Once again two counties in Florida held up definitive results for weeks after the votes were supposed to have been counted. And, once again, all the legal wrangling, recounting, and boxes full of uncounted ballots mysteriously discovered did not change what we knew about the big races on election night. Perhaps even more notable than the election results is the fact that a modicum of accountability may have finally materialized in Broward County as the Elections Chief was forced to resign after what the New York times (generously) called her "series of failures."
When the dust settled, the GOP had picked up a net two Senate seats to maintain control of the chamber. The DEMs flipped the House with a ~37 seat gain. The net change in Governorships stands at six for the DEMs. Although NPR has declared the elections a "Big, Blue Wave" the logic behind the claim seems a bit tortured, as it rests on a comparison to other Dem turnouts against a GOP President. 37 seats is a shift to be sure. But, in contrast to 63 seats that Obama lost in 2010 and the 52 seats Clinton lost in 1994, the first midterm for this President doesn't seem worthy of the "wave" title. ElectoralMap's analysis for a "wave" election would have seen the Dems pick up a few Senate Seats and nearly fifty House seats, so it's hard to see breaking out the surfboards on this one.
Nov 5, 2018
The final forecast for the mid-term races is posted. The short version is this: the GOP will hold the Senate, picking up one seat, but lose the House as the Dems gain roughly 30 seats. The Dems will gain between eight and eleven Governor's mansions. It could take a while for the dust to settle after this election, especially if the polling is accurate with 30 seats in the house coming down to a photo-finish. I probably won't post results and analysis until control of the House is settled. Until then, go vote if you haven't already!
Oct 31, 2018
Today's forecast is the last for the month of October, as we draw down to the final week before the election. The polls are about as useless today as they were at the start of the season, and five states still show a dead heat in the Senate races. So, the question is, are these five states really this close? Will the spread in the actual results really be within the margin of error in these polls? I'm skeptical, but I suppose it's possible that these races are really this tight. The bad news for the Dems is that they can run the table on the dead heats right now and still not take the Senate. They would need that, plus one other pickup in a tossup state. If the blue wave is really a crimson tide, they could be facing a GOP senate 57 seats strong.
Oct 26, 2018
Forecasts tighten slightly
The Senate forecast is down to one dead heat in Missouri, with five tossups leaning toward their incumbent and two others showing a flip. If the election were held today, the Democrats would have to win all the tossup seats to gain 51 seat control of the chamber. Missing just one would mean an evenly divided Senate, with Mike Pence casting tiebreaking votes. The tightening in the polls also took the Republicans' best case scenario from +7 seats down to +5.
In the Governor races, Democrats made gains in Florida, Iowa, and Wisconsin which are now predicted as flips -- although just barely. In the US House of Representatives the polling was virtually unchanged, shifting by just one seat in favor of the Democrats. To gain control of that chamber, the Dems would need to win roughly 49% of the tossups. That should be more than achievable if the "blue wave" has any basis in reality, and it's statistically likely if they are truly tossup seats. Nate Silver currently gives the Dems an 84.5% chance of taking the
Oct 9, 2018
Forecasts barely budge
The Kavanaugh confirmation is a done deal, and both sides continue to lay claim to an energized base --- and this has yet to manifest in any change in the forecast. The Senate moved slightly redder and the House moved slightly bluer, by one tossup each. This is indistinguishable from noise in the polling data for all practical purposes. The Nate Silver prediction continues to run noticably bluer than the RCP poll average.
Oct 1, 2018
All Forecasts Updated
The first major polls "post Kavanaugh hearings" have posted and they are reflecting next to no change to the forecast. There was a slight bump toward the Republicans in the Senate, and a slight bump toward the Democrats in the House. The governor forecast remains unchanged. Both parties are putting forth the claim that the embattled SCOTUS nomination is energizing their base, which may or may not be true. But the polls seem to be implying that the "energy" in question is being found in people who were already energized.
Sep 26, 2018
House Forecast Posted
I've aggregated roughly ten polls for the competitive districts in this year's House races. The polls chosen probably have a slight Dem-bias built in, as most of them are showing a "bluer" forecast than indicated by the RCP aggregation. Surprisingly, however, these polls are redder than Nate Silver's 538 forecast, which seems to be heavily skewed Democrat this year.
The Democrats need 23 seat flips to take control of the House of Representatives, which appears fairly likely at this point. If the polling is even remotely accurate right now the Dems can expect to pick up up at least 5 seats, and as many as forty if the "blue wave" is a reality. We can apply some symmetry to 2014, which was a red-wave year in which the GOP picked up 63% of the seats in the "take-all-tossups" scenario. If we assume that the blue wave will more or less materialize, then awarding them 63% of their best-case scenario would give them 29 seats. That's a much closer margin for seizing the house.
Sep 24, 2018
Senate & Governor forecasts largely unchanged
In a tumultuous week of politics surrounding the Kavanaugh nomination, the forecasts for the Senate and the Governors' races has remained largely unchanged. It seems that which party people believe to be telling the truth depends largely on their political affiliation -- which I guess shouldn't be that surprising for a he-said-she-said accusation brought at the 11th hour. There was a slight nudge in favor of the Republicans in the polling data, but it probably can be attributed only to noise.
Stay tuned, the house forecast is in the works and will be posted soon.
Sep 17, 2018
Governor's forecast posted
The Senate map was updated today with the latest polling data, and I've added a map for this year's gubernatorial races. The current polling suggests a strong year for the Democrats, with flips in New Mexico, Illinois, Michigan, and Maine. In Alaska, the Repblicans will likely wrest control from liberal Independent Bill Walker.
In the 2014 season, the polls were heavily skewed in the Democrats' favor resulting in a forecast of a one seat pickup for the Dems -- when in reality the Republicans picked up two Governorships. It's difficult to say if that skew will reappear this year, as the parties in power have traded places. Usually the party out of power has a stronger showing, which was the case in 2014.
Sep 12, 2018
First forecast of the season posted
The first Senate map is up, and it's rather....boring. The aggregation right now shows a seat gain in Nevada for the Democrats, and one in North Dakota for the Republicans, with four seats sitting at a coin-toss. This is pretty reflective of the most likely outcome in the Senate this year, which is a net change in power too small to flip the chamber. This season I've chosen roughly ten polls to aggregate, with a net bias (in my subjective judgement) in favor of the Democrats. This doesn't reflect my own personal bias, but rather the fact that I believe it will be a high turnout year for an energized base that absolutely despises President Trump.
You can use the radio buttons above the map to give tossup races to either party. In the 2014 mid-term races, the Tossups go the the Republicans scenario turned out to be a dead-on accurate forecast of what actually happened on election day. There are two ways to attempt to translate that fact to this year's race: One, the parties in/out of power are reversed - so maybe this year the Dems will claim the tossups. Two, the polls being used are very similar, so perhaps the tossups should go to the GOP because of the inherent bias towards the Dems. You decide!
Check back for updates, as I will likely be adding some data about house and governor races.