question of proportionality and ‘‘fairness’’ of a system. These are the main building
blocks of an electoral system. But electoral laws have many more components than
the system itself. Rules on ballot access, political advertising, or theWnancing of
parties all aVect, or at least we think they may aVect, the performance of parties in
elections. And the list of laws and rules governing features of elections is much
longer than these three examples (Massicotte, Blais, and Yoshinaka 2004 ). The
kinds of eVects pointed out by Duverger may well explain what happens within a
district but these broader kinds of supporting laws may regulate both the kinds of
coordination that may occur across districts or, more simply, shape the playing
Weld for parties in diVerent ways. Laws on elections, then, may have similar, albeit
milder, eVects as electoral systems.
More importantly perhaps, they may be easier to change than electoral systems.
A fairly simple model of electoral system change in which self-interested politicians
jockey for advantage would—absent uncertainty and any costs to change at least—
have us suppose that electoral systems are in a constant state ofXux. This is, as the
evidence of Colomer shows, most deWnitely not the case. Models of institutional
change driven by rational actor models would probably need to build in compon-
ents of uncertainty and costliness as a means of slowing down the changes. On the
other hand, changes in electoral laws may well be easier to change and confer the
same kinds of advantages as electoral manipulation, if on a smaller and more
modest scale.
Some evidence of changes in laws does seem to support that pattern, as Birch
shows for the Eastern European cases: the gradual inching up of some thresholds,
some tightening of nomination procedures, and/or the introduction of a ‘‘deposit’’
for candidates tended to beneWt bigger parties (with more resources) but, more
importantly, raised the barriers to entry for newcomers (Birch, Millard, and
Williams 2003 , 189 – 91 ). But these tendencies were not seen everywhere. Some
countries did not increase the vote threshold and in some places nomination
requirements were relaxed.
One question is whether we see evidence of this attempt to keep out new
entrants and protect current players in more established democracies as would
be consistent with models grounded in rational self-interest. The available evidence
on that is decidedly mixed. If anything, changes in some of the supporting electoral
laws have seen a relaxation of barriers that could favour new entrants or at least
smaller parties (Bowler, Carter, and Farrell 2001 ).
Even these kinds of changes have their own uncertainties and their own costs. It
may not be worth changing entire electoral systems or even laws for the sake of one
or two seats, notwithstanding the fact that politicians are motivated by seat
maximization. Furthermore, too blatant a manipulation of electoral rules of any
kind could well discredit the result from any new set of laws or rules in the eyes of
voters or even some politicians. The choice of electoral rules may thus not be
entirely unconstrained.
590 shaun bowler