One catch-all vote every few years is a very noisy metric for measuring people's agreement with one specific issue. Even more so when the choice is quite limited (e.g. two-horse races, between mostly similar horses, like many US votes between slightly-more-right-wing-authoritarian versus slightly-less-right-wing-authoritarian)).
