These specific assumptions about voting behavior can be generalized in several
ways, without altering the nature of the results. Two specific aspects are central
to our argument, however. First, voters are not fully intertemporally rational,
and not modeled as strategic players. We are not too apologetic about ruling out
strategic voting. Instances of this phenomenon may certainly be observed in the
real world. Given the low individual stakes for a single, atomistic voter, however,
sophisticated strategic voting is not necessarily more plausible than simple retro-
spective voting. On the other hand, we believe that it is much more important to
model professional politicians as fully rational, strategic actors.
The second central aspect is that voters reward only their own party (rather
than the whole coalition) when they are pleased with the government performance.
Again, the specific way this behavior is modeled is not so important. The central
idea is that at least some voters are ideologically attached to a party, and that
their voting behavior discriminates between parties in a coalition government.
But if the parties merge, voters become unable to discriminate between them. As
stressed by Bawn and Rosenbluth (2002), the idea that voters can discriminate
between parties in a coalition government, but not between groups inside a single
party, may be at the core of why coalition governments behave differently than
single-party majorities.
2.2.5. Electoral rules
When votes have been cast, they are translated into seats for the next legislature
according to the electoral rule in place. Under proportional elections, all voters
belong to a single national district, and the electoral formula is PR. Thus, each
party receives a seat share in the next legislature directly proportional to its vote
share in the national district. Under majoritarian elections, voters are distrib-
uted in a continuum of single-member districts indexed by d, and the electoral
formula in each district is plurality rule. Thus, district d has one seat in the next
legislature, and the seat is won by the party with the highest vote share in d.
2.2.6. Political-economic equilibrium
A full politico-economic equilibrium is an equilibrium of this multi-stage game.
More precisely, an equilibrium is:
a) a party system resulting from optimal choices by the primitive groups of
legislators at the party formation stage, taking into account the electoral rule and
the expected equilibrium outcomes at stages b)-d).
11