members gives their party. The post-Second World War era was particularly low,
and the contemporary period (as in the nineteenth century) is considerably higher.
Even at lowest ebb, however, party had by far the largest eVect on the casting of the
roll call vote (Weisberg 1978 ). He put the scholarly challenge to be to take party
voting as the base line, with the theory tested by seeing how much it improved on
party-line voting. This was at a point when the Democratic majority was divided,
with nearly as many votes being cast with a ‘‘conservative coalition’’ (a majority of
northern Democrats opposed by a majority of Southern Democrats and a majority
of Republicans) as were being cast along party lines. To be sure, claiming a vote is a
‘‘party vote,’’ when a simple majority of one opposes a simple majority of the other
party, set a modest standard, even though the eVect of party on the individual vote
was stronger than that aggregate pattern suggested. Still, if congressional voting
was primarily ‘‘party plus,’’ it was nonetheless the case that party was much less
consequential in shaping legislative choice than in virtually all other legislatures.
The above reXect, in eVect, a parallelism between citizens’ and legislators’ voting
choices. In both cases, party served as a strong base line, but there was more. In
both cases, the role of party reached a low point at about 1970 , climbing back to a
more historically precedented high level more recently.
Congressional theory sought to explain the variation in levels of party voting
and, at least indirectly, answering the question of why the US Congress lagged its
European counterpart, even at its contemporary higher levels of party voting. The
theoretical literature poses three explanations (for reviews see, e.g., Aldrich and
Rohde 2000 ; Cox and McCubbins 2005 ). One is that the observed levels of party
voting revealed very little to do with the role of party in Congress. Championed
most vigorously by Krehbiel ( 1991 , 1993 ), his argument is that the pattern of party
voting in Congress mirroring that of party voting in the public is no coincidence.
Legislators’ votes reXect the wishes of the public asWltered through the goal of
reelection. To be sure, legislators do not simply vote the views of their constituency,
but the role of the party organization and leadership in Congress is, at best,
marginal. And this makes a sharp contrast with their European brethren.
Cox and McCubbins take a diVerent view ( 1993 , 2005 ). In their view, party is the
primary organizing device of Congress. Congress is thus organized to fulWll the
collective interests of the majority party, and one important aspect of that is to
ensure that the majority party structures Congress so that it does not put the
reelection chances of the duly elected members of the majority at risk. The party
thus shapes the agenda so that members can vote for what the majority party’s
members want without voting against their constituents’ wishes often or on
important matters. Thus, the majority party is pleased to have its members on
committees that serve their constituents’ concerns and can reward their constitu-
ents with distributive beneWts. But while it provides room for its members to serve
their constituents, it also provides room for its members to act on their collectively
shared interests, all in the name of assisting their members’ reelection chances.
political parties in and out of legislatures 567