6 of 36
In summary, previous research has not presented a coherent story on whether
money influences votes, but suggested a sporadic relationship. It is thus more appropriate
to focus on those potential cases rather than to study the general pattern. Furthermore,
there is also a maturation of methodologies. Given the endogenous nature of campaign
contributions, which depend among other things, on the expectation of votes,
simultaneous equation models are preferred.
Research Design
Model
In this paper, the legislators’ voting decisions are conceived as a result from the
balancing of a number of factors. Politicians are subject to the influences from their
constituencies, party, interest group, and their own belief on the issue. Among these
factors, party, ideology, and the district’s interests tend to be fixed and would produce a
fixed position for their votes. At issue is whether the interest groups’ influences have
been translated into any deviation from this position. This model is consistent with a
number of previous academic works
10
(Chappell 1981, 1982; Welch 1982; Wright 1990;
Fleisher 1993; Davis 1993).
The model is also supported by Denzau and Munger (1986), who argued that a
vote represents a delicate balance of organized interests, voters, and legislator (p89):
“organized interest groups offer campaign contributions to improve their own wealth,
voters offer votes to obtain outcomes closer to their most preferred outcomes, and
legislators seek both campaign contributions and votes to obtain reelection.”
10
Some works either dropped the partisanship or ideology probably due to the multicolliearity problem,
and others did not incorporate the district’s interests. However, the logic is still the same.
Convention