What the democracy indices measure
This site cites democracy indices throughout the contemporary sections. A reader who cites an index without knowing what it measures is easy to discredit, so here is what each one actually is.
V-Dem
The Varieties of Democracy project measures over 600 attributes of democracy for 202 countries and territories from 1789 forward, using more than 4,200 country experts whose codings are aggregated through a measurement model that produces scores with explicit confidence intervals.claim type: fact Its headline Liberal Democracy Index combines electoral democracy with liberal constraints: courts, legislature, civil liberties. Strengths: historical depth, transparent methodology, uncertainty estimates, and an episodes approach that identifies autocratization and democratization as processes with start and end dates. The U-turns research this site cites repeatedly is built on that episodes framework.claim type: fact Limits: expert coding is judgment, however carefully aggregated, and year-to-year movements inside confidence intervals should not be over-read.
Freedom House
Freedom in the World scores political rights and civil liberties on a 100-point scale from analyst assessments against a checklist, yielding the Free, Partly Free, and Not Free categories. Strengths: long continuous series since 1972, country narratives, accessibility. Limits: coarser scale than V-Dem, a methodology more dependent on individual analysts, and a US-based institutional perspective its critics note.
The EIU Democracy Index
The Economist Intelligence Unit scores sixty indicators across five categories, producing the full democracy, flawed democracy, hybrid regime, and authoritarian regime labels journalists reach for. Its “flawed democracy” band is wide enough to contain very different situations, which is exactly why the label circulates so well and informs so little.
Bright Line Watch
A US-focused academic project surveying political scientists and the public on the same battery of democratic-performance questions since 2017. Its value is granularity on one country: which specific norms experts see eroding, and how expert and public perception diverge. It is a perception instrument, not a structural index, and honest citation says so.
How to use them
Where the major indices agree, and on the direction of movement in the large recent cases they do, the convergence of independent methodologies is itself evidence. Where they disagree, the disagreement is usually about thresholds and categories, not raw facts.claim type: inference The practice this site follows: cite the specific finding, not the brand; prefer the index whose method actually measures the thing being claimed; and never let a single-year category change carry an argument that the underlying data will not.