Incumbents do better when voters perceive the economy as improving and worse when they do not.
The finding is stable in direction and unstable in magnitude. Perceived national conditions matter more than personal ones, recent months more than the whole term, and the effect shrinks where coalition government makes it unclear whom to blame. Perception rather than measurement is doing the work, which is why partisanship colours the input to the mechanism.
In practice
- Two economies with identical growth produce different election results, and the difference is which one voters believed was improving.
- A chief executive is judged on the last two quarters of a five-year tenure.
- Possible explanations
- One explanation is retrospective evaluation, in which voters use recent conditions as a cheap summary of competence. Another is attribution, since the effect depends on whether voters see the government as responsible.
- In its favour
- It is among the most studied regularities in political science, and the direction holds across countries, periods and methods.
- Against it
- Effect sizes vary widely with clarity of responsibility, media framing and which economic measure is used, and voters weight the months before the election far more heavily than the whole term, which is hard to square with competence assessment.
- Standing
- One of the most studied regularities in political science and among the most variable: effect sizes depend on clarity of responsibility, on whether voters judge national or personal conditions, and on which few months they weigh.
- Further reading
- Wikipedia
- Subject
- Law & Politics
- Essay
- What can be proved and measured about power