Traditional models of accountability suggest that a good performance increases support. This relationship is a cornerstone in democratic societies because otherwise political elites become untethered from the public they are supposed to serve. Yet as much as this model of accountability is normatively desirable, research from the last decades show that if people like their leaders, then they tend to approve of all sorts of performances. That is, the relationship runs opposite of the traditional model. Such findings are often attributed to the string of tendencies that underpin directional motivated reasoning. The purpose of this paper is to examine the relationship between institutional trust and performance evaluations of the institutions during the Covid-19 pandemic. Based on a multi-wave panel data with over 10,000 Swedish respondents, we show that there is a reciprocal relationship between evaluations of how the institutions handled the pandemic and institutional trust. In other words, both accountability and motivated reasoning occurred during the pandemic. Among the two, the motivated reasoning effect appears stronger. We primarily rely on dynamic panel models, but the results are robust to different estimation techniques such as 1-way or 2-way fixed effects models.