The theory of affective intelligence posits that an individual's emotions help govern a reliance on political habits or, alternatively, deliberation and attention to new political information. Some of the evidence adduced draws on the fact that voters who are anxious about their own party's candidate do not rely blindly on their partisanship but instead consider policy and personality when they vote. In a provocative paper, Ladd and Lenz (2008) argue that emotions reflect an evaluative judgment, akin to likes and dislikes, that has little to say about attention and habit. Here we examine the ANES data from 1980 to 2004 and find that the affective intelligence theory's original findings remain statistically robust. On closer examination, we also learn that Ladd and Lenz reformulated the theoretical test by using a different operationalization of affect and a different dependent variable and found results at variance from ours. We find it an inappropriate test. In the end, we agree with Ladd and Lenz that cross-sectional data cannot crisply test the short-term impact of emotions on attention and habit and concur that ultimately experiments will move the debate forward. We further observe that Brader's (2005, 2006) powerful field experiments explicitly test the special effect of emotions on attention and judgment and support the affective intelligence model.