So, now that we have the census data saying how many people each congressional district needs to gain or lose, I took a shot at it and think I came up with something darn near workable. Redistricting experts, let me know how close I came to something the courts would accept!
First, the basic principles I tried to follow, based on what I’ve read about previous guidance from the courts:
- Prefer contiguous, reasonable shaped districts
- Avoid splitting counties & cities where possible
- Value coherent regional / cultural groupings
- St. Paul and Minneapolis must be in separate districts
- The Iron Range / Arrowhead and Red River valley must be in separate districts
- Minimize change from previous configuration
- Impact on current incumbents is not a consideration, separate from principle of least change
Here’s what I came up with:
And a Metro detail:
Now, an explanation of my changes from the current configuration:
Currently CD4 includes all of Ramsey County and three tiny slivers of Washington and Dakota Counties, and needs to grow by 48,367 people. It seemed immediately clear that any sensible solution was likely to include expansion into Washington County. This makes sense from a grouping perspective now, as places like Mahtomedi, Oakdale, and Stillwater have taken on the feel of suburbs, rather than removed outposts separate from the Twin Cities. I removed those existing bits from Washington and Dakota Counties, and instead extended CD4 eastward into Washinton County to the St. Croix River, using I-94 as the southern boundary, and basically just extending the northern border of Ramsey County on the north end. (You could also use Wa. Co. Hwy. 10 and then either the railroad track south of Big Carnelian Lake or Square Lake Trail north of Big Carnelian Lake to the state border, to use a physical divider that follows basically the same line.) That northern boundary within Washington County would be the leading contender for where fine-tuning adjustments should happen here.
For CD5, it’s currently Minneapolis proper and some stuff around it, including the weird dangling piece of Anoka County, and needs to grow by 46,509. I accomplished this by keeping all of the existing areas, and simply extending the westernmost border north and south to the edges of Hennepin County. Those northwest and southwest areas would be where fine-tuning would happen, just as in the previous arrangement.
For CD3, it’s currently the part of Hennepin County that’s not in CD5 and a bit of Anoka County, and needs to gain 12,806. I did this by dropping the part in Anoka County and adding the entirety of Carver County. Fine-tuning would primarily take place along the border with CD5, which in turn also allows you to move the split in Anoka County as needed. If really necessary, an incursion into CD2’s territory in Scott County could be done, although that should be avoided if possible.
In CD2, it’s currently a bunch of counties along the southern side of the Twin Cities, and needs to lose 69,524. I did this by first gaining the piece that was in CD4 in Dakota County and some of what was in CD6 in Washington County, and then losing all of Carver County to CD3. The border with CD4 in either Ramsey or Washington County (but preferably not both) would be the fine-tuning line here.
CD1 is the strip along the southernmost part of the state, and needs to gain 18,204. I did this by adding all of Sibley County (15k) and then the little piece of McCleod County that juts down into Sibley County, basically drawing a straight line across the top of Sibley County instead of its current U shape. That division of McCleod County would be the fine-tuning line.
CD6 is currently serving the role of the one oddly shaped district that nobody knows quite how to group, encompassing the city of St. Cloud and various areas north of the Twin Cities, and needs to change the most, shedding 96,487. St. Cloud complicates matters, as the court doesn’t like splitting up cities much, it’s the main population center for this district, and it also happens to straddle three separate counties. Currently the border is a split down Stearns County. I started trimming here with the stuff in Washington County, giving some to CD2 and some to CD4. Once that was done I was able to cut off roughly the right amount more by moving the division of Stearns County eastward, such that it lined up with an extension of the western edge of Wright County, and just grazes the western city limits of St. Cloud. This means that I was able to keep all of St. Cloud in one district, trim off some of the awkward sticking out bit of Stearns County, and make the district less sprawling by cutting off the south-central portions of Washington County, and still maintaining CD6 as a buffer between the Twin Cities districts of CD4 and CD5 and the Iron Range’s district, CD8 (because having those bordering directly would just be wrong). The Washington County split with CD4 and the Stearns County split with CD7 are the obvious adjustment points here.
For CD7, that’s the northwestern district with the Red River valley, and needs to gain 37,479. I did this by first losing a small number to CD1 in Sibley and McLeod Counties and to CD8 in Red Lake County, and then gaining the large chunk from CD6 in Stearns County. Again, the adjustment points are in McLeod and Stearns Counties where there is a county split.
In CD8, it covers the northeastern part of the state, with the Iron Range and Arrowhead, and needs to gain 2,649. it looked like I could get about the right number simply by gaining the rest of Red Lake County from CD7. If this is too much, Red Lake County can remain split and only acquire part of it for CD8.
Here’s what it looks like overlayed on the current districts:
Like I said at the beginning, the courts don’t consider protection of incumbents to be a valid value for drawing lines, but we can still analyze the impact of what I came up with.
First, Betty McCollom (D-4, St. Paul/Ramsey) and Keith Ellison (D-5, Minneapolis/Hennepin) are obviously still in their same districts. Similarly, Erik Paulson (R-3, Eden Prairie/Hennepin) is still good – my line between CD5 and CD3 would be on Eden Prairie’s eastern border. John Kline (R-2, Lakeville/Dakota) is also still in his district – we avoid some drama here by him not living in Carver County, which would have shoved him in with Paulson had it been the case. But, he’s in Lakeville, so nothing to see here either. Tim Walz (D-1, Mankato/Blue Earth) is smack in the middle of his district still as well. Collin Peterson (D-7, Detroit Lakes/Becker) is also still safely in his district. While he is on the southern edge of his, Chip Cravaack (R-8, Lindstrom/Chisago) is also still in his district, as it didn’t change there.
The one obvious exception here, and thing that may ruffle feathers about this map, is that Michele Bachmann (R-6, Stillwater/Washington) would NOT still be in her district. As I alluded to earlier, this seems very difficult to avoid, so I suspect we’ll see a number of proposals that include this situation. Frankly, it always was pretty strange to have Stillwater grouped in with St. Cloud, and CD6 has had a far less than ideal shape which will undoubtedly become more regular with any proposal that cuts off the almost 100,000 people it needs to lose. My map then has Bachmann’s current residence in the same district as McCollum, and while Bachmann could challenge her there, I would consider that unlikely, as CD4 would likely remain safe for Democrats (although marginally less so than before), making it a long shot, and because Bachmann’s main support lies in the parts of CD6 far from Stillwater anyway. The more likely scenario I think would be for Bachmann to move just a little ways north/northwest to still be able to run in the smaller district. I’m sure she’s been thinking about this possibility for some time, as the likelihood of Stillwater being removed from CD6 in one manner or another has been quite apparent for some time. (This is of course unless she ditches that idea altogether to focus on a presidential run.)
So, while incumbent protection is not supposed to be part of drawing districts, my plan still ended up keeping 7 out of 8 in their current districts, with the last one being rather unavoidable (and being more of a question of whether Stillwater would be joining CD4 or CD8). The significance of this is that since we still have districts drawn by partisan figures in the legislature rather than a non-partisan panel, there’s a higher probability of this map being agreeable to many of the people trying to draw one.
I certainly do not know the precinct-by-precinct details, but in terms of political makeup I’ll try to hazard a few generalizations as well.
CD4 and CD5 both have to grow out away from the cities proper of St. Paul and Minneapolis. This means they should become slightly less of a DFL stranglehold, but given how enormous DFL margins are in those districts currently don’t take that to mean they’ll suddenly be competitive – merely less not-competitive, if that makes sense. CD3 is a Republican district gaining more Republican territory in Carver County, so I don’t expect it to get any more competitive. CD2, like CD4 and CD5, likely will become closer to competitive, but not actually approaching toss-up territory. The reason for this is that it’s losing solid Republican support to CD3, and gaining back pieces of CD4 and CD6 that are probably more Democractic-leaning than the district as a whole. Sibley County appears to be a bit of a toss-up area, meaning things are likely to remain close in CD1, so at least we have one district that we can count on getting good show out of consistently. CD6 I’m not sure I can quite predict. It’s losing more Democratic-leaning voters in Stillwater, but also more Republican-leaning voters in central Stearns County, and in roughly similar numbers. My suspicion is that this will remain solid Republican territory almost all the time, but there’s a chance that if the DFL put up a better campaign here than they’ve been doing in a year where there was a significant shift in their favor nationally they could take it for a term or two. My map does not have CD8 making the drastic southerly incursions people have suggested it could do, since with a gain of 2,649 that’s just not necessary. That means that it’s not going to suddenly become more conservative, so while Cravaack’s still there for now, I would suggest that he can expect a tough re-election campaign still. CD7 is gaining Republican voters from CD6, strengthening the same story it’s been for a while – it’s probably still safely Democrat for now, but I’d expect it to flip once Collin Peterson retires.
All in all, I’m feeling pretty happy with what I came up with. I’m not sure if the world of partisan bickering would like something that was done only considering the simple math and cartography of it all, rather than weighing the political opportunities and risks, but hey – this is how it should be done. Now, if there’s someone with a bit of spare time, I’d love to hear how close I came on the actual numbers, since I had to make a few somewhat educated guesses here and there.