A dynamic model of party membership and ideologies
| Date | 01 April 2019 |
| Author | Bilge Öztürk Göktuna |
| DOI | 10.1177/0951629819833185 |
| Published date | 01 April 2019 |
| Subject Matter | Articles |
Article
Journal of Theoretical Politics
2019, Vol.31(2) 209–243
ÓThe Author(s) 2019
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/0951629819833185
journals.sagepub.com/home/jtp
A dynamic model of party
membership and ideologies
Bilge O
¨ztu
¨rk Go
¨ktuna
Galatasaray University, Turkey
Abstract
We analyze theone-dimensional electoral competition between two parties whenthe ideology of
each party is endogenously determined. The parties are composed of two factions: the ‘opportu-
nists’ and the ‘militants’. The ideology of each party is determinedby the preferences of the med-
ian citizen supporting the party. Under the proportional system, where parties are represented
proportionally to the share of their votes, we first study the short-term political equilibria. We
then introduce a dynamic setup that endogenizes the composition of the parties, in order to ana-
lyze the stability of these equilibria.We make explicit the stability conditions for the two equilibria
where all the opportunists belong to the same party and for the unique equilibrium where they
are distributed between both parties. The conditions involve the rates of party switching and of
ideological adjustment. This coupled adjustment process makes it possible for party competition
to sustain proportional representation, fluctuation in party positioning, and some degreeof policy
divergence.
Keywords
Electoral competition; evolutionary dynamics; opportunist; proportional system
1. Introduction
Political parties are thought to be characterized by their ideologies. However, in
reality, parties are loosely formed around broad political ideologies; within any
party, there are people who hold a variety of opinions and have a variety of moti-
vations. Moreover, be it for ideological or for purely electoral concerns, there may
sometimes be serious discrepancies between politicians of the same party, and
between the politicians in power and the people.
Corresponding author:
Bilge O
¨ztu¨rkGo
¨ktuna, GalatasarayUniversity, Istanbul, Turkey.
Email: goktunabilge@gmail.com
A striking example of the first case is the situation of Barack Obama in 2015,
illustrated by this quotation:
President Barack Obama went toCapitol Hill on Friday morning to make a personal plea
for his own party’s support.
Democrats ignored him.
And now, the prospects for the Trans-Pacific Partnership, the biggest free trade deal in
history, to be finalized and adopted are grim – unless Democrats can be convinced to
change their minds or Republicans can find another way to revive the bills and rescue
Obama’s biggest second-term legislativepriority.
(Bradner and Walsh, 2015)
An example of the second case is the reaction of Turkish political class to the
launching of the US war in Iraq in 2003:
Turkey’s parliament failed to pass a proposal Saturday to allow more than 60,000 U.S.
troops to operate from Turkish bases and portsin the event of a war with Iraq.
[. . .] The proposal has little popular support in Turkey. Hundreds of thousands of pro-
testers rallied Saturday in downtown Ankara. Public opinion polls show more than 90
percent of Turks oppose a U.S.-led war against Iraq.
The United States has offered $6 billion in economic aid to offset fears that war could
devastate Turkey’s economy. Refusal to participate could have severely limited Turkey’s
role during a war and in a post-war Iraq.
(Amanpour, 2003)
So, in reality, party positions cannot be reduced to pure selfish re-election con-
cerns; nor can they be reduced to simple ideological positioning. In this formal
study, we will attempt to grasp the complexity of party decisions by considering
that politicians are of two categories: either ideologically motivated or selfishly
interested in winning office, and that the composition of each party is not fixed but
can evolve under endogenous pressure.
We attempt to model these aspects of the political process through an electoral
game. In the short run, before each election, the two parties simultaneously and
independently announce their policies. There are different types of candidates in
both parties and the policy proposals are determined through the aggregation of
different political motivations. We attempt to explain the dynamic patterns of the
policy proposals of the parties through the behavior of the candidates. The candi-
dates motivated by winning an office do not have any loyalty to the party since the
party is a means to achieve the goal. In the long run, they can switch parties if they
see that another party can offer a better chance of achieving their goals. The candi-
dates motivated by representing their electors cannot accurately compute the pre-
ferences of their electoral base before the election results. They have only past
elections’ results to decide about their proposals. Given these, they adjust their
ideology to better represent their electorate. The decision rules of these different
candidates create a coupled dynamics.
The received literature on the role of political parties is mainly developed within
the spatial models of electoral competition, and has included a variety of
210 Journal of Theoretical Politics 31(2)
assumptions concerning the objectives of the party elites. The first is that of
Downs (1957), where the candidates are office- or vote-oriented, and the second is
given by Wittman (1983), where the candidates are policy-oriented with the ideolo-
gical preferences of the parties taken to be exogenously determined. The concep-
tion of parties as unitary actors with preferences fixed during the legislative term is
misleading: ‘‘It is more productive to consider parties as being comprised of a num-
ber of different elements, or faces, each of which potentially interacts with all of
the others.’’ (Katz and Mair, 1995). Sartori (1976) states that a party is made up
of rival groups (‘the issue is, then, how the unit ‘‘party’’ is articulated, or disarticu-
lated, by its sub-units’) and proposes that each of these groups have an impact on
the mix that is the party and that ‘their presence alters the mix, i.e., the relative
weight of each element of the amalgam’.
This view of parties taking positions through an internal factional process led
Roemer (1999) to combine the two previous approaches about political motiva-
tions in a game theoretic model and argue that parties are composed of three fac-
tions: the opportunist faction (inherited from Downs (1957)), the reformist faction
(inherited from Wittman (1983)) and the militant faction, focused on the publicity
of the ideal policy of the party
1
. As Roemer (1999) shows, the reformist faction has
no strategic influence; we thus limit the analysis to two factions, the opportunist
and militant ones.
In reality, the militant faction is not homogeneous. In this study, we shall con-
sider that the militant faction, instead of promoting an exogenously defined ideol-
ogy, determines the ideology of the party endogenously. The political position
promoted by the militants of a party will be the median of the political positions of
the supporters of that party. In this way, the militants promote an ideology that
would be supported by a majority of the party voters at the equilibrium of a delib-
eration process akin to a closed primary. Thus, the ideology of the party is endo-
genous and the ideologies of the parties come into being from the preferences of
the voters. This procedure is taken from Roemer (2001), who adapts the notion of
citizen-candidate and includes parties and commitment.
The fact that politicians have ideological concerns is highly understandable and
acceptable. After all, today’s democracies are based on political representation: cit-
izens elect politicians who will represent them.
2
Citizens express their political opi-
nions and raise their voices through political parties (Almond et al., 1993). Parties,
on the other hand, reconcile varying interests of their supporters and form their
political agendas (Budge et al., 1987).
3
This link has given rise to the model of
responsible party government in the sense that a party is conceived as responsible
for representing the policy preferences of its supporters, not the individual
legislators.
The relationship between the expectations of citizens and the parties is recog-
nized by political scholars but remains a question of empirical research. Inspired
by the responsible party government model, Dalton (1985) provides cross-national
evidence on the representation process using the amount of agreement between citi-
zens and politicians as a measure of effective representation. He shows that, taken
overall, there is substantial agreement between citizens and politicians and that,
O
¨ztu
¨rk Go¨ktuna 211
To continue reading
Request your trial