Legislative Bellwethers: The Role of Committee Membership in Parliamentary Debate

• Date: 01 May 2019
• Author: Max Goplerud,Miguel Won,Jorge M. Fernandes
• Published date: 01 May 2019
• DOI: http://doi.org/10.1111/lsq.12226

307

LEGISLATIVE STUDIES QUARTERLY, 44, 2, May 2019

DOI: 10.1111/lsq.12226

JORGE M. FERNANDES

University of Lisbon

MAX GOPLERUD

Harvard University

MIGUEL WON

INESC-ID

Legislative Bellwethers: The Role

of Committee Membership in

Parliamentary Debate

Political parties and legislators use legislative debates to establish their

reputation, challenge rivals, and engage in coalition management, among many

other tasks. Yet, existing theories on parliamentary debates have abstracted away

from the need for information and expertise, which are costly to acquire. Drawing

on the “informational” perspective on legislative organization, we address this

problem by arguing that party leaders use committees as training arenas for their

backbenchers. They task their assigned members with acquiring specific expertise

and then rely heavily on those members during the corresponding debates. We

turn to the Portuguese legislature, from 2000 to 2015, to discuss how saliency,

government dynamics, and party size affect the use of experts. We test this theory

using a novel approach to classify speeches that leverages the texts of legislation

as training data for a supervised approach.

“[A legislative bellwether] is, in a word, one whom his col-

leagues regard as relatively more ‘expert’ than they, and yet suff i-

ciently akin to be trustworthy.”(Dahl 1950, 60)

Political actors need specialized information and skills to

operate effectively in legislatures. These resources, however,

are costly to acquire and disseminate. In some contexts, legisla-

tors are flooded with information and face difficulty in extract-

ing what they need. In others, exogenous actors—for example,

interest groups or the bureaucracy—withhold information on

purpose, creating diff iculties for legislators to map their intents

© 2018 Washing ton University in St. Louis

308 Jorge M. Fernandes, Max Goplerud, and Miguel Won

into consequences (Krehbiel 1991). Parties have an essential role

in creating structures to help digest information insofar as they

promote hierarchical structures and functional specialization

(Saalfeld and Str⌀m 2 014).

Speechm aking is one of the most prominent activities i n leg-

islatures. Exist ing scholarship on legislative debate s focuses on the

impact of electoral system incentives to uphold a cohesive party

brand (Proksch and Slapin 2012), on the ideological distance be-

tween legislators and the party leadership (Diermeier et al. 2012;

Maltzman and Sigelman 1996), and how speaker selection has an

impact in vote-s eeking politics (Herzog and Benoit 2015). Yet, the

analysis of the mechanisms through wh ich parties and legislators

obtain information for speechmaking is mainly absent from the

literature. Indeed, most speeches are about complex and techni-

cal legislation, interacting with the party’s brand and objectives.

While some speeches can be delivered without much knowledge

of the issue, effective sp eechmaking requ ires highly informed a nd

trained political actors.

The contribution of this ar ticle consists in maki ng a rigorous

empirical t est of the relationship between comm ittee systems and

speech making activities. We explore the consequences of com mit-

tee assignments, which have been largely overlooked in favor of

committe e assignment politics. We argue that com mittee systems

serve as a training ground for speechmaking activities.1 Political

parties us e their prerogatives as gatekeepers of b oth committ ee as-

signments and floor time to articulate an optimal strategy whose

goal is to give the floor to legislators who have gained expertise

on a specif ic topic from the corresponding committee arena.

To account for the pivotal role of political parties in parlia-

mentary democracies (Müller 2000), we depart from Krehbiel’s

view of members’ exper tise being used as a colle ctive good for the

chamber as a whole. Instead, we suggest that committees are are-

nas for members to acquire useful information for their parties.

In return, party leaders rely on these experts as their preferred

actors to take the f loor and speak on behalf of the party on those

topics. By and large, this s erves to establish the part y’s reputation

and policy position on a par ticular jurisdi ction and signal the par-

ty’s positions to actors in the political environment, for example,

fellow parties, the electorate, and interest groups. Furthermore,

our theoretical argument suggests that the intersection between

309Legislative Bellwethers

committee-system specialization and floor specialization should

be more pronounced the more salient the issue is for the party.

In doing this, we also c ontribute to the g rowing literatu re on

text-as-data in political science. Existing research has mostly re-

lied on unsupervised methods—for example, topic models (Blei,

Ng, and Jordan 2003; Grimmer and Stewart 2013)—for uncov-

ering topics in legislative speech. Our contribution introduces a

different approach relying on supervised learning methods. We

use bills to train our sup ervised model and then use t he estimated

model to predict the labels attached to speeches. The use of bills

in conjunction with parliamentary debates opens up a new fron-

tier to apply text methods to critical questions in parliamentary

democracies as it makes supervised learning methods feasible as

bills provides a plausible set of training data. Our innovation is

particularly useful given that whenever parliamentary debates

exist, there are also bills. Many sources of legislative data both

historical ly and comparatively are likely to contain the nec essary

information to use th is supervised procedure.

We test our hypotheses in a causally credible way by using

a fixed effects approach. We include controls for backbencher-

committee pairs and thus leverage changes in speaking behavior

by comparing backbenchers’ behavior befo re and after being as-

signed to a particular comm ittee jurisdiction. Thus, unobserved

time-invariant omitted variables do not undermine our results.

We turn to the Portugues e case, over the period of four legisla-

tures (1999–2015). Portugal offers an ideal i nstitutional environment

to test our claims because it allows holding constant existing rival

explanations. Indeed, Portugal’s closed-list proportional electoral

system, with associated high levels of party loyalty and cohesion,

washes out potential competing explanations from electoral rules

and intraparty dissent (Proksch and Slapin 2012). Furthermore,

in Portugal, rules of procedure give parties a monopoly on deter-

mining com mittee assignments and make them gatekeepers of the

floor (Fernandes 2016). Such an institutional setting permits us to

explore how party leaders make decisions to use their committee

delegates to speak on their behalf in the plenary.

Legislative Speeches in Parliame ntary Democracies

Speeches are one of legislators’ most important tools in

modern democr acies. They serve a variet y of functions that, in re-

turn, are intimately linked with legislators’ goals. First, they help