PS 686 Reading Summaries, 10.6.09
Paul Sniderman, James Glaser, and Robert Griffin.
"Information and Electoral Choice," in
Information and Democratic Processes
, ed. Ferejohn and Kulkinski.
Central question: do well-informed voters use a different decision process to make
political choices than low-information voters?
If so, can the process used by low-
information voters approximate that of the high-information voters?
Answer: yes, the two groups do use different processes.
There exist two principal models of how voters make decisions about for whom to
The first is a retrospective approach, which focuses on evaluating the
incumbent and then asking “is the way things have been going is good enough?”
the answer to that question is yes, then the voter will choose the incumbent; if not, he
will choose the challenger.
The other model involves comparing competing
candidates based on their policy preferences in comparison to each other and to those
of the voter, and placing those preferences in a broader ideological and/or partisan
Well-informed voters should tend to use the latter, while low-information voters
should use the former.
This approach builds on a model developed by Shanks and Miller in which a
sequence of variables, arranged in temporal order, affect the voting process, from
most removed in time (fixed personal characteristics, like gender and race) to least
removed in time, which has both strengths and weaknesses:
It provides an effective way to compare the policy calculus approach to the
performance evaluation model
It does not include issue proximities, or how far voters estimate candidates are
from their own positions (Sniderman, et. al. see this as a useful simplification,
since most voters don’t have stable preferences)
It assumes that ideological predispositions precede party identification
temporally, which may not be true
It assumes that the variable sequence is the same for all voters, which gets back to
Sniderman, et. al.’s original critique of the literature on decision making