由于许多曲折，2024年的总统竞选可能会成为历史上最难分析的选举之一。一个值得注意的挑战将是衡量被大肆宣传的“会议反弹候选人在民意调查中的支持率有时会随着其政党备受关注的正式集会而上升。通常情况下，看看每次党代会后一两周的民调变化，就能让我们知道该事件是否给了该党一个提振。

　　但事实证明，这在2024年会变得更加棘手。想想看旋风这一次，这不是一个夸张的术语，在过去两个月的政治事件中，我们需要考虑我们如何解读公众舆论的任何变化:

　　天哪，伟大的火球——这让美国人接受了太多。不仅如此，这些事件中的许多在近期内相互重叠，这使得人们几乎不可能隔离出其中任何一个事件可能对公众舆论产生的影响。例如，特朗普在共和党大会后获得的任何提振可能也与大会前两天对他的暗杀企图或他在大会第一天选择万斯作为竞选伙伴所引起的轰动有关。然后，当然，拜登在7月21日退出，这意味着竞选的民意调查实际上重置了就像旧的超级任天堂游戏机一样就在共和党大会结束三天后——而且吹进弹药筒不一定会使屏幕上的图像更容易理解。

　　所有这些都意味着，尽管哈里斯和特朗普的情况明显不同，但在解读他们在大会后的立场时，他们都有复杂的故事。特朗普的得票率改进者1点多一点从大会开始到拜登退出的民调中。尽管如果拜登继续竞选，这一趋势可能会继续下去，但竞选活动的重新设定意味着我们只有三天时间来直接评估特朗普大会的民调影响。最终，特朗普在民意调查中的地位自那以来一直停滞不前，他在大会后的小幅上升现在成了历史的注脚。

　　对她来说，理论上，哈里斯并没有真正得到反弹:在8月19日民主党大会开始时，她的民调支持率略低于47%在538的全国平均水平在大会结束一周后，她的支持率刚刚超过47%，只有0.5个百分点的微小变化:

　　哈里斯的缺乏反弹符合最近的趋势，即提名人在大会后的改善越来越小。总体而言，自1968年以来，从大会开始到结束后一周，候选人的投票份额平均增加了约2个百分点。但是，虽然从1968年到1996年，16名候选人中有11名在大会后至少上升了2个百分点，但从那时起，只有14名候选人中有5名可以这么说。A更加两极分化的选民是罪魁祸首对于这种趋势，作为可供选择的选民越来越少现在比过去，即使在选民还在了解那些在国家舞台上时间不长的候选人的情况下。

　　然而，哈里斯在大会后没有反弹并不意味着她最近在民意调查中没有得到改善。7月24日，当538开始我们对哈里斯-特朗普对决的平均民调时，哈里斯的支持率约为45%，几乎与特朗普的44%持平。但到8月19日她的大会开始时，哈里斯的支持率已经提高了约2个百分点，达到约47%，而特朗普的平均支持率继续徘徊在44%左右——这也是他们两人在我们最近的民调中的平均支持率。

　　哈里斯的支持率上升可能是因为她的竞选团队经历了许多通常伴随全国大会的情况在几周内以前真正的聚会。毕竟，拜登的退出，哈里斯的候选人资格和她选择了明尼苏达州州长蒂姆·沃尔兹作为一个竞选伙伴提供了很多同样的覆盖面和增强党的基础这是一个现代的惯例旨在产生。民主党方面的所有这些发展可能还会削弱特朗普可能已经开始看到的任何蓬勃发展的反弹——创造一种类似于在共和党活动后立即而不是一个月后举行民主党大会的局面。考虑到人们对7月底和8月初竞选活动的狂热关注，哈里斯在全国代表大会前反弹也就不足为奇了。

　　这个故事的很大一部分是美国人对哈里斯的看法发生了相当大的变化，现在他们越来越熟悉她作为一个政治实体，而不是她作为拜登副总统的角色。7月21日，38%的美国人对哈里斯有好感在538的平均值中，而52%的人持反对意见。但截至东部时间8月30日下午2点，哈里斯的观点几乎平分秋色，46%的人持赞成观点，47%的人持反对观点。与她在民意调查中的地位类似，自民主党大会结束以来，哈里斯的平均支持率仅上升了一点点——约1个百分点，但自拜登退出竞选以来，她的支持率上升了8个百分点。

　　最后，这场颠倒的竞选在民主党全国代表大会后一天又经历了另一场冲击，这可能将一些注意力从哈里斯身上转移开，并可能影响民调:独立候选人小罗伯特·F·肯尼迪中止了他的竞选并在8月23日支持特朗普。这将如何影响当时表示计划支持肯尼迪的大约5%的选民的选择仍有待观察。

　　然而，538对民调的分析肯尼迪离任前的数据表明，他的离任可能没那么重要。我们发现，特朗普在从涉及肯尼迪的多名候选人竞争转向与哈里斯的直接竞争时，仅获得了0.2个百分点的优势——这个微小的数字大约相当于538民调平均值的平均绝对每日变化。不像1992年，独立的罗斯·佩罗最后一天撤退在7月份帮助比尔·克林顿(Bill Clinton)的民主党大会上，肯尼迪在退出时的民调支持率远低于他，这降低了他的离开对大会后民调的潜在影响。(而肯尼迪让门半开着对于今年11月投票给他的人来说，他极不可能像佩罗那样取得同样的成就经过正式重新参加竞选10月初和赢得19%的选票全国人民投票的结果!)

　　在一个经历了一系列无与伦比的竞选活动的夏天，民主党大会更典型地证明了它对民意调查的微不足道的影响。然而，这部分是因为哈里斯的加入已经如此戏剧性地改变了比赛。进入劳动节，这是竞选活动冲刺阶段的传统开始，哈里斯保持着对特朗普的微弱优势——尽管我们相信这场竞选活动仍将面临许多曲折。

　　Harris got a polling bounce before her convention, not after

　　The 2024 presidential race may go down in history as one of the most difficult elections to analyze thanks to its many twists and turns. One notable challenge will be measuring the much-ballyhooed "convention bounce," whereby candidates' support in the polls sometimes increases following their party's intensely-watched formal gathering. Normally, a look at how the polls have changed a week or two after each party convention could give us a sense of whether or not the event gave its party a boost.

　　But that's proving to be a lot trickier in 2024. Just think ofthe whirlwind— for once not a term of exaggeration — of political events in the past two months that we need to consider in how we read any changes in public opinion:

　　Goodness gracious, great balls of fire — that's been a lot for Americans to take in. Not only that, but the near-term overlap of many of these events has made it mostly impossible to isolate the effect that any one of them may have had on public opinion. For instance, any boost Trump enjoyed after the Republican convention may have also had to do with the attempt on his life two days before the convention or buzz generated by his selection of Vance as his running mate on the first day of the convention. Then, of course, Biden dropped out on July 21, meaning that polling of the race effectively resetlike on an old Super Nintendojust three days after the Republican convention ended — andblowing in the cartridgewon't necessarily make it easier to interpret the images on screen.

　　All this means that, while their situations clearly differ, Harris and Trump both have complicated stories when it comes to reading their post-convention position. Trump's share of the voteimproved bya little over 1 pointin the polls from the start of his convention through the day Biden dropped out. And while it's possible that trend may have continued had Biden stayed in the race, that campaign reset means we only had three days to directly gauge the polling impact of Trump's convention. Ultimately, Trump's standing in the polls has remained stagnant since then, and his small post-convention bump is now something of a historical footnote.

　　For her part, on paper, Harris didn't really get a bounce: At the start of the Democratic convention on Aug. 19, she was polling a tad shy of 47 percentin 538's national average, and a week after the convention ended, she's sitting just above 47 percent — a negligible change of around 0.5 points:

　　Harris's lack of a bounce fits in with the recent trend of smaller-to-insignificant improvement for nominees after their conventions. Overall, candidates since 1968 have averaged about a 2-point increase in their polling vote share from the start of their convention to a week after its conclusion. But while 11 of 16 candidates from 1968 to 1996 saw at least a 2-point post-convention bump, only 5 in 14 since then could say the same. Amore polarized electorateisa primary culpritfor this trend, asfewer voters are up for grabsnow than in the past, even in cases when voters are still getting to know candidates who haven't long been on the national stage.

　　Yet the absence of a post-convention bounce for Harris doesn't mean she hasn't enjoyed a recent improvement in the polls. When 538 kicked off our polling average for the Harris-Trump matchup on July 24, Harris stood at about 45 percent, almost even with Trump's 44 percent. But by the time her convention kicked off on Aug. 19, Harris's standing had improved by roughly 2 percentage points to about 47 percent, while Trump's average continued to hover around 44 percent — which is around where both of them sit in our most recent polling average as well.

　　Harris's polling growth likely happened because her campaign experienced many of the conditions that typically accompany a national conventionin the weeksbeforethe actual party gathering. After all, the hullabaloo over Biden's withdrawal, Harris's candidacy andher selection of Minnesota Gov. Tim Walzas a running mate provided much of thesame wall-to-wall coverageandenergizing of the party basethat a modern conventionaims to generate. All these developments on the Democratic side of the ticket might additionally have diminished any burgeoning bounce that Trump may have been starting to see — creating a situation akin to holding the Democratic convention immediately after the GOP event instead of a month later. Considering the rampant attention paid to the campaign fireworks of late July and early August, it's not surprising Harris got a bounce ahead of her convention.

　　A big part of that story is the sizable shift in how Americans perceive Harris now that they are becoming more familiar with her as a political entity separate from her role as Biden's vice president. On July 21, 38 percent of Americans held a favorable view of Harrisin 538's average, while 52 percent held an unfavorable view. But as of Aug. 30 at 2 p.m. Eastern, opinions of Harris were almost evenly split, with 46 percent holding a favorable view and 47 percent holding an unfavorable one. Similar to her standing in the polls, Harris's favorability average has only ticked up a tad — by about 1 percentage point — since the end of the Democratic convention, but she has seen an 8-point improvement in her favorability since Biden dropped out of the race.

　　Finally, this topsy-turvy race experienced another shock just one day after the Democratic convention that likely drew some attention away from Harris and could affect the polls: Independent candidate Robert F. Kennedy Jr.suspended his campaignand endorsed Trump on Aug. 23. How that will influence the choices of the approximately 5 percent of voters who said they planned to back Kennedy at the time remains to be seen.

　　However,538's analysis of pollsfrom before Kennedy's departure suggests that his exit probably won't matter that much. We found that Trump gained just 0.2 points in margin when shifting from a multi-candidate contest involving Kennedy to head-to-head race against Harris — a tiny amount that's about equivalent to the average absolute daily change in 538's polling average. Unlike in 1992, when independent Ross Perot'swithdrawal on the final dayof the Democrats' convention that July helped boost Bill Clinton, Kennedy had far less support in the polls at the time of his exit, reducing his departure's potential impact on post-convention polls. (And while Kennedyleft the door ajarfor people to vote for him this November, he is extremely unlikely to pull off the same feat as Perotbyformally reentering the racein early October andwinning 19 percentof the national popular vote!)

　　In a summer that's seen an unparalleled series of campaign events, the Democratic convention proved more typical in its negligible impact on the polls. However, that's in part because Harris's entry had already altered the race so dramatically beforehand. Heading into Labor Day, the traditional start of the stretch run of the campaign, Harris maintains a slight edge over Trump — though we're sure this campaign will see many twists and turns still ahead.