Measuring democracy, expert by expert
The V-Dem Institute, in the Department of Political Science at the University of Gothenburg, Sweden, is the home of one of the world’s most-used democracy datasets. The Varieties of Democracy project — co-hosted with the University of Notre Dame’s Kellogg Institute for International Studies — measures democracy through the judgments of country experts: hundreds of scholars worldwide, with country coordinators and regional managers, coding indicators that are then aggregated through a measurement model into indices. The result is a dataset its working papers describe as covering some 30 million data points for 202 countries from 1789, freely available to researchers and the public.
The Institute was established in 2014 by Professor Staffan I. Lindberg, according to secondary accounts — primary confirmation from the Institute’s own site is still pending, and the year is given here with that caveat. The wider data project held its launch conference in 2013 and released its first full public dataset in 2016.
A boundary to keep straight: “V-Dem” names both the Institute and the dataset it houses. This page is about the Institute. The dataset has its own record in the site’s Dataset Directory, and the two must not be conflated — the Institute produces the data; the dataset is the product.
What it produces
- The V-Dem dataset: expert-coded indicators aggregated into five core democracy indices — electoral, liberal, participatory, deliberative, and egalitarian democracy — with DOI’d versioned releases, downloadable data, and a codebook at v-dem.net.
- The annual V-Dem Democracy Report, the project’s flagship publication.
- The V-Dem Working Paper series, published openly.
For this site’s purposes, the V-Dem indices are the standard quantitative comparative-framing source for China’s regime classification and global democratization trends.
The method’s limits
The dataset’s core inputs are expert judgments aggregated through measurement models — subjectivity and coder-selection effects are its best-known methodological limitation, acknowledged in its own methodology and the scholarly literature, and not adjudicated here. V-Dem covers China across its full time series, but expert-coded scores for a closed polity depend on external observers and carry higher uncertainty than scores for open polities; no specific China score is asserted on this page.
Secondary accounts describe funding from government organizations, the World Bank, and research institutions; primary confirmation from the Institute’s funding statements is pending. No claim is made here about funding effects on measurement. The current director was not verified from the Institute’s site in this pass and is not named here.
No Chinese name for the Institute is confirmed on its own site, and none is used here.
Using the V-Dem Institute in research
Use the V-Dem indices as comparative framing with the methodology caveat stated alongside the numbers — expert-coded, model-aggregated, higher uncertainty for closed polities. Do not merge or rank V-Dem against Freedom in the World, Polity IV, or the EIU Democracy Index: different producers, different methods, different things measured. The full dataset record sits in the Dataset Directory; this institution page is its producer context.