Discusses 13 research papers prepared for the Griffiths subcommittee, generally and under three headings: (1) Program Impacts on Behavior, (2) Measurement, Mix, and Equity Issues Stemming from Welfare Benefits, and (3) Design and Administration Issues. All confirm the general impression of administrative complexity, inefficiency, and inequity. Assumptions disconfirmed by research results: Income maintenance has little effect on men's work effort. The specific pattern of benefit levels, guarantees, "tax" rates, etc., makes little difference. Job training programs are seldom effective. Excluding low income families with working fathers from AFDC is a logical incentive to family breakup, but the evidence shows no such effect. Administrative simplicity is impossible in a federal/state/local system. Local desires tend to conflict with federal policy. The path out of the welfare mess is not clear, but meanwhile current gross inequities could be eliminated. 27 pp.