2010 was a great year for Republicans. In November we won five Class 3 seats, plus won the vacancy left by Barack Obama to gain 6 on the Democrats. On top of that we won the special election to replace Ted Kennedy in January. The Senate thus went from a 60-40 supermajority, to a barely-functional 53-47 Democrat advantage.
When you win that many seats, it's hard to keep them all. Let's see if the GOP can defend its current 54-46 majority on this terrain.
Chart of the landscape
I've prepared here a chart of the 2016 Senate landscape. It's a big chart with a lot of numbers on it. Let me walk you through the data I've gathered for this.
First, I've looked at how these seats fared in 2004, the last Presidential year they were up, and listed the Republican lead (or deficit) in each race that year. Next, I did the same thing for the 2010 races, which allowed me to check what the "wave" differential was between 2004 and 2010. It turned out to be an average of 11.7 points, excluding anomalies*. Finally I looked at the 2012 Presidential in those states.
Weak incumbents in solid states
For most Senate seats these results all go for the same party, and those seats are generally safe. The incumbents potentially at risk despite that would be Richard Burr, Michael Bennet, John McCain, Patty Murray. I say potentially, but for Burr (NC, +12 in 2010), McCain (AZ, +24 in 2010), and Murray (WA, +4 in 2010), you'd need a really good candidate to beat them.
The only incumbent of those four who's likely under threat is Bennet of Colorado. He beat Pete Coors by 4 points, Ken Buck by 1 point, and Obama only won the state by six. Murray is the only other non-retiring incumbent to win by single digits, but Obama won her state by 15. A good candidate is needed against Murray, but the Colorado race is competitive regardless. So that's one vulnerable Democrat.
As for open seats, Florida (Marco Rubio retiring) and Nevada (Harry Reid retiring) are big fights coming, one held by each party. I don't see California, Indiana, Louisiana, or Maryland being big threats to switch, though bad candidates for either party can change that. I say the open seats are likely to net no partisan change.
The marginal GOP wave seats
And that's where the GOP's good news ends. The following Republicans won by the average 2010 wave or less: Mark Kirk (IL, +2), Pat Toomey (PA, +2), Ron Johnson (WI, +5).
On top of that, Kelly Ayotte did win by 23, after Judd Gregg won by 32 in 2004, but Obama won New Hampshire by 9. Ayotte could lose, as New Hampshire throws out incumbents all the time. Rob Portman won by 18 after George Voinovich won by 28, but we've lost the last two Presidential races in Ohio. Portman could lose, but at least he's an incumbent. That's five vulnerable GOP incumbents.
If we take every seat not listed above as a hold, Republicans start with a 46-43 advantage. Add in the seats I think are fairly safe (Burr, McCain, Murray), and the GOP starts up 48-44. That means of the 8 remaining seats, the GOP must win 2 for a tie, and 3 to retain the majority. Two of those are probably gone (IL, WI), so realistically the GOP needs to win three out of six.
It's early, and the results of the primaries (including the Presidential) can change everything in a moment. If Donald Trump wins all bets are off, and bad nominees or scandals can throw away solid seats for either party. But assuming no nationwide catastrophes for either party, and the close ones going evenly, I expect the GOP is set to retain a bare 51-49 advantage in the next Congress.
Get out there and vote though. The primaries matter.
* I excluded Alaska from this calculation due to the sour grapes run by Lisa Murkowski, as well Idaho and South Dakota because Mike Crapo and John Thune ran unopposed in 2004 and 2010 respectively. The 'wave' swings in these races are too contaminated by those special factors, to be useful for us here.