# Senate Surprises: What to Look Out For in the Election

**Potential Surprises **

I created the above map starting from a 270-to-Win Senate projection based on current polls.

270-to-Win had Alaska and South Carolina as tossups. I felt that was a bit far-fetched although it is "possible". Even Texas is "possible" but I discount anything under a 20% chance.

**Meaning of Circles **

- A red circle on a brown state designates a tossup state but one that I expect Republicans will win.
- A green circle on a brown state indicates a state that I believe can easily go either way.
- A red circle on a blue state indicates that I believe the Democrats will win the state. The state is in play for the Republicans, but not close enough to call a tossup.
- A blue circle on a red state indicates that I believe the Republicans will win that state. The state is in play for the Democrats but not close enough to call a tossup.

**Iowa Senate Polls **

All polls from 538 Political Polls.

**Iowa Analysis**

Iowa represents the Democrats best chance to win the Senate outright.

538 has the Iowa Republican Senate Forecast as 53-47 for the Republicans but I rate the state a 50-50 tossup because all recent polls of "likely voters" give Greenfield a lead with by an average of 2 percentage points.

Why does 538 have Ernst ahead? I believe it is because 538 expects Iowa to vote for Trump by a 50.5 to 48.2 margin.

By the way, that is close enough that Iowa could easily land in Biden's column.

**Georgia Senate Polls**

Georgia has two Senate races in play. Republicans will easily hold the race between Perdue (R) and Ossoff (D).

In the Special Election there are 6 candidates, three of which have no chance. It is between two Republicans (loeffler and Collins), and Democrat (Warnock).

**538 Georgia Forecast **

The Special Election Rules are such that if no candidate gets 50%, there will be a another vote on January 5.

538 expects Georgia to vote for Trump by a 50.7 to 48.5 margin.

That puts Georgia in play for Biden. It also puts Warnock in play in the January 5 election.

Silver's 17% estimate is thus way too low. The special election will also be influenced by turnout in January adding more uncertainty.

Republicans dismayed by Trump's loss might not show up to vote in January.

I give Warnock a 40% chance, more than double Silver's estimate.

**Kansas Polls**

That's the only Kansas poll recent enough to discuss. It looks like a tossup but Data for Progress is a sponsored poll, I suspect by Democrats.

I accept Silver's assessment that Republican Marshall has a 79% chance of winning.

**Montana Polls **

Once again there is only one poll recent enough to discuss. This time, however, the poll is by A+ rate Siena College.

The poll shows Democrat Bullock trails Republican Daines by a mere one percentage point.

Wendie Fredrickson is the Green Party candidate.

Given that Fredrickson would likely take votes away from Bullock, it's somewhat a mystery this race is close.

Silver has Silver Republican Baines a 66% to 34% favorite over Bullock. Silver also has Montana going to Trump by a 54.2 to 43.7 margin.

Given that Montana is not really in play for Biden, I think a 34% chance for Bullock is too high. I give Bullock about a 20% shot.

**Maine Senate Polls **

**Maine Senate Discussion**

Maine has more polls to discuss than any other state. Clearly the spotlight is on Sue Collins who is expected to lose.

## RECOMMENDED ARTICLES

But note that only one poll has Gideon above the 50% mark. This is similar to no poll having Hillary over 50%.

More Information has the poll tied. I doubt it. Pollsters marked with an "*" are sponsored.

Silver gives Collins a 41% chance.

But Silver also projects Biden will win Maine and Maine1, the former 55.3 to 43.2 and the latter 60.4 to 38.2.

It will take huge split ticket voting if those numbers are accurate.

I suspect Collins has about a 33% chance of winning.

**North Carolina Senate Polls**

**North Carolina Discussion**

Tillis must be truly despised to be trailing in every poll by 4 to 11 percentage points.

Yet Silver gives Tillis a 36% chance.

I suspect the key reason is Silver expects Biden to win North Carolina by a mere 50.0 to 49.2 margin.

Given the polls, I suspect a 30% chance for Tillis is about right.

**South Carolina Senate Polls**

South Carolina is considered so much of a shocker that few believe it. Color me sceptical as well.

Pollster "Brilliant Corners" is sponsored, most likely by Democrats and possibly Harrison. Silver lists sponsored polls with an "*" but does not list the sponsor.

The polls from August are too stale to consider.

The best poll of the lot is by highly respected Quinnipiac University who rates this even.

Nate Silver has the odd of Harrison winning at 21%. The reason is clear to see. Silver projects South Carolina will go to Trump by a 53.8 to 45.3% margin.

It will take enormous split-ticket voting if Silver has the presidential forecast correct.

Polls show Trump leads Biden in South Carolina by 6 to 10 percentage points.

If that lead holds, it will be very difficult for Harrison to pull this off.

**Nate Silver 538 Senate Projection**

Nate Silver's Senate Projection has the Democrats a slight (62-38) chance of winning the Senate.

Here is my assessment.

**Possible But How Possible?**

To hold the Senate, Republicans will need to win a clean sweep of Iowa, Montana, Kansas, South Carolina, and the Georgia Special Election, plus Maine or North Carolina or some unknown surprise.

**If **the above map holds and it comes down to Maine and North Carolina then Republicans should hold the Senate as winning one of those two states is likely.

The problem is the the "**IF**". The above map is "possible" but very unlikely based on a more realistic assessment.

**Expected Results**

I expect Democrats will win all the blue states, the Republicans all the red states plus all the tossup states with red circles.

That makes the map 50-49.

That's a win for the Democrats if they win the White House as is increasingly likely.

**Odds of Democrats Winning At Least One of Iowa or Georgia**

- If we assume the Democrats odds of winning Iowa are 50% and Georgia 40% then the odds of the Democrats winning
*at least one*is (0.50 + 0.40) - (0.50 * 0.40) = 0.70. - If you go with Silver's model, the odds of the Democrats winning
*at least one*are (0.47 + 0.17) - (0.47 * 0.17) = 0.56.

Thus even Silver's has a better than 50% chance of the Democrats winning *at least* one of those two states.

Democrats chances of winning the Senate are substantially higher than Silver projects. Georgia and Iowa are the keys to understanding why.

Mish