make it possible for legislators to realize gains unrealizable in unorganized, state-
of-nature assemblies. The relative balance of agenda control residing in legislative
committees, directory boards, and presiding oYces varies across legislatures. In
parliamentary systems, these powers are generally vested in cabinet ministries—
technically part of the executive branch, but which themselves areWlled from
among members of the legislature, and are dependent on its conWdence for
survival. The key point is that, in almost all democratic systems, parties are the
gatekeepers of the formal oYces that control the legislative agenda. Moreover,
Carroll, Cox, and Pachon ( 2004 ) demonstrate that, as democracies mature,
parties expand their control over the oYces that determine the legislative agenda,
and the distribution of these oYces among parties grows increasingly regular. In
short, as party systems stabilize, so do the key partisan elements of legislative
organization.
- 2 Parties and the Legislative Agenda
How does partisan agenda control provide decisiveness? Diverse accounts of
legislative politics converge around the idea that parties reduce the potentially
inWnite number of policy options to a limited set, primarily by establishing
platforms or manifestos that advertise party positions to voters, and then by
disciplining legislators to constrain their voting in line with these party positions
(Aldrich 1995 ). Comparative studies of roll call voting suggest that legislative
agendas are strongly limited in ways consistent with the idea that parties produce
procedural order. Cox, Masuyama, and McCubbins ( 2000 ) demonstrate that
the long-dominant LDP in Japan used its control over the parliamentary
agenda to prevent proposals that might divide its governing coalition from coming
to theXoor. Amorim Neto, Cox, and McCubbins ( 2003 ) provide evidence that
multiparty legislative coalitions in Brazil acted similarly, as cartels that limit
legislative proposals to protect the policy interests of member parties. In both
cases, the point is that parties—sometimes as partners in coalitions—both limit the
policy alternatives among which legislatures formally choose, preventing cycling,
and ensure that some alternatives enjoy procedural advantages that prevent
bottlenecks.
Other empirical evidence also highlights the relative orderliness of voting in
legislatures, in contrast to the theoretical prospect of majority rule cycles. The most
widely used method for estimating legislator ideal points suggests that agendas
across a wide range of legislatures show remarkably limited dimensionality (Poole
and Rosenthal 2001 ; Rosenthal and Voeten 2004 ). That is, across various legisla-
tures in quite diVerent political systems, and in the US Congress throughout most
of its history, legislators’ voting patterns can be accurately mapped using only a
legislative organization 445