Research preview. Scores are automated LLM extractions. Country means track survey-based affective polarization (r = 0.35 across 37 democracies, see Validation); item-level audits continue. Only country-years with at least 100 out-party mentions, and countries with at least five such years, are shown. Do not cite these numbers without checking with the author.
🦀 Parliamentary Polarization Map

Elite polarization,
measured from parliamentary speech

An interactive map of how legislators talk about their own and rival parties, built from millions of political mentions extracted across national parliaments. It includes the world map, country trajectories, cross-national comparisons and the underlying data.

How the map is built

This project aggregates parliamentary data worldwide to measure elite polarization by analyzing what politicians say about one another inside parliaments. It employs an agent-AI-driven framework for parliamentary data collection, homologation, and analysis, orchestrated through the OpenClaw framework.

The system collects transcripts from academic archives across many countries. When corpora exist only on parliamentary websites, it uses web-scraping techniques to acquire them. Because many parliaments publish proceedings as PDF scans, the pipeline separates born-digital PDFs, where text is extracted directly, from scanned-image PDFs, where optical character recognition is applied. For video-only records, it extracts official captions or subtitles; otherwise, it tests automated speech recognition on a sample before transcription.

OpenClaw coordinates the workflow across collection, normalization, affiliation recovery, extraction, validation, and escalation. Collected transcripts are normalized into a unified format in which each row is one speaker’s turn, with country, date, speaker name, party, and speech text. Since speaker–party affiliation is often missing, a recovery algorithm uses registry and metadata joins, deterministic date-window rules for known affiliations over time, and fuzzy matching against biographical and parliamentary records. Unresolved cases then fall back to language-model lookup.

The extraction stage uses the Google Gemma 4 26B LLM to identify passages where a speaker evaluates another named politician, distinguishing attacks or praise from neutral and procedural references. Each mention is coded as in-group or out-group using recovered coalition affiliation. The output is a cross-national dataset of elite-hostility relationships, aggregated by party-pair and time period, covering over 120 million speeches across 139 countries.

How the corpus was acquired

139 corpora · 119.6 million speeches (+ 3.9 million broadcast-caption segments), grouped by how each parliament’s record was obtained.

Ready-downloadincl. ParlaMint releases
105
101.8M speeches
Web-scrapecustom site scrapers & APIs
13
11.8M speeches
PDF / OCRborn-digital + scanned-image text
18
6.0M speeches
Video captions / ASRbroadcast captions & speech recognition
3
3.9M segments

World map

Choropleth of any indicator, pooled across all years (mention-weighted) or for a single year.

Country profile

Each party's trajectory within a country — out-party evaluation by default. Countries without a party breakdown show the national aggregate.

Compare indicators

Scatter two indicators against each other for a given year. Bubble size = corpus volume; colour = region.

Heat map

Country × year grid for one indicator. Grey cells have no data.

Data table

The underlying panel. Switch between country-level and party-level rows; filter, search and sort.

Can these numbers be trusted?

The map is built by an AI model reading parliamentary speeches. Below are the checks we ran on it, each as a question with a short answer. Terms are explained at the bottom of this box.

What the terms mean

All checks with sample sizes (technical table)
CheckSampleResult

Do politicians' words match what voters feel?

Each dot compares two things for one election. Across: how harshly members of parliament spoke about other parties in the years before that election (higher = harsher). Up: how much voters disliked the other parties in the election survey, measured by Orhan's (2022) Affective Polarization Index. If the dots rise from left to right, harsher parliaments go with more divided voters.

Left: one dot per election. Right: one dot per country, averaging its elections. Across: how harshly members of parliament spoke about other parties in the years before the election, as the average hostility of their mentions of other parties (0 = neutral; higher = harsher). Up: how much voters disliked the other parties in the election survey (Orhan's index). The dashed line shows the overall trend. Hover over a dot for the country and year.

Countries

Countries compared with the trend line. "Voters vs parliament" is how far the country sits above or below the line: a positive number means voters are more divided than the tone of their parliament would suggest; a negative number means the parliament is harsher than its voters.

CountryElections Hostility in parliamentVoters' dislike (survey)Voters vs parliament
Survey values from Orhan (2022, Democratization); parliamentary data as of . Countries whose parliaments we do not yet hold for the survey years are not shown.