100 Senate districts...with a twist

[President Punxsutawney Phil]

You have became a special master whose job it is to draw 2 Senate districts in every state. The only instructions you have been given are - 1) preserve counties when possible and 2) [most important] try to make the districts in each state be as distant from each other as possible in terms of partisanship. You are given a deviation band of 10%. Contiguity is required, but that's about it. The VRA does not apply. Compactness is something you can aim for, but it's far from necessary.

How well do you do?

[CookieDamage]

Did NJ and PA.

NJ is split into

One urban district centered on Essex, Hudson, Bergen counties, as well as Paterson, New Brunswick, and other areas. Ultra safe Dem.

The remaining state is put into a swing district. This district has blood red areas like Ocean county, Sussex, and Warren counties, including very red, populated areas like Lakewood and Toms River. However, it also has Camden, Trenton, Princeton, and Burlington county. It was a Trump-Murphy-Hugin-Biden district.




PA was simple. One district contains Philly + its burbs + Harrisburg, Scranton, York, Lancaster, Allentown. Safe D.

The other was basically Pennsyltucky plus Pittsburgh. Safe R.

[CookieDamage]

I also have a suggestion. It would be more challenging to try to make each district as politically similar as possible. The issue with making them as far apart is that it'll almost always become the urban+suburban blue district with the larger, more rural, republican district. Making them instead both competitive seats will be harder but more interesting.

EDIT: In that vein, Pennsylvania becomes more interesting. Both seats were decided by under 4 points.

[President Punxsutawney Phil]

I also have a suggestion. It would be more challenging to try to make each district as politically similar as possible. The issue with making them as far apart is that it'll almost always become the urban+suburban blue district with the larger, more rural, republican district. Making them instead both competitive seats will be harder but more interesting.

EDIT: In that vein, Pennsylvania becomes more interesting. Both seats were decided by under 4 points.

This is an acceptable variation and both are welcome to be under the umbrella of this thread.

[bagelman]

https://davesredistricting.org/join/c237e188-b79c-43e0-b57b-f18a96e0d853

Maryland splitting only one county.

https://davesredistricting.org/join/3ad632f5-8c6a-4ded-ad16-355c5e281319

Missouri, but this map doesn't respect counties.

https://davesredistricting.org/join/1ed13870-7619-432e-8bbd-21426f254b94

Tennessee, but this map doesn't respect counties.

https://davesredistricting.org/join/99d083b6-50a9-436f-a402-2bfeb8c52deb

Georgia

[I spent the winter writing songs about getting better]

Super easy, barely an inconvenience.

[bagelman]

https://davesredistricting.org/join/d893a68d-4de1-4d7b-830c-a0fde793db0e

Here's a Georgia with both districts almost exactly the same. The northern district is usually just a teeny bit more conservative but it's such a small difference and doesn't really show up for 2020 pres.

[Left Wing]

Here's an IN map that doesn't split any counties, the red district is Republican to comical level with Rainwater coming within ten points of surpassing Myers' vote total in 2020. The blue district isn't solidly Democratic federally (Clinton +.5, Biden + 4.1), but should hold up considering the larger margins other Dems have won by (Gregg +9.8, Donnelly +13.2). It doesn't look amazing as a map but it could be uglier.

[Adam Griffin]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

[Adam Griffin]

^^^ If you don't care how things look and are willing to use county point contiguity in the Macon metro, then you can rather easily get a 66.2-point difference without county splits and with 17k deviation:

https://davesredistricting.org/join/c7a1436e-4576-4306-9f79-43838b38d6ef

CD-1 (Yellow)
66.1-32.6 Biden (5,347,442)

CD-2 (White)
65.7-33.0 Trump (5,364,466)




And for the "similar districts" alternative:

No county splits; 8k deviation:

https://davesredistricting.org/join/c8a8a832-6ce4-4cfe-b93b-a4ead752b1b2

CD-1 (North Georgia)
49.8-48.9 Biden (5,360,051)

CD-2 (South Georgia)
49.6-49.2 Trump (5,351,857)

[Schiff for Senate]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

What shocks me is that the 2nd district is only 65% Republican (and that the 1st is only 65% Democratic), though I guess it makes some sense since it includes the entire GA02, which is bluish.

[Schiff for Senate]

I did NC, based on 2016 presidential election results (2020 isn't that much different, honestly, though some areas did shift rightward in 2020 while others shifted leftward).

https://districtr.org/plan/102547

1st: The slightly lesser-populated of the two, with 5,142,000 people, the district's voting age population is 65.9% white, 17.8% black, 8.9% Hispanic and 3.6% Asian. It starts off in southwest NC and straddles the western border of the state, careful to avoid Asheville, and after reaching the VA border turns east, taking generous portions of the state's northern area, taking in Winston-Salem, Greensboro, Durham, and Raleigh and most of its suburbs. While parts of the district stretch further along the VA-NC border, it also turns southeast to take in a chunk of land centred around Goldsboro. Comparative to the state, it appears to have shifted slightly leftward in more recent years.

2nd: This district has 5,297,000, and is the more diverse of the two (61.4% white to 65.9% white). As a percentage of total VAP, the districts have identical Hispanic populations (8.9%), but the 2nd has noticeably more African-Americans (21.9% instead of 17.8%) and slightly less Asians (it has 2.7% instead of 3.6%). This district is more polarized, southern, and eastern/coastal in character, though it does include Asheville, taking in Charlotte, Fayetteville, the entirety of the NC coast and then some northeastern chunks of the state including some of Raleigh's northern suburbs. Comparative to the state, it appears to have shifted slightly rightward in more recent years.


Below are given some different races (all are 2016 or earlier, unfortunately) - note that I used 2016 presidential as the baseline of comparison, and the other races are afterthoughts - and the GOP share of the two-way vote in each district in each race, as well as the difference (if more than 0.4%, the race is bolded). Of note is that these districts voted the same way in every election given except the ultra-close 2016 gubernatorial election, and had virtually zero difference in the 2016 presidential race (the 1st voted to the left of the 2nd by a hair).

2016 PREZ: 1st is 51.97%, 2nd is 51.98% (diff.: -0.01%)
2016 SEN: 1st is 53.1%, 2nd is 52.94% (diff.: 0.16%)
2016 GOV: 1st is 49.45%, 2nd is 50.47% (diff.: -1.02%)
2014 SEN: Data seems incorrect; in actuality Tillis won narrowly but it shows as Tillis having lost to Hagan, so not going to include this one
2012 GOV: 1st is 56.2%, 2nd is 55.52% (diff.: 1.08%)
2012 PREZ: 1st is 51.73%, 2nd is 50.41% (diff.: 1.32%)
2008 GOV: 1st is 48.48%, 2nd is 48.11% (diff.: 0.37%)
2008 SEN: 1st is 46.15%, 2nd is 45.17% (diff.: 0.98%)

[President Punxsutawney Phil]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

What shocks me is that the 2nd district is only 65% Republican (and that the 1st is only 65% Democratic), though I guess it makes some sense since it includes the entire GA02, which is bluish.

A 2010s iteration of that would probably run all the way to Savannah. Population growth in Metro Atlanta has made an arm to take in Bulloch County both harmful and unnecessary.

[Adam Griffin]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

What shocks me is that the 2nd district is only 65% Republican (and that the 1st is only 65% Democratic), though I guess it makes some sense since it includes the entire GA02, which is bluish.

For what it's worth, if you draw 2 non-contiguous districts simply based on whole counties and how D/R they are within the deviation provided, you can only get to 67-68% D/R for each: not a whole lot more room to move relative to the above proposals.

However, if you go by precinct and make it non-contiguous, then you can get at max something like 72-27 Biden & 69-30 Trump for the 2 areas.

[Schiff for Senate]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

What shocks me is that the 2nd district is only 65% Republican (and that the 1st is only 65% Democratic), though I guess it makes some sense since it includes the entire GA02, which is bluish.

For what it's worth, if you draw 2 non-contiguous districts simply based on whole counties and how D/R they are within the deviation provided, you can only get to 67-68% D/R for each: not a whole lot more room to move relative to the above proposals.

However, if you go by precinct and make it non-contiguous, then you can get at max something like 72-27 Biden & 69-30 Trump for the 2 areas.

Fair enough, but I thought even with contiguity the districts could deviate more. Of course if you aren't going to be contiguous and not care about county boundaries you have pretty much a free rein, but I figured even within the said constraints you could have more polarized results.

[Adam Griffin]

A 2010s iteration of that would probably run all the way to Savannah. Population growth in Metro Atlanta has made an arm to take in Bulloch County both harmful and unnecessary.

Just a bit, yeah. Definitely would eat up a cut chunk of the NW corner of Chatham County.

[Adam Griffin]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

What shocks me is that the 2nd district is only 65% Republican (and that the 1st is only 65% Democratic), though I guess it makes some sense since it includes the entire GA02, which is bluish.

For what it's worth, if you draw 2 non-contiguous districts simply based on whole counties and how D/R they are within the deviation provided, you can only get to 67-68% D/R for each: not a whole lot more room to move relative to the above proposals.

However, if you go by precinct and make it non-contiguous, then you can get at max something like 72-27 Biden & 69-30 Trump for the 2 areas.

Fair enough, but I thought even with contiguity the districts could deviate more. Of course if you aren't going to be contiguous and not care about county boundaries you have pretty much a free rein, but I figured even within the said constraints you could have more polarized results.

I think it's easy for GA to look more polarized or homogenously populated than it truly is to those who haven't spent years understanding it. It's simple enough to see a ton of counties and a ton of partisan variance among them and think the polarization across such boundaries is immense, but in actuality, a lot of Republicans live inside Democratic counties and (albeit to a lesser degree) vice-versa; it's hard to contiguously and geographically segment the state into 2 equal and sensible pieces while exceeding two-thirds in favor of any one party.

The main reason is that most of Georgia is, well, empty (relatively speaking). This kind of illustrates things: the colored areas comprise the equivalent of 3 CDs (21% of 2020 state population), with 2 contiguous CDs centered around the least populated regions, followed by a third that fills in the remaining least-populated gaps across the state. The vast majority of Georgians either live in the ATL metro or in a scattering of tiny metropolitan areas across the state (which based on the map below of the 11 remaining CDs worth of population, Biden would have won by the same amount he won Virginia; 11.1 points).

[Schiff for Senate]

Georgia:

No county splits; somewhat geographically-relevant. A 63.4-point difference between the 2 (a much tidier, no splits 62.1-point version with 5k deviation can be found here).

CD-1 (Atlanta, Macon, Athens, Augusta & I-20 Corridor)
65.1-33.7 Biden (5,356,060)

CD-2 (Everywhere Else)
65.4-33.4 Trump (5,355,848)

https://davesredistricting.org/join/426c6941-cc1f-4907-87a3-bc3652199901

What shocks me is that the 2nd district is only 65% Republican (and that the 1st is only 65% Democratic), though I guess it makes some sense since it includes the entire GA02, which is bluish.

For what it's worth, if you draw 2 non-contiguous districts simply based on whole counties and how D/R they are within the deviation provided, you can only get to 67-68% D/R for each: not a whole lot more room to move relative to the above proposals.

However, if you go by precinct and make it non-contiguous, then you can get at max something like 72-27 Biden & 69-30 Trump for the 2 areas.

Fair enough, but I thought even with contiguity the districts could deviate more. Of course if you aren't going to be contiguous and not care about county boundaries you have pretty much a free rein, but I figured even within the said constraints you could have more polarized results.

I think it's easy for GA to look more polarized or homogenously populated than it truly is to those who haven't spent years understanding it. It's simple enough to see a ton of counties and a ton of partisan variance among them and think the polarization across such boundaries is immense, but in actuality, a lot of Republicans live inside Democratic counties and (albeit to a lesser degree) vice-versa; it's hard to contiguously and geographically segment the state into 2 equal and sensible pieces while exceeding two-thirds in favor of any one party.

The main reason is that most of Georgia is, well, empty (relatively speaking). This kind of illustrates things: the colored areas comprise the equivalent of 3 CDs (21% of 2020 state population), with 2 contiguous CDs centered around the least populated regions, followed by a third that fills in the gaps across the state. The vast majority of Georgians either live in the ATL metro or in a scattering of tiny metropolitan areas across the state.

I disagree with the bolded section (though it obviously depends on what is defined as 'a lot') and have evidence. I find that out of (by my count) 30 Biden counties, 2 (Clayton and DeKalb) crossed the 80% mark, and out of what should come to 129 Trump counties, some 24 gave him north of 80% (including Brantley and Glascock, which gave him 90% of the vote each). So GA is actually quite polarized. I think what does make the results as they are is that some suburbs of Atlanta, included in the blue district, are still quite moderate, and the entire GA02, where blacks outnumber whites, is in the red district, thus making it more moderate.

[Adam Griffin]

I disagree with the bolded section (though it obviously depends on what is defined as 'a lot') and have evidence. I find that out of (by my count) 30 Biden counties, 2 (Clayton and DeKalb) crossed the 80% mark, and out of what should come to 129 Trump counties, some 24 gave him north of 80% (including Brantley and Glascock, which gave him 90% of the vote each). So GA is actually quite polarized. I think what does make the results as they are is that some suburbs of Atlanta, included in the blue district, are still quite moderate, and the entire GA02, where blacks outnumber whites, is in the red district, thus making it more moderate.

Obviously Georgia is (racially) polarized, but the real polarization isn't necessarily at its greatest on a county-by-county basis.

35% of Trump's votes in GA came from counties Biden won. 29% of Biden's votes in GA came from counties Trump won.

I'm not sure what the national share is by comparison, but the fact that Trump got more than 1/3 of his votes from counties that only comprise 53% of the 2020 population in one of the most racially polarized states still makes the geographic argument hold. After all, many people forget that rural Georgia is still >20% black; this is why (along with the fact that vast swathes of the state that go heavily-GOP are very empty, and quite a few that are empty still go quite D for the aforementioned reason) you can't draw 2 equally-sized segments that go 75%+ for either candidate.

[Schiff for Senate]

I disagree with the bolded section (though it obviously depends on what is defined as 'a lot') and have evidence. I find that out of (by my count) 30 Biden counties, 2 (Clayton and DeKalb) crossed the 80% mark, and out of what should come to 129 Trump counties, some 24 gave him north of 80% (including Brantley and Glascock, which gave him 90% of the vote each). So GA is actually quite polarized. I think what does make the results as they are is that some suburbs of Atlanta, included in the blue district, are still quite moderate, and the entire GA02, where blacks outnumber whites, is in the red district, thus making it more moderate.

Obviously Georgia is (racially) polarized, but the real polarization isn't necessarily at its greatest on a county-by-county basis.

35% of Trump's votes in GA came from counties Biden won. 29% of Biden's votes in GA came from counties Trump won.

I'm not sure what the national share is by comparison, but the fact that Trump got more than 1/3 of his votes from counties that only comprise 53% of the population in one of the most racially polarized states still makes the geographic argument hold. After all, many people forget that rural Georgia is still >20% black.

Mostly unrelated, but interesting fact about GA: In 2020, exit polls showed Hispanic men only voted for Biden by 3 points!

Anyway, to what you said, the reason Trump got such a high percentage is because the blue counties of metro Atlanta are very populous and still produce the most raw votes for Trump. And many of them are only light blue, going for Biden by under 20 points and therefore not being that strongly Democratic to begin with. Biden's 29% can be explained because there were a lot of counties in the southwest with sizable black populations that narrowly voted for Trump but gave Biden a decent number of votes (though to be fair, they are quite rural). Further east, a similar story is in the suburbs and exurbs of Augusta.

[Adam Griffin]

I disagree with the bolded section (though it obviously depends on what is defined as 'a lot') and have evidence. I find that out of (by my count) 30 Biden counties, 2 (Clayton and DeKalb) crossed the 80% mark, and out of what should come to 129 Trump counties, some 24 gave him north of 80% (including Brantley and Glascock, which gave him 90% of the vote each). So GA is actually quite polarized. I think what does make the results as they are is that some suburbs of Atlanta, included in the blue district, are still quite moderate, and the entire GA02, where blacks outnumber whites, is in the red district, thus making it more moderate.

Obviously Georgia is (racially) polarized, but the real polarization isn't necessarily at its greatest on a county-by-county basis.

35% of Trump's votes in GA came from counties Biden won. 29% of Biden's votes in GA came from counties Trump won.

I'm not sure what the national share is by comparison, but the fact that Trump got more than 1/3 of his votes from counties that only comprise 53% of the population in one of the most racially polarized states still makes the geographic argument hold. After all, many people forget that rural Georgia is still >20% black.

Mostly unrelated, but interesting fact about GA: In 2020, exit polls showed Hispanic men only voted for Biden by 3 points!

Anyway, to what you said, the reason Trump got such a high percentage is because the blue counties of metro Atlanta are very populous and still produce the most raw votes for Trump. And many of them are only light blue, going for Biden by under 20 points and therefore not being that strongly Democratic to begin with. Biden's 29% can be explained because there were a lot of counties in the southwest with sizable black populations that narrowly voted for Trump but gave Biden a decent number of votes (though to be fair, they are quite rural). Further east, a similar story is in the suburbs and exurbs of Augusta.

A lot of those SW and eastern non-urban counties have next to nobody in them, for what it's worth. Nevertheless, there are fewer Ds (both %-wise and nominally) in Trump-won areas than there are Rs in Biden-won areas because the aggregate of GOP turf in Georgia is more R than the aggregate of DEM turf is D (though because there are fewer places of population that are heavily-R, you invariably have to dip into areas that aren't heavily-GOP to make a 2-district map).

BTW and in relation to my previous post, I did the math: 43% of Trump's national vote came from counties Biden won; 30% of Biden's came from Trump-won counties. That basically means Trump America is just as R as Trump Georgia (at the county level), but obviously Biden America is more D than Biden Georgia.

At any rate, I think we've made this enough about Georgia - since this is a national thread, I'm gonna pipe down now!

[President Punxsutawney Phil]

I disagree with the bolded section (though it obviously depends on what is defined as 'a lot') and have evidence. I find that out of (by my count) 30 Biden counties, 2 (Clayton and DeKalb) crossed the 80% mark, and out of what should come to 129 Trump counties, some 24 gave him north of 80% (including Brantley and Glascock, which gave him 90% of the vote each). So GA is actually quite polarized. I think what does make the results as they are is that some suburbs of Atlanta, included in the blue district, are still quite moderate, and the entire GA02, where blacks outnumber whites, is in the red district, thus making it more moderate.

Obviously Georgia is (racially) polarized, but the real polarization isn't necessarily at its greatest on a county-by-county basis.

35% of Trump's votes in GA came from counties Biden won. 29% of Biden's votes in GA came from counties Trump won.

I'm not sure what the national share is by comparison, but the fact that Trump got more than 1/3 of his votes from counties that only comprise 53% of the population in one of the most racially polarized states still makes the geographic argument hold. After all, many people forget that rural Georgia is still >20% black.

Mostly unrelated, but interesting fact about GA: In 2020, exit polls showed Hispanic men only voted for Biden by 3 points!

Anyway, to what you said, the reason Trump got such a high percentage is because the blue counties of metro Atlanta are very populous and still produce the most raw votes for Trump. And many of them are only light blue, going for Biden by under 20 points and therefore not being that strongly Democratic to begin with. Biden's 29% can be explained because there were a lot of counties in the southwest with sizable black populations that narrowly voted for Trump but gave Biden a decent number of votes (though to be fair, they are quite rural). Further east, a similar story is in the suburbs and exurbs of Augusta.

A lot of those SW and eastern non-urban counties have next to nobody in them, for what it's worth. Nevertheless, there are fewer Ds (both %-wise and nominally) in Trump-won areas than there are Rs in Biden-won areas because the aggregate of GOP turf in Georgia is more R than the aggregate of DEM turf is D (though because there are fewer places of population that are heavily-R, you invariably have to dip into areas that aren't heavily-GOP to make a 2-district map).

BTW and in relation to my previous post, I did the math: 43% of Trump's national vote came from counties Biden won; 30% of Biden's came from Trump-won counties. That basically means Trump America is just as R as Trump Georgia (at the county level), but obviously Biden America is more D than Biden Georgia.

At any rate, I think we've made this enough about Georgia - since this is a national thread, I'm gonna pipe down now!

Let's broaden this a bit. How does Georgia compare to other states in the Deep South here?

[President Punxsutawney Phil]

This is an effort at Texas. Compactness was thrown out the window. El Paso, Brownsville, Houston, Dallas, Austin, and San Antonio are all placed in one district. Biden won the Dem district 60-39. Trump won the other district 63-35. Over 900,000 more votes were cast in the R district than the D one.
https://davesredistricting.org/join/bd966bd4-b8cc-4bc6-aa8c-e9d9c03b01d7

[I spent the winter writing songs about getting better]

Wisconsin is incredibly easy for having the largest distance between them. No county splits:



Biden 59.94%-Trump 38.34% in the blue district, Biden 39.72%-Trump 58.55% in the red district. Interestingly it gets even more polarized with the 2018 Governor numbers: Evers 60.53%-Walker 37.35% in blue, Evers 38.88%-Walker 59.20% in red.

[Schiff for Senate]

Wisconsin is incredibly easy for having the largest distance between them. No county splits:



Biden 59.94%-Trump 38.34% in the blue district, Biden 39.72%-Trump 58.55% in the red district. Interestingly it gets even more polarized with the 2018 Governor numbers: Evers 60.53%-Walker 37.35% in blue, Evers 38.88%-Walker 59.20% in red.

The converse (2 politically identical districts for WI), with 2020 presidential as the baseline:

1st: the smaller (area and population), eastern district. Major cities include Milwaukee (though much of WOW is given to the 2nd) and Green Bay. It is much more diverse than the 2nd - it has more Hispanic-origin people, more African-Americans, more Asians...though the 2nd has slightly more Native Americans.
2nd: significantly larger and more western, a good chunk of the district's population resides in Madison, and even more live in Southeast Wisconsin - much of WOW and Kenosha and Racine Counties are in here to help balance ultra-liberal Madison, but also included is the Driftless Area (it has all of western WI), and it also stretches across WI's northern boundary (it also straddles WI's southern boundary).

2 way vote percent margin in various races (and analyses by me):

2020 PREZ (baseline): TIE - both were D+0.64
2018 GOV: big difference, actually, but it makes sense given western and rural WI shifted more to the right in the Trump years but is more welcoming to Democrats at a state/local level (to put it another way Walker is more establishment and less populist than Trump, making him more popular in eastern WI and less popular in western WI) - 1st was R+1.3; 2nd was D+3.4
2018 SEN: similarly large difference, but it makes sense since the 1st is more urban/suburban and if they tied for the presidential race it means in non-Trump, more local elections (the 2nd WI also likes charisma and progressivism/populism, helping Baldwin, while the 1st is more bland/establishment and probably prefers someone like Biden to Baldwin), the 2nd will probably be a bit bluer than the 1st - 1st was D+9.3; 2nd was D+12.3