The Sound of Rebellion: Voting Dissent and Legislative Speech in the UK House of Commons

• Published date: 01 May 2020
• Author: Jonathan B. Slapin,Justin H. Kirkland
• Date: 01 May 2020
• DOI: http://doi.org/10.1111/lsq.12251

153

LEGISLATIVE STUDIES QUARTERLY, 45, 2, May 2020

DOI: 10.1111/lsq.12251

JONATHAN B. SL APIN

University of Zurich

JUSTIN H. KIRKLAND

University of Virginia

The Sound of Rebellion: Voting

Dissent and Legislative Speech in the

UK House of Commons

Using data on the content of debate associated with votes in the UK House

of Commons from 1992–2015, this article examines how government party MPs

employ language in legislative speech when they vote against the party line. We

find a robust statistical association between dissent on votes and the use of first-

person pronouns, simpler language, and giving longer speeches. Using a random

forest algorithm for classification, we find that these language covariates are pre-

dictive of rebellion. The use of simpler, first-person language has implications

for political representation and offers new insight into how MPs use votes to

distinguish themselves from their party, perhaps reflecting attempts to connect

with constituents.

Members of the British House of Common s partici-

pate in legislative debate to r epresent citiz ens, build pe rsonal

brands, support thei r party, and occa sionally to express d issent.

Comparative research on parliamentar y behavior suggests t hat

MPs’ participation in debate is a function of par ty oversight and

ideology (Herzog and Benoit 2015; Proksch and Slapin 2015), coa-

lition particip ation (Martin and Vanberg 2008), and gender (Bäck

and Debus 2018; Bäck, Debus, and Müller 2014), among other

variables. Research also uses leg islative debate to understand the

topics underpi nning legislative activity (Quinn et al. 2010) and to

measure constit uency service (Marti n 2011). But less work has ex-

amined the link betwe en the nature and conte nt of MPs’ speech

and their voting decis ions (but cf. Schwarz, Traber, and Benoit

2017). Do MPs use differe nt language when voting again st their

party than when t oeing the party line? Do rebell ious MPs use di f-

ferent rhetorical tur ns than their loyal col leagues? Literatu re on

rebellion in th e House of Commons sugge sts that MPs, especially

© 2019 Washington University in St. L ouis

154 Jonathan B. Slapin and Justin H. Kirkland

those in the governi ng party, use rebell ion to connect with vot-

ers, garner na me recognition, and differentiate the mselves from

the rest of their par ty (Benedetto and Hix 2007; Campbell et al.

2016; Kam 2009; Kirkland and Slapin 2019; Slapin etal. 2018),

but it has not examined the la nguage they use i n speeche s when

rebelling. W hen voting against t heir party, MPs likely look to use

debate to communicate a particular position or sentim ent to vot-

ers, a task that may requi re different lang uage than expres sions

of support for the party position.

Importantly, the existenc e of such “rebel rhetoric” may

have consequences for repre sentation and politica l engagement.

Recent studies have argu ed, for example, that simpler lang uage

is associated with the expansion of the fra nchise in the 19th

United Kingdom, as cabi net minist ers, in part icular, attempted

to make themselves compr ehensible to new voters (Spirling 2016).

Others have demonstrated th at populist politicians use less c om-

plex language to dist inguish themselves from other politicians

and to reach voters (Bischof and Senninger 2017; Brosius et al.

2017; Lin and Osnabrügge 2018). And, of course, much literature

in the American context suggest s that the tone and rhetoric of

politics and camp aigns can impa ct voter engagement and tur n-

out (e.g., Ansolabehere and Iyengar 1995; Lau and Pomper 2001).

Knowing whether such rhetori c exists in the House of Commons,

and whether it is associat ed with voting against one’s party, can

help us to understand the motivations for rebellion among back-

bencher MPs.

This arti cle uses data on d ivisions and related pa rliamen-

tary debates i n the UK House of Commons to exa mine whether

certai n linguist ic character istics are as sociated with “rebel”

speech— speech es given by governing par ty backbencher MPs

during debates when t hey vote against the m ajority of their p arty.

We present a series of exploratory hyp otheses taken from existing

literature about the li nguistic cha racteristi cs we might assoc iate

with rebellious s peech. In particular, we are intere sted in whether

MPs who vote against their par ty are more likely to us e different

sentiment and li nguistic complex ity in their speeches compared

with loyal MPs—variables oft en associate d with populist rheto -

ric, and thatothers have shown to affect politica l representation

and voter engagement. Lastly, using a random forest algorithm ,

we examine whether th ese speech character istics are predictive of

rebellion, rather th an simply assoc iative. We find that they are.

These resu lts give us confidence that we could u se these linguistic