A more obvious method than this, having electoral thresholds—such asGermany’s 5 percent threshold—places hurdles in front of smaller parties that
they mayWnd hard to overcome. In fact it is, in part, the intent of these thresholds
to stiXe smaller (and often more extreme) political parties, parties that can make
coalition bargaining even more diYcult.
In general terms, then, the electoral system shapes the number of parties but, like a
pebble thrown into a pond, the eVects of the number of parties ripples through the
political system. Often these eVects are of great consequence to our normative
understanding of representation, of accountability, and of governability. Changing
an electoral system is not something to be done lightly. However, when either analysts
or politicians consider changes to the electoral system it seems that the main topics of
debate really focus on the distributional eVects of the system. That is, even though
changing the electoral system does involve many changes on ideas of representation or
accountability, a lot of the discussion focuses on who wins and who loses (see, e.g.,
Benoit 2004 and Bawn 1993 for discussion of postwar Germany).
In eVect, this means that many of the studies of change in electoral systems focus
on the central point of who is doing the choosing. Colomer has an especially
thorough statement of the way in which parties themselves may or may not choose
to invoke the eVects of Duverger’s Law. His insight is to note that parties choose
electoral systems and not—as a mechanical reading of Duverger suggests—vice
versa (Colomer 2005 ; Blais, Dodrzynska, and Indridasan 2005 ). Electoral system
choice is, as Colomer notes, made by parties and so, to a large extent, Duverger’s
eVects are endogenous to the initial choice of institutions, a fact that actors in early
post-cold war Eastern Europe knew well. The choice of electoral systems in the newly
democratic Europe was neither easy, straightforward, nor without controversy
precisely because actors recognized that electoral systems do have eVects on political
parties (Colomer 2005 ; Birch, Millard, and Williams 2003 ). So for these authors
Duverger may have it the wrong way round to some extent: a party system with a few
big parties will choose an electoral system that favors keeping a few big parties and
not want to change that. Similarly, a party system with lots of small parties is likely to
want to keep a system such as list PR that favors small parties.
3 Changing Electoral Systems?
.........................................................................................................................................................................................
Clear-cut points of choice of electoral systems are fairly rare, although in principle
countries can consider reforming their current system at any time. Electoral
institutions change relatively infrequently and the major changes of recent
582 shaun bowler