The Theory of Electoral Systems

Date01 December 1970
Published date01 December 1970
Subject MatterArticles
Syracuse University
N ELECTORAL LAW authoritatively prescribes the manner in which the
political preferences of a community are to be expressed and ordered. The
relative merits of the various systems -
plurality, party list, single trans-
ferable vote, ballotage, etc. -
are debated with alacrity by politicians and scholars.
Indeed, the main features of the classic dialogue between Mill and Bagehot are
discernible in modern research. J. F. S. Ross,l expanding on Mill’s argument, insists
that any system other than proportional representation (with preferential voting
and quota counting) is simply not &dquo;civilized,&dquo; since the principle of representation
demands the fullest possible expression of preferences. F. A. Hermens,2 injecting
Bagehot’s argument for the single-member constituency with contemporary data,
warns that proportional representation leads inexorably to multipartism and
chronic instability. Perhaps the best known analysis is that of Maurice Duverger,-9
who claims that proportional representation has a multiplicative effect on an other-
wise dualistic political universe.
To the extent that this debate concerns &dquo;stable&dquo; government versus &dquo;fair&dquo;
representation, it is obvious that there are some value choices to be made and
defended. Yet it is equally clear that the contrary positions bifurcate from a com-
mon factual premise about the power of the electoral system as an independent
variable. My purpose here is to consider its status as an important dependent
variable. To turn the traditional argument around, it is plausible that a certain
type of party system tends to produce a certain type of electoral system.4 After all,
an electoral system redistributes’ power shares so as to determine who shall occupy
the positions of authority. Political elites have a stake in this process, and the
selection of an electoral law is bound to be deliberative, not Delphic. The nature
of this critical calculation is a neglected area of comparative political research.
The inquiry will be conducted within the framework of Anthony Downs’
deductive model of political behavior,5 with the incorporation of some useful
corollary statements from Riker6 and Buchanan and Tullock.7 The model serves
NOTE : I am fortunate to have had Professor Frank Munger’s constructive criticism during
the initial drafts of this paper. I owe thanks also to Professor Frank Marini for his
sage professional advice. Any shortcomings are my responsibility.

Elections and Electors (London: Eyre & Spottiswoode, 1955).

Democracy or Anarchy? (Notre Dame, Indiana: Abbey Press, 1941).

Political Parties, trans. B. and R. North (New York: Wiley, 1966). Unfortunately Duverger
ascribes to this relationship the status of a "sociological law" without adequately deal-
ing with several glaring exceptions.
Causal symmetry is suggested by J. G. Grumm, "Theories of Electoral Systems," 2 Midwest
Journal of Political Science (1958), 357-76; L. Lipson, "The Two-Party System in
British Politics," 47 American Political Science Review (1953), 337-58; A. Downs,
An Economic Theory of Democracy (New York: Harper & Row, 1957), 124.

Op. Cit.
W. H. Riker, The Theory of Political Coalitions (New Haven: Yale U. Press, 1965).
J. M. Buchanan and G. Tullock, The Calculus of Consent (Ann Arbor: U. of Michigan
Press, 1965).

both as a construct for making sense out of the decision-making process and as a
source of non-obvious propositions which may be tested empirically. We assume,
after Downs, that political parties are rational: they seek to minimize inputs and
maximize outputs.8 Party members are motivated by the income, prestige, and
power which derive from holding office. While they share the same ultimate goal,
however, they may disagree on the most efficient means to it, and thus the party
will behave like an imperfect coalition. This situation obtains because of uncer-
tainty, the absence of reliable knowledge about past, present, future, and hypo-
thetical events. Information is costly, and party members possess it unequally.
Party rationality in Downs’ model means that parties, given the goal of win-
ning elective offices, attempt to maximize votes. In order to deduce party behavior
regarding electoral law change, we shall want to say that a party seeks to maxi-
mize not just votes, but also legislative seats. This requirement increases the utility
of the model in this particular investigation without altering its logical structure.
We shall reconsider the maximization rule at one point below, but this should not
involve serious difficulties.
We know from correlational studies and simple mathematical necessity that
electoral laws tend to produce disproportional effects: with few exceptions, parlia-
mentary party systems are less fractionalized than elective party systems.9 More-
over, the various electoral systems produce a wide range of disproportional results.
If we think of electoral laws in terms of the cost (in votes) of a parliamentary
seat, then the evidence indicates, for example, that a plurality system requires
much higher &dquo;payments&dquo; from smaller...

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT