(α - αi) / (1 - αi) is the probability that an outsider will gain access to a public
sector job, so the expected utility of an outsider is
Uo
Г ' ʌ Ug +
1- αi
1-
α a - ai y
\ 1 - ai / _
Up
a a. — ai ∖ z —ʌ a a — ai \
⅛ ) u(W+rk>+ 1 - (r—⅛ )
u(w + rk) + g (G) (2.21)
where we recall that Up is the utility of a private sector employee and that those
outsiders who do not get a public sector job (the number of which is 1 — a) all
end up finding employment in the private sector, due to flexible adjustment of the
private sector wage rate w.
When choosing a fiscal policy package (W, G, τ), politicians face the technolo-
gical and market constraints (2.4), (2.13) and (2.14) plus the government budget
constraint which requires that the revenue from capital taxation must cover the
cost of the wages to public sector employees:
τ (1— a) k(r+τ)=aW. (2.22)
Moreover, in order to be able to attract workers to the public sector, these workers
must be offered a utility level at least as high as that enjoyed by workers in the
private sector. This recruitment constraint in turn requires that
W ≥ w. (2.23)
Our assumption that public sector insiders have full job security also implies that
fiscal policy must satisfy the ‘non-firing constraint’ a ≥ ai . In the analysis below
we assume that this constraint is never strictly binding.11
Our parsimonious model obviously relies on strong simplifications. First, in a
more elaborate political economy framework politicians might try to dole out the
marginal high-paying public sector jobs in return for political support. Second,
the model feature that campaign contributions are zero in equilibrium derives
from an implicit assumption that all voters are equally well informed. As shown
by Baron (1994), when voters have different information sets it may be optimal
for lobbies to offer positive campaign contributions in equilibrium to influence
11If it were binding, we would have a relatively uninteresting scenario with an exogenous
allocation of labour between the public and the private sector.
15