It Only Takes One
Data is fun. I wanted to thank everyone who read my first article on election data for the 2020 cycle. I had a lot more traffic than I expected. If you enjoyed it, please spread it around! But the data from my first post reminded me of something: people often like precise articles with digestible information, especially when it’s focused on data. Moving forward, I hope to post more with articles that are not “10 minute reads”, according the the Medium algorithm.
On to this week’s update!
As always, we will cover any changes to the map based on events and recent polling. Right now, I am only including polling agencies with top notch performance (2% MoE) from actual results over the last 4 cycles.
Colorado (9 EV): From Likely Biden to Safe Biden
Since the consolidation of the primary field, Biden has seen his lead growth to margin like we see in New Jersey. Colorado has one of the highest levels of college and postgraduate educated voters (58%). This state on a national level looks out of range.
Nevada (6 EV): From Likely Biden to Lean Biden
The first poll since Biden became the presumptive nominee, and while was sponsored by a slightly right leaning organization, Biden has a small enough lead (and only the one recent poll) to proceed with caution here.
Montana (3 EV): From Safe Trump to Likely Trump
Just for funsies, I’m giving Trump his first “Likely” rating this season. He had been on both extremes, somewhat fitting for where his comfort and strength lie with the electorate. Montana is interesting right now, in February he led all candidates by 20+ points, but that lead is down to 6 in May. There is a surprisingly close U.S. Senate race with the current Democratic governor (with a 66% approval rating, as a Democrat, in Montana) making a huge push against an incumbent. The rare potential for a down ballot seat to help the top of the ticket. But I’m not holding my breath, as Montana is a 61% Republican state for Presidential races.
Iowa (6 EV): From Safe Trump to Likely Trump
There is nothing about Iowa that makes it a swing state anymore. However, with the first state poll in 3 months and 10+ point swing within the MoE for Trump and I’ll move this reluctantly.
Other Polling Notes: Two straight Texas polls put Biden +1 and TIE. It’ll take a lot more for me to move Texas from its current perch in Lean — especially since I won’t us a Toss-Up category.
Lastly, Georgia continues to trend into complete unknown. Of the 7 polls conducted, Trump has a aggregate lead of +4, so roughly 0.5%, with the most recently polls Trump +1 (May) and Trump +2 (April). I still can’t see Georgia voting blue.
My plan was to finish the other 6 states that will decide the election, but I wanted to keep it short. My plan this week was to cover Georgia and I had been compiling information. Needless to say, this is a good week to cover this state.
It is still very early for any news story to define a national election. But a local story can certainly start defining the state campaign. We essentially elect a President based on 50 local elections anyway, so these things matter and the Ahmaud Arbery murder has the potential to upend Georgia politics.
Combined with Stacey Abrams receiving more total votes than any other Democrat in the history of the state in 2018 has to have Republicans in that state on edge.
In 2000, Bush wins by 13%, over performing the national spread by 14. In 2004, Bush wins Georgia by almost 17%, over performing his national win by 15. Then McCain wins in 2008 by 5% and Romney in 2012 by 8%. With Obama’s convincing national vote win, this was still a 12 point over performance. In 2016, Trump wins by the same margin as McCain, but the over performance is only by 7 points. And then Abrams almost shocks the country by winning in Georgia.
Why?
It’s not just race. Though, the 10% dip in white vote from 70% to 60%, while the black vote has increased up to 30%. And, unlike in 2000 and 2004, when Bush was making headlines about reshaping the GOP and monopolizing American politics because he won a majority of the Latino vote around the country, they went in the opposite direction on that theory.
In 2004, Georgia was very rural with an urban/suburban/rural split of 20/44/36. That has massively shifted to now being 22/63/15. Democrats continue to make massive gains in the suburbs, but both Kemp and Trump hung on that demo, by slim margins. Trump performed better than Kemp among younger voters.
Abrams performed 12 points better (40–28) than Clinton did among white college graduates in the state. Locals often perform better, but this is a huge different, especially factoring what the data continues to show about female and minority voters among white and white male voters.
Georgia should come around and stick in the Trump column even though it will likely be the slimmest margin since Bill Clinton won the state in 1992. However, if Trump continues to slip with other demographics, even Georgia could get interesting.