Following recent work by Rosenzweig (1999), this paper,re-examines the effe
ct of Aid to Families with Dependent Children (AFDC) benefits oil nonmarita
l childbearing through age 22. Unlike most previous work, Rosenzweig finds
a statistically significant and quantitatively large positive AFDC effect,
Using data from the Panel Study of Income Dynamics, we replicate his analys
is and explore the reasons his findings differ from earlier research findin
gs. We are able to reproduce his main finding in a model that includes stat
e and cohort fixed-effects; we find that control for state effects increase
s the estimated AFDC effect. When we examine fertility separately by age, w
e find no AFDC effect on teen non-marital births, bur a targe effect on the
behavior of women in their early 20s.