Variables
The dataset covers 6,358 legislative processes from the 7th up to the 21st legislative term of the German Bundestag (1972–2026). Below are the variables of the raw dataset, followed by the fields the Data Explorer derives from them.
Category values — subject areas, committees and initiators — are the official German designations of German institutions and are shown untranslated, so that every value can be traced back to the source.
Variables of the dataset
| Variable | Description |
|---|---|
id | Unique identifier of the bill in the Documentation and Information System for Parliamentary Processes (DIP) |
title | Title of the bill |
gsig | Parliamentary-archive signature, if present |
start | Date of the first procedural step counted towards the beat calculation (usually the government or parliamentary bill) |
end | Date of the last resolution of the legislative process counted |
startEpoch | Start date as epoch days (numeric value for calculations) |
startYear | Calendar year of the start date |
wp | Legislative term of the German Bundestag in which the bill was handled |
duration | Length of the process in calendar days, including the start and end dates |
steps | Number of weighted procedural steps counted towards the beat calculation |
beats | Average number of weighted procedural steps per calendar day (steps divided by duration) |
tillTermEnd | Calendar days between the start of the process and the end of the respective legislative term |
committeeInv | Total number of committee referrals; referrals to several committees are each counted individually |
plenary | Number of plenary deliberation stages counted; combined deliberations are weighted by the stages they contain |
mediationEv | Number of mediation-procedure steps counted (e.g. mediation or conciliation proposals) |
committeeOut | Number of committee outputs counted, in particular recommendations for decisions and committee reports |
divGov | Measure of the party-political alignment of the governing majorities in Bundestag and Bundesrat at the start of the process |
iniGroup | Initiator group: Federal Government, Bundestag party group, Bundestag committee, Bundesrat/Land, or mixed |
zustimmung | Normalised consent requirement: yes, no, inconsistent, or not recorded |
mediation | Boolean: true if a mediation procedure took place |
sachgebiete | Array: subject areas assigned to the bill according to DIP |
initiativen | Array: initiator(s) of the bill (e.g. the Federal Government, the Bundesrat, or Bundestag party groups) |
ausschuesse | Array: all committees the bill was referred to within the period considered |
zustimmungRaw | Array: the unnormalised consent verdicts (raw) |
Fields derived in the Data Explorer
These are not part of the raw dataset; they are computed when the data is prepared for the explorer.
These fields are available in the raw dataset and may be normalised:
| Field | Description |
|---|---|
iniGroup | Group formed from the set of initiativen: Federal Government (BReg), Bundestag party group (BT), Bundestag committee (BTA), or Bundesrat / Land (BR). Processes with initiators from more than one of these groups are labelled mixed, without any entry are labelled not recorded. |
zustimmung | Normalisation of zustimmungRaw: each entry is reduced to the “ja” or “nein” before its first comma. A process carrying both verdicts is labelled inconsistent (278 processes). |
mediation | Boolean: true where mediationEv is greater than 0. |
Multiple values
Four columns hold lists rather than single values: sachgebiet, initiative,
ausschuesse and zustimmungsbeduerftigkeit. Roughly half of all processes are
assigned to more than one subject area, and the median process passes through
five committees.
That matters for any aggregation, so the explorer offers two counting modes:
- Primary value (default) — only the first listed value counts. Bar totals therefore equal the number of processes, and relative shares add up to 100 %.
- All values — each process counts in every one of its categories. Totals then exceed the case count, and shares are relative to mentions rather than to processes.
Processes without an entry (such as the 864 with no subject area) form their own not recorded category in both modes, so they never disappear from the distribution.
Independently of this setting, filters always match against all values of a process: filtering on the Finanzausschuss finds every process that committee touched, not only those where it happens to be listed first. Several selected values within one field are combined with OR, different fields with AND.
The data is not yet available for download — if you are interested in using it, please get in touch.