7
Recent work strongly suggests that this institutional factor may be of importance to
economic voting. One important study by Norpoth (2002) looks at aggregate presidential
election outcomes between 1872 and 2000, and shows that macroeconomic conditions exert
no significant effect on incumbent party support when the incumbent president does not
appear on the ballot. Other studies have also shown that retrospective economic and policy
voting is muted when the incumbent president is not running (Miller and Wattenberg 1985;
Nadeau and Lewis-Beck 2001).
2
These researches do not take into account the potential influence of political
sophistication, but there are reasons to expect term limits to be of consequence with regards to
the mediating effect of sophistication on retrospective economic voting. We may expect
economic voting in general to be less prominent in open-seat presidential races, but this
phenomenon should be particularly blatant among the more sophisticated segments of the
electorate. High sophisticates are better able to make appropriate causal attributions for
economic performance (Gomez and Wilson 2001). In effect, we should expect this group of
voters to be unlikely to use retrospective economic assessments when deciding which
presidential candidate to support in open-seat races since it may be easier for them to
understand that the incumbent party’s candidate has very little to do with the current state of
the economy or with their own personal economic condition. By contrast, low sophisticates
might still attribute responsibility for the incumbent party’s candidate, whether he is the sitting
president or not.
2
Some accounts of Al Gore’s 2000 election defeat only partially agree with this institutional interpretation,
preferring instead to blame Gore’s campaign for failing to convince voters to attribute causal responsibility to
him for good national economic conditions (see Rudolph and Grant 2002; Fiorina, Abrams and Pope 2003).
Convention