Let’s review what this means historically.
The charts in this post, except where noted are from a data download of Gallup’s Quarterly Party Affiliation Survey from 2015 Q1 through 2026 Q2.
Lead Chart Question
In politics, as of today, do you consider yourself a Republican, a Democrat or an independent? (As of today, do you lean more to the Democratic Party or the Republican Party?) Shown are % party identifiers + leaners.
Gallup Party affiliation 2015-2026
Affiliation Poll Question
In politics, as of today, do you consider yourself a Republican, a Democrat or an independent?
From the above data I calculated democrat minus republican spreads.
Gallup Party affiliation Spreads
Current Party Affiliation Spreads
- Including Leans: 10.0 percent
- Ignoring Leans: 6.0 percent
Both spreads are the largest since 2015.
National Election Results & Polling Analysis Tracker
| Year | Dem Actual (%) | Rep Actual (%) | Margin Actual | Current Spread Including Leans | Error Direction |
|---|---|---|---|---|---|
| 2016 | 48.2% | 46.1% | D+2.1 | D+3.2 (2016 Q4) | R+1.1 (Poll was 1.1% too Dem) |
| 2018 | 53.4% | 45.0% | D+8.4 | D+7.2 (2018 Q4) | D+1.2 (Poll was 1.2% too Rep) |
| 2020 | 51.3% | 46.9% | D+4.5 | D+6.5 (2020 Q4) | R+2.0 (Poll was 2.0% too Dem) |
| 2022 | 47.8% | 50.6% | R+2.8 | R+5.5 (2022 Q4) | D+2.7 (Poll was 2.7% too Rep) |
| 2024 | 48.3% | 49.8% | R+1.5 | R+4.0 (2024 Q4) | D+2.5 (Poll was 2.5% too Rep) |
| 2026 | TBD | TBD | TBD | D+10.0 (2026 Q2) | TBD |
Error Direction Explanation
- R+ indicates the poll over-estimated Democrats (the actual result was more Republican than the poll).
- D+ indicates the poll over-estimated Republicans (the actual result was more Democratic than the poll).
The largest directional errors since 2015 were in 2022 where Republicans were 2.7 percentage points too high, and 2024 where Republicans were 2.5 percentage points too high.
The largest directional error for Democrats was in 2020 when the spread was 2.0 percentage point too high.
Even if one doubles that Democrat directional error, you are looking at a +6.0 national percentage point spread favoring Democrats.
This assumes the polls do not change before the election. Let’s use Q2 polls for all years.
Historical Q2 Polling Baseline vs. Actual Election Results
| Year | Dem Actual (%) | Rep Actual (%) | Margin Actual | Q2 Spread Including Leans (From Your Image) | Polling Error |
|---|---|---|---|---|---|
| 2016 | 48.2% | 46.1% | D+2.1 | D+4.8 (Row 7: 47.6% – 42.8%) | R+2.7 (Poll was 2.7% too Dem) |
| 2018 | 53.4% | 45.0% | D+8.4 | D+5.0 (Row 15: 46.7% – 41.7%) | D+3.4 (Poll was 3.4% too Rep) |
| 2020 | 51.3% | 46.9% | D+4.5 | D+6.6 (Row 23: 48.2% – 41.6%) | R+2.1 (Poll was 2.1% too Dem) |
| 2022 | 47.8% | 50.6% | R+2.8 | D+0.3 (Row 31: 45.3% – 45.0%) | R+3.1 (Poll was 3.1% too Dem) |
| 2024 | 48.3% | 49.8% | R+1.5 | R+2.0 (Row 39: 44.3% – 46.3%) | D+0.5 (Poll was 0.5% too Rep) |
| 2026 | TBD | TBD | TBD | D+10.0 (Row 47: 49.0% – 39.0%) | TBD |
Largest Errors Using Q2 Data
- 2018 Midterms: 3.4 percent too Republican when Republicans held the White House
- 2022 Midterms: 3.1 percent too Democrat when Democrats held the White House.
Political Context
- In both Midterms the Q2 results overcounted the incumbent party
- In 2018 Q4 the poll estimated a D+7.2 environment, but the actual result was a D+8.4 “blue wave”.
- In 2022 Q4, Democrats held the White House. The poll estimated a heavy R+5.5 wave, but the actual result was a much narrower R+2.8, meaning the historical midterm penalty for Democrats wasn’t as severe as the polls anticipated.
Expect a Blue Wave
Expect another Blue Wave in November. The incumbent party rates to get hurt badly even in a normal cycle.
This is not a normal cycle. Trump is one of the most unpopular presidents in history.
Gerrymandering will save some Republican net House seats but Dilution always weakens something, somewhere else.
A Blue Wave on Polymarket only requires winning the House and a pickup of two Senate seats. I believe this is a given.
Right now, I expect Democrats to win 230 to 240 House seats.
How the Popular Vote Translates to Senate Seats
I drew that blue box and that’s where things are now. Even +5 to +6 is a Senate pickup of three.
It certainly could get worse for Republicans. I put margins of error in that box only to the downside of Democrats.
From a Gallup starting point of +10, the possible upside is +12 or +13, in a wide range centered between +6 and +12.
Silver has Ohio, Maine, and Michigan as “Lean Democrat” with North Carolina as “Likely”.
That’s a pickup of three, not counting Alaska and Texas which Silver calls tossups.
Senate Outlook
In the Senate, Republicans are fielding very weak candidates in Texas, Ohio, Iowa, North Carolina, and Maine.
Alaska has a weak incumbent Republican facing a screwy ranked voting system that will pit one Democrat against three Republicans (two of them with the same name, one the incumbent).
A Blue Wave only requires a pickup of two Senate seats. I expect Democrats to win four of those, with a clean sweep of all six in play.
Pollsters call Alaska a tossup. I don’t.
Based on ranked voting and the current political environment, I think Polymarket Alaska Senate is right with Democrat Mary Peltola at 64 percent.
Polymarket has Ohio Senate at 56 percent Democrat. Assuming Democrats hold in the Senate and flip Maine that’s a Senate takeover, and my minimum expectation right now.
The Trump Factor
Trump is likely to make matters worse for Republicans, perhaps much worse.
The war in Iran is unpopular, tariffs are unpopular, diesel is at record highs and climbing heading into Christmas shipping and Autumn crop harvesting.
Factor in series of inane proclamations on Truth Social, the trade war with Canada, and deportations.
To top it off, Trump wants to personally appear to support candidates. Other than a desperate Paxton in Texas, no Republican wants to be seen with a toxic Trump.
The preceding items point to stunning Democrat numbers with young voters, independents, and Hispanics.
Combine that with lackluster Republican turnout and you are looking at a potential +6 Senate win.
I have been discussing that possibility since January when it looked crazy. Now it doesn’t.
My current forecast is Democrat +4 or more with about a 60 percent confidence rating. Senate +3 or better is about 75 percent.
Related Posts Plus a Diesel Update
September 2, 2026: Is it Idiotic to Fight a War Over Something Worthless in Two Years?
My vote is yes, but that’s what Bessent says we are doing.
September 3, 2026: The Price of Diesel Hit a New Record on Gas Buddy, AAA Tomorrow
Diesel jumped nearly 10 cents today after 5 cents yesterday.
AAA did hit a new record on September 4 of $5.8500.
That record fell again today. As of September 5, the price of diesel is a new record $5.8819.
Expect to see $6.00 soon.
I repeat my August 2 warning: So, Don’t Be Shocked if Democrats Take Six Republican Senate Seats
The grim setup for Republicans in the midterms keeps building.
what are the chances that 16 republicans grow a set of Massey balls and we get an impeachment that matters. Are there 16 republican senators that are not pedo protectors or pedos i dont think so but one can dream
A few years ago I moved far away from the big city and now live in a farm community. Which changes my algorithm.
The issue of farm inputs, fertilizer, getting more expensive is not a theory. The concern now is the absolute shortage of fertilizers for the planet. The amount of estimated sulfur production has been cut in half. Sulphur is required to make potash fertilizer.
doesn’t the usa export about 30% or so of our food. couldn’t most of flyover country folks stand to lose a few chins by not eating so much? the bread basket of central CA will rule the empire. perhaps it’s time to bring our baby killing military back stateside to be a republic and not a world wide warmongering empire. don’t be frightened, most amerikans can just buy some new belts not fitted for fatsos.