We have reached the point where "too close to call" does not cut it.  Unfortunately, there are probably more "too close to call" Senate races than in any other recent cycle.  It is also the case that sometimes the House and Senate go in different directions even when there is a strong movement toward one party or the other.  In 2018, the Democrats picked up 40 House seats but lost Senate seats in Florida, Missouri, Indiana, and North Dakota, while picking up seats in Arizona and Nevada.  In 2020, Republicans gained more than a dozen House seats but lost a net 3 Senate seats (4 in total, losing 2 in Georgia, Colorado, and Arizona, and regaining Alabama).

At the moment, the Senate is divided 50-50, and the House has 220 Democrats and 212 Republicans with 3 open seats — 2 formerly Democrat-held seats in Florida and one formerly Republican-held seat in Indiana.  The GOP needs a net pickup of 5 House seats to gain majority control, assuming they win the open Indiana seat.

Pretty much every serious analyst of the House races believes that the GOP is favored to regain control.  Most of those who look at individual races — Cook Political Report, Larry Sabato, and 538 — are forecasting GOP gains of around 20 seats.  About half of those gains are expected to come in seats now regarded as tossups (many more seats in the tossup category are now held by Democrats than Republicans).

RealClearPolitics (RCP) believes that the GOP will win about 30 net seats, assuming that the two parties split those in its tossup category.  The fact that Democrats are playing defense and pouring in late money and surrogates in many places where they won comfortably in 2020 suggests GOP strength and the potential to have bigger gains if the party wins a large share of close races, as it did in 2020. 

The every-ten-year redistricting after the Census resulted in a small net gain for the GOP.  This gain was smaller than the net effect of court-ordered redistricting in Pennsylvania and North Carolina before 2020, which moved about a half dozen seats toward Democrats in the two states.  The most gerrymandered states at this point are Illinois, by and for the Democrats (14-3), and Ohio, by and for the Republicans (12-3).  In New York State, an attempt to create a 22-4 map for the Democrats was so egregious that a state court threw it out.  With the strong challenge by Lee Zeldin in the governor's race, Republicans could now win 8 or 9 House seats in the state.

My own likely range for GOP pickups is 20–25.  This may seem small, but combined with the gains in 2020, the GOP will have effectively regained close to 40 seats, about what they lost in 2018.  You don't gain 40–60 seats in one cycle when you are already almost at parity.  The generic ballot has moved toward Republicans by 3–4 points the last month or two.

For a period during the summer, Democrats led in the generic ballot following the Dobbs decision.  Abortion is still a wild card in the election.  I expect that in states where there is an abortion referendum on the ballot, Democrats will do better, as they did in Kansas.  If turnout is much larger for women than men, this could be a good sign for Democrats.

It is worth noting that Democrats won the national House vote by 3% in 2020.  A generic ballot lead of almost 3% today for Republicans today is a 6% shift from 2020, when Republicans won almost half the House.

While abortion may work for Democrats, inflation, high interest rates, declining stock market and home values, the flood of 5 million illegals over the southern border (you know, the border that is "secure"), and rapidly rising violent crime are all issues working for Republicans.  The crime issue in particular seems to be resonating in the suburbs, with a big shift since 2020 among white women, a group that moved away from Republicans when Trump was on the ballot.

Democrats have continued to focus on abortion, but also pivoted to the alleged threat to democracy created by MAGA Republicans.  This is an issue identified by voters only when prompted and useful mostly to stimulate turnout among committed base Democrats.

One factor that could help Republicans is polling errors, which were large in the two years when Trump ran for president (more muted in 2018).  Polls can influence turnout and fundraising, which can be decisive in close elections.  One of the major networks had a poll showing Biden winning Wisconsin by 17 points two weeks before Election Day in 2020.  Biden won the state by 0.6%.  Polling firms say they now get to conduct their surveys with between 1 in every 100 they contact and as low as 1 in every 250.

Given the sharp political division in the country, and the silo effect of people not trusting "the other side" or the "media," there is the possibility of some systemic undercount of GOP-supporters.  One possible indicator of this is that some firms say they are polling a decent number of people who claim to be Republicans and voted for Republicans, including Trump in the past, but plan to vote for Democrats up and down the ballot this year.  Seeing GOP defections when Trump was on the ballot was not that surprising, but it is possible that this year, this is an attempt to "game" the polls.

How many Liz Cheneys are really out there?  Some pollsters who had very good forecasting records in 2016 and 2020, such as Trafalgar, believe that the GOP will have bigger gains in both Houses than the conventional wisdom at this point from the networks and better-known forecasters.  I remain more cautious, though I think Republicans will gain control of both Houses of Congress.

As for the Senate, I think the trendline is towards the Republicans almost everywhere, but many races are still in doubt.  Florida, once a primary battleground in every statewide race, is now a solid GOP-leaning state.  Part of this has come from a shift among Hispanic voters, also seen in several other states with large Hispanic populations.  Both Marco Rubio and Governor Ron DeSantis should win re-election easily.

In Ohio and North Carolina, Republicans J.D. Vance and Ted Budd are likely to win their races, probably outside the margin of error.  Two other GOP-held seats are closer.  Ron Johnson has a lead in Wisconsin, but it is a modest 2–3 points over Mandela Barnes, and this is one race where a GOP lead has shrunk a bit the last week or two.  I think there is a general unhappiness among the electorate, which can damage all incumbents to some extent.  Democrat Mandela Barnes is not a great candidate for Wisconsin, suggesting that both parties can have candidate quality problems.  One thing in Johnson's favor is that he outperformed his final poll numbers both times he won against Russ Feingold.  I think Johnson will win, but close.

And then there is Pennsylvania.  John Fetterman's debate disaster should have been the end of his chances.  Then again, if the Democrats had any self-respect and decency, they would have replaced him as the nominee after his stroke.  Instead, his doctors and the media have covered up his disability and shamed the one honest journalist who reported on his cognitive issues.  Of course, Joe Biden was hidden in a basement for most of 2020, so his mental decline would not be as obvious.  Now, two years later, he is incapable of performing any role other than serving as the angry finger-pointing reader of divisive texts prepared for him denouncing MAGA Republicans..

GOP nominee Mehmet Oz has worked diligently around the state and is about even with Fetterman after trailing badly following his razor-thin primary win.  Oz's opponent for the Republican nomination, David McCormick, would likely be in better shape in this race were he the nominee.  Oz has struggled with some voters about his move from New Jersey, his Turkish citizenship, and his Muslim religion.  For Pennsylvania, he is a bit foreign.  Polls are quite dispersed on this race.  Some show Oz taking the lead by 2–3 points; others show the race tied; some have Fetterman ahead by 5–6 points.

This is a state where there was a lot of early voting, mostly by Democrats.  This was the case in 2020 as well, with Biden winning the mail voting and Trump winning big on Election Day, and Biden winning overall by just over 1%.  When Republicans win Pennsylvania, it is always very close.  Pat Toomey won twice by 2% or less.  Rick Santorum had a similar record in his two Senate victories.  Trump won in 2016 by 0.6%.  My gut tells me that Fetterman is a slight favorite to eke out a win because of his banked early votes from before the debate.  However, all the late visits by Democratic party leaders suggests they think it is slipping away.  This is one race I think is a total jump ball. 

At one point, Chuck Grassley in Iowa and Mike Lee in Utah seemed to be at some risk, but both have opened up much bigger leads this last week. 

Hence, only Pennsylvania is a tossup, meaning that the GOP base is 49 before any pickups, or Pennsylvania. 

There are six Democrat-held seats in play. 

I think Adam Laxalt is in good shape in Nevada against Senator Cortez Masto in a state where Democrats routinely win close statewide races.  Harry Reid is no longer around to run the get-out-the-vote operation.  John Ralston, an experienced political writer in the state, thinks Democrats underperformed in early voting.  If the GOP win Nevada, they are at 50 without Pennsylvania.

I think Herschel Walker will have the most votes in Georgia after Election Day.  It is not clear, though, if he can get to 50% and avoid a runoff.  David Perdue won by 2% in November 2020 but just missed the 50% level.  He then lost in the runoff.  Governor Kemp is likely to win re-election against "co-incumbent" and longtime "election denier" Stacey Abrams.  Kemp will likely not be on a runoff ballot, which could hurt Walker a bit.  The success of the Georgia football team and the ad for Walker made by his former coach Vince Dooley right before Dooley's death will help Walker.  I rate Walker a slight favorite over Senator Warnock, though I likely will get to call this race again in a December runoff.

If Republicans win Georgia and Nevada, they will be at 51 without Pennsylvania.

Four other Democrat-held states are possible pickups, though Colorado with Joe O'Dea is a long shot.  Arizona; New Hampshire; and, amazingly, Washington State are all possibilities.

Senator Mark Kelly of Arizona has led almost every poll, though the margins have dropped.  This is a state like Nevada, where Democrats have had recent success in close statewide races.  Blake Masters did very well in the one debate.  Republican Kari Lake, running for governor, could help Masters a bit.  On balance, Republicans win this race, and Pennsylvania, if this turns out to be a very strong Republican year.  If it is a more modest wave, Democrats will likely win both.

It is shocking that Don Bolduc could spring an upset in New Hampshire.  He was nominated only two months ago and trailed incumbent Maggie Hassan by 10%.  Now the race is very close, 1 or 2 points in most polls.  Bolduc is selling authenticity.  Hassan has outspent him well over 10 to 1.  I rate Hassan as a small favorite, but Bolduc has a shot.  New Hampshire has very limited early voting, so there is no banked lead for Hassan for Bolduc to make up on Election Day.  Governor Chris Sununu will win easily.  If his margin gets close to 15 points, he could pull Bolduc over the line.  This is one of the races where Democrats spent heavily in the GOP primary to destroy the candidate they did not want to face, and thereby get a weaker Republican nominee easier to beat.  It would be poetic justice for Bolduc to win.

Finally, we have the best single candidate of the cycle: Tiffany Smiley in Washington.  Running against five-term senator Patty Murray, best known for getting re-elected, Smiley has pulled to within a few points.  Smiley is a nurse married to a veteran blinded by a suicide bomber in Iraq.  She became a strong advocate for veterans and has focused her campaign on inflation and crime.

This is still uphill in the final days, but she is close enough that winning is a possibility.

My final guess is Republicans win 51 or 52 Senate seats, meaning they win 2–3 among Pennsylvania, Arizona, Nevada, Georgia, New Hampshire, and Washington.  Nevada and Georgia are the most likely, but the other 4 are close enough that a bigger Senate majority is possible.  It is possible that Senate control will again come down to a Georgia runoff.

Governors: For the competitive races:

• Michigan: Whitmer (D) wins 
• Nevada: Lombardo (R) wins
• Arizona: Lake (R) wins
• Wisconsin: Michels (R) wins close race
• Minnesota: Walz (D) wins
• Oregon: Drazan (R) wins
• Kansas: Kelly (D) wins
• Oklahoma: Stitt (R) wins
• Texas: Abbott (R) wins easily
• Florida: DeSantis (R) wins easily
• Georgia: Kemp (R) wins
• Pennsylvania: Shapiro (D) wins easily 
• Maine: Mills (D) wins
• New Hampshire: Sununu (R) wins easily
• Illinois: Pritzker (D) buys one more election with his inheritance
• Ohio: DeWine (R) wins easily
• New Mexico: Michelle Lujan Grisham (D) wins
• New York: If I base my pick on how panicked Democrats in the state are about this race, then Zeldin wins.  The fundamentals still seem daunting for him.  Trump won below 20% in New York City.  Zeldin says he needs to double that to win statewide.  However, if Dems are unenthusiastic about Hochul (is anyone enthusiastic about Hochul?), then turnout may be lower, and Zeldin can reach his targets more easily.  Nate Silver thinks Hochul has 97% chance of winning and projects a margin of 12%.  I think these numbers are excessive, though Hochul remains a favorite.

Photo credit: Rick Obst, CC BY 2.0 license.