I (i.e., David Rothschild) live in constant regret for spending way too much time in 2016 worrying about who was going to win, and way too little about what information that I could produce that would actually be useful for a well functioning democracy. That is why for 2018, I (along with my partner Tobias Konitzer) shifted the focus of PredictWise completely: less horse-race and more policy. We spent 2018 studying what the public wanted, informing both the public and political leaders on public opinion. And, how to reach the public more effectively, to educate and information them. Here is a quick overview of some of that work. That being said, now it is Election Day, so what am I watching (post published at 12 PM ET on Election Day):


1) Polling error is just a reflection of what the polls missed about voters. In this election we are tracking voter turnout among the young and minorities. And the preferences of white women. A high younger vote, high minority vote (both of which we have anticipated), and a sharp move for white women to Democrats, would mean the difference between a blue wave and a narrow Democratic loss.

2) Democrats are 75% to win house, but higher probabilities of big win or small loss, than middling. There are 33 toss-up seats, but I expect 25+ to go one or the other party, as they are highly correlated. In 2016 Democrats won 194 seats and Republicans 241 seats. 218 seats control the house. Going into 2018 the Democrats are very solid in 181, solid in 9, and good in 18 = 208. Republicans are very solid in 129, solid in 17, and good in 48 = 194. Which means the situation is very tilted in the Democrats favor, but not a slam-dunk.

3) Thus, due to the shifting map, every meaningful race is in a seat currently controlled by the Republicans!

4) Democrats will dominate the two-party vote share, without question. Only reason this is competitive is because of gerrymandering (both natural and malicious).

5) Key Races: here are some of the key races I am following today. In green are the 33 toss-ups, plus a handful of lean Republican and lean Democratic. Again, I predict 25+ of these races go in one direction or the other.



1) House has every seat up every two years, but with six year terms, only 1/3 of senate up each time. This year’s landscape was insanely hard for the Democrats. Technically the Democrats have 23 seats and Republicans 42 seats heading into the election (with 35 seats contested: 33 regular elections and 2 special elections. Obviously there are some easy elections for both sides: CA, CT, DE, HI, MA, MD, NM, NY, RI, VA,VT, WA for the Democrats were never expected to be close, and MS, NE, UT, WY for the Republicans.

2) Democrats have already locked up a lot of races that could have been tough: ME, MI, MN, MN-S, OH, PA, WI. I have these seven elections as basically over before they start, despite Trump winning in MI, OH, PA, WI, and getting close in MN and ME. Including this the Democrats have 42 to the Republicans 46. That leaves 12 really meaningful contests.

3) Four states are leaning hard Republican: Mississippi’s special and North Dakota are almost out of reach for the Democrats. North Dakota is a sitting Democratic senator. Pickups in Texas and Tennessee will be tough, but possible. Four states gets the Republicans to 50. Missouri, Indiana are Democratic holds that are toss-ups, with Arizona (open) and Nevada (Republican incumbent) as possible pick-ups. Finally Florida, Montana, West Virginia, and New Jersey are all likely holds, but certainly vulnerable.

4) Most likely outcome is 51 or 52 seats for the Republicans.

5) Democrats will be favored to pick-up 2-3 seats in 2020, so every seat is precious.


1) Democrats will pick up a lot of state mansions this year, only question is how many. This includes safe pick-ups in Illinois and New Mexico. But, includes continued loss in Democratic states as well: Massachusetts, Maryland, New Hampshire, and Vermont (yes, Vermont).

2) So many interesting races: Georgia is where Republican Secretary of State Kemp is running on a mix of racism, voter suppression, and deception, but is still 70 percent to beat Abrams. Kobach is well known for both his xenophobia and voter suppression and he is a tight race in Kansas. This follows Kansas’ disastrous run with hard-core Republican economics. Ohio is leaning Democratic for Cordray, who ran the Consumer Finance Protection Bureau that the Republicans basically shut down 2017. Walker is on the ropes in Wisconsin after his massive corporate welfare scam blew up very quickly. Gillum is facing another racist in Florida and looking good. Lamont is making a comeback in Connecticut after losing the 2006 senate election to Lieberman (who then went on to squash the option to buy Medicare, undermining Obamacare and killing 10,000’s of Americans). Brown is a liberal governor in Oregon in a tighter fight than expected. After a disastrous run, including the Flint water disaster, Michigan looks like it will turn its state mansion over to a Democrat, Whitmer. Finally, Maine has been run by Trump’s double for last few years, and looks like it will break that.


Voting Rights Restoration for Felons Initiative would restore voting rights to 1.5 million ex-felons in Florida. Needs to get 60% plus and it is currently 70 percent likely to pass.