Repeating the same steps as in section 3.1.1, we can rewrite the expected seat
share of party P =12 in the next legislature, given that it forms a single-party
government today, as:
E (sp ) = 2 + φ XX (V J - V ‘J ) .
(3.6)
I=1
As before, the expected seat share is a linear function of the weighted utilities of
all the voters, but now all economic groups receive the same weights.
It is easy to see that in equilibrium the expected seat share for the party in
government at the party formation stage entering the continuation payoff, N WSP
must be E(sP) = 2, since in equilibrium VJ = V‘J for all J. By symmetry, 2 is
also the expected equilibrium seat share of a single party in opposition. If the
opposition has two parties, they split the remaining expected equilibrium seats
equally; i.e., E(iiisp) = 1 in шWP
Equilibrium policy When the government is supported by a single party, con-
flicts of interest inside the government disappear, because voters can only punish
or reward the party in government. Whereas they could discriminate between
party 1 and party 2 in a coalition government, such discrimination is no longer
possible between the groups of legislators forming the large party. Even though
the voters of groups 1 and 2 still have conflicting interests, this conflict remains
latent in policy formation.13
Specifically, equilibrium policy now results from the maximization of (2.4) with
regard to all policy instruments, subject to (2.1), (2.2), and (3.6), taking the value
of expected future seats, RP as given. Repeating the steps in section 3.1.1, the
optimality conditions imply:
gS‘J =Hg-1(1),J=1,..,4 .
(3.7)
Comparing (3.7) and (3.3), we find that the overspending on groups represented in
government disappears. Intuitively, the type of government matters for spending
13If politicians were not opportunistic, but instead motivated directly by all aspects of policy
(e.g., as ”citizen candidates”), there would be a conflict between the legislators inside a single-
party government. But the differences between single-party and coalition governments induced
by voting behavior would still remain. With conflicting interests between legislators in merged
parties, it would also be central to model within-party bargaining (and its differences from
within-coalition bargaining).
16