2018 was a mid-term, but the Ds got their 4 year electorate to come out. They had a 10M vote advantage, which is about a 9% advantage, which explains most of their advantage for the year; 2018 was a D+8 year. Wanna see the math? Here it is:
2014 (midterm): 40M R, 35M D
2016 (4 year): 63M R, 61M D
2018 (midterm): 50M R, 60M D
The numbers are total House votes each of those years. Unlike Presidency/Senate, the House is the same; every 2 years, all seats are up for reelection. This makes it a good standard metric to judge changes by.
Democrats got their 4 year electorate to show up in a mid-term (2018), while the GOP was only half-way between their normal mid-term and their 4 year electorates.
10M / (50+60M) = 9.09%
The D advantage in 2018 was +8. Notice that ~9.1% was due to having their 4 year electorate vs a semi/enhanced mid-term for the GOP.
Without that advantage, it'd have been a R+0.9 year.
UNLIKE 2018, 2020 is a 4 year/Presidential election year, just like 2016. This means polls that were more accurate in 2016 will likely be more accurate in 2020. Polls that missed in 2016 but got 2018 right will likely over sample D, between 5-8 points.
Make sense?
Courtesy of Halo 3: https://www.youtube.com/watch?v=TQi4hDz0s9E
(Or search YouTube for "Then it is an even fight")
2018 was a mid-term, but the Ds got their 4 year electorate to come out. They had a 10M vote advantage, which is about a 9% advantage, which explains most of their advantage for the year; 2018 was a D+8 year. Wanna see the math? Here it is:
2014 (midterm): 40M R, 35M D
2016 (4 year): 63M R, 61M D
2018 (midterm): 50M R, 60M D
The numbers are total House votes each of those years. Unlike Presidency/Senate, the House is the same; every 2 years, all seats are up for reelection. This makes it a good standard metric to judge changes by.
Democrats got their 4 year electorate to show up in a mid-term (2018), while the GOP was only half-way between their normal mid-term and their 4 year electorates.
10M / (50+60M) = 9.09%
The D advantage in 2018 was +8. Notice that ~9.1% was due to having their 4 year electorate vs a semi/enhanced mid-term for the GOP.
Without that advantage, it'd have been a R+0.9 year.
UNLIKE 2018, 2020 is a 4 year/Presidential election year, just like 2016. This means polls that were more accurate in 2016 will likely be more accurate in 2020. Polls that missed in 2016 but got 2018 right will likely over sample D, between 5-8 points.
Make sense?
Courtesy of Halo 3: https://www.youtube.com/watch?v=TQi4hDz0s9E
(Or search YouTube for "Then it is an even fight")
And we’re R+1 now.
Gosh I hope so.
I go back and forth between being nervous about this one and thinking Trump's going to win in a blowout.
All the cards are stacked against it, and the left's hand is on the table. They're leaving nothing on the field. But neither is the Right.
Now we see who's stronger and hope to God the eternal fence-sitters get this right...
2018 -- Trump wasn't on the ballot and the voters were angry with Republicans for not working with Trump.
Low turnout = lost House.
Wanna know the great irony of that?
2018 the GOP did 50% better than their normal midterm. 50M vs the 40M they got in 2018. There WAS a large surge in GOP voters in 2018. No one remembers it because the Democrats had their 60M 4 year electorate.
People always want to say that Republican turnout in 2018 was low and the nation was turning on Trump and the nation is shifting left, etc...
...but the GOP had 10M more votes and 50% more turnout than a normal mid-term.
Not exactly a damning indictment, imo