Participatory Budgeting Preferences Data Set

The data set includes information about 125 elections (configuration settings and descriptive statistics), 1499 projects and 125k anonymized voters and their budget preferences. Preferences were sollicited with different elicitation methods (K-approval, knapsack, K-ranking and K-token). For some elections, voters provided also preferences under a secondary elicitation method, resulting in vote pairs from the same voter on the same budgeting question but with a different elicitation method.

This data set was collected as anonymized votes from the Stanford Participatory Budgeting Platform. The source code for this platform is available on https://github.com/StanfordCDT/pb . The data set (exported on 22 May 2023) includes 124 primary ballots (79 K-approval, 21 knapsack, 13 K-ranking) and 38 secondary ballots (2 K-approval, 18 knapsack, 18 K-ranking).