Explaining Legislative Leadership Influence

• Date: 01 September 2013
• Published date: 01 September 2013
• DOI: 10.1177/1065912912458369
• Author: Christopher Z. Mooney
• Subject Matter: Articles

458369PRQXXX10.1177/10659129124
58369Political Research QuarterlyMooney
Regular Article
Political Research Quarterly
66(3) 559 –571
Explaining Legislative Leadership
© 2012 University of Utah
Reprints and permissions:
sagepub.com/journalsPermissions.nav
Influence: Simple Collective Action
DOI: 10.1177/1065912912458369
prq.sagepub.com
or Conditional Explanations?
Christopher Z. Mooney1
Abstract
Why are the leaders in some U.S.-style legislatures more influential than others? This study uses individual-level data
on lawmakers’ perceptions of their leaders’ influence to test three general theories of legislative power delegation:
legislative leaders have no real power, simple collective action theory, and Conditional Party Government theory. These
perceptions of speakers’ legislative influence are modeled with varying intercept, multilevel, ordered probit models.
The analyses strongly support the simple collective action problem explanation of legislative leadership influence,
in particular suggesting that collective problems caused by the internal dynamics of the legislative process drive the
delegation of influence to leaders.
Keywords
legislative studies, state politics and policy
Why do members of U.S.-style legislatures voluntarily
lawmaker pursued his or her goals without regard to his
give up significant control over lawmaking to the top
or her colleagues or the chamber as a whole, nothing
majority party leader in their respective chambers?
might ever be accomplished. Without some coordination
American lawmakers in a given chamber arrive at the
of effort, even the personal goals of many lawmakers
capitol after each election as constitutional equals, with
might not be met. But in the 1990s, Krehbiel challenged
significantly more personal power and freedom of action
this line of thought, claiming that congressional parties,
than their counterparts in most parliamentary systems
leaders, and other legislative institutions have no impact
(Loewenberg 2011). But on what is usually the first offi-
on legislative output, which is determined solely by the
cial vote of the session, the Senate president, the speaker
median preferences of a chamber’s members (Krehbiel
of the House of Representatives, or whatever the title,
1993, 1998; Schickler 2000). In an attempt to refute
the internally elected leader of a chamber’s majority
Krehbiel and to explain with more nuance why legisla-
party is typically given significantly more control over
tive leaders are powerful, scholars began exploring
the legislative process and outcomes than any other
whether this delegation might depend on certain political
member. This delegation of influence poses a significant
conditions, such as the level of policy problems facing a
puzzle to legislative scholars. Indeed, the questions of
legislature (Calvert 1987, 1992; Dewan and Myatt 2008)
how much control a sovereign group delegates to its
or the political conditions of the two parties (Aldrich
leader or leaders and under what conditions are among
1995; Rohde 1991; Aldrich and Rohde 2000; Cox and
the most significant and general in political science
McCubbins 2005).
(Bendor, Glazer, and Hammond 2001; Calvert 1992;
However, a thorough exploration of these important
Kiewiet and McCubbins 1991).
general questions of legislative influence delegation
The delegation of legislative control to leaders is one
was stunted by the lack of variation in both leadership
of the central questions in an important and long-
standing theoretical debate arising out of the literature
1University of Illinois, Springfield, IL, USA
on the U.S. Congress. Initially, scholars attempted to
explain what was seen as over-time fluctuations in lead-
Corresponding Author:
ership influence by simple collective action explanations
Christopher Z. Mooney, University of Illinois, Institute of Government
and Public Affairs, One University Plaza, PAC 451, Springfield, IL 62703-
(Calvert 1987, 1992; Shepsle 2006; Olson 1965). A leg-
5407.
islature can only act formally as a collective, so if every
Email: cmoon1@uis.edu

---

560
Political Research Quarterly 66(3)
influence and the potential causes of it reduced scholars’
of the subject that use aggregate-level indicators of state
ability to test these theories in the national body. To
legislative leaders’ tools and influence (Battista and
address this empirical problem, beginning with Richard
Richman 2011; Richman 2010). These findings have
Clucas’s seminal 2001 article, several scholars have
implications not only for the literature on legislative lead-
turned to state legislatures to test and generalize these
ership, but for the study of collective action and principal-
theories of legislative leadership delegation. While state
agent relationships generally.
legislatures have most of the same basic institutional
constraints and incentive structures as Congress, they
Legislative Leadership
also have significant variation on the collective prob-
lems and political conditions that are hypothesized to
“Power”—Tools versus Influence
affect leadership delegation (Squire and Hamm 2005).
Before developing hypotheses about the causes of legis-
The study of these state bodies allows scholars to
lative leadership delegation in depth, we must consider
explore this process by changing the research question
this process in the context of one of the most elusive
slightly from “do legislators delegate legislative control
concepts in the study of politics—power. Political scien-
to their leaders?” to “why, where, when, and to what
tists have long recognized that this is a difficult concept
degree do legislators delegate legislative control to their
to understand and even more difficult to operationalize
leaders?” This literature has made significant strides
(Dahl 1968). But scholars of legislative leadership power
with this approach (Clucas 2001, 2007, 2009; Battista
have tended not to consider their conceptual dependent
2011; Richman 2010; Battista and Richman 2011).
variable very carefully. While typically treated inter-
This article adds significantly to our understanding of
changeably in this literature, at least two fundamentally
the delegation phenomenon by using for the first time
different concepts are involved here: tools and influence.
individual-level data on lawmakers’ perceptions of their
By differentiating between these two, we can come to a
leaders’ influence to test these theories of leadership del-
much better understanding of the leadership delegation
egation. Heretofore, scholars of legislative leadership
process.
have relied mainly on institutional indicators of the
Most legislative leadership “powers” are formal rules
powers—or “tools”—available to majority party leaders
designed to give top leaders control of a chamber or cau-
in a chamber.1 While the assessment of the variation
cus. To avoid the confusion so often found in this litera-
across chambers of these institutional tools can tell us a
ture, I call these legislative leadership “tools.” Leadership
great deal about leadership delegation, it is at best cer-
tools are typically adopted as a matter of rule3 by a cham-
tainly an indirect measure of the actual influence of a
ber and bestowed on a specific office. They are struc-
leader. I argue in the next section of this article that lead-
tured institutions, relatively exogenous4 to the process,
ership tools and influence are very different things, and
robust, and persistent (Shepsle 2006). These tools can
as such, they should be studied independently. Indeed,
include, for example, a leader’s right to appoint members
lawmakers’ individual-level perceptions of their leaders’
of standing committees, to refer bills to those commit-
influence is probably the most direct and valid indicator
tees, and to determine the order of debate on the floor of
yet used in this literature to assess how much control a
the chamber, among other specific grants of authority.
given leader has over the legislative process in his or her
Studies have shown that these tools can enhance leaders’
chamber.
control over the policy output of a chamber (Cox,
I use the variation in the perceived influence of speak-
Kousser, and McCubbins 2010; Jenkins 2008; Lazarus
ers of 48 state Houses of Representatives derived from a
and Monroe 2007; Hasecke and Mycoff 2007), although
2002 survey (Carey et al. 2002) to examine the funda-
not all scholars agree with this assessment (e.g., Krehbiel
mental question of the leadership delegation in U.S.-style
1993, 1998).
legislatures.2 Using varying intercept, multilevel, ordered
On the other hand, the “power” legislative leaders are
probit models of these perceptions, I test hypotheses
said to have in the legislative process refers to something
drawn from the three theories of leadership power dele-
qualitatively different from these tools; I refer to this as
gation. I find, first, that state House speakers are believed
their “influence.” While leadership tools are policy
by their colleagues to be quite influential and that this
instruments established formally, a leader’s influence
perceived influence varies systematically across states
refers to his or her ability to control the outcome of some
and lawmakers. Second, I find that speakers have the
aspect of the process. Thus, while leadership tools are
most influence when problems relating to their respective
relatively...