TANF Data Samples

This repository includes sample synthetic data in the initial TANF data model structure. The TANF data model is comprised of two datasets, one at the case level and one at the member level. The table below briefly describes the TANF data model components. This model is intended to be flexible depending on state interests and data availability, but it is generally designed to facilitate analyses around caseload dynamics, churn, and recidivism.

Core Data Elements	Purpose	Examples
Individual identifier	Identify members and link to cases	client ID, member ID
Case identifier	Identify cases and link to members	case ID
Case type	Create categories meaningful to program administrators	basic cash assistance, child only
Geographic indicator	Classify by jurisdiction	county name, region
Basic client demographics	Characterize client population	DOB/age, gender
Additional Data Elements
Case status indicators	Analyze specific issues; adjust for abnormalities in the data	paid late, timed out, recipient of specific intervention
Other benefit receipt	Observe other program participation	SNAP receipt, Medicaid receipt
Head of household identifier	Identify case head of household	HOH client ID
Additional client demographics	Characterize client population more finely	Marital status, race/ethnicity, education attainment, citizenship, refugee

Full documentation for the data model structure is available in an FSSDC research brief.

The R script sample_data_introduction.R reads in the sample TANF data and provides a basic introduction to this data model.