This article introduces the State Executive Approval Database, a dataset of gubernatorial approval ratings that updates and adds to data previously collected by Beyle et al. In addition to the survey marginals, the dataset presents continuous quarterly and annual measures of the latent level of governor approval that are amenable for time series analysis. After evaluating how survey data availability varies across states and over time, I use the data to evaluate whether governors receive a honeymoon. While new governors do not have higher than expected levels of approval, the public expresses comparatively low levels of disapproval for new governors. This honeymoon is largely restricted to their first quarter in office and only occurs when they are elected to their first term. Governors who take office after their predecessor resigned get a slightly longer and more sustained reprieve from disapproval. Governor approval is also significantly shaped by unemployment levels in their state. These data will provide scholars with new opportunities to study accountability and representation at the state level.

In this paper, we analyze how citizens evaluate their president, especially focusing on why voters lower their evaluations at an individual-level perspective. We assert that citizens raise their evaluations of a new president when their expectations are met and lower their opinions when his or her performance disappoints them. Furthermore, the evaluations of the president are not only affected by a government's economic and diplomatic performance, but are also influenced by individual awareness of salient political issues, the contents of the news, and their own perceived influence on politics. We test the theoretical hypotheses using Taiwan's Election and Democratization Study (TEDS) panel survey data. The statistical results support our theoretical hypotheses and show that when President Ma's performance did not meet respondents’ expectations of unification and economic prosperity, the respondents updated their evaluations of Ma. In addition, it also shows that the extent of respondents’ media exposure and political efficacy are significantly associated with the change in their opinions of Ma.