Many researchers blame voter registration requirements for inequalities in
turnout rates across various groups in U.S. society. The number of states,v
ith election-day registration (EDR) of voters doubled between the 1990 and
1994 elections, providing a unique opportunity to examine its impact on tur
nout inequality across demographic groups. The adoption of EDR is found to
be associated with large and significant improvements in the turnout rates
of young persons relative to older persons, and of recent movers relative t
o nonmovers. Turnout inequality by income class also declines with EDR adop
tion, but not by a significant amount in multivariate tests. The adoption o
f EDR does not improve equality of representation across educational levels
.