Theoretically, legislators with information about the preferences of their constituents
have incentives to take positions that are congruent with those of their constituents. However,
when legislators are uncertain about the positions favored by many of their constituents, they are
likely to present an ideologically inconsistent voting record. Based on this theoretical framework
of legislator uncertainty about their constituents, we posit that (1) legislators who face substantial
changes in the makeup of voters from one election to the next and (2) legislators who lack
knowledge about their constituents are more likely to vote in an inconsistent manner.
Empirically, we examine the effect of increased uncertainty on legislative voting
behavior. The dependent variable is a legislator’s voting record inconsistency. Fenno (1996,
325) noted that a senator’s consistency in policy voting can be critical in establishing strong
representation connections with the constituency. Specifically we measure the inconsistency in a
legislator’s voting record by the percentage of votes incorrectly predicted by a legislator’s
revealed preference based on roll calls (estimated by the NOMINATE algorithm).
This study departs from previous work that examines why legislators change their voting
records in one ideological direction or that examines the correlates of legislative roll-call voting.
Also, with the exception of Morton (1993) and Geer (1996), little work has been done that
examines the effect of incumbent uncertainty of constituency preferences on legislative voting.
Morton (1993) and Geer (1996), though, examine the effect of uncertainty on ideological
positioning and not inconsistency. Ideological inconsistency due to legislator uncertainty has
never been empirically examined—neither with experimental data nor with “real-world” data.
This paper contributes to the growing literature that theorizes about legislative
positioning. Various theories of candidate uncertainty about voter locations suggest divergence
or convergence under different conditions (e.g., see Alesina and Cukierman 1990; Berger,
2
Convention