The data underneath Constitute are prepared as open-linked data in order to provide for efficient human and machine consumption, following the standards of the Semantic Web. For those unfamiliar with the standards and the rationale for such data, the W3C site and the Wikipedia entries on "Linked Data" and the "Semantic Web" are good places to start.

We provide Linked Data dumps in four different forms:

  • The countries.xml contains our data on countries; names, languages, borders, etc.
  • The topics.xml contains our data on topics, to allow tagging of content with specific labels and descriptions.
  • The chronology.xml contains our data on constitutional events over time.
  • The metadata contains our metadata about each constitution in a central file.


Analysts can also access "tabulated" versions of the excerpt data from the site itself. After "pinning" an excerpt, an entire search, or a comparison, visit the Pinned page to select and export custom slices of the data as .csv files (as well as .pdfs and Google Docs).


Developers and data enthusiasts can access Constitute's underlying data through an API, the documentation for which is here.


All of the data are from, or were collected for, the Comparative Constitutions Project. Some English translations are used with permission from HeinOnline and the Oxford Constitutions of the World. Many of the Arabic texts were prepared by International IDEA, in partnership with the Constitute team. Some of the Spanish texts were prepared in partnership with the Human Rights Lab of the University of Los Andes.