GGSR expects authors to make the evidence underlying their conclusions as transparent as ethical, legal, and contractual obligations permit. Governance research may draw on administrative records, interviews, surveys, archival documents, legal sources, digital traces, and analytical code. The appropriate form of access depends on the material and the risks of disclosure.

Data-availability statement

Every research manuscript should include a statement identifying the evidence supporting the study and how it can be accessed. Provide repository links and persistent identifiers where available. Explain any embargo or restriction and the legitimate procedure for requesting access. If no new data were generated, or the study relies entirely on public sources, state this and identify those sources.

The statement should cover relevant code and documentation as well as data. A general assurance that material is available on request is insufficient when access depends on conditions that readers cannot identify. Explain who can authorize access and any restrictions affecting independent verification.

Documentation suited to the design

• For quantitative work, provide variable definitions, source and version information, inclusion and exclusion rules, transformations, and the code needed to reproduce reported analyses where sharing is feasible.
• For qualitative work, explain the source or participant selection, interview or observation procedures, and coding or interpretive approach. Share instruments, coding frameworks, or suitably protected excerpts when these can be released responsibly.
• For documentary and institutional analysis, identify the document series, archival references, collection dates, and relevant versions. Explain how evidence was selected and interpreted.
• For computational work, describe software, dependencies, parameters, and data-processing steps sufficiently to assess the results.

Use a suitable institutional or subject repository where practicable. A deposit should include a readable description, file inventory, and usage conditions. Cite datasets and software as research outputs rather than only mentioning them in a footnote.

Confidentiality and restricted access

Open sharing must not expose participants, confidential institutional records, copyrighted content, or material covered by consent and contractual restrictions. Removing names may not prevent identification when combinations of roles, places, dates, and quotations reveal a person or organization. Authors should assess these risks before depositing data.

Where public access is inappropriate, explain the restriction and consider feasible alternatives such as metadata, instruments, aggregate results, synthetic examples, or controlled access. Editors may request supporting records for confidential assessment when appropriate. Any proposed transfer must respect the relevant authorization and safeguards.

Accuracy and later updates

Check that shared files correspond to the reported analysis and that links work at acceptance. If a dataset or code error is discovered later, notify the journal and identify the affected results. Updating a repository does not replace a correction to the article when the published interpretation changes. See Data Integrity and Corrections.