Explaining Electoral System Change: The Adoption of Proportional Representation in Belgium and Its Rejection in Great Britain Nina Barzachka Third Annual Graduate Conference Woodrow Wilson Department of Politics University of Virginia Spring 2007
Explaining Electoral System Change: The Adoption ofProportional Representation in Belgium and Its
Rejection in Great Britain
Nina Barzachka
Third Annual Graduate ConferenceWoodrow Wilson Department of Politics
University of VirginiaSpring 2007
“No party, even in America, voluntarily resigns a sure hold on power for the uncertainties of a new system simply on grounds of abstract justice.”
Thomas Harrison Reed, A.B., LL.B.Professor of Political Science, University of Michigan,
in Government and Politics of Belgium, 1924
In 1899, Belgium became the first country in the world
to introduce proportional representation (PR) for national
legislative elections of its lower chamber. It was Prime
Minister Beernaert who first pushed for PR in 1894 and
ultimately sacrificed his political career for it, knowing
fully well that the new electoral system would significantly
diminish the overwhelming majority (104 out of 152 seats) of
his Catholic Party. “And when, five years later, the reform
was implemented, it was again with the initiative of a
Catholic government, despite the fact that the majority [of
the governing party] had increased further [with 112 out of
152 seats]” (Barthélemy 1912 p.536). By contrast, the
adoption of proportional representation was considered and
rejected by portions of both the Liberal and the
Conservative parties in the British Parliament in 1885 and
1918.
1
The fact that proportional representation for national
legislative elections was first introduced by a political
party that commanded formidable parliamentary majority and
expected continued future electoral returns under the
existing majoritarian system and plural vote has been has
not been emphasized by the recently emerged literature on
electoral system change. The Belgian case seems to
contradict not only common-sense expectations, but also the
literature of electoral system change and the literature on
institutional change, in general. The observation that
electoral rules result from strategic calculations of the
major political actors (parties and party leaders) has been
made in the works of notable political scientists (Reed
1924; Rokkan 1970; Mainwaring 1999; Boix 1999; Benoit 2004;
Colomer 2004; Colomer 2005; Andrews and Jackman 2005). It is
now clear that large parties that expect to continue winning
election under majoritarian or plurality rules have no
incentive to switch to proportional representation, while
smaller parties prefer PR (Boix 1999; Benoit 2004; Colomer
2005). Similarly, Knight (1995) refines the conventional
2
rational choice understanding of institutional change by
acknowledging power asymmetries between actors. Knight
points out that the most powerful actors tend to impose
their institutional preferences. The conditions that led to
the consecutive rejections of proportional representation in
Great Britain, where both Gladstone and Lloyd George opposed
PR in 1884-1885 and 1916-1918 respectively confirm these
conclusions. In that light, the Belgian anomaly appears even
more puzzling. What motivated the leaders of the Catholic
party in Belgium to support proportional representation
despite expected electoral losses under the new system? Was
this a foolish sacrifice or a matter of political necessity?
What are the implications of the Belgian case for the
electoral system change literature and the institutional
change literature in general? It is the goal of the present
paper to answer these questions, as well as to identify a
causal mechanism of change from majoritarian electoral
system to a system of proportional representation.
The case studies reveal several major variables that
contributed to the adoption of PR in Belgium and to its
3
rejection in Britain. These include: 1) the nature of
existing electoral institutions and especially the existence
of the plural vote and the size of electoral district; 2)
the level of disproportionality created by existing
electoral institutions; 3) the degree of institutional
stability; 4) the perception of internal crisis; 5) very
strong and consistent pressure form the Left (mass strikes
and demonstrations); 6) dangerous overlap of economic,
political, religious, linguistic and ethnic cleavages.
Electoral institutions such as the plural vote and large
districts in combination with the sharply divided society
(the cleavage structure) created highly disproportional
election results in Belgium. This disproportionality
discredited these very same electoral institutions and
underrepresented actors began to demand change. Strong
extra-parliamentary pressure from the Left interacted with
cleavage structure to create a sense of national crisis i.e.
the disintegration of the young country appeared as a very
real threat. Perception of internal crisis and institutional
instability interacted and aggravated each other, thereby
4
increasing the uncertainty in which actors operated and
creating an opening for institutional change. At this
crucial point politicians were compelled to compromise in
adopting proportional representation. The presence of many
large multi-member districts1 and the rejection of
redistricting made PR a viable option. And while actors
preferred the type of electoral system, which they believed
would benefit their own party in so far as that the
consequences of the new electoral system were foreseeable
(i.e. actors had relatively short-term horizons and
electoral system change was risky), their partisan goals
cannot simply be described as seat-maximization.
Furthermore, in the face of crisis, certain key politicians
were willing to compromise in order to preserve the unity of
the country. By contrast, the case of Great Britain
illustrates that when the predominance of the single-member
district is accompanied by a high degree of institutional
stability, lack of crisis and a simpler structure of
cleavages, the adoption of proportional system of
1 The presence of many large multi-member districts made PR was necessary, though not sufficient for the introduction of PR.
5
representation becomes unlikely. PR was not introduced in
Britain in 1883-1885 and 1916-1918 because circumstances and
actors failed to create opening for this type of
institutional change. In both countries, actors attempted to
change the institutions, as they were simultaneously
constrained by them.
I identify this causal mechanism based on some
assumptions derived from the literature on institutional
change but, primarily, through a detailed historical
analysis and process tracing of the circumstances that led
to the adoption of proportional representation in Belgium
and its rejection in great Britain. Unlike the most recent
works of scholars of electoral system change Boix (1999),
Cusack, Iversen and Soskice (2003), Colomer (2005), who
employ primarily quantitative methods and formal modeling, I
rely on both secondary and primary sources and focus on the
preferences (and reasons for these preferences) of the main
political parties and party leaders, as well as on the
effects of existing electoral institutions and social
conditions. My approach is largely inductive. In that
6
respect, my paper provides detailed examples and explanation
of the causal mechanism that are missing in existing
quantitative work.
I have selected as my case studies four episodes when
PR was considered in Belgium and Great Britain. Both are
Western European countries with high levels of
industrialization in the late 19th and early 20th century.
Both were unitary parliamentary systems2, undergoing the
process of democratization (and more specifically suffrage
extension) during this period. The two Belgian episodes are:
the period 1831-1893, with a special focus on the conditions
that led to the Constitutional Reform of 1893; and the
period between the reform and the introduction of PR (1893-
1899). The British episodes are: the reforms of 1885 and
1918. In three of the four episodes, PR was rejected; it was
adopted in Belgium in 1899. This approach allows me to add a
longitudinal dimension to the comparison across countries
and present a dynamic causal mechanism.
2 This controls for other institutional arrangements, including federalism.
7
I begin with a brief overview of some of the main
theoretical explanations of institutional change. I then
examine consecutively the two Belgian cases and the two
British episodes. I conclude with an answer to the question
why proportional representation was introduced in Belgium
but not in Great Britain.
Theoretical Approaches to Institutional Change:
The circumstances surrounding the introduction of
proportional representation in Belgium and its rejection in
Great Britain can provide useful insights into the
mechanisms of electoral change in particular and
institutional change in general. Institutions are understood
by both rational choice scholars and historical
institutionalists as established formal and informal rules
and procedures that guide conduct (Bates 1998; Orren and
Skowronek 1994; Thelen and Steinmo 1992). Electoral laws and
proportional representation in particular are one of the
clearest examples of institutions.
The main difference between the manner in which
rational choice and historical institutionalism explain
8
institutional change remains in identifying the sources of
this change. While most rational choice theorists view
actors (individuals; groups; parties) as the agents of
change, historical institutionalism emphasize the causal
importance of existing institutions. Finally, Paul Pierson’s
notion of path dependence attempts to create a bridge
between the two perspectives by incorporating both structure
and agency as part of the causal mechanism (Pierson 2004).
Rational choice theorists employ a macro-analytical
approach and game theory (Bates 1998; Scharpf 2000; Greif
1998; Knight 1995) in order to study institutional change.
They focus on individual actors as agents of change and
consider institutions to be part of the structure, which
provides constraints and incentives to the actors. Rational
choice theory assumes that: 1) actors are rational (and
always act in ways that allow them to maximize utility and
minimize cost); 2) actors operate in an environment of
perfect information (they know what their interest is and
they are aware of the costs and benefits of the alternative
options); 3) all actors posses the same motivations and act
9
in a similar fashion (actor homogeneity). Consequently,
institutions are created as a result of negotiations between
actors who are pursuing their self/interest (Scharpf 2000
170). In that respect institutions constitute equilibria.
Actor co-operate and uphold the status quo as long as it is
beneficial to them. Exogenous shocks can alter the
perceptions of actors. Once co-operation is no longer
perceived as beneficial by some actors, they have the
incentive to defect. The old institutional arrangement is
destroyed and a new institution can be formed if a new
equilibrium is found. This mechanism accounts for the
creation and demise of institutions as well as for periods
of institutional change (Scharpf 2000). The majority of the
rational choice literature (among them Bates 1998; Greif
1998; Knight 1995) focuses on exogenous shocks as factors
that alter actors’ perceptions.
Historical institutionalism differs from the rational
choice approach in that it emphasizes the causal
significance of earlier institutions. For instance,
historical institutionalists question the rational choice
10
assumption that actors are all-knowing and rational. In
cases when actors do not posses all relevant information and
might be unable to rationally evaluate all options, they are
much more likely to rely on existing rules (institution) in
deciding their course of action (Thelen and Steinmo 1992 8).
“By shaping not just actors’ strategies (as in rational
choice), but their goals as well, and by meditating their
relations of cooperation and conflict, institutions
structure political situations and leave their own imprint
on political outcomes” (Thelen and Steinmo 1992 8). Orren
and Skowronek, by contrast, emphasize the agency of
political institutions much more forcefully. Because
different institutions are established at different points
in time, there is an overlap of institutional structures,
which may co-operate or be in conflict with each other. In
as much as institutions are “purposeful and intentional”
they will strive to affect the creation of other
institutions (Orren and Skowronek 1994 326-8). This
constitutes a major point of difference between rational
choice theory and historical institutionalism. As for the
11
particular mechanism of institutional change, the historical
institutionalist approach relies on several different
explanations. In Krasner’s model of punctuated equilibrium,
periods of stability are interrupted by points in which
exogenous shocks cause dramatic change or even overhaul of
existing institutions (Thelen and Steinmo 1992). In that
respect, punctuated equilibrium accounts for both periods of
stability and times of drastic exogenous change. It becomes
evident that historical institutionalism does not have a
single, coherent theory that explains institutional change.
Furthermore, the representatives of this school of thought
seem to employ a definition of institutions that is broader
than the understanding of institutions as mere procedures3.
The main problems with the pure rational choice and
historical institutionalism explanations of institutional
change constitute a mirror image of each other. Pure
rational choice approaches tend to underestimate the
importance of previous institutional arrangements, while
traditional historical institutionalism perspectives
3 After all, how can electoral laws in themselves be purposeful and intentional?
12
undervalue agency. It is in examining the origins of
specific institutions that the inadequacies of these two
approaches become the most clear. Actors attempt to alter or
destroy some institutions but they remain constrained by
existing institutional arrangements. This is a tension that
cannot be escaped or ignored. Recently, efforts have been
made to reconcile this dichotomy. The rational choice
institutional approach is one such example (examined by
Norris, 2004). The recent literature on electoral system
change (Boix 1999; Benoit 2004; Colomer 2004; Colomer 2005;
Andrews and Jackman 2005) appears to fit in this category.
Another important effort in the direction of bridging
the gap between rational choice and historical
institutionalism is Paul Pierson’s (2004) concept of path
dependence4. According to the model of path dependence,
existing institutions continuously provide restrictions to
the options available to actors in the future. In the long
run, certain institutional arrangements become entrenched.4 While considered part of the school of historical institutionalism, path dependence is accepted by some rational choice scholars. Furthermore, it is Pierson’s goal to move in direction of reconciliation of rational choice and historical institutionalist approaches to institutional change.
13
Change can result only when an opening is created through an
exogenous shock.
While this model is dynamic and particularly suitable
to the study of electoral institutions (Pierson 2004), it
suffers from several shortcomings. It becomes deterministic
the longer one remains on the same path. In that respect,
the model allows great degree of agency in the early stages
of the path, but agency disappears at the latter stages.
Path dependence also operates under the assumption that
change is always exogenous. However in practice, it is
possible for an opening to be created endogenously through
the actions of agents who want to change existing
institutions. The causal mechanism, derived from the case
studies in this paper, demonstrates that even when path
dependent logic is at work, actors can endogenously create
such institutional openings.
The Introduction of Proportional Representation inBelgium:
14
The history of electoral system of independent Belgium
can be dividend into three main periods, namely the period
of limited suffrage or régime censitaire (1831-1892); the period
of the universal male suffrage accompanied by the plural
vote (1893-98); and the era of equal universal male
suffrage5 (1919 to present). Proportional representation was
introduced in 1899, as a corrective measure to the system of
the plural vote (Caramani 2000; Gilissen 1958; Barthélemy
1912; Mahaim 1900). For the purposes of this paper
developments that led to the adoption of PR can be divided
into two episodes: the period, which clearly demonstrated
the gross disproportionality of régime censitaire (between 1831
and the Constitutional Reform of 1893); and the period
between the Constitutional Reform and the introduction of
proportional representation (1893-1899).
5 While authors like Cusak, Iversen and Soskice (2003) consider 1894 theyear of introduction of universal male suffrage, their interpretation ignores the plural vote. Here I employ the conventional definition of universal suffrage as equal universal male suffrage (Collier 1999; Caramani 2000) and consider 1919 as the year of introduction of universal male suffrage in Belgium.
15
Initially, Belgium had a two-ballot majoritarian
system and 41 electoral districts6. “Elections were partial:
half of the representatives of the Chambre de Représentants (or
Kamer der Volksvertegenwoordiger) were elected every two years to
a four-year term. General elections were carried out only if
the parliament was dissolved” (Caramani 2000 151). Under
the system of limited suffrage men were required to pay a
fee (cens), originally between 20 to 100 florins in order to
vote. This was a common feature of the voting systems in
most of Western European countries at the time. In Belgium,
however, instead of being included in ordinary electoral
law, the system of régime censitaire was inscribed in Article 47
of the Belgian Constitution (Barthélemy 1912; Gilissen 1958
91; Mahaim 1900). While electoral laws could be revised by a
simple majority in an ordinary parliament, a constitutional
revision could only be made by 2/3 majority in a lower
chamber, especially elected and authorized to conduct such
revision (Gilissen 1958 123; Mahaim 1900). This made
constitutional reform particularly difficult.
6 At first, they were both single (9) and multi-member districts (32). Gradually districts became larger. (Barthélemy 1912 513).
16
The question of proportional representation was first
brought up in Belgium in the early 1860s by the progressive
Liberal Count Goblet d’Alviella and was supported during the
next thirty years by both Liberal and Catholic members of
the bourgeoisie (Barthélemy 1912 532). In his account of the
early years of the PR question in Belgium, Barthélemy
highlights two interesting points, namely that PR was an
elitist idea, which enlisted the support of members of both
of the country’s main parties (Barthélemy 1912 524). The
issue of proportional representation encountered strong
opposition by the older and more reactionary, yet very
influential figures from both parties, including Frère-Orban
(leader of the Liberals) and Woeste (Catholic Party). In the
1870s, Hare’s system of PR made its appearance in Belgium
(Barthélemy 1912 533). The fact is significant because it
illustrates that PR proponents in Belgium followed closely
the developments abroad and that the events in Belgium did
not take place in isolation. Barthélemy concludes, however,
that the influence of Hare’s system was not substantial
because in 1878, a Belgian professor, Victor d’Hondt
17
proposed a new PR system (Barthélemy 1912 534). The question
of proportional representation was discussed on several
occasions (in 1866, 1871 and 1878) in the lower chamber of
the Belgian Parliament only to encounter hostility and
ridicule.
Nevertheless, in the 1870s the system of régime censitaire
began to create disproportionalities between votes and
seats, disproportionalities that overwhelmingly favored the
Catholic Party. The electoral districts (diverse in size)
were the same as the administrative districts inherited from
the Dutch. In addition, the requirement that there should be
one representative per every 40,000 inhabitants was
inscribed in the Constitution. As the major industrial
centers grew in population, the size of their districts also
increased.7 This majoritarian electoral system accompanied
by multi-member districts led to gross overrepresentation of
the countryside (the stronghold of the Catholic Party) at
the expense of urban areas. “[S]pread throughout the entire7 According to Barthélemy: “in 1831, the district of Brussels elected only 7 deputies, Gand [elected] 6, the Anvers and Liège [elected] 4 each” (Barthélemy 1912 513). By 1899, only 40 years later these number had changed drastically, “Brussels elected 18 deputies, Liège and Anverseach [elected] 11, Gand 9, Chaleroi 8” (Barthélemy 1912 513).
18
surface of the country” and representing “only a cultivated
and intelligent minority” (Barthélemy 1912 518-519) the
Liberal Party was clearly at a disadvantage. It must be
mentioned that the majoritarian system8 favored the rural
areas, that were not only conservative and thus, affiliated
with the Catholic party, but also predominantly Dutch-
speaking (Gilissen 1958; Tyson 1904).
After the “scandalous” election of 1880, the
shortcomings of the existing system, the need for electoral
reform and PR as a viable alternative became blatantly
obvious. In 1880, “the Catholics, with 1240 votes less than
[the Liberals], won 14 more seats9” (Barthélemy 1912 534).
As a result, PR supporters from both parties formed The
Reformist Belgian Association for Proportional
Representation in 1881, which sponsored numerous conferences
and brochures explaining PR. In 1882, d’Hondt published a
redefined version of his system and in 1888 the Association
proposed a bill based on it (Barthélemy 1912 535; Gilissen8 Barthélemy considered this system dangerous as it focused on regional rather than national matters and favored extreme parties at the expense of the moderate ones (Barthélemy 1912).9 “The liberals received 22,000 votes and only 26 seats, while the Catholics received 20,979 votes and 40 seats.” (Barthélemy 1912 522).
19
1958). Catholic Prime Minister Beernaert “declared his
sympathy for the principle but added that because of its
complexity, the project needs to be carefully studied”
(Barthélemy 1912 535). It is important to note that the idea
of proportional representation did not become salient until
particular events demonstrated the problems with the
previous electoral system and put certain political actors
(in this case the Liberal Party) in a particularly
disadvantageous position. In this light, the fact that it
encountered the sympathy of many Catholic politicians
appears somewhat surprising. The results of the 1880
election, however, were not sufficient to lead to the
adoption of PR.
In 1893 the Belgian constitution underwent a major
reform that had a serious impact on the electoral system of
the country. The Constitutional Reform of 1893 implemented
several innovations including: 1) reform of the Senate; 2)
the obligatory vote for which Belgium is famous; 3) the
province as a new electoral unit, and, most importantly, 4)
the substitution of limited suffrage with universal male
20
suffrage tempered by the plural vote (Barthélemy 1912;
Gilissen 1958). The last institutional change is of
particular importance to the question of PR introduction, as
it aggravated the disproportionality between votes and seats
and further strengthened the dominant position of the
Catholic Party in the lower chamber. It is, thus, necessary
to examine the conditions and events that led to the 1893
Constitutional Reform.
The question of universal suffrage is the main factor
that necessitated the constitutional reform. The matter of
universal suffrage was in turn fueled by the democratic
movement in Europe and strong pressure form the Left
(Gilissen 1958). At the time, Belgium was the fastest
industrializing nation in Europe, a country with a rapidly
growing working class (Mahaim 1900 382; Barthélemy 1912).
Influenced by the first International in 1864, “in 1866 a
Manifesto of the Workers [was] launched, calling for ‘the
abolition of the voting fee [cens]’ and for ‘equality in the
right of suffrage’” and in 1885 POB (Belgian Worker’s Party)
[was] constituted” (Gilissen 1958 120). In 1886 Belgium was
21
shaken by mass workers protest that threatened to turn into
a civil war (Barthélemy 1912 268-269; Mahaim 1900) and which
“showed the gravity of the social situation and the strength
of the working class” (Gilissen 158 121). This strength and
the importance of universal suffrage were further emphasized
four years later, when in 1890 in Brussels “several tens of
thousands protestors” demanded universal suffrage (Gilissen
1958 21). Furthermore,
“In 1891, under considerable pressure from the POB, the Catholicprime minister and the king both let it be known that they mightaccept some type of electoral reform. The POB followed up the nextyear with a general strike, which ended only when the governmentannounced the forthcoming elections of a Constituent Assembly torevise the electoral articles of the constitution” (Collier 199990).
Two important observations need to be noted here.
First, disenfranchised Belgians
were pushing for universal suffrage that was to lead to a
profound change in the electorate. Second, the party system
was transformed by the creation of the Belgian Worker’s
Party. This confirms the observations of Colomer (2004) that
changes in the party system took place prior to the
introduction of proportional representation .
22
As for the position of the parties that enjoyed
parliamentary representation, the Liberals were divided on
the issue of universal male suffrage, a division which,
according to Gilissen, cost them the elections of 1884. Paul
Janson, a progressive Liberal and a supporter of universal
suffrage and PR, had introduced in vain several propositions
for constitutional reform. Under the strong pressure from
POB in the early 1890s, however, it became apparent to all
sides that the issue could no longer be avoided. The
Catholic Party consented to reform in an effort to pacify
the Left (Gilissen 1958 121-122; Collier 1999). In what
Barthélemy describes as an astute political move Prime
Minster Beernaert quietly lobbied the Catholic Party for
support of a constitutional revision, while allowing the
radical elements to bring forth the matter to the
parliamentary tribune (Barthélemy 1912 269). In February in
1892, Paul Janson’s proposition for reform of Article 47 was
accepted by Parliament (Gilissen 158 122). According to the
provisions governing constitutional reform, elections for
23
constitutive assembly were held in 1892. Gilissen describes
the results of the election in the following manner:
The constituent chamber, elected on the 14th of July 1892consisted of 92 Catholics and 60 Liberals; the former lacked 9seats to capture two thirds of the votes and realize the reform ontheir own; they had to take into consideration the opposition; Thelatter was nevertheless divided approximately in half between thedoctrinaires and the radicals or the progressives (Gilissen,p.123).
All members of the constituent chamber agreed that
constitutional reform was necessary but could not reach a
settlement on the type of reform. Most Liberals and
Conservatives, however, were suspicious of universal male
suffrage (Gilissen, 1958; Barthélemy 1912). A week after
Paul Janson’s bill for the introduction of universal male
suffrage (suffrage universel pur et simple) was voted down, POB
organized a mass national strike. It was under this strong
extra-parliamentary pressure from the Left that the two
parties represented in Parliament, the Liberals and the
Conservatives, were forced to consider a compromise with
POB: the introduction of universal male suffrage tempered by
the plural vote (Barthélemy 1912; Gilissen 1958 124; Collier
1999 90). Under this system all men could vote but those
24
amongst them who were over a certain age; possessed a
certain income and property; and/or had university education
could obtain one or two additional votes, for a maximum of 3
votes (Gilissen 1958 124; Mahaim 1900 387; Barthélemy 1912).
The constitutional reform of 1893 is important for the
study of the introduction of PR in Belgium for several
reasons. It demonstrates the difficulty with which universal
suffrage with the plural vote was in Belgium and emphasizes
the role of the Left in electoral reform. The events of 1893
paved the way for PR by introducing conditions that
enfranchised many workers and thereby allowed a third party,
POB in the electoral competition (Mahaim, 1900; Barthélemy
1912; Gilissen, 1958). This situation contributed
significantly to the disproportionate results of the 1894
election and the anomalies that could be produced when three
parties compete under a majoritarian system. Moreover, as a
result of 1893 revision, “the province became the electoral
unit, a reform which would permit the introduction of
proportional representation without a new revision”
(Gilissen, p.125).
25
The matter or PR, which Prime Minister Beernaert
supported, was also briefly considered and rejected as part
of the constitutional reform. Beernaert, however was
instrumental in ensuring that the type of electoral system
was not included in the Constitution, but rather left to be
decided by an electoral law (Barthélemy 1912 538). This was
an important precondition because adopting a new electoral
law was easier than embarking on another Constitutional
Reform. Mr. Beernaert was one of the prominent advocates for
PR during the parliamentary discussions in 1894. According
to both Mahaim and Barthélemy, the Prime Minster had become
genuinely convinced that minority representation is
necessary and that PR would prove to be the most just
system. At the same time both authors underscore the fact
that Beernaert’s actions were not taken without regard to
what was best for the Catholic Party. Mr. Beernaert knew
that the inflated parliamentary majority only masked the
weaknesses and internal divisions of the Catholic Party and
believed that, while diminishing its majority, PR would
bring strength and cohesion to the party. Mr. Beernaert
26
resigned amidst strong criticism from his own party and
after PR for the legislative level was rejected by
Parliament in 1894 (Mahaim 1990 78-79; Barthélemy 1912 538).
However, in 1895 a law regulating local elections,
introduced the use of PR (with d’Hondt’s system) as a
tiebreaker. This contributed to the adoption of PR for
parliamentary elections because it demonstrated that the
system worked well and allowed people to familiarize
themselves with it (Mahaim 1900 78; Barthélemy 1912 558;
Gilissen 1958). It served to decrease the uncertainty which
surrounded PR.
After 1894 the need for reform of the existing
electoral system became apparent. The results of the 1894
election showed that the majoritarian system, combined with
the plural vote increased the disproportionality between
votes and seats even further. For comparison, with a total
of 962,000 votes the Catholics had 104 seats, the Socialists
(POB) received 310,000 votes and 28 seats, while the
Liberals remained the most underrepresented with 537,000
votes and 20 seats (Gilissen 1958 127; Mahaim 1900 77).
27
The members of Parliament, however, could not agree on
what the reform should entitle. While the progressive
representatives of the Liberal and the Catholic parties
preferred PR, their more reactionary counterparts (again
both Liberals such as Bara and Catholics like Woeste and
Schollaert) opted for redistricting. They claimed (in error)
that the substitution of all multi-member districts (scrutin
de liste majoritaire) with single-member ones (scrutin uninominal)
would solve the problem. In January 1899, “the King asked
the cabinet to prepare a project that would cut Belgium into
152 circumscriptions” (Barthélemy 1912 528). Catholic Prime
Minister de Smert de Naeyer, a moderate and supporter of
PR, resigned in protest. The Liberals and Catholics who
opposed the bill did so because they realized that the
introduction of scrutin uninominal for all districts would not
preclude disproportional representation. Moreover, they were
particularly worried that the division of districts could
lead to gerrymandering as in the United Stated at the time.
The bill was rejected. (Barthélemy 1912 528). This
28
demonstrates that actors were trying to learn from the
electoral experiences of other countries.
Mr. Smet de Naeyer was replaced by Mr. Vandenpeereboom,
“protector of a reactionary, ultra clerical regime”
(Gilissen 1958 129), who proposed the partial introduction
of proportional representation only in large multi-member
districts. This meant that PR was to be introduced in large
cities, where the Socialists and Liberals typically won. By
contract, the smaller, rural and Dutch-speaking districts,
in which the Catholic Party dominated and its rivals
received no representation were to remain under the
majoritarian system. It was not difficult to predict that
the reform was going to increase further the Catholic
majority and the level of disproportionality (Mahaim 1900
83; Barthélemy 1912 540; Gilissen 1958 129). This is a
fragrant example of the Prime Minster attempting to
manipulate electoral laws in order to increase the strength
of not only his own party. It is an example of seat
maximizing motivation par excellence.
29
The reform was unacceptable to the Liberals and the
Socialist, who did not have enough representatives in order
to stop it. Instead, they decided to block parliamentary
debates and a possible vote by disruptive behavior. The
Socialists went as far as utilizing trumpets and other wind
instruments (Mahaim, 1900 84) and throwing paper balls at
the ministers (Barthélemy 1912 541). Mass protests sponsored
by POB erupted in Brussels and other large cities and as the
government sent mounted police (Mahaim, p. 84), “civil war
was about to burst” (Barthélemy 1912 541). Under this
pressure, Vandenpeereboom withdrew his proposal and
resigned.
Restored as Premier, Smet de Naeyer sponsored the bill
for the adoption of proportional representation in all of
Belgium. The law passed the lower chamber on November 24th,
1899 (Barthélemy 1912 542) “by 65 Catholics and 5
Proportionalist Liberals, against 35 Catholics and 28
Socialists and Liberals” (Mahaim 1900 85). According to
Mahaim, PR became a reality, “because they [its opponents]
were unable to devise a system of voting to replace the
30
existing system with which all were discontented” (Mahaim
1900 85). The socialists surprisingly rejected the reform,
claiming it is acceptable only if accompanied by universal
suffrage. Barthélemy explains that they used this pretext to
cover their concern that with PR, the Liberals would no
longer depend on the POB. Catholics who voted against PR
were unconcerned with electoral justice, while those
Catholic who supported it, though interested in justice,
acted primarily out of fear that a strategic alliance
between Liberals and Socialists would hurt the Catholic
Party (Barthélemy 1912 547).
So, what does the behavior of Catholic proponents of
proportional representation suggest? First of all, it
underscores the fact that parties (especially very large
ones) are not unitary actors. Different politicians within
the same party are often guided by different motivations.
Some, like Vandenpeereboom, Woeste and Schollaert, were
clearly seat maximizers and yet, seat maximization was a
goal that led them to pursue different strategies. While
Vandenpeereboom wanted a selective introduction of PR only
31
in areas where the Catholic party was weak, Woeste and
Schollaert completely opposed the new electoral system and
advocated redistricting. The motivations of staunch PR
supporters in the Catholic Party such as Beernaert and Smet
de Naeyer appear to be more complex. They were clearly
motivated by partisan concerns but the very fact that they
were willing to adopt a system that was to diminish their
parliamentary majority precludes one from classifying them
as seat-maximizers. However, it is difficult to consider
Beernaert and Smet de Naeyer as politicians who deliberately
wanted their party to loose power. Rather, it appears that
for them preserving the plural vote (as a check on universal
suffrage) was more important than preserving the
majoritarian electoral system. It is possible (though I
possess no direct evidence of it10) that the Catholic
leaders who supported PR calculated that they would still be
able to gather the most votes between the three parties even
10 In his book Barthélemy includes a table, designed by one M. Hector Denis, which calculates the results of the elections between 1870 and 1898 under the plural vote majoritarian system and under plural vote PR.It is unclear when exactly this table was created (though certainly between 1898 and 1912), and whether Catholic Party leaders had access toit or similar calculations.
32
under a system of proportional representation. In fact, in
the first election after the introduction of PR, the
Catholic Party was still able to preserve a majority of 86
out of 152 seats or 56.6%, even if its seats decreased by 26
in comparison to the previous election (See Appendix A). The
preservation of the plural vote was only possible with
support of those Liberals who were in favor of PR and were
concerned of the rising power of the Left and the Liberal
Party’s dependence on POB. In that respect, the
introduction of the question of PR could be considered as a
brilliant heresthetic move on behalf of Beernaert and his
colleagues.
On the other hand, it is clear that Beernaert’s concern
about the future of the Catholic Party was much more than
preserving majority in the lower chamber of the Parliament.
Beernaert realized the need for improvement of the party’s
discipline and unity even at the expense of loosing seats.
Furthermore, Beernaert and other Catholic leaders who
supported PR wanted to break the alliance between Liberals
and Socialist by introducing a system that ensured the
33
independent survival of the Liberal Party. They also
believed that the Catholic party should improve its image
after Vandenpeereboom’s fiasco, that it should“ rehabilitate
itself in the eyes of the public and take credit for an
equitable reform” (Barthélemy 1912 546). These are all
partisan concerns that do not fit within a simple model of
seat-maximization.
It is also necessary to consider the possibility that
Catholic leaders who supported PR clearly realized the
internal crisis that Belgium faced and acted to preserve the
unity of the country. This motivation fits with what Benoit
calls governablitity motives for electoral change (Benoit,
2004). Last but not least, it is important to note “the idea
of justice” or the desire on behalf of Catholic PR
proponents to introduce a system that allows for the
representation of minorities. The two motivations seem
closely interrelated but should be distinguished for
analytical reasons. Barthélemy praises Beernaert and his
supporters as acting in the interest of the country, “they
have had the wisdom to take up the task to implement reforms
34
which maybe could have been done without them, against them,
or maybe also […], against the general interest of the
country” (Barthélemy 1912 549). Other authors including Reed
(1924); Mahaim (1900); Gilissen (1958) share Barthélemy’s
belief that the behavior of the Catholic Party was also
driven by noble motives and was, thus, exceptional. I
remain a bit puzzled as to why the idea of justice should
stop only at PR with the plural vote.
Nevertheless, the events surrounding passage of
proportional representation do not diminish its positive
effects. It preserved the Liberal Party and contributed to
the stability of the nation. (Mahaim 1900; Barthélemy 1912;
Gilissen 1958; Humphreys 1908). PR decreased “the perilous
coincidence which tended to establish itself between
political divisions and racial and linguistic divisions,”
“diminish[ed] the virulence of the electoral campaign” and
strengthened the Belgian parties and their party discipline
(Tyson 1904 159-160).
Proportional representation in Belgium was introduced
as a result of slow, incremental changes of the electoral
35
institutions over the period of thirty years. These changes
escalated the disproportionality of election results and
level of conflict between the main parties. During the first
episode (1831-1893) the combination of industrialization and
the constitutional provision of minimum representation led
to increase of district magnitude; larger districts in turn
interacted with the cleavage structure to create very
disproportional election results. As the problem of
disproportional representation became obvious in 1880,
politicians from both parties responded by turning to the
idea of PR. It was not until the Left organized several
major strikes to demand universal male suffrage, that the
Constitutional Reform of 1893 was implemented. The new
institutional arrangements created conditions conducive to
the adoption of PR. On one hand, some institutions played a
permissive role. The fact that the majoritarian electoral
system was not inscribed in the Constitution in 1893
facilitated the consideration of PR in the future. In
January 1899, the distinction between PR and redistricting
and the refusal of Belgian MPs to introduce single-member
36
districts, left the door open for PR. On the other hand,
other electoral institutions continued to create problems.
Universal male suffrage with the plural vote in combination
with old institutions (large districts) and existing
cleavage structures further increased electoral
dispororptionalities and the power of the Catholic Party.
When extremely conservative actors of the Catholic party
attempted to further manipulate electoral institutions to
their advantage, the Left responded by mass protests. This
caused an internal crisis that ultimately led to a
compromise: the adoption of PR. This mechanism is very
similar to the idea of path dependence. However, the main
difference here is that the crisis is endogenous: generated
by a combination between the actions of politicians under
conditions created by institutional effects and social
structures. Pressure from the Left and the increasing degree
of disproportionality remain the key factors that
contributed to the necessity of institutional reform.
The Question of Proportional Representation in Great
Britain:
37
In Great Britain, like in Belgium, the question of
proportional representation for national elections of the
lower chamber was closely connected to the extension of the
suffrage. The issue of PR in Great Britain was seriously
debated twice during the late 19th and early 20th centuries.
The first time PR was considered was in the period between
1883-1885 in the context of the Reform Bill of 1884. The
second time was between 1916-1918 during the Speaker’s
Conference and in the discussions of the Representation of
the People Bill.
In Britain, as in Belgium, proportional representation
was supported by both Liberals and Conservatives. Similarly,
some of its more pragmatic proponents considered it a
possible defense against the rise of Labor, while others
deemed PR a suitable method for just representation of
minorities. In his book Democracy and Redistribution, Boix (2003)
underscores the conservative case for proportional
representation. He explains that in the 19th century Europe,
“Given their position on the income scale, middle-class
individuals prefer[ed] a restrictive democracy (in which
38
only the wealthy and the middle class vote) to either a
regime controlled only by the wealthy or a universal
suffrage system” (Boix 2003 47). The early years of the
development of proportional representation in Great Britain
could be perceived in such light. Paul Kent argues that
Hare, one of the people who invented the single transferable
vote (STV) system of proportional representation in 1857,
was guided primarily by his distaste of full democracy and
his fear that as the franchise was made more inclusive,
upper and middle class voters would become minorities.
Hare’s primary motivation in developing a system of
proportional electoral was the belief that representation of
upper and middle classes must be insured, if Britain was to
be protected from the rise of the proletariat. The same
could be said about John Stuart Mill who supported extension
of the suffrage but was worried about the quality of
leadership and the influence of the lower classes. Mill
accepted Hare’s system for it provided means for
reconciliation between his belief in the extension of the
suffrage and his concerns that Left forces could come to
39
power as a result of such extension (Kent 1972). Both Hare
and Mill believed in the necessity to educate and co-opt the
lower classes (Kent 1972; Bromund 2001 78). Gradually, the
idea of proportional representation gathered more
supporters, especially when another suffrage extension
became more likely.
The Proportional Representation Society (PRS) in Great
Britain was formed in 1884 to counter what was perceived in
1883 as the Liberal government’s intention to sponsor a
suffrage reform and Joseph Chamberlain’s open hostility to
PR. Britain had already implemented two franchise reforms
(in 1867 and 1832), yet electoral reform was part of the
1880 election campaign of the Liberal Party and, once in
power, they had to demonstrate that they were willing to
fulfill their promise (Hart 1992 100; Bromund 2001 78). “The
Proportional Representation Society was founded in January
1884 by Sir John Lubbock, a noted financier, naturalist and
political economist, and was led by Lubbock, Henry Fawcett,
Leonard Courtney and Albert Grey” (Bromund 2001 78). They
were all MPs from the Liberal Party. It is interesting to
40
note that both in Britain and Belgium, the individuals who
first took interest in the issue of proportional
representation were members of the Liberal Party (the second
largest party, most vulnerable to the rise of the Left).
While Jennifer Hart emphasizes the idealistic and forward
thinking of these men, Bromund points out that they were
primarily moved by the fear that “if this democracy was not
accompanied by PR, it would endanger the unity, stability
and good government of Britain” (Bromund 2001 78). Bromund
explains that the leaders of the PRS believed that
proportional representation would gradually: co-opt
minorities and make them part of the system; “help unify the
nation and the empire” (Bromund 2001 81); decrease party
influence; and ensure the quality of parliamentary members
(Bromund 2001 81; Hart 1992). Nevertheless, the supporters
of proportional representation among the British elites were
not only motivated by narrow electoral and partisan
concerns. Bromund points out that the leaders of PRS were
guided by “the coherent ideology of communitarian
liberalism.” (Bromund 2001 82) and can be considered “true
41
believers and moral reformers” (Bromund 2001 82). As in
Belgium, the members of the PRS attracted supporters among
both the Liberals and the Conservatives (Hart 1992).
In 1884, the Proportional Representation Society made
its first attempted to introduce PR. As the government
prepared the Reform Bill in the winter of 1884,
representatives of the PRS met with Prime Minister Gladstone
in order to convince him to consider proportional
representation as part of the reform. Gladstone rejected the
inclusion of PR on grounds that it is a matter not of
suffrage, but of seat redistribution, which was supposed to
be discussed in a separate bill. Trusting that a
redistribution bill would be introduced during the next
session, and concluding that the government would not allow
PR to be brought up at this moment “the PRS decided not to
propose an amendment to the franchise bill” (Hart 1992 105).
The Reform Bill was passed in the Commons. However, the
proponents of proportional representation found support in
the House of Lords, which decided that the bill should not
become law “unless ‘accompanied by provisions for so
42
apportioning the right to return members of Parliament as to
insure a true and fair representation” (Hart 1992 107).
While some of the Lords who returned the Bill to the Commons
did so because they believed in the necessity of minority
representation (Conservative Lord Salisbury being the most
prominent example11), many used the issue in order to block
the suffrage reform. “During this period there was a
vehement campaign in the country against the Lord’s refusal
to pass the franchise bill before a redistribution bill had
been enacted” (Hart 1992 111).
In Great Britain (unlike Belgium) government remained
hostile to the issue of proportional representation. As
Gladstone openly rejected Lubbock’s attempts to bring
attention to the matter of PR in the Commons, the government
prepared the Redistribution of Seats Bill (proposed on Dec.
1st, 1884) without informing the Proportional Representation
Society. The Bill included the removal of the limited vote
of 1867, as well as the creation of single-member districts
in most of the country. “In announcing that one-member11 Both Hart and Bromund agree that Lord Salisbury’s position was motivated primarily by his concern that the Conservative party might become a minority in the future. He supported PR out of pragmatism rather than idealism (Hart, 1992; Bromund, 2001).
43
districts would be the general and prevailing rule,
Gladstone justified the arrangement as the best on the
ground that it was very economical, very simple, and went a
long way towards ‘what is roughly termed the representation
of minorities” (Hart 1992 114). The bill was approved by the
Lower House. In the Upper House, the only PR supporter at
the time, Lord Salisbury, was compelled by the author of the
redistricting bill, Charles Dilke, to not bring up the
concept of minority representation during the debates. Both
Hart and Bromund explain that Salisbury agreed with the Bill
because he believed that single-member districts would be
enough to ensure future representation of the Conservative
Party12. (Hart 1992 113; Bromund 2001 92). The Proportional
Representation Society attempted lobbying the Lower and
Upper House but its efforts were in vain. Lubbock made one
last attempt to address the matter of proportional
representation with the single transferable vote:
On the first day of the committee stage of the Redistribution ofSeats Bill (2 March 1885) in order to get the subject discussed,
12 In addition, Bromund contends that “Salisbury shied away from proportional representation, as Leslie Stephen remarked, because it did nothing to ensure that the working classes would not dominate the system. Indeed it guaranteed that the class with the most votes would have a majority in the Commons” (Bromund, p.92)
44
Lubbock moved ‘that the committee be instructed that they have thepower in all cases where an elector is entitled to one vote only,to enable the elector to nominate more than one candidate to whomunder certain circumstances that vote might be transferred in amanner indicated by the elector (Hart 1992 117).
Though it acquired the support of some of the Liberal and
some of the Conservative MPs the motion was rejected by 134
to 31. On June 25th 1885, the Lower House voted the
Redistribution of Seats Bill (Hart 1992 118-9). As a result,
the single-member district became the predominant electoral
unit. “Only 27 constituencies (23 boroughs, the City of
London and 3 Universities) were allocated two members by the
bill” (Hart 1992 113). By creating the false impression of
solving the matter of minority representation, the
Redistribution of Seats Bill of 1885 significantly decreased
the possibility of PR introduction in Britain. “Nothing
could have sapped support for PR more effectively than
electoral districts” (Bromund 2001 93).
In contrast, the multi-member province was the
electoral unit in the 1893 revision of the Belgian
Constitution. Furthermore, Belgian supporters of PR argued
successfully and correctly that redistricting would not have
45
solved the problem of disproportional representation. One
should also note that the majority of electoral districts in
Belgium were considerably larger than their British
equivalents. The problem of disproportionality was much more
acute in Belgium than in Great Britain. In Britain the idea
of seat redistribution appeared as a much more reasonable
solution than the introduction of a completely new electoral
system.
In addition to the predominance of single-member
district, four other factors contributed to the rejection of
PR in 1883-1885. First, the organization of the Proportional
Representation Society was not efficient. The members of PRS
made several mistakes. They lacked funding and delayed
selecting a particular type of PR (Hare’s single
transferable vote) until after the introduction of the
Redistribution of Seats Bill. The PRS settled on the issue
of the exact method of PR only after the single-member
district was brought up. Redistricting seemed to be a much
easier task than the STV, which includes a complicated
procedure for the transfer of votes into seats. In addition,
46
some of the members of the organization, Courtney in
particular, exhibited a condescending attitude and excessive
loquaciousness on the floor. They ultimately utilized the
wrong type of arguments. In short, the Proportional
Representation Society failed to demonstrate that PR was a
better solution than seat redistribution (Hart 1992; Bromund
2001).
Second, Bromund emphasizes the difficult position of
the organization in regard to settling on what arguments in
favor of proportional representation it should employ. The
Proportional Representation Society was a heterogeneous
group that failed to rally the supporters of PR. It included
both idealistic believers in PR and pragmatists who were
suspicious of the ability of fancy franchises to hold back
the advent of the masses. As a coalition the PRS sought “to
appeal to all its supporters and so keep the pressure on the
government” (Bromund 2001 93). It was betrayed by both sides
of its followers when it came down to the crucial votes.
On the other hand, the Liberal government did
everything possible to stop the advent of PR, mainly because
47
Prime Minister Gladstone did not believe in the principle of
minority representation (Hart 1992). It is interesting to
note Gladstone’s ability to maneuver in such a way as to
prevent the PRS from even putting up the issue of
proportional representation on the agenda in 1884. Finally,
Bromund points to British conservatism as another reason for
the cold reception of proportional representation, “PR was
profoundly untraditional, and the supposedly immemorial
English constitution was never worshipped more reverently
than in the mid-Victorian era, when it seemed to symbolize
the political stability and grandeur of Great Britain”
(Bromund 2001 91).
The next key moment for PR in British history came when
the Speaker’s Conference of Oct. 1916-Jan. 1917 “suggested
that STV should be adopted for all boroughs constituencies
which elected three or more members, and the [alternative
vote] AV for single-member constituencies” (Chadwick 1999
195). The Conference also recommended that the remaining
multi-member districts should not have more than 5 seats
(Hart 1992 191). The Conference had been convened in order
48
to address the demands for extension of the suffrage created
by World War I. As the war raged on, it no longer seemed
just to deny the right to vote13 to people who fought for
their country (Close 1977 890; Chadwick 1999 195; Hart 1992
178). As Hart explains some members of the Speaker’s
Conference were true believers in the fairness of PR, while
others acted out of concern that without a mechanism of
minority representation, the impending franchise bill would
prove to be extra advantageous to the recently founded
Labour Party (Hart 1992 183).
The Liberal government, led by Lloyd George, disagreed
with the Conference’s recommendation for PR. The bill, with
the proposal for partial introduction of proportional
representation, nevertheless, went to the House of Commons,
where it encountered persistent opposition. “[T]he Lords
reinstated proportional representation in it three times,
but their proposals were always turned down by the Commons”
(Hart 1992 185). The Representation of the People Act of
1918:
13 About half of the working class was disenfranchised at the time (Close, p.890; Chadwick, p.195; Hart, p.178).
49
“did not prescribe that elections in any constituenciesexcept some University seats should be held under theprinciple of PR, but provided that a Royal Commission was toprepare a limited scheme to apply it to one hundred seats.Any proposals emanating from the commission would only beimplemented if approved by both Houses of Parliament withintwenty-one days” (Hart, p.185).
The Lower House voted against the Commission’s Scheme (Hart
1992 185) in 1918.
Especially noteworthy in the events of surrounding the
reform of 1916-1918, is the fact that the Liberals and the
Conservatives were split on the issue of PR in both the
Lower and the Upper House. In the Commons most Conservatives
voted against proportional representation because of what
Close deems, their “grater pragmatism and devotion to
continuity” (Close 1977 906). The Liberals and Labour
originally favored proportional representation, “until
alienated by the Lord’s decisions” (Hart 1992 186). Hart
identifies two reasons for the apparently contradictory
behavior of these two parties. First, some MPs were
irritated by the possibility that the mode of their election
could be imposed by the Peers. Second, and more importantly,
they suspected the Upper House in using PR to stall the
50
franchise bill (Hart 1992 186), as it had done during the
1884-85 reform.
The majority of Conservatives in the House of Lords
supported the introduction of proportional representation
because they believed in the so called “conservative case
for PR” i.e. that in the long turn this electoral system
would ensure the survival of their party and prevent the
domination of the proletariat (Hart 1992; Close 1977). As
for the liberal representation in the Lords, Hart explains
that they were split evenly but does not elaborate on their
motivations (Hart 1992).
The factors that contributed to the failure of PR in
1916-18 were similar to the ones that were detrimental to
the issue during the first campaign. As in 1884-5, there
were several problems with the organization and strategy of
the PRS, including the presentation of PR as a panacea for
all problems from the allegedly deteriorating quality of MPs
to Britain’s involvement in World War I. Such allegations
pushed away many politicians (Hart 1992). In addition, just
like during 1884-85, the Proportional Representation Society
51
was slow in identifying the STV the method of PR that they
supported. Furthermore, during the second campaign the
organization was not able to prevent the simultaneous
consideration of the AV and the STV. The conflation of the
two issues confused further many legislators who where not
familiar with the technical aspects of the two systems and
considered them very complicated. This seriously hurt the
chances of PR implementation (Hain 1986 5). Second, Hart
makes a point that is similar to Bromund’s contention that
the coalition nature of the PRS proved to be harmful to its
cause. “The trouble with what can be called ‘the
conservative case for PR,’ was that it failed to win over
enough Unionists to secure majority in the Commons, while it
also alienated some Liberals and many Labour MPs” (Hart 1992
191).
In third place, as during the previous PR campaign, the
Proportional Representation Society was not able to secure
the support of the Prime Minister. Though in 1916-18
conditions were a bit more favorable because the government
was divided on the issue of PR, the Premiere Lloyd George
52
deeply disliked the concept of minority representation and
acted to hinder the initiatives of the PRS. While both
Gladstone and Lloyd George were primarily motivated by their
ideological stances against minority representation, Lloyd
George had a much more pragmatic, seat-maximizing
motivation. He was afraid that PR might damage the Liberal
Party’s majority in Wales (Hart 1992). Finally, British
conservatism expressed by the refusal of many politicians to
believe in the shortcomings of the first-past the post
system; their lack of desire to understand the STV (which
was in itself a complicated procedure); their ideological
resistance to and often irrational fear of minority
representation were also factors that affected both reform
campaigns (Hart 1992). In Britain PR was an “innovative
idea” (Bromund 2001), advocated by individuals who shared
“unconventional” religious and philosophical beliefs (Kent
1977) and who were less prominent than most of the PR
opponents (Hart 1992).
In short, in Great Britain proportional representation
remained just an idea. It was an idea that never acquired
53
enough salience to be seriously considered by the majority
of British politicians. By the late 19th century, Great
Britain was already a country with strong constitutional
tradition and stable electoral system. It did not have as
strong and hostile Left, nor ethno-linguistic divisions14
that were as deep and as polarized as the ones in Belgium.
Consequently, Great Britain did not experience the internal
crisis that necessitated the change of the electoral system
in Belgium. Furthermore, while extension of the suffrage
continued, the seat-redistribution reform of 1885, which
made single-member districts predominant became a major
institutional hurdle to PR. This too seems to be consistent
with the concept of path dependence, except for the
endogenous nature of the shock. Circumstances and actors in
Great Britain failed to create an opening for electoral
system change.
Explaining Electoral System Change: In Lieu of Conclusion
14 The Irish question is one possible exception. In the case of Irish Home Rule, proportional representation was, in fact, introduced for Ireland. However, the majority of British politicians were convinced that Ireland is an exception and that PR is not necessary in Great Britain, a county with strong constitutional tradition and electoral institutions that work well. (Hart, 1992).
54
A comparison of Belgium and Great Britain reveals
several factors that contributed to the introduction of
proportional representation in Belgium, and to the rejection
of this type of electoral system in Great Britain. These are
the following: 1) existing electoral institutions
(especially the interaction between district size and the
plural vote); 2) different levels of institutional
stability; 3) the perception of internal national crisis; 4)
very strong and consistent pressure form the Left in the
form of mass protests; 5) dangerous overlap of all major
societal cleavages, which created two distinct ethno-
linguistic and economic groups. It is necessary to consider
how these factors interact with each other in order to lead
to the introduction of proportional representation.
The historical episodes explored in this paper
demonstrate that new institutional arrangements (in this
case, the change from majoritarian/plurality systems to
proportional representation) result from the process of
negotiation between key political actors. While it is actors
(politicians and parties) that manipulate electoral rules,
55
actors themselves are constrained by existing electoral
institutions.
It must be noted that in both Belgium and Great
Britain, the concept of proportional representation was
developed and popularized in the context of suffrage
reforms. In that sense, both countries experienced the rise
of worker’s political parties and changes in both the
electorates and party systems. The starkest difference
between developments in the two countries is the fact that,
while proportional representation came to be viewed as an
important alternative in the Belgian Chamber of
Representatives, it did not merit serious consideration in
the British Parliament 1884-5 and 1919. Despite the efforts
of the members of the Proportional Representation Society
and the backdrop of suffrage reform, the credibility of the
first-past-the-post electoral system for national
legislative elections in Britain was not questioned. No
opening was created for the adoption of PR in Great Britain.
By contrast, such institutional opening did exist in
56
Belgium. The crucial question is what explains this
difference?
There are two aspects to the answer, namely, the
perception that there are serious problems with the existing
electoral system and the solutions proposed. Problems with
the existing electoral system stem from the manner, in which
electoral and constitutional institutions interact with pre-
existing social conditions. In Great Britain, due to smaller
electoral districts and a simpler cleavage structure there
was less electoral disproportionality. In that respect, when
existing institutions and social conditions conspire to
create gross electoral disporportionalities, the likelihood
of change to PR increases. Second, a simpler cleavage
structure and comparatively less hostile Left did not lead
to major internal crisis in Great Britain15. Arguably, World
War I also created a sense of crisis in the British society,
a crisis that prompted the consideration of another suffrage
extension in 1916. Nevertheless, this was an external threat
that served to bring political groups and the nation, in
15 It must be noted again that the Irish question is the one major exception (Hart, 1992).
57
general, closer together (Hart, 1992). It must be noted that
the since the high degree of disproportionality and strong
Left pressure are present when PR is adopted and absent when
it is not, the present study does not allow one to
distinguish which of these factors is more important.
Further inquiry into this issue is necessary.
On the other hand, institutional alternatives also
affect institutional choices. Several factors contributed to
the consideration of PR as a credible solution in Belgium.
First, because Belgium was a relatively young country,
politicians were more open to consider electoral systems,
other than the existing majoritarian one. This becomes
particularly clear when compared to the situation in Great
Britain, where most influential politicians, PMs included,
refused to attempt to even understand PR. When the
institutional stability is low, the constraints are weaker
and the possibility of very different institutional
arrangement increases (Belgium); when the institutional
stability is high, the structure exercises greater
constraints on the actors (Britain). Second, the specific PR
58
formula proposed is important. Actors in both countries
preferred PR formulas that were developed by their
compatriots and were thus, more familiar to them. However,
the single transferable vote developed by Hare in Britain
was considerably more complicated than d’Hondt’s formula. In
Belgium, PR appeared to be a much more feasible option.
Third, actors in both countries operated in conditions of
heightened uncertainty - the effects of PR systems were
known in theory but not in practice. The positions which
Gladstone and Lloyd George adopted in regard to PR
demonstrate Thelen and Steinmo’s (1992) point that when
actors posses imperfect information they are more likely to
rely on existing institutions. The feasibility of this
electoral system in Belgium became clearer after PR was
introduced for local elections in 1895. The practical
application of PR (albeit in somewhat different context)
increased politician’s confidence in the system and lowered
some of the uncertainty associate with it.
Forth, existing institutions can serve as veto points
that increase the likelihood of adoption of other
59
institutions, or can prevent it entirely. This is because
once a particular institutional arrangement takes root, it
becomes part of the overall institutional structure and,
short of a complete overhaul of the system, could prove
conducive to certain institutional arrangements and
unfavorable to others. The size of the electoral district is
such an institution. Once single-member districts become
introduced and are considered as the norm, they prove to be
a major hurdle to the introduction of proportional
representation. In that respect, multi-member districts are
a necessary condition for PR. The existence of an upper
chamber can also serve as a veto point.
Nevertheless, when an opening is created actors can
maneuver by suggesting existing/possible options. For
example, the major reason why PR was considered a credible
solution in Belgium but never in Great Britain is that the
issue of electoral system type was never distinguished from
the issue of seat redistribution16. Gladstone made a cunning
herestehtic move when he managed to convince the members of
16 While the two issues are obviously connected (i.e. one cannot have PRand single-member districts), theoretically these are distinct concepts.
60
the PRS to not introduce the question of proportional
representation in the Reform Act of 1885. Gladstone linked
PR to seat redistribution, but when single-member districts
were introduced in most of England, the question of PR
became irrelevant. By contrast, Beernaert managed to
convince the Parliament to not inscribe the type of
electoral system in the constitution, thus leaving a wider
institutional opening that allowed the possibility of PR
adoption via a regular law. The fact that the Belgian
Premiers supported PR, while the British ones, rejected it
hints to the importance of certain political actors and
serves to further strengthen Knight’s argument about power
asymmetries.
The outcome, change from majoriatrian/plurality system
to proportional representation, is the result of
negotiations and compromise between actors. Actors matter
both in the process of negotiations once an institutional
opening is created, and also in creating such institutional
opening. However, actors’ actions alone were not sufficient
to create opening without the existence of certain
61
institutions. In that respect, the mechanism presented above
is consistent to a large extent with the concept of path
dependence. A major difference between the two is that in
path dependence institutional change results as a
consequence of exogenous shocks, whereas here the shocks are
endogenously created. This suggest the possibility that the
notion of path dependence should be altered as to
incorporate endogenous change.
Finally, the case of Belgium demonstrates that party
leaders are not necessarily motivated by the narrow concern
of seat maximization. Nevertheless, PR was introduced in
Belgium as a defense mechanism against the left as Rokkan
and Boix claim. The adoption of proportional representation
is a compromise struck between actors on their way to
introducing equal universal male suffrage. In that respect,
Belgium is as much less of an anomaly than it appears to be,
based purely on an analysis of the power distribution
between parties in parliament. This demonstrates that the
notion of threat, understood by Boix (1999), Benoit (2004),
62
Colomer (2005), and Andrews and Jackman (2005) in very
narrow electoral terms, should be broadened.
63
Appendix A: Composition of the Belgian Chamber ofRepresentatives
First Election Under Universal Male Suffrage and Plural Vote(Majoritarian System) - 1894
Votes Seats in ParliamentPolitical Party
Number of Votes
Percentage ofVote:
Number of Seats
Percentageof Seats
Catholic Party
962,000 53.2%
104 68.4%
Liberal Party
537,000 29.7%
20 13.2%
Belgian Worker’s Party
310,000
17.1%
28 18.4%
TOTAL: 1,809,000 100% 152 100%Sources: Gilissen, 1958; Mahaim, 1900.
Last Election under Universal Male Suffrage and Plural Vote(Majoritarian System) – May 22nd, 1898 Partial Election (Group 1)
Votes Seats in ParliamentPolitical Party
Number of Votes
Percentage ofVote:
Number of Seats
Percentageof Seats*
Catholic Party
N/A N/A 112 73.7%
Liberal Party
N/A N/A 12 7.9%
Belgian Worker’s Party
N/A N/A 28 18.4%
TOTAL: N/A N/A 152 100%Sources: Tyson, 1904; Barthélemy, 1912; Caramani, 2000.
Belgium 1900: First Election (General Election) Under PRVotes Seats in Parliament
Political Number Percentage of Number of Percentage
64
Party of Votes
Vote: Seats of Seats*
Catholic Party:
N/A N/A 86 56.6%
Liberal Party:
N/A N/A 33 21.7%
Belgian Worker’s Party:
N/A N/A 32 21.1%
“DemocraticChristian – a dissident from the Catholic party.” (Tyson 1904 158)
N/A N/A 1 0.66%
TOTAL: N/A N/A 152 100%Sources: Tyson, 1904; Barthélemy, 1912.
*Results rounded to nearest tenth of a percent.
** I am in the process of compiling additional data.Appendix B: Result of British Parliamentary Elections
General Parliamentary Election: 24th of November – 9th of December1885
Votes Seats in ParliamentPolitical Party
Number of Votes
Percentage of Vote:
Number of Seats
Percentageof Seats*
ConservativeParty:
2,020,927 44.6
N/A N/A
Liberal Party:
2,199,998 47.44
N/A N/A
Nationalists(Irish): 310,608 6.7
N/A N/A
Other Parties: 106,702 2.3
N/A N/A
Labour Party - - - -
65
Independent Labour Party
- - - -
TOTAL: 4,638,235 100
N/A N/A
Source: Caramani, 2000.
General Parliamentary Elections during 1910Votes in June 1910 Votes in October 1910
Political Party
Number of Votes
Percent of Votes:
Number of Votes
Percent of Votes
ConservativeParty 3,104,407 46.6 2,420,169 46.2Liberal Party 2,866,157 43 2,293,869 43.8Labour Party 505,657 7.58 371,802 7.1Nationalists(Irish) 126,647 1.9 131,720 2.5Other Parties 64,532 1 17,678 0.3Independent Labour Party - - - -TOTAL: 6,667,400 100.08 5,235,238 100Source: Caramani, 2000.
*Results rounded to nearest tenth of a percent.
** I am in the process of compiling additional data.
Bibliography:
66
Andrews, Josephine and Robert Jackman. “Strategic Fool: Electoral RuleChoice Under Extreme Uncertainty.” Electoral Studies. Vol. 24 (2005) pp.65-84.
Barthélemy, Joseph. L’Organisation du Suffrage et L’Experience Belge. Paris: M.Giard & E. Brière, 1912.
Benoit, Kenneth. “Models of Electoral System Change.” Electoral Studies. Vol.23 (2004) pp. 363-389.
Boix, Carles. “Setting the Rule of the Game: The Choice of ElectoralSystems in Advanced Democracies.” The American Political Science Review. Vol. 93No.3. (Sept., 1999) pp. 609-624.
Brew, David. “The United Kingdom.” European Electoral Systems Handbook. Hand, Geoffrey, Jacques Georgel and Christoph Sasse, eds. London: Buttersworths London, 1979.
Bromund, Ted R. “Uniting the Whole People: Proportional Represnetation in Great Britain, 1884-5, Reconsidered.” Historical Research, Vol.74, No.183 (Feb. 2001) pp.77-94.
Caramani, Daniele. Elections in Western Europe Since 1815: Electoral Results byConstituencies. London: Macmillan Reference, 2000.
Chadwick, Andrew. Augmenting Democracy: Political Movements and Constitutional Reformduring the Rise of Labour, 1900-1924. Vermont: Ashgate Publishing Ltd, 1999.
Close, David H. “The Collapse of Resistance to Democracy: Conservatives,Adult Suffrage and Second Chamber Reform, 19911-1928.” The Historical Journal Vol.20, No.4 (Dec., 1977), pp. 893-918.
Collier, Ruth Berins. Paths Toward Democracy. Cambridge: Cambridge University Press, 1999.
Colomer, Josep M. “It’s Parties that Choose Electoral Systems (or,Duverger’s Laws Upside Down).” Political Studies. Vol. 53. (2005) pp. 1-21.
Cusack, Thomas, Torben Iversen and David Soskice. “Specific Interestsand the Origins of Electoral Institutions.” Paper prepared for theAnnual meeting of the American Political Science Association, ChicagoSept. 2-5 2004.
Gilissen, John. Le Régime Représentatif en Belgique depuis 1790. Bruxelles: La Renaissance du Livre, 1958.
67
Hain, Peter. Proportional Misrepresentation: The Case Against PR in Britain. England: Wilwood House Limited, 1986.
Hart, Jennifer. Proportional Representation: Oxford: Clarendon Press, 1992.Humphreys, John H., “Proportional Representation in Belgium,” Contemporary Review, 94 (July-Dec., 1908) pp. 437-451.
Kern, Paul B. “Universal Suffrage Without Democracy: Thomas Hare and John Stuart Mill.” Review of Politics, 34:3 (July 1977). Pp. 306-322.Knight, Jack. “Models, Interpretations, and Theories: ConstructingExplanations of Institutional Emergence and Change,” in Jack Night andItai Sened (eds.) Explaining Social Institutions. Ann Arbor: University ofMichigan Press, 1995.Lakeman, Enid. How Democracies Vote: A study of Electoral Systems. London: Faber andFaber Ltd., 1955.
Mahaim, Ernest, “Proportional Representation and the Debates Upon the Electoral Question in Belgium,” Annals of the American Academy of Political and Social Science Vol.XV (Jan-May, 1900) pp. 381-404.
Mainwaring, Scott P. Rethinking Party Systems in the Third Wave of Democratization: TheCase of Brazil. Stanford, California: Stanford University Press, 1999.
Norris, Pippa. Electoral Engineering: Voting Rules and Political Institutions. Cambridge:Cambridge University Press, 2004.
Pierson, Paul. Politics in Time. Princeton: Princeton University Press, 2004
Powell, G. Bingham, Jr. Elections as Instruments of Democracy:Majoritarian and Proportional Vision. New Haven & London: YaleUniversity Press, 2000.
Reed, Thomas Harrison. Government and Politics of Belgium. New York: World Book Company, 1924.Scharpf, Fritz W., “Institutions in Comparative Policy Research,”Comparative Political Studies 33:6-7 (August 2000) pp. 762-90.
Thelen, Kathleen Ann. “How Institutions Evolve: Insights fromComparative Institutional Analysis.” Mahoney and Rueschemeyer (eds.)Comparative Historical Analysis in the Social Sciences. Cambridge: Cambridge UniversityPress, 1992.
Thelen, Kathleen and Sven Steinmo. “Historical Institutionalism inComparative Politics.” Structuring Politics: Historical Institutionalism in ComparativeAnalysis. Cambridge: Cambridge University Press, 1992.
68