This document outlines the pre-analysis plan for the Federal Reserve Bank of Minneapolis’ analysis of the City of Minneapolis’ Guaranteed Basic Income (GBI) Pilot. The study contains 530 households drawn from 9 ZIP codes in Minneapolis. Treated households receive $500 per month for 24 months. Control households receive $150 for completing surveys at 6 months, 12 months, and 24 months. Participants were randomly assigned to treatment within 8 strata defined by poverty, presence of children, and two ZIP code groups. Early enrollment data showed significant differential attrition. The plan addresses biases from selective attrition by (a) conditioning on age, baseline income, and education (b) controlling for the baseline outcome variable and (c) estimating treatment/control contrasts for the change in outcomes (difference-in-differences).
Experimental design
Stratified randomized controlled trial.
Design details
Randomization method
Randomization done by office on a computer. Random list within strata generates first 199 treatment offers, 200th spot and any treatment refusals filled from a randomized list of the remaining participants.
Randomization unit
Household.
Clusters
530 participants at baseline, 288 usable observations at 6-month follow-up.
Housing Quantity (INDEX_HQUANT), Use of Low-Cost Credit (INDEX_CRED), Healthcare Utilization (INDEX_HUSE), Healthcare Access (INDEX_HACC), Respondent employed in the week before the survey, Respondent works multiple jobs, Respondent’s household could cover a $400 expense, Respondent reported or implied hourly wage (among workers), Respondent’s household provides financial support to others, Respondent’s household not flagged on common housing instability screener, Respondent transportation access, Respondent school/training attendance, Points in the cumulative distribution of annual household income (asked only at annual intervals)
Cited by
No citation data available for this registration yet.