My feeling is that an increase in Trumps approval will correspond to better congressional poll results for Republicans.
Here's my theory. Trump has been cast as a Radical Republican. I think people see congressional lawmakers as more centrist and appropriate. If more people approve of Trump that indicates to me that people are not just leaning Republican, but becoming base supporters of Trump. As a consequence, Republicans will have a larger base to work with during the mid-term elections.
I think the key number is 47 for Trump's approval ratings. That would represent a 6 points swing and save the Republicans from any meaningful losses. Remember that Trump won the election with 46% of the vote and congressional Republicans dominated. If he starts to exceed 48, the Democrats are in big s##t trouble. Over 50% and it becomes a disaster.
You didn't see this number correspond with Obama's approval rating, because Democrats don't show up in big numbers during mid-term elections and Democrats were not viewed so favorably in many congressional districts. Obama seemed reasonable and centrist to the blue collar voter, but Democrats did not. We have the opposite situation working today in which people view Republicans more favorably than Trump.
This is a myth. In truth, turnout is usually in flux and depends on the electorate base of the parties. With Republicans bleeding educated voters, who turn out in good numbers historically, the opposite is easily possible in a midterm with a president like Trump. In fact, most elections we've seen so far since 2016 have seen massive swings against Republicans in what are supposed to be extremely low turnout elections.
We'll see what happens in 2018, but you're getting ahead of yourself here.
Oh I admit it. I'm totally getting ahead of myself. There are so many variables at play with Trump and Anti-Trump politics. I'm just theorizing based on the data that pops out to me. Now I'm gonna go super nerdy with my theory by pointing to data I think will be important and selectively pick data that support my theory.
Let's analyze Generalissimo's 1994 poll results. Democrats were tied with Republicans right before the election. In Mid-August Clinton's approval rating was 39%, but bumped up to 46% before the mid-term elections (aka the Republican Revolution). Clinton's numbers averaged out to 42.5%. The Democrats only received 44.8% of the votes. In 2010, Obama's approval rating was around 44-46% before the election, and Democrats received 44.9% of congressional votes. In both these elections, the Republican Party received 52% of the vote, which is lower but pretty close to the Presidential disapproval level.
In 2006, Bush's approval rating was between 37% and 44% before the election. That averages out to 40.5%. Republicans received 44% of the vote, while Democrats received 52%, despite a disapproval of 51-58%.
In the last election, Trump garnered 46% of the popular vote and Republicans won elections by good margins. This indicates to me that Trump's poll numbers and approval do not correlate Republican voter share at the congressional level. Now I don't think its fair to compare Presidential elections years with mid-term elections, but I think Trump's approval is likely a baseline for Republicans.
In sum, I think Presidential Approval polls are nice indication of the incumbent party's base support in a midterm election. In terms of voter share, Democrats and Republicans seem capped at 52-53%, despite high approval rating, but Republicans receive a greater benefit from this cap because of districting. Generally, I would add 2-2.5% more points to Trump's average approval rating a couple of months before the election. Let's call it the Trump Rules. I would also add an additional .75% to it for every additional point Trump falls below 43% to account for political climatology.
If I had to put my chips all-in blind, I would call it 51-53% to 46-47.5% in favor of the Democrats in terms of voter share.
In November, I will look back at this post and conclude that I must have been hammered at 4:55PM on a Thursday.
