Field experiments document near-zero marginal effects of most campaign advertising on vote choice in US general elections. Some interpret this finding as evidence of "partisan intoxication"---that contemporary American voters remain loyal to their parties even when confronted with new information. We present new evidence consistent with an informational interpretation of this finding: that voters are rarely persuaded by additional information about candidates they know a great deal about, but are more open to persuasion about candidates about whom they know less. The 2020 US Presidential election represents an opportunity to test these competing perspectives due to the presence of one candidate about whom most Americans are very familiar by virtue of his four years in office, Donald Trump, and another about whom Americans know less, Joe Biden. We conducted survey experiments (n=113,742) exposing each individual in a treatment group to two of 291 unique pro- or anti- Trump or Biden messages. Our results are consistent with an informational interpretation of many persuasive effects in campaigns and their absence. We show that vote choice in the 2020 US Presidential election changes in response to both pro- and anti-Biden messages, but that genuine effects of pro- and anti-Trump messages were between much smaller and non-existent. Further consistent with an informational interpretation, we show that vague messages about Biden are significantly less effective than those that offer specific information about him, and that evaluations of Biden are also significantly more malleable than evaluations of Trump. Positive information about Biden also causes Republican voters to cross party lines and say they would support him. These results would likely change if campaigns were to better inform voters about Biden, but raise a puzzle of why nearly all Democratic campaign advertising in the 2020 US Presidential election has focused on Trump instead of Biden.