The paper investigates the impact of political events on consumer conf
idence. In doing so, we distinguish between the announcement of electi
ons, the election results and government changes. We argue that such p
olitical events alter the expectations of consumers on future policies
. In a two-party political system, election results lead straightforwa
rdly to a new government so a government formation will not be of inte
rest to consumers. In a multiparty political system, however, the gove
rnment will usually be made up of several parties; i.e., be a coalitio
n government. The election results do not, therefore, resolve the poli
cy uncertainty. Hence, a government formation contains important infor
mation about future policies. We test the theory on Belgian consumer c
onfidence data. The results suggest that unexpected elections and new
governments affect consumer confidence; ideology is not important. In
the tests we control for the effects of economic variables. These resu
lts imply that political events by themselves can indeed affect econom
ic variables. Caution is therefore required when interpreting tests of
political business cycles.