Economic Voting: Enlightened Self-Interest and Economic Reference Groups
This research tests a new theoretical perspective on economic voting. There is a longstanding
debate on whether voters are:
`sociotropic' voters', i.e., basing their vote on the state of the
national economy; or `pocketbook' voters, i.e., basing their vote on the state of their own nances
(Kiewiet 1983, Kinder & Kiewiet 1979). We believe that this debate can be reduced to asking
what information voters use to form expectations about their own pocketbooks in the future. We
argue that voters use information about the economic fortunes of their own economic reference
group, rather than the national economy, to form expectations about the impact of government
on their own economic fortunes. This allows voters to evaluate both the economic competence
of incumbents, as well as the distributive tendencies of incumbents. Allowing voters to evaluate
distributional consequences of alternative parties in power is consistent with research showing that
left and right parties pursue dierent economic policies with dierent distributional consequences
(Hibbs 1977, Alesina, Roubini & Cohen 1997). Thus it allows for a theoretically richer model of
voter behavior; and allows us to synthesize the distinct literatures on sociotropic voting and political
business cycles. This work is motivated in part by the divergence of wages for dierent groups of
workers since the 1970s. As variance in economic performance increases across groups, we would
expect to see more reliance on economic reference groups and less on the national economy as an
indicator of the incumbent's likelihood of providing favorable voter-specic economic performance
in the future. We examine presidential approval over time across dierent demographic groups of
voters, and show that those approval ratings are in
uenced both by national economic performance
and by group economic performance measured by the change in the group's mean hourly wage.
April 14, 1999
Jonathan Nagler
Harvard University