Measuring State TANF Policy Variations and Change After Reform *

Objectives. Our objectives are to describe the policies adopted after PRWORA, which vary across states, to test for common underlying policy concepts, demonstrating how these policies are interrelated, and to examine whether policy stringency diffused to neighboring states results in greater policy stringency across all states over time. We convert textual TANF welfare guidelines into empirically derived policy dimensions and use the derived quantitative scores to describe variation and change in welfare policy dimensions across status during the 1996–2003 post‐welfare‐reform period. Methods. Utilizing the Urban Institute's Welfare Rules Database, we apply a factor analytic methodology to 78 unique state policy guidelines that were coded on a lenient‐to‐stringent continuum. Regression analyses, employing spatial contiguity weighting, are used to describe policy diffusion. Results. The results identified 15 underlying first‐order post‐welfare‐reform policy dimensions, which for scientific parsimony were further reduced to three second‐order underlying dimensions representing rules governing eligibility: eligibility requirements for groups, behavioral responsibilities for maintaining eligibility, and eligibility time limits and exemptions. Analysis of the quantitative scores showed that by 2003 states had become more lenient regarding eligibility criteria for groups but decidedly more stringent regarding behavioral guidelines for maintaining eligibility and eligibility time limits and exemptions. Spatial clustering is not found globally but is significant for some states at the local level. Spatial diffusion is apparent only for behavioral rules. Conclusions. Our results suggest that TANF policy variations across states go beyond payment levels to include salient eligibility rules. The patterns of variability in change scores across states do not support a pervasive “race to the bottom” conclusion.