ positions often form the basis for further deliberations.
Committee meetings are closed to the public, no minutes are recorded, only final recommendations with proposals and remarks.
The recommendation includes:
- A brief description of the case
- Comments from the committee and/or its factions
- A proposal for a decision
- Summaries and document references
Some cases pass through the committee with unanimous support, but in many cases, there is disagreement. When this happens, the committee splits into two or more factions. The disagreement is formally recorded in the recommendation as a faction remark, a written statement expressing the faction’s position.
Background
The archive was initially compiled by Hilmar Rommetvedt at Rogalandsforskning as part of a European research project. NSD took over in 1985. Originally covering 1945–1977.
Data Quality
Text-based data is challenging to code consistently. Since the phrasing of dissent varies, coders must apply careful judgment. Sikt emphasizes accuracy and consistency when coding and reviewing data.
Use
Since each remark is coded by party support, the archive can be used to study coalition patterns and party alignment in committees. By analyzing agreement and disagreement, researchers can calculate political distance over time and across topics or committees.
Examples of Faction Remarks and Calculation of Political Distance
Example 1:
"The committee’s majority, all except the members from the Progress Party, points out that the conflict over the development of the Alta-Kautokeino watercourse in the 1980s made it clear that there was a great need in Norway to clarify the state’s relationship to Sámi culture and legal status (...)." (Recommendation O No. 80 (2004–2005), Standing Committee on Justice – Finnmark Act)
In this example, all parties except the Progress Party (FrP) support the statement. This is recorded as agreement between H, Ap, KrF, and SV – and disagreement with FrP.
Example 2:
"The committee members from the Socialist Left Party and the Centre Party believe that the industrial use of natural gas as a raw material should be given favorable framework conditions (...)." (Recommendation S No. 135 (2004–2005), Standing Committee on Energy and the Environment)
Here, SV and Sp are behind the remark. The other parties in the committee – KrF, FrP, Ap, and H – do not support it. This is recorded as disagreement between SV/Sp and the other parties.
How Political Distance is Measured
To calculate political distance between two parties (A and B), four possible scenarios are considered for each remark:
- Both parties participate (positive agreement)
- Neither party participates (negative agreement)
- Only party A participates
- Only party B participates
Scenarios 1 and 2 represent agreement, while scenarios 3 and 4 represent disagreement and form the basis for calculating political distance.
| Participation pattern | Number of remarks |
|---|---|
| Both participate | 2,346 |
| Neither participate | 2,906 |
| Only Sp participates | 1,163 |
| Only SV participates | 1,213 |
| Total | 7,628 |
Political distance is calculated as the share of disagreements (3 + 4) of the total: ((1,163 + 1,213) × 100) / 7,628 = 31.2%. This means that Sp and SV disagreed on 31.2% of the remarks they could have participated in during this session.
Comment on the Distance Measure
Rommetvedt’s party distance index originally includes cases of negative agreement (scenario 2), but this can distort results: two parties not supporting a remark does not necessarily mean they agree on the issue. Research Director Hilmar Rommetvedt therefore proposed an alternative method: only cases of positive agreement and disagreement are included.
The formula calculates:
- How often party A supports remarks that party B does not support
- How often party B supports remarks that party A does not support
- Then the average of these two percentages is taken
Example: Rommetvedt’s Index for Sp and SV (2004/2005)
RI_Sp = (1,163 × 100) / (1,163 + 2,346) = 33.1%
RI_SV = (1,213 × 100) / (1,213 + 2,346) = 34.1%
Rommetvedt Index (RI_SpSV) = (33.1 + 34.1) / 2 = 33.6%
This method provides an adjusted and more precise measure of disagreement based solely on active participation (positive agreement).
Sources
- Committee recommendations (published in Storting proceedings, vol. 6)
- 1945–1996: Manually registered
- From 1996/97: Downloaded from stortinget.no
About the Datasets
- Fraksjon_Parti and Fraksjon_Person: quantitative data on party and representative level
- Fraksjon_Merknadstekst: full-text remarks by paragraph, linked to party/representative data
Variable Description
The table has one entry for each committee recommendation. Recommendations to the Storting and Odelsting, both with and without comments, are included. The table currently covers the period from autumn 1945 to spring 2005.
Identification number where the first digits indicate the session number, the next digit shows the recommendation type and the last three digits are the serial number. Recommendation type is documented in the table i_innstillngstype, and id_innstilling 1421026 thus becomes recommendation no. 26 from session 142.
Internet address for the recommendation at www.stortinget.no.
Numeric code for the type of recommendation, for example recommendation to the Storting regarding a Storting proposistion. The codes are explained in the table i_innstillngstype.
The Storting's consecutive numbering of the recommendations within each recommendation type.
Reference to Storting document in connection with the recommendation. Example: "St.prp.nr.68 (2000-2001)".
Consecutive numbering of parliamentary sessions. See the table i_stortingssesjoner for details.
Date of the recommendation.
Numeric code for which Storting committee has made the recommendation. See the table i_stortingskomite for details.
Number of pages in the recommendation. May also indicate that the recommendation is composed of multiple documents.
Contains the text of the recommendation, divided into paragraphs and with reference to the factions behind any comments. The table currently covers the period from autumn 1996 to spring 2005.
Identification number where the first three digits indicate the session code, the next digit shows the recommendation type and the last three digits are the serial number. Recommendation type is documented in the table i_innstillingstype, and id_innstilling 1421026 thus becomes recommendation no. 26 from session 142.
Consecutive numbering of the paragraphs in the recommendation.
Numeric code for what kind of text the paragraph is about, for example 7 'Fraksjonsmerknad - første avsnitt' ('Faction remark - first paragraph'). The codes are explained in the table i_innst_avsnittype.
Sections with a text type that indicates a remark (code 7 or 8) also have a number that links the text to the remark information for the party and person. This number is repeated in the relevant tables. For other text types, this variable is NULL.
The actual text of the paragraph.
Remark participation at party level with emphasis on the role the party has played in the committee. The table currently covers the period from autumn 1996 to spring 2005.
Identification number where the first three digits indicate the session code, the next digit shows the recommendation type and the last three digits are the serial number. Recommendation type is documented in the table i_innstillngstype, and id_innstilling 1421026 thus becomes recommendation no. 26 from session 142.
Consecutive numbering in each recommendation of the paragraphs with remarks. Corresponds to the remark number in sto_frkmrk_innstillingstekst.
Is 1 if the faction remark has a majority in the committee, 0 if the remark belongs to the minority. Note that this coding differs from the older polsys table Fraksjon_parti, where the values were 2 and 1, respectively. When it is not known whether the remark belongs to the majority or the minority, the variable is NULL.
Numeric code for the party behind the remark. The code is explained in the table i_parti.
The party's role in the remark. Among parties, we count those that were represented in the Storting in the relevant session. The values are given in the table i_merknadsdeltakelse.
Contains faction remarks at representative level. For each paragraph with a remark in the committee report, participation or not for all committee members is shown here. The table currently covers the period from autumn 1996 to autumn 2004.
Identification number where the first three digits indicate the session code, the next digit shows the recommendation type and the last three digits are the serial number. Recommendation type is documented in the table i_innstillngstype, and id_innstilling 1421026 thus becomes recommendation no. 26 from session 142.
Consecutive numbering in each recommendation of the paragraphs with remarks. Corresponds to the remark number in sto_frkmrk_innstillingstekst.
Enumeration of the members of the committee.
Consecutive internal identifier from the table i_politikere.
The value 1 means that the representative did not participate in the comment, while 2 indicates participation.
The table shows the number of recommendations for each session, the number of recommendations with remarks, the number of remarks distributed by recommendation type and total. The table currently covers the period from session 90 (autumn 1945) to session 149 (spring 2005).
Consecutive numbering of parliamentary sessions. See the table i_stortingssesjoner for details.
Number of Storting recommendations.
Number of Odelsting recommendations.
Number of budget recommendations.
Number of Storting recommendations with remarks.
Number of Odelsting recommendations with remarks.
Number of budget recommendations with remarks.
Total number of remarks in Storting recommendations.
Total number of remarks in Odelsting recommendations.
Total number of remarks in budget recommendations.
The table documents the different types of recommendations: Storting recommendations, Odelsting recommendations and budget recommendations.
Data quality is generally good.
Sequence number for the type of recommendation.
Explanation of the code.
Numeric code for the type of text the paragraph is about.
1 Identifying paragraph
2 Title paragraph
3 Error
4 Heading paragraph
6 General content paragraph
7 Factional remark - first paragraph
8 Factional remark - continuation
Sequence number for the paragraph type.
Explanation of the code.
Any additional information.
Explanation of the codes for the different forms of participation in remarks.
0 Not represented/present in the committee
1 The party does not participate in the remark
2 The entire party faction participates in the remark
3 The majority of the party's faction participates in the remark
4 Half the party's faction participates in the remark
5 The minority in the party's faction participates in the remark
The variables are as follows:
Sequential number for the type of participation.
Explanation of the code.