A few days ago, the New York Times published an article online that examined the results of a new paper by political scientists B. Pablo Montagnes, Zachary Peskowitz and Joshua McCrain of Emory University. The paper looks at the rate of support for a president and how that may affect those who self identify as members of that same president’s party. The short of it is that the more popular a president may be, generally speaking, the more likely poll respondents are to self identify as members of the same party. The more unpopular a president may be, the more likely poll respondents are to self identify as members of the opposing party, a third party, or independent.
So, what does this mean for Trump? Well, for starters, Trump remains the least popular president since polling began after the Second World War. Trump entered the White House with a low approval rating and has continued to sink from there. There have been spikes, where his approval has risen, but it has never attained the same sort of numbers that new presidents typically get, and it has never cracked 50%, let alone a 50%+1, the smallest of majorities possible.
At the same time, however, Trump’s popularity among self identified Republicans has stayed steady, never dipping below 70%. Indeed, his support among self identified Republicans has largely remained in the 80 to 90% range. So what, you may ask. Doesn’t this just mean that as long as Republican voters continue to support him, the Republicans in Congress will never do anything to hold him accountable other than issue statements declaring their “concern” for his actions? Well, yeah, that’s true. But the other thing that can be gleaned from the data is that that base of support may not be as large as they think it is, even if it is extremely fervent.
As Montagnes et al point out, part of what can be surmised by the change in self reporting in party affiliation can be found in the way the questions are asked. In Gallup polls, for instance, they found that “the presidential approval question is the very first substantive question in the survey” and it is then followed by the question about party affiliation. Because the cat is out the bag on how they feel about the president, and because most people will identify with a perceived winner, many people will report as being in the other party if they have already made it known that they are not a fan of the president.
This is backed up by also pointing out how the ideology question, that is if a respondent self identifies as the having the same ideological leanings as the president, which comes much later in the survey, shows little to no change despite the president’s approval ratings, even if the party affiliation does. In other words, self identified conservatives are more likely to report that they are not Republicans if the Republican president is performing poorly, while they may self report as Republicans if the Republican president is doing well.
There is further evidence for their assertion in that, according to Gallup, self identification with the Democratic Party has gone up since the election, and while it has held around the same number since right after the election, self identification as a Republican has decreased not only since the election, but again since Trump took office. Charles Franklin, Prof Emeritus of UW-Madison has also noted this trend. It can also be found in other polls as well, though the degree to which self identification with the Republican Party has dropped varies somewhat based on which pool, and ultimately which random sample is utilized. Nonetheless, the trend holds true across all polling that respondents are less inclined to self identify as Republicans both since the election, and again since Trump took office.
This isn’t just a right wing phenomenon either. This can be found among those on the left and their self reporting as Democrats during a Democratic presidential term. It’s worth noting that this may help to explain why Clinton was performing well in polls among self identified Democrats even as her chances of winning were bottoming out. Simply put, Democrats were supporting her, as the number of self identified Democrats were shrinking. This gives a false view of the support. Clinton could hypothetically have had 100% support among Democratic voters, but if those voters were a smaller and smaller share of the electorate in swing states, it wouldn’t really matter.
The flip side to this though, is that it helps to explain why Trump’s approval ratings continue to dip generally, while his approval ratings among Republicans remains nearly unmoved. We aren’t adding new people to the electorate, after all. What we could be seeing through this data is that while those who continue to support Trump continue to do so with vehemence, the amount of people overall who are supporting him is dropping.
Republicans in Congress will no doubt continue to use Trump’s ratings among the base as justification for dragging their feet in holding him accountable. They may even use this high approval rating among the base as justification for their own policy positions and their reluctance to show up for town hall meetings and shirk their responsibilities and duties elsewhere. But, they should be aware that even though the base loves Trump, and they really, really do, that base appears to be shrinking, and with it so is the support for themselves.
Image credit Getty Images