Replication Data for Overpromising Populists: Party Manifesto Promises and Government Data

Replication data for a 2026 study on political party promises published in the European Journal of Political Research. The dataset includes files for statistical analysis, robustness tests, and validation of machine predictions. It was authored by Mathias Bukh Vestergaard and hosted on Harvard Dataverse.

Use Cases

• Replicate statistical analyses on party promise-making based on the provided do-files and data.
• Test the validity of machine learning predictions for political text using the included test_pred files.
• Analyze issue categories in party manifestos using the CMP-codes categorization spreadsheet.
• Examine U.S. government (president) data in relation to party promise patterns.

Strengths

• Includes replication code (main.do) and a script (dsl.R) for running robustness tests, facilitating reproducibility.
• Contains specific validation files (test_pred) for testing machine prediction models mentioned in the description.
• Last updated on 2026-06-27, indicating recent maintenance.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and overall dataset scale are unknown, which may limit suitability assessment.
• Description metadata is limited; actual data quality and completeness require manual inspection.

Provenance

Source

Harvard Dataverse, authored by Mathias Bukh Vestergaard.

Collection Method

Likely compiled from sources like the Comparative Manifestos Project (CMP) and U.S. government data for replication.

Freshness

Last updated 2026-06-27 15:09:24; freshness should be verified.

Geography

Includes data for the United States; broader geographic scope is suggested but unspecified.