Nate Silver weights every new poll based on if the firm leans more Democratic of Republic. Since the Kavanaugh hearings the landscape has changed, but Silver keeps adjusting for new polls that come out but we're run before Kavanaugh.
If you just look at polls on his list that we're run for the most part in October, the GOP is 22-2 in the seven races they need to get from 47 to 54 Seats. In all 7 races the GOP candidate leads the majority of polls:
47 seats definitely GOP
48. North Dakota 1-0 (ahead in one poll, not behind in any) by an average of 10.4 points.
49. Tennessee 3-0, 7.5
50. Texas 7-0, 6.5
51. Nevada 3-0, 4.0
52. Arizona 4-1, 1.6
53. Florida 2-0, 0.3
54. Missouri 2-1, -0.4
Indiana, Montana and West Virginia have not been polled and look tougher for the GOP, but gives a ceiling if 57.
Below are old notes:
and here is how the Senate looks based on the trend of polling over time. It looks like the GOP's chance to expand their edge in the Senate is including after the Kavanaugh charges and defense. All four key races in which Kavanagh has been polled show his confirmation supported and thus making a Republican win more likely.
October 7 update - The day after Kavanaugh became a member of the Supreme Court, it appears the hearings gave the GOP a very solid 50 seats, with North Dakota, Texas and Tennessee all moving strongly to the Republican. The next 6 seats certainly look close, and I grade the GOPs chances as a "B" to take Missouri to get to 51, a "BC" to probably sneak out close wins in Florida and Indiana to get to 53. They may fall just short in Arizona, Montana and Nevada, (Grade CD below) but winning those as well would take them to 56 seats.
Only one poll in Arizona shows the GOP winning, but the fact that support for President Trump, Kavanaugh, ICE and particularly the GOP candidate are all strong, it would seem she has a decent chance to sneak across the finish line. The fact that Senator Flake endorsed McSally before the primary is an interesting dynamic, as she was one of the few to overcome a Trump-endorsed candidate in a primary. It would seem Trump campaigning in Arizona to add to Flake's support might complete the picture for her.
State | ND | TX | TN | MZ | FL | IN | AZ | MT | NV | WV | NJ | MN |
---|---|---|---|---|---|---|---|---|---|---|---|---|
Jan-Apr | -3 | 1 | -21 | |||||||||
May | -4 | |||||||||||
June | 4 | -8 | -7 | |||||||||
July | -3 | -10 | -2 | |||||||||
Aug 1-10 | 0 | -12 | ||||||||||
Aug 11-20 | 4 | 4 | -6 | -6 | ||||||||
Aug 21- | 1 | -2 | 0 | 0 | -6 | -8 | ||||||
Sept 1-10 | 4 | 1 | 2 | -1 | -9 | |||||||
Sept 11-20 | 4 | 9 | -1 | 1 | 0 | 2 | -5 | -2 | -2 | -12 | -7 | |
Sept 21-30 | 2 | -3 | -4 | -4 | ||||||||
Oct 1-10 | 12 | 6 | 7 | 0 | -1 | -2 | -1 | -4 | -9 | |||
GOP if wins | 48 | 49 | 50 | 51 | 52 | 53 | 54 | 55 | 56 | 57 | 58 | 59 |
GOP chances | A | A | A | B | BC | BC | CD | CD | CD | D | F | F |
State | ND | TX | TN | MZ | FL | IN | AZ | MT | NV | WV | NJ | MN |
I used bold italics for numbers that include CNN polls, which has been so much skewed toward the Democrats than other polls.
I start by putting the polls (or the average of polls) listed in chronological order from the oldest poll on top to most recent on the bottom, then if there is a Kavanagh poll from after the charges made against him I include that at the bottom.
Republicans get to 47 seats based on simply seats not up for re-election or seats in which the Democrat has never been within single digits.
GOP looks strong for 48th and 49th seats based on trend.
48. Texas. As much as the Democrats hope to take out Ted Cruz, and the Republican was only a 1 points ahead in late August. However, that went to +4 in early September and +9 in mid-September, so it looks to be on a strong trend line for Cruz.
49. North Dakota. Likewise, after trailing by 3 points early in the year (-3), the Republicans led by 4 by June and were still up by 4 in mid-September, so it looks like North Dakota will give the GOP 49 seats. Oct 2 Note - this looks more likely with a new poll showing Cramer now +10, and the state supporting Kavanaugh 60-27.
GOP also on course for three more challenger wins to get to 52.
With Flake leaving, Republicans would feel very strong if they got to 52 votes for times when they lose a vote on the floor in Maine and Alaska, particularly if there are other US Senate votes. The trend for three other challengers seem to give them a good shot at three.
50. Indiana. Trailing by 12 in early August (-12), then 6 in late August (-6) and now up 2 by mid-September (+2) is the most lightning fast shift in favor of the GOP and seems to give them a very strong shot at the 50th seat and control of the Senate.
51. Missouri. Going from a tie to a one-point win by mid-September gives the GOP a tight edge for the 51st seat. Oct 2 Note - CNN polled Democrat back ahead, but their polls are getting much more Democratic results this year then all others . A state poll showed Hawley moving ahead and more voters saying McCaskill opposition to Kavanaugh will make them LESS likely to vote for her.
52. Florida. Almost the exact same scenario plays out in Florida except that it has been an exact tie in every polling period except for a 1-point lead in early September.
The Republicans do have a very realistic shot at stretching it to 54 seats, but the next two are tougher. Oct 2 update - a couple of recent polls show this edging Democrat so I might drop this to 54 and move the next two up a spot.
53. Tennessee. I will put Tennessee ahead of Montana only because they did lead by 4 points and it is an open seat, however, the 4 point lead in mid-August dropped to a -2 in late August, though it did get back down to -1 in mid-September - so very tight. Oct 2 - Nate Silver does have this as a GOP lean, though he is more bullish on the Democratic candidates in a number of others.
54. New Jersey. After making up almost an entire -21 in New Jersey, the GOP did slip back a little to a -6 in mid-August, so an upset is possible but trend has not continued as of now. October 2 Update: With a new poll showing Republicans within 2 points, I am starting to believe the GOP might win this race in the shocker of the night.
55. Montana. The one on the stronger trend toward a Republican win is Montana, where Tester's lead eroded from 8 points in June, to 3 points in July to 2 points recently. A big debate on Montana PBS could be key but the trend here is toward a GOP win. Oct 2 update - I wonder if north Dakota's overwhelming 60-27 support is close to the number in Montana, and if so if Teaters recently announced opposition to him makes this closer.
56. Nevada. This looks like a seat the GOP will give up even though it's always been within 2 points either way. However, the GOP's +1 early in the year turned into a -1 in early September and -2 in mid-September, so the trend is small but definitely toward the Democrat.
57. Arizona. Going from a 2 point lead to a -5 between early and late September definitely shows this trend away from Republicans, but certainly still winnable.
58. West Virginia. The GOP has polled barely within single digits a few times, but no steady trend and the recent polling is a -12, so this one looks very difficult for the GOP despite being overwhelmingly pro-Trump. October 2 - I moved this up one spot only because the support for Trump and Kavanaugh is so strong and it is possible he could vote with his fellow Democrats against the nomination.
59. Minnesota. Each polling period has the GOP within single digits, but no consistent trend so looks pretty solid for the Democrats to hold.
To get to 60, of course, the Republicans would need to win all the previously listed races AND win some other race in which they had never been within single digits - so just as the GOP seems to start with 47 in the bank the Democrats seem to start with 41 in the bank.