November Research Update

It's been a while since I updated this. I am still doing research on whether association/support for Donald Trump affects perceptions of conservatism. Some highlights:

The CCES data from 2020, as shown below, does indicate that endorsements from Trump- or endorsing him- leads to greater perceived conservatism, even when controlling for DW Nominate first dimension scores. However, candidate running for office don't have DW Nominate scores. The CAN endorse Trump, but their endorsements are less important than, say, a sitting incumbent who is known to the public. Thus, they leave us with less data unless we want to develop sophisticated measures of how much their social media accounts or their TV ads praise Trump. However, we can look at them a bit more qualitatively. Here are the top 10 most conservatively rated candidates in the CCES 2020 data. 

1. Lauren Witzke (DE)
2. Jason Lewis (MN)
3. Corky Messner (NH)
4. Cynthia Lummis (WY)
5. John James (MI)
6. Roger Marshall (KS)
7. Karen Handel (GA)
8. Daniel Gade (VA)
9. Mark Ronchetti (NM)
10. Lauren Boebert (CO)

The most interesting aspect of this last is that 2 candidates here- Lauren Witzke and Lauren Boebert- have ties to QAnon. In theory, Qanon has nothing to do with conservatism. It is a mass delusion that sounds even crazier than something on the cover of a supermarket tabloid from the 1990s. It is divorced from the role and size of government. But nonetheless, these two candidates are rated in the top 10. Jason Lewis, rated the second most conservative candidate, is a far right-wing radio host who has been referred to as "mini Trump" by a number of local MN and national news outlets. There ARE candidates here- like John James, Daniel Gade and Mark Ronchetti- who are not overly associated with Trump- at least not comparatively. But the fact that 3 exhibit highly Trumpian qualities is indeed interesting and worthy of further scrutiny. A conjoint experiment that lists conspiracy theories as one of many candidate traits might be a good route to explore in future analyses. 

Speaking of experiments, I ran an experiment on Mechanical Turk in which the only aspect of a candidate profile that varied was their support for- or opposition to- Donald Trump generally. Candidates who supported him were rated more than 1 full point for conservatively on a 7 point scale compared to a baseline candidate. This is a LARGE effect size., Candidates who repudiated him were seen as no different from a baseline candidate. More coming on that soon....