15 November 2022

Midterm Congressional Race Results Updated

After the midterm U.S. Senate races, Democrats have 50 seats, Republicans have 49 seats, and while the incumbent Democrat in Georgia got more votes than the Republican, because he got only 49.4% of the vote which is just 0.6 percentage points short of a majority, in part due to about 2.1% of the vote being cast for a Libertarian. So the GA Senate race is going to a runoff in in December. Still, Democrats will control the Senate either way in 2023 because the Vice President who resolves tie votes, is a Democrat. So, the situation in the U.S. Senate will be at least as good as the status quo and maybe better.

In the House, a majority is 218 seats and Democrats went into the midterm with 222 seats. 

As I write, there are 205 seats that have been called for Democrats and 215 that have been called for Republicans, with Republicans flipping a net ten seats so far (four of which were in New York, where the state supreme court through out the legislative redistricting map for one that was much less favorable to Democrats).

There are 15 U.S. House races that haven't been called yet

Eleven of the remaining uncalled races are in California (four leaning GOP - one of the five GOP leaning races in image was called for the Republican - and seven leaning towards Democrats). Honestly, none of those races are razor thin and none looks likely to flip from the current leads, but the vote counting there is painfully slow, with as little as 52-53% of the vote counted a week later in some races.

There are two uncalled races in Colorado. CO-3 is held by incumbent Republican Lauren Boebert in a Trump +8 district in which she is leading by a fraction of a percentage point in at the moment, but probably just enough to win given the likely makeup of the uncounted votes remaining. CO-8 is a close race in a newly created open seat, but the Republican in that race has already conceded. Really CO-3 is the only race left where a Democrat has any meaningful change of winning a Republican leaning race. The new Colorado delegation to Congress will probably be two Democrats as U.S. Senators and five Democrats in the U.S. House, and three Republicans in the U.S. House (one of relatively safe Republican winning by only a nose). 

The Denver Post reports that there are 4,685 votes left to be counted in the CO-3 race in which incumbent Republican Lauren Boebert leads by 1,222 votes (50.2% of the vote). So, Democratic challenger Adam Frisch needs about 63% of the remaining vote (mostly from Pueblo County which leans Democratic) to win, which would be a stretch, but is not outside the realm of possibility.

Democratic leaning races in Maine and Alaska are delayed due to the extra time it takes to determine a winner with rank choice voting.

The New York Times has called 217 seats for Republicans so Republicans need to win only one of four seats not yet called for a House majority, which seems very likely:

So, Republicans are on track to win a net five to eight flipped seats. Absent several lucky breaks, Republicans will have a thin 218-221 seat majority in the House. 

Pre-election polling had suggested a "red wave" election that would have shifted 30-40 seats to the Republicans, and it is looking like they will have made only 20-25% of those gains.

Still, it is quite frustrating. A shift towards Democrats of less than a percentage point in a few key races and/or a win in the New York State redistricting court case could have shifted at least three House seats, leaving Democrats with a bare majority in the House and complete control of Congress. 

A mere 0.4 percentage points strong showing in the Georgia Senate race would have made the Georgia Senate seat a clear win in the first round that will now be a hard fought and uncertain outcome dependent upon turnout in December (although Democrats won two such clutch runoff races in Georgia in 2020).

The impact of political campaign efforts and tactics is overrated, but at the margins in races with a fraction of a percentage point margins, they can make a difference.

Democrats won only about one in eight faces that neither Presidential candidate won by five points or more.

Republicans may have underperformed relative to the polls, but Democrats came up just a little short of what was needed to avoid two years of certain gridlock in Washington and didn't have an awe inspiring effort either.

The U.S. House results map also illustrates just how decisively Republicans control rural districts outside of New England and majority-minority (or nearly majority minority) rural areas in the "black belt" in the South, southern Texas, southern Arizona, and New Mexico (Alaska while it has lots of land, is actually quite heavily urbanized in addition to the population in rural areas being heavily native Alaskan). Put another away, only two House districts that touch the Pacific Coast (adjacent districts in Southern California) went to Republicans.


An in between visual depiction that captures both location and population accurately:
Some other observations:

1. The basically complete lack of a midterm defeat for President Biden is closely tied to his historic complete lack of coattails in 2020. There were few candidates dependent upon that who were swept out of the office by the lack of Presidential race momentum.

2. After the 2020 redistricting, there were 211 districts that Biden won by five percentage points or more and 190 districts that Trump won by five percentage points or more. There were 34 seats that neither Presidential candidate won by five percentage points or more. Democrats only needed to lose no more safe seats than the Republicans did, and to win 7 out of 34 competitive house districts (as defined by the 2020 Presidential election vote) to keep a majority in the House.

In the most likely scenario they will win 4 of them or so (figuring in net flipped safe districts as well). By that measure, Democratic performance was pretty mediocre. Republicans won a very large share of the winnable Congressional district races. Democrats have failed to win even as many seats in the House as supported Biden in 2020.

For example, Republicans are currently leading in CA-22 which was a Biden +13 district and CA-27 which was a Biden +12 district, and won in CA-45 which was a Biden +6 district. Something was very wrong with the Democratic party campaigns in these districts and those races alone might even have been enough to flip the House.

At a minimum, it is hard for Democrats to legitimately complain about gerrymandering as a factor in the overall control of the House even though it was clearly a factor in some individual states (probably hurting Democrats in Texas, New York and Wisconsin). Query, however, if legislative redistricting rather than independent commissions used more in blue states than red ones could have tipped the balance more in the favor of Democrats.

3. Biden’s lack of dynamic leadership and effective use of the bully pulpit, which has lead to his low approval rating, and his failure to tame the two moderate Senators in his own camp which limited his legislative achievements contributed to Democrats performing less well on fundamentals than he might have performed. Inflation he couldn’t do anything about, the jobs situation and GDP growth and stock market were decent. But he could have controlled message discipline and been more bold. Biden is not great on the campaign trail even though he managed to win and to win the primary, and the Democrats could run someone much better in 2024 if they could mobilize to do it. Biden’s campaign lacked the critical component of “hope” that Obama and Clinton and Reagan had and that Democrats need to win.

4. There is no doubt that Dobbs helped democrats in the 2022 election.

5. Polling once again was materially inaccurate, although this time in favor of Democrats, probably due to a failure to reflect the effects of Dobbs (and to a lesser extent, January 6) in their voter turnout model. Turnout was good, at least among Democratic leaning demographics, for a midterm election. We should trust polling less in 2024 than we have from 2016-2022.

6. While there are only a tiny handful of moderates in Congress (in both houses combined), there are lots of moderate Governors in both parties who won races in territory where members of their own party fared far worse in federal elections.

7. Candidate quality matters, but only a little. The abysmally bad Lauren Boebert in CO-3 is in or close to recount territory although she is leading in seeking re-election, in a Trump +8 district. Dr. Oz lost in PA, but not by much and not by significantly worse than par for the course given the partisan breakdown in PA. Walker did as well in the Senate race in GA as other much less toxically bad candidates have. About a dozen incumbent election deniers weren’t re-elected either due to primary fights or general election losses, which is about 7%. This isn’t a huge hit, but incumbent members of Congress usually do much better.

8. Florida was long seen as a partially Northern enclave in the South politically. There are large swaths of Florida where young people born there don’t even have Southern accents. And, for several election cycles is was the pivotal swing state. But Florida seems to have completely succumbed to its Southern roots again. Meanwhile, migrants to GA allied with non-white voters there have made GA competitive again, and NC is heading in the same direction due to migration to the economically thriving parts of the state like the research triangle from outside the South. Meanwhile, Ohio, which was once a swing state, is now solid red, and PA is getting pulled in that direction.

9. Not a single Republican was elected to the House from New England, New Mexico, Hawaii, or Alaska.

10. Republicans have been shedding the upper middle class, but this doesn’t seem to have resulted in mission critical shortages of campaign funding for them yet, or in decisive edges for Democrats in campaign financing from its new more affluent coalition.

11. The Senate gave us gridlock from 2000-2022. The House may add to it in the next session of Congress if the Democrats can’t keep a majority there even if they can make it to 51 Senators allowing just one converted moderate Senator to provide a majority vote.

12. I remain stunned that our country is split 50-50 by the current rules despite the fact that the GOP has gotten so extreme. There is a deep divide not just at the top. It isn’t an artificial divide. And a lot of the country has radicalized towards the far right.

4 comments:

Guy said...

In a maximally competitive environment, wouldn't you expect a 50/50 divide? If it was 49/51 then there was some platform element, or proxy, or dirty trick that the 49% party would have changed/played to get to the even split.

Tom Bridgeland said...

My understanding is that the Rs won a popular vote majority nationwide. Weren't you railing recently about how unfair it was that the minority party might win an election?

andrew said...

@TomBridgeland

It is impossible to really evaluate this in the case of the Senate in any given election cycle. You'd have to tote it up over six years.

In the House, the Republicans have a majority, mostly likely with two seats to spare although not all the votes have been counted, flipping a net seven seats. I don't know what the popular vote totals are in the collective house races so far (if you could point me to someone who could save me the math, that would be great). I don't doubt that it was close and was the midterm election generally.

As I noted in a post today, the marginal Republican in the House had a margin of victory of 0.69 percentage points, the marginal Democrat in the Senate had a margin of victory of 0.9 percentage points, and a one percentage point national shift in popular opinion would have produced a Democratic House majority and 52 Democrats in the U.S. Senate.

Still, aggregate House votes aren't really strictly comparable to aggregate Presidential popular votes, because a significant number of House seats are uncontested suppressing the vote of the party without a candidate in a lopsided race.

I continue to favor a system of voting for Congressional seats in which states with more than one seat would have voters vote for a political party and allocate the state's seats accordingly, and in which single seat states would have ranked choice voting or a majority requirement with a runoff election. This would eliminate spoiler effects and gerrymandering from the equation, while breaking it up by state would limit the ability of fringe parties with tiny percentages of support to win seats in Congress in most states.

andrew said...

@Guy

There is a theorem in academic political science whose name escapes me, which states that in a two party political system, coalitions tend to be developed in a manner that brings the two parties as close as possible to an even division of power given the rules that apply in that political system. The mechanism by which this balance is secured is primarily by redrawing the coalition boundaries of the major political parties, which are always fluid between elections.

For example, Democrats recently basically traded a lot of their working class white voters to the Republican Party and got a lot of college educated white voters in exchange. Likewise, Democrats have trades a lot of their male voters to the Republican Party and go a lot of female voters in exchange.

Time will tell how long the existing coalitions will last. For example, Republicans are trying hard to pick up conservative and religious Hispanic voters, while Democrats are picking up a continually growing share of young voters.

The midterm elections support the conclusion that this theorem is working perfectly in the current U.S. political system.

The problem is that if the rules of the game favor one side or the other, this 50-50 point will differ from the median national voter's position. In political systems that are not biased in favor of one political party or the other, the median voter theorem (which is a distinct and different hypothesis) states that policy will tend to track the opinions of the median voter.

Another problem is that coalitions need to consider more than just a division of voters.

Each party needs a corp of activists with time but not money to work political campaigns. Historically, Democrats relied upon unions and civil rights activists for this, and Republicans relied upon religious people. Both sides are seeing those resources wane.

Each party also needs money for its operations. Historically, Democrats relied upon upper middle class professionals for money and Republicans relied upon rich business people and big business for money. These balances are somewhat out of balance in the current coalitions.