Electoral and partisan cycles between US economic performance and presidential popularity: a comment on Stephen E. Haynes

In this paper we discuss a recent paper by Stephen E. Haynes in which he relates electoral cycles in political support to electoral cycles in economic variables. Haynes finds that the cycle in support for Republican presidents is explained by the cycle in economic variables, whereas the cycle in support for Democratic presidents is not. In our opinion this shortcoming is due to his specification of the popularity function. Haynes estimates a popularity function which incorporates the notion that voters reward the incumbent for favourable outcomes (score hypothesis). Our popularity function combines the score hypothesis and the notion that voters cast their ballots for the party that best fits the current economic situation (issue hypothesis). We show that the electoral cycle in popularity of both Republican and Democratic presidents is explained very well by the cycle in economic variables.