Approaches to Evaluating Welfare Reform: Lessons from Five State Demonstrations. 1. Issues


This section discusses two issues related to the use of samples defined by events since random assignment:

  1. Should impacts be estimated when administrative or survey data are incomplete or clearly incorrect for a large fraction (one-fifth or more) of the sample?
  2. Should impacts also be estimated for subgroups defined by employment or program participation decisions since random assignment? If so, how?