1
Does the allocation of “pork” projects to constituents affect legislative position-
taking? Are legislators who deliver substantial amounts of federal largesse more likely to
diverge from their campaign opponents and from their constituency medians?
Empirical research on the incumbency advantage has suggested the opposite. Legislators
engaging in substantial district attention are often the most marginal legislators (Erikson
1971), and thus are likely to be very close to their districts’ preferences. Other scholars
of congressional representation, in contrast, posit that as legislators build trust with
constituents (via more district service and project allocations) the legislators can deviate
directly from the wishes of their constituents in a monotonic fashion.
Theories by Ansolabehere and Stewart (2000) and Groseclose (2001), among
others, suggest that legislative candidates with high valence advantages (see Stokes 1963)
are likely to lead to election campaigns where candidates do not necessarily converge to
the median voter in their constituency, though this relationship is not monotonic. In some
instances, where a candidate has a slight valence advantage, the advantaged candidate is
likely to be close to the median of the district relative to a candidate with no valence
advantage. However, as the valence advantage grows, the candidate is likely to diverge
from the district median. Examples of candidates’ possible valence advantages are
numerous, but can generally be conceived as characteristics that clearly give the
legislator or candidate a clear advantage in their elections (e.g., charisma, “pork” project
allocations, constituency service, and name recognition).
In this paper, we conceive of an incumbents’ valence advantage as larger “pork”
project allocations in Congress. Interestingly, little empirical work has examined the
effect of project allocations on position-taking by legislators. However, the competing
Convention