uninformed voters.12 Third, like other static probabilistic voting models, our
model neglects the potential time inconsistency problem in the political process,
implicitly assuming that reputation mechanisms keep politicians from defaulting
on their campaign promises.
In addition to ensuring analytical tractability, the above simplifications allow
a precise definition of the popular concept of ‘political distortion’. In the analysis
below we shall thus measure the degree of political distortion by the following
parameter (using the specifications from (2.20)):
pi - po
δ ≡ ai ∙ --------
po
= αi2h2ψ.
(2.24)
The political distortion is the product of the predetermined size of the public sector
lobby (αi ) and the relative increase in votes a political candidate may expect
to gain by catering to the economic interests of insiders rather than outsiders,
(pi - po ) /po .Themoreδ exceeds zero, the greater is the political influence of
public sector insiders relative to that of other voters. It is intuitive that the
political distortion is greater the larger the lobby and the greater the impact of
lobby efforts on voter preferences (the higher the value of h). We also see that a
smaller dispersion of ideological preferences (a higher value of ψ which reduces the
interval over which ideological preferences are distributed) increases the political
distortion. When ideological preferences are fairly similar across a large number
of lobby members, an increase in the economic benefits offered by one party to
lobby members will induce many of them to shift their vote in favour of that party,
and hence the lobby becomes more influential.
Note how our political setup tries to account for the views of those advocates
of tax competition who argue that the public sector tends to employ too many
people on overly generous conditions: First, because economic benefits offered to
public sector insiders generate more votes than benefits offered to outsiders (as
reflected in the fact that pi > po), our model includes an incentive for politicians
to offer rents to public sector workers. Second, when a political candidate offers
high public sector wages, he may also be inclined to promise more jobs in the
12 Lorz (1998) also offers a political economy model with positive lobbying activity in equilib-
rium. In his setting tax competition causes a welfare-increasing drop in lobby activity because it
reduces the ability of the government to redistribute income, thereby diminishing the expected
gain from lobbyism. However, unlike the present paper, Lorz (op.cit.) does not provide an
explicit description of the voting process.
16