The fact remains: Party ID moderates how voters respond to economic change

Recent research suggests that party identification biases voters' economic perceptions in general, but that all voters respond to economic change at the same rate. This implies that voters update their economic perceptions in parallel and are able to hold governments to account. But this has two problems. First, it contradicts evidence of partisan-motivated information processing. Second, parallel-updating does not imply unbiased information processing and is normatively less appealing than if partisans' economic perceptions were to converge at economic extremes. In this article, I argue instead that party identification does moderate how voters' economic perceptions respond to economic change. I test my argument on data from one ordinary and one extraordinary period in Britain's recent economic history using competing Bayesian multilevel ordered-logit models. I show that economic change does lead to changes in voters' economic perceptions. But I also show that party identification moderates this process. As such, voters update their economic perceptions along separate, not parallel, paths.