U.S. Social Security benefits are widely recognized as redistributive
across class, race, and gender lines. However, this assessment is base
d on the impact of retired worker (contributory) benefits. The distrib
utional effects of noncontributory spouse and widow benefits, through
which two-thirds of older women make their claims, are markedly differ
ent. Class-based theorists examining the stratifying effects of welfar
e state programs focus on sources of bias that are disadvantageous to
lower-class workers while feminist accounts highlight sources of bias
that are disadvantageous to women. Using published and unpublished sum
mary data made available by the Social Security Administration, I demo
nstrate that while historical sources of gender bias linked to contrib
utory benefits have dissipated somewhat, race, class and marital statu
s biases linked to noncontributory benefits persist. Ultimately, benef
its linked to marital status are potentially as exclusionary for some
groups of women as are benefits linked to contributions for other grou
ps. I reject arguments that spouse and widow benefits may be justified
as rewards for unpaid domestic work or as mechanisms to target the po
ol: Noncontributory benefits are unrelated to domestic labor and syste
matically are most advantageous to middle-class and upper-class marrie
d White women who have never contributed to the system.