Talk Elections

Michigan is one of several states who are changing their redistricting laws this cycle, although Michigan is certainly the state changing their laws the most. With their new commission, Michigan’s previous redistricting history matters very little and instead the history from other commission states deserves greater attention. The rules, influencers, and geographic breakdown of this commission matter more than people’s previous actions and the party’s goals. This examination is therefore going to be a bit different from ones that have come before.

Link to 2010 Atlas Discussion (https://uselectionatlas.org/FORUM/index.php?topic=127927.0)

Redistricting History

Michigan had Republican trifectas in both 2000 and 2010, and both times they drew maps to increase the amount of accessible GOP seats. Both times the GOP kept to redistricting guidelines adopted in the 1990s when control was divided between the parties. They had every reason to stick to them - these guidelines favored the ‘modern’ Republican coalitions geographic breakdown. The times had changed, and what had once been a good-govt proposal now was abused by the GOP. Counties and municipal lines needed to be respected as much as possible, and that various local lines that were smaller than a congressional district needed to be inside a single district. These guidelines incentivize cutting the larger counties and keeping the multitude of GOP rural counties whole.


Of course, the goals of the two different GOP governments were divergent because the various entrenched democratic interests were in different parts of the state. Bart Stupak was locked into the northern first district, which in the 1990s resembled its present incarnation and contained the GOP stronghold of Traverse City. The Detroit region and its suburbs were also carved up to favor to democrats, particularly in Macomb and the near side of Oakland. That needed to change. The big 1st was made into a new rural pack for the north of the state to free up what GOP suburbs were in the region. This 10th was entirely new based out of the red rural thumb and Republican suburbs. The 12th was designed explicitly to pack in the near-side spillover from Detroit and make the suburban seats safer. The rest of the map was pushed around to maximize Republican safety.

Now, the Democrats won a special election on the old map, but that special was held concurrent with a new election on the new map which the Republicans won. David Curson’s 2-month tenure is one of the shortest in congressional history, since he had only been elected to fill a vacancy.


2010 was an entirely different beast. The previous map had held up well until the democratic waves came crashing through, and the democrats flipped control of the state delegation. The 9-6 was back in 2010 though, only this time Stupak had been traded for the 9th in Oakland. Once again, the GOP moved against the Detroit region, but the rest of the map was reinforced to protect their incumbents. The AA Detroit seats now tentacled out into the suburbs to pack in growing minority communities. The 5th packed in even more of the Tri-Cities, but also cracked the upstate in case rural Dems came back. The 1st got Traverse city back out of a similar desire to lock down the peninsula against a potential working-class democratic comeback. Like before, the rest of the state was carved up in order to spread out democratic cities and ensure no seat would ever truly be in danger.

Since 2011

I won’t really bother with much here because the previous process is changing. The democrats tried to crack the old map where Obama was strongest, and they consistently failed. It required the 2018 wave for the democrats to break through the gerrymander and make the map 7-7. However, the gerrymanders have held at the legislative level, even as the democrats win the statewide vote.

The most important thing to happen this decade around redistrict was addressed in the header. In 2018 Michigan overwhelmingly passed proposal 2, which creates an independent commission responsible for future redistricting efforts. The GOP keeps mounting legal challenges, but so far, the commission remains intact and appears to have full authority next year. The whims of legislators will no longer directly decide Michigan’s maps.

2021

So, let’s talk about this Michigan commission. (https://www.michiganradio.org/post/redistricting-proposal-passes-michigan) The commission has 13 members: 4, Democrats, 4 Republicans, and 5 Independents. These commissioners cannot be employed or connected to any government employee or lobbyist. The test for these commissioners is adequate self-identification with one of the three relevant groups. Now, self-identification would normally be a problem especially when compared to measurable tests of partisanship like party ID or past vote records. However, Michigan appears to have anticipated these fears. At the start of the process, the Michigan SOS (Jocelyn Benson, Democrat) will randomly mail at least 10K voters to select them for the commission. The SOS has stated she sent invitations to 250K voters (https://www.michiganradio.org/post/many-michigan-registered-voters-seeking-serve-new-redistricting-commission), and already over 1,000 have been successfully processed their application. This is random selection, not self-selection like in California. From here, the SOS will select 60 partisans of each color and 80 independents of adequate quality to potentially serve on the commission. The majority and minority leaders of each chamber can strike up to 5 candidates from the pool of 200 each (reducing it to 180), in order to remove visible turncoats. From there, the rest of the applicants will be put in their individual pools and randomly drawn to fill the four/four/five seats for that pool.

Michigan’s entire process is steeped in random selection to ensure the commission is legitimate. Even if that fails, the passage requirements should prevent malicious actors. The maps need only be passed by a simple majority (unlike a supermajority in some commission states), but that majority needs to include at least two Republicans, two Democrats, and two unaffiliated. Therefore, it is clear that the partisan commissioners need to work together. Rather than partisan infiltrators or malicious actors, we should perhaps fear ignorance instead. The shear amount of randomness means that we may end up with uninformed or sheepish commissioners who will follow the brash personalities of parochialists or a single commission clique. We don’t know, and frankly, that’s just part of the commission process.

Looking at California, Washington, Arizona, and other states that cared about public input in 2010, we get a general idea of the process to a commission-drawn map. The committee will schedule public hearings to meet with citizens across the state. Through this process, hypothetical maps may be released to guide the discussion on potential pairings of COIs. Protecting communities of interest, race, and local geography is important under the mandate of the law. Sometimes discussion may be guided by the media or outside influencers trying to ‘push’ the commission towards a particular outcome. Republicans for instance got the idea of a <50% white district in Washington into the public discourse, and the California commission was famously lobbied unsuccessfully by an “soundproofed neighborhood group” which was designed to twist LA districts to an incumbent’s benefit. Eventually, they will hire a firm and give them guidelines on what should go where and what things to look for in their lines. The map will then be presented before the public, and may face subsequent changes depending on public demands. One such call that came up a bit ago was the pairing of the rural thumb with upstate, along the shores of Lake Huron through Bay City.


So, what might this new map look like? The first question that needs to be addressed is that of the lost district. Taking a look at the current growth estimates, things are rather patchy. Detroit is shrinking, as usual, but her suburbs are growing. Outside of the metro area we see that while some rural and small-town communities are losing population, others are gaining. The western shore, Grand Rapids, and Lansing in that regard are doing well for themselves. Now, for the past two decades the GOP forced the cut district upon Detroit. This makes sense in the context of their partisan goals, and when you consider the regions population losses, but the side effect is that the whole area ends up with less districts than its pop would normally demand. It is partially why on the present map you have multiple GOP districts grabbing random parts of the metro area. Perhaps therefore the 2020 cut will be forced upon the rest of the state outside of the Detroit metro area, though this is a commission so who knows.


We then move on to the more interesting discussion: Communities of Interest. Michigan has a lot of them, and they will likely be bickering about who deserves to go where and why. I’m just going to list off some of these communities, some are obvious, others are less so. The two AA districts are likely to shrink into Wayne for compactness reasons and drop below 50% AA (but still high enough to dominate a primary) but what communities will go in said districts are a open question. Is Ann Arbor going to be paired with Wayne or it’s more college educated neighbors in the suburbs? Do North Macomb and Oakland constitute a COI (potentially shared with other areas) even though it crossed county line? Should the Grosse Pointe’s go with their more likeminded brothers in Macomb, rather than their natural neighbors in Wayne? Should there be an ‘arab access’ district with areas like Plymouth, Livonia, Dearborn, and Hamtramck? Should the shores of Lake Michigan and Huron be considered huge COIs?  If they do, they conflict with other COIs like the Flint/Saginaw/Midland/Bay City ‘tri-cities’ region in the east and the Grand Rapids metro region in the west. Is the tri-cities even a good region since some cities have strong AA presences and others do not? What about Mount Pleasant – should it go with the nearest similar university in Lansing? How about Kalamazoo and Battle Creek? These questions and countless more will be answered during public hearing.

What’s left to Decide

N/A. The committee does not directly consider incumbent residencies or incumbent partisanship as part of their mandate. It’s very likely some representatives will have to change their address if they want to continue to have a chance at serving in Washington.

I've been working on some potential state legislative maps for Michigan, as those are reasonably easy to predict if you're following good government grounds of minimising city and county splits. I'll try to type those up and put them here.

Ottawa County under 2010 numbers is sufficient for one SD by itself but under 2016 block estimates it is no longer so.

https://davesredistricting.org/join/21a90af8-a1a5-4a7f-b51c-ef28e70cbfe6
https://davesredistricting.org/join/db73d662-b1fc-4fb4-acf6-905ad66fef2a
2 potential maps.  Detroit area gets shifted a lot once drawing a map based on compactness rather than tendrils for partisan reasons.  The only main difference between my 2 maps is whether Flint and Lansing are in the same district.  A decade ago that would've been seen as a Dem pack but looking forward it makes a Clinton+14 (ideal margin, enough to be safe but not so lopsided it wastes Dem votes) seat outside of Detroit metro.  Without combining the 2 cities, it's difficult to create a safe dem district outside of Detroit, due to political geography.  My second map gives Flint and Lansing each their own districts, but both districts are competitive, one leaning R one leaning D.  It is true Trump wins a majority of seats on both maps, but that is basically inevitable due to political geography on any fair map.  Dems are heavily packed into Wayne County (which can't be cracked due to the VRA) and sprinkled out throughput the rest of the state.  Any map where Clinton wins 7 or even 6 seats is probably a Dem gerrymander, compensating for Dem's geographical disadvantage rather than drawing fair lines.  The real debate will on exactly what each side wants to compromise on, and whether each side values more safe seats or more competitive seats. 

I think it's pretty clear that no Democratic-favouring map is going to seek to put Washtenaw and Wayne in the same district. Based off the 2018 population numbers, Wayne, Macomb and Oakland have a combined entitlement to only just over 5 congressional districts, which frees up Washtenaw to soak up Republican turf in South East Michigan.

Alternatively, if that's not happening then I would expect Flint to be paired with Saginaw and Bay in a Democratic map, given that Kildee doesn't look to be vulnerable on the present lines.

As the Arizona independent commission has proved, independent commissions actually don't draw bipartisan gerrymanders or give any consideration to the political interests of the parties. The Michigan Independent Commission has very clear criteria which they must abide by in this order:
Equal Population, Geographical Contiguity, Communities of Interest, Partisan Fairness, favouring or disfavouring incumbents, respecting county and municipal boundaries and compactness.
And I'll just quote the Michigan Constitution on the fourth criteria
The commission will draw a map similar to those the AZ commission draws. Prioritising COIs, Compactness and Partisan Fairness (so an 8-5 map is clearly not acceptable) and not giving any consideration to the concerns and demands of political parties.

I had a quick go at designing what a Democratic congressional map might look like. The places where the thumb is put on the scale are pretty obvious, but some of them are probably things that could be justified in an actual map. I've used the population estimates for counties and county sub-divisions for 2018. They don't quite match up to what the figures will be for 2020, but it's probably close enough.

()

()

MI-1: C 35.6% T 58.8% - safe R.
MI-2: C 37.5% T 56.5% - safe R. Probably more Moolenaar's district than Huizenga's, though neither lives there.
MI-3: C 42.9% T 50.2% - likely R. Maybe you could make this competitive if you combined Grand Rapids with Muskegon or Kalamazoo and stripped out the suburbs, but that's never realistically going to happen.
MI-4: C 47.8% T 46.1% - lean D. A Lansing-based central Michigan district ought to exist under any fair map, but if you top it out with counties to its south then Democratic prospects are much better than if you go north.
MI-5: C 48% T 46.9% - likely D. Possibly a worry if the competitiveness on the presidential level starts to manifest itself at a congressional level, but if that happens Democrats are screwed in Michigan anyway.
MI-6: C 38.9% T 54.9% - safe R. A more aggressive gerrymander would strip out Kalamazoo, but I wanted halfway clean lines away from Detroit.
MI-7: C 53% T 41.6% - safe D. Possibly only likely, but the Democratic base here is pretty inflexible. Keeping Washtenaw and Wayne separate ought to be a key Democratic objective.
MI-8: C 47.7% T 47.6% - lean D/toss-up. Livingston is drowned out with Democratic-trending bits of Oakland. Very vulnerable if Republicans improve their performance in highly-educated areas, but as it is Slotkin has nothing to complain about.
MI-9: C 52.9% T 42.6% - likely D. The most obviously gerrymandered district, because I didn't want to concede a seat in Macomb. If it drops Southfield to bolster MI-11 or MI-13 then you could make it less obvious, but it would be much more competitive.
MI-10: C 32.6% T 62% - safe R.
MI-11: C 61.8% T 34.8% - safe D. 50.1% white, 42.8% black. It's perhaps pushing it whether this counts as a VRA district, but the Democratic primary is certainly black-majority so it's probably OK.
MI-12: C 54.3% T 40.8% - safe D. This is designed as the white Wayne County district, so I can send both the VRA districts across county lines to soak up white Republican voters.
MI-13: C 65.9% T 30.6% - safe D. 46.7% white, 40.6% black, 9.8% Hispanic.

I think something similar, albeit likely less gerrymandered to your MI-11 could be one of the focal points of discussion. Sending one of the VRA seats north into Macomb to grab the near suburbs does wonders to ease the burden of the map when it comes to partisan balance, at least in PA-2018 style of thinking. There however are  two big problems with it. First, it prevents macomb from having a seat nested out of the county, which could be a swing seat - especially if you consider the Grosse Pointe's extensions of Macomb culturally. Second, 8-mile road is arguably the clearest COI divider in the state and we should probably avoid cracking it whenever possible.

8 Mile Road has certainly been a COI divider historically, but is that still true to the same extent? Eastpointe isn't far off an African-American plurality at this point and similar changes can be seen in the south of Warren, so it's arguable that that dam has finally broken there.

I tried out a light D-favoring map as well and came up with the below. It's still only a 7D-6R map, and the 7D seats aren't all super-safe. Sort of a tough map for the Democrats to get something good out of, especially given the constraints on map-drawing. I ended up with a very different approach from the map above, and maybe that map is better. I felt bad about stranding Saginaw in a Safe R district, e.g. Might try fiddling around with this more.

https://davesredistricting.org/join/96a9fba0-1c69-4fea-bec5-d31421ca6b6f

Edit: Actually, inspired by other maps on here, here's what I think is a much better map. This one could be 9D-4R if the Democrats are lucky (the Grand Rapids district voted for Trump by less than 1,000 votes) and has 8 Clinton districts:

https://davesredistricting.org/join/96a9fba0-1c69-4fea-bec5-d31421ca6b6f

I tried to make the most "plain" map that I could.  I really didn't favor either party or take PVI into account at all (except I guess the Detroit Metro).  I'd think the commission would go with something close to this, if they favor COI's and not splitting counties and so on.

()

()

https://davesredistricting.org/join/86c3575b-47c5-455b-bdef-ba4557967108

What would actually really help is only drawing one majority AA district in the Detroit metro.   It allows the remaining AA vote to be distributed much more efficiently.   Other than that just a typical map where each major metro gets it's own district.

I actually really like the MI-2 (purple) district here.  I grew up in the area and the west coast area does have it's own community.  It has way more in common than the MI-1 area (dark green), which is more country-bumpkinish.   Both are safe R though.

You are talking about drawing a map without consideration of partisanship. Partisan fairness means drawing a map that represents the politics of the state, and it does mean drawing lines to achieve that, as Independent Commissions already do. Trump 7-6 Clinton seems like a reasonable assumption, with one competitive D seat and one or two competitive R seats.

Illegal map, need 2 AA seats in Detroit.

I don't think it'll even be possible to draw two AA majority seats in Detroit anymore,  the numbers aren't there anymore.

But once you account for COIs, 2 vra seats in Detroit, ect it's difficult to get 6  Clinton seats. It is possible with 4 in Detroit, 1 in Flint, and 1 in Lansing, but it's risky for Dems because margins in the Detroit and Lansing seats would be razor thin.  In order to pass a bipartisan commission, you can't draw a Dem gerrymander.

Once again, it isn't a bipartisan commission. It's an independent commission. There is a difference between them that you don't seem to understand.

doesn't have to be a majority, look at southeast VA.

There are members of both parties, and people from each party need to agree on a map.

https://davesredistricting.org/join/1d117936-4d28-4a7a-9dfe-529e010bef54
Ok I actually was able to make a 7-6 map that isn't a blatant Dem gerrymander (like cracking Macomb) and follows the VRA (unlike a certain other map).  It also maintains COIs well.  Some Dem commission members might not like the Clinton margins in the western MI seats, but with a fair map you can draw either 1 solid dem seat or 2 tilt dem seats in western MI.  In this map I chose the latter.

Not sure how likely the commission is to draw a map like this but here is my attempt.

()

It somehow ended up as quite a Republican gerrymander (even if I did not look at partisan data and tried to generally keep counties whole). However most of the R leads are quite small. Only 4 districts are decided by more than 11 points. So it can also work as an R dummymander maybe.

Also, this map keeps 2 VRA districts in Detroit, at 53 and 48% black VAP

I think this map is perfection:

()

()

https://davesredistricting.org/join/86c3575b-47c5-455b-bdef-ba4557967108

Two AA districts now (I doubt the population numbers by 2020 will be correct though,  probably both very under-populated).

MI-3 (Red) really should go west from Grand Rapids, not east,  the metro itself extends west into Ottawa, makes the most sense.

Love the Flint and Lansing seats, both competitive, both respect COI's in the area.

Tons of competitive seats on the map overall,  only five seats were won by either party by more than 10% (including the two AA ones),  with 3 within 5%.

This is my favorite so far.

()
https://davesredistricting.org/join/f1e1bee7-6029-41ce-b898-8ba4e67cd46f (https://davesredistricting.org/join/f1e1bee7-6029-41ce-b898-8ba4e67cd46f)
1 - R+10
2 - R+3
3 - R+15
4 - D+3
5 - R+2
6 - R+11
7 - D+2
8 - R+13
9 - R+2
10 - D+1
11 - D+10
12 - D+31 (49% Black)
13 - D+24 (49% Black)

If the Commission decides to prioritise partisan fairness and competitiveness. COIs still respected. 6 seats have a D PVI, 7 seats have a R PVI. 6 seats voted Clinton, 7 seats voted Trump. 4 seats are safe R, 3 seats are Safe D, and the remaining 6 are competitive, with 3 leaning left and 3 leaning right. Only 3 townships + Detroit are split.

There should only be 13 districts in a 2020 map. You may also be using outdated population totals. That would explain why the two black districts are possible if you are using 2010 data and 14 districts.

I think double-splitting Wayne and Oakland between MI-9 and MI-12 as on this map is illegal under the Michigan rules. You can't have two districts that both split the same two counties.

Should be solvable by putting Pontiac in MI-09 and pushing MI-12 down through Mexicantown, allowing MI-13 to take up the rest of Wayne from MI-09.

Perfect?!?  You unnecessarily pull Grand Rapids out of Kent County severing the suburbs and how you drew Detroit was a blatant Dem gerrymander by cracking republican leaning areas.  You made a 7-6 map and so did I: https://davesredistricting.org/join/1d117936-4d28-4a7a-9dfe-529e010bef54
mine actually respects COIs.

I wasn't aware of the Michigan requirements.    How about this then?   If all you need is two Black plurality seats, it almost makes the Detroit Metro a Dem gerrymander (only that all the seats are Lean D now, no tossups).

()

()

https://davesredistricting.org/join/86c3575b-47c5-455b-bdef-ba4557967108

2016:
1: 33.8%D - 60.8%R
2: 37.1D - 56.7R
3: 42.6D - 50.6R
4: 45.0D - 48.9R
5: 48.6D - 46.3R
6: 43.1D - 50.8R
7: 36.1D - 58.5R
8: 51.0D - 44.1R
9: 51.5D - 43.2R
10: 33.0D - 62.1R
11: 68.0D - 28.6R (47.4% AA)
12: 72.3D - 24.5R (48.6% AA)
13: 55.3D - 39.9 R

3 Safe D
2 Likely/Lean D
2 Tossup
2 Likely/Lean R
4 Safe R

Overall 6 Clinton and 7 Trump

By 2022 I'd expect MI-3 (Red) to trend to either tossup or maybe Lean D.  MI-5 (Flint/Tri-Cities) will trend to Lean R maybe.

Why split Kent County?  Also the 3 way cut of Genesee is unnecessary and unfair to the Flint community.

If you split Ottawa and Kent, then that's actually breaking up a COI.   The counties east of Kent have very little in common with Grand Rapids, and the metro extends westward into Georgetown/Hudsonville.  

The main focus of the Detroit metro was working around the two AA seats,  nothing else was done deliberately.  

And please - Kalamazoo is a COI with Monroe county?  Or Lapeer with Livingston?  Grand Rapids is better served with Mecosta county? You have county chains/groups in that map that make no sense.

You obviously just packed the Dems into a few Detroit districts and then spread the remaining metros out with Rural areas.

Why not read the description. This map is merely an example of what a commission could do if they prioritised competitive districts.

gotta love that 20-mile wide fajita strip COI that stretches like 200 miles

Competitive districts can be achieved without dicing up communities.

The competitive map respects COIs just as much as your map does.

my map respects COIs far better.

LOL
>"Ottowa and Kent counties being in different districts breaks a COI"
>proceeds to separate Grand Rapids from half its suburbs by unnecessarily breaking Kent in 2.

Also I love how compact districts in Detroit that don't cut 20 miles into the suburbs count as packing.  You literally shred Northern Oakland County for purely political purposes.

violates the VRA, just reach the second black district into southern Oakland County and you'll get the numbers you need.  A tendril into Pontiac isn't even needed.

here I edited my map for more compactness and evening out the black populations in vra districts.  Still 7 Trump 6 Clinton.  4 safe R (Huron Shore, UP, west coast, Detroit exurbs),  2 likely R(South MI, Macomb), 1 Lean R(Grand Rapids),  2 lean D(Flint, Lansing), 1 likely D(Oakland),  3 safe D(2 Detroit, Ann Arbor).
https://davesredistricting.org/join/1d117936-4d28-4a7a-9dfe-529e010bef54

how so? michigan goes down to 13 CDs and detroit has shrunk significantly relative to the state

But there are still sufficient black electors for it to be possible for them to elect two representatives of their choice and they're geographically concentrated enough to pass the Gingles test. Instead you've packed them into a district that must be about 75% AA. That's as obvious a violation of the VRA as you'll ever see.

well i'm sorry i was just following city lines and compactness

why not this, compact, 10 county splits, COI preservation and 7-6 trump

()
MI-01: Trump +23
MI-02: Trump+9
MI-03: Trump +24
MI-04: Trump +21
MI-05: Clinton +1
MI-06: Trump +8
MI-07: Clinton +16
MI-08: Trump +4
MI-09: Clinton +42
MI-10: Trump +31
MI-11: Clinton +3
MI-12: Clinton +9
MI-13: Clinton +54

your map still consists of many strangely configured districts, ex. mackinack bridge all the way to metro detroit, monroe to kalamazoo, stretching grand rapids out into rural counties instead of keeping it an urban district

The vast majority of the Grand Rapids suburbs are in the MI-3 district,  the city portion of Kent county actually cuts off very quickly after Ada, which is in the district.   Nowhere close to "half" the suburbs are broke off, the portion of Kent that isn't in MI-3 only has about 148k people, compared to the 278k in Ottawa.

Kent and Ottawa have too many people for a single congressional district. If you believe they ought to go in the same congressional district (and it's an eminently fair argument if you look at a map) then you have to remove some of the outlying portions of one or both of them. It's not evidence of a gerrymander, just evidence of Kent being a large county.

I guess you could do it this way (All of Kent, some of Ottawa),  I just hate the wrap around district you end up with for MI-2

()

https://davesredistricting.org/join/86c3575b-47c5-455b-bdef-ba4557967108

This map is decent overall, but you have two Wayne-Macomb districts and two Oakland-Livingston districts. Under the Michigan rules, you generally can't double-split counties like that. (I could *maybe* see a commission making an exception for the black seats if it seemed like the double-split was designed to preserve them.)

There's no maybe about it - the current map does so, so there's no realistic question but that it's legitimate to secure two AA-majority seats.

Non Detroit metro is fine but in Detroit the districts are a bit odd, particularly sending the black districts deep into suburban territory.  One black district can remain entirely in Wayne, the other can go into southern Oakland to get enough blacks to get to 49-50%.  Generally speaking it's good to split counties once if at all.  There should be a Macomb district, it would end up being aswing district since Macomb went Obama in 2012, Trump in 2016, then Dem in 2018 for governor and senate.

if i do what you say it changes the whole configurement of the map. i don’t want to do a 9-4 gerrymander lmao. from a COI standpoint the suburbs that get paired with detroit are blue collar as well which at least makes sense, all while having two wealthy COI suburban districts (7 and 11) and i already have an entirely wayne-based blue collar one (12).

Well since everyone here seems intent on drawing a partisan map that would be unlikely to get support from the Republican commissioners it needs for passage, I'll get in on the game too.  Here's a 9R-4D map where Trump wins all the red districts by 10 or more.  Very clean and decent on COIs.  Interesting how it's less brazenly partisan than the current map yet better for Republicans at the same time.  I have to thank you guys for the idea of shoving the VRA districts deep into the suburbs, that's how I could make this map, just shoving them into different suburbs ofc.  This shows how the vra districts are a double edged sword that can be used as a weapon by either party with respect to shoving them into the northern suburbs.
https://davesredistricting.org/join/5184cd9e-e518-41f5-9a1f-8eab1fa5412a

Yeah, because Battle Creek is clearly a COI with southern Wayne, and putting Lansing and Ann Arbor together wouldn't make either upset.

And the Thumb to Saginaw to Eaton district...yeah.

Guess it's my job as thread OP to get us back on track. Posted this on my twitter a while back. 5/5 safer, 3 competitive. Whitmer won 9, Snyder '14 won 8 because he of course bombed in Flint.

()

Neither AA seat is >50% but both are plurality AA in 2016 data. MI-12 tries to get all the big arab communities together. I prefer cutting as few counties as possible, so someone (Isabella) drew the short straw. Would have liked to keep the Lansing region together, but needed to get an equitable distribution statewide.

I made the map to make a point

You can still keep Lansing together and get a Clinton seat, like I did on my map. https://davesredistricting.org/join/1d117936-4d28-4a7a-9dfe-529e010bef54 
Also a lot of those county breaks seem a but random, you could clean it up a bit without losing a 7-6 breakdown.  And what did Isabella county ever do to you? HAHAHA

()

The three core Detroit counties (Oakland, Macomb, Wayne) are slightly over 5 districts.  There should be one solely in Macomb and one solely in Oakland.  One AA district can take the excess Macomb population and the other can take Southfield, leaving one district to take the southern and western tier of Wayne, plus most of the rest of Oakland.  

The 10 counties in the Detroit CSA are just shy of 7 districts, which logically means one district from Monroe to Washtenaw and then one district from Flint to St Clair, splitting Livingston between them.  It's pretty perfect, actually.  

The rest falls into place pretty naturally: a Grand Rapids district including suburbs in Ottawa, a Lake Michigan district, a Kalamazoo-Battle Creek-Jackson district, a Lansing-centered district, a Midland-Saginaw-Bay City-thumb district centered on Saginaw Bay, and the north.  

This map has 5 Clinton districts (the 4 in Wayne/Oakland + the Ann Arbor district) but the Lansing district is only Trump +2 and is more Democratic downballot, and the Flint district is Trump +10 but is basically even if you add 2012 to 2016.  

We have already had this discussion at length elsewhere but it comes down to personal preference over cuts. Is cutting three counties once better than cutting one county three times? In my opinion, no, because you carve up three easily defendable COIs, when you could only carve up one. However, everyone is entitled to their own opinion on that matter. I would say a map that cuts only six counties to achieve pop equity, 3 of which need to be cut because they are overpopulated, therefore does a  good job.

Of course you should also try and observe other COIs while you are at it, stuff like pairing Flint with the northern suburbs doesn't make cultural sense even though it might pop-wise. In that regard, I only failed Lansing, since the map needs true partisan equity (safe R = Safe D) for a even state.

https://davesredistricting.org/join/75dbead5-5961-4005-b824-7ec359d690ca

Comments on this map? I tried to keep cities and counties whole while being as nonpartisan as possible. That said, I'm not knowledgeable on Michigan COIs, so there could be a major mistake here. It's a 7-6 R map with 3 competitive R-leaning districts (3, 4, and 6) and 1 competitive D-leaning district (7).

I would say small cuts into 3 counties are better than cutting 1 county into 4.  Also, Isabella county isn't Wayne or Oakland, keeping it whole is easy.  I did not pair Flint with the suburbs on my map, I agree it doesn't make sense culturally.   As for partisan fairness, it isn't explicitly defined relating to specific criteria, just one of many factors.  Basically the point is so the commission doesn't draw a blatant partisan gerrymander, doesn't mean 1 safe seat has to always be matched by another safe seat.  It is obvious you are compensating for Dem's geographic disadvantage since you are drawing a tilt dem map, like you always do.  Many factors go into drawing a map and the commission may decide COIs are more important.  Not saying they'll draw a R gerrymander, but the map might reflect the geographic reality of the state.  If I were a republican commissioner I wouldn't try to draw an unfair map, but I'd reject any dem attempt to draw lines to benefit them without a concession given to me.  With commissions like this it's all about give and take.  If I'm a commissioner maybe I agree to a Lansing to Kalamazoo district, then you agree to keeping Macomb whole.  Something like that.

This  is basically a Macomb whole district, I just tend to think the grosse pointes belong in with Macomb, and they are hardly paragons of Dem partisanship, plus it allows one to nest the 8th. I wasn't attacking you with Flint, just giving an example.

Also If you think I draw Dem tilting maps...oh baby. I take fault with Stephan Wolf because I think he draws Dem tilting maps. I will concede that most of my write-ups so far have been in states where things will either get better for Dems or the fundamentals of the game favor them, so that might give the impression. If we were to get a thread going on Wisconsin, Florida, South Carolina, or prod me in Ohio, you would see my other side. Hell, you already saw the 4-way cut in Tennessee, the 8-1 in Indiana, and you can go on my Twitter for things like my Oklahoma City spiral. Remember, I'm the only guy who also expects the GA02 cut.

Good except for cracking republican areas in Oakland and Macomb.

Flint really belongs with Saginaw, maybe not Midland and Bay, but at least it should have Saginaw in with it.  If there's 1 real COI on the whole map outside the Detroit metro, it would be Flint with the Tri-Cities area.

not really.  You just want them together so the seat doesn't vote Trump.  Saginaw and Flint are different cities, not a single COI.  Now if I'm a Republican on the commission, I might still agree to a Flint-Saginaw district, it's a small concession.  But the other side would need to cooperate in other areas.  It's inevitable some districts will be drawn in a way that disproportionately favor one party, but the whole map can't be drawn with subtle decisions that all happen to favor 1 party.

I'm not saying you are the most biased, but in your "fair" maps have some subtle bias.  On your map the tipping point district is in Macomb, and Macomb leans more dem downballot, you know this.  Your map would likely be 7D-6R.

That's ridiculous.  Putting Flint with Saginaw isn't a concession.   The two urban areas are extremely similar.   They both even have significant Black populations.   If that's how Republicans are going to be than the commission is doomed to fail.   You have to agree to basic level COI guidelines,  not just making sure no line benefits Democrats without concessions.

This isn't warfare.

2 separate metros that happen to be similar doesn't make a COI.  That's not how COIs work.  With that logic Flint and Lansing could be a COI too.  I'd be fine with that, you wouldn't.  My point is that I'd be fine with a Flint-Saginaw district, but the rest of the map can't all be drawn to favor dems as well.

I believe two things, both in regards to 2020 and next decade:

1) The presidential topline is absolute. It may just take 4-8 years for the opposition and govt to switch and then see the opposition take all their lined up gains. Marginal presidential seats will be marginal presidentially, Safe seats will be safe, and seats that have moved between categories will move.

2) Strong or reoccurring trends are going to continue to influence the future.

This also may make my maps seem 'favorable' to the D's or R's depending on whose perspective I take. I almost always believe that a D/R pack and R/D pack is better than two/three D/R seats closer to the median. The twin packs will survive all ten years while the cracks may flip depending on the environment. This includes regions moving hard towards the left/right, they deserve to be packed just as well as the established regions.

Genesee's median income is $39,000,  Saginaw's is $41,000,  Ingham (Lansing) is $54,000.

Also Genesee is 20.7% Black, Saginaw is 18.8% Black, Ingham is 11.8% black.

From 2010 to 2018,  Genesee and Saginaw have both declined in population by 4.4% and 4.7% respectively.   Ingham has grown 4.2%.

I could go on and on here, but the point is clear.

Flint and Saginaw are both urban, and have way more in common with each other than the rural counties around them (or Lansing).   If there's going to be any discussion at all about "COI's" (like you've brought up dozens of times in this thread) then there has to be some sort of framework to go by on what that is and what the goal is to be,  otherwise any district drawn anywhere can be considered a partisan gerrymander.

Nice stats on 3 cities!  Doesn't make a coi.  Again I wasn't arguing Flint and Lansing are a coi, just used that as an example of how you line of thinking could be used against you. 

()

The Detroit CSA is an objective measure of COI, but if you insist, here's a version with Genesee + Saginaw + Genesee-adjacent leftovers from neighboring districts.  It's Trump +2.  

https://davesredistricting.org/join/878acdf9-d9e1-49cb-840a-fc56c0e492b8
here's a competitive map that slightly favors republicans.  While dems would win about half of the seats in a year like 2018 maybe a majority, a decent r year could be a slaughter.

The commission isn't made up of political apparatchiks or rabid party hacks. All the members are selected at random from a pool of independent applicants. And to serve on the commission you can't actually have any political ties whatsoever (no position within a party, staffer, lobbyist, consultant etc), merely that you registered as a member of a party on voter rolls. Just look at the Arizona commission for what the membership will be like. All lawyers, most with doctorates and additional degrees, and with no actual political links.
The aim of the commission is not to draw a bipartisan gerrymander. It's to draw a fair map that prioritises COIs while making sure it doesn't advantage either party (and yes that means adjusting for the geographic disadvantage).

Yes, I thought I covered this well in the breakdown, but I guess it deserves to be repeated. This isn't a NJ style commission where the commissioners have constant chats with the party in question, leading to wheeling and dealing. A commission like this listens to public input and then passes off their idea of approximate COIs to a chosen mapping firm. Once preliminary maps are presented, the commissioners may haggle over smaller details, however the core plan was selected by the people (or whomever decides to try and make an argument before the commission in their public hearings) and the demographics of the state. This is what happened in CA, and MIs commission is a mirror version of CAs.

Flint and Saginaw are a natural combination, as they're both manufacturing cities with similar sets of economic issues.

However, if you choose not to pair them, then there's one possible combination nobody else has tried yet, namely a Lansing-Saginaw district.

https://davesredistricting.org/join/fd58f3f7-afcc-4211-b3ce-3bb4d1df620a

It's not a particularly natural pairing, but it works better than a Flint-Lansing district, as Saginaw is small enough for them to room for both it and Eaton and Clinton counties.

Flint then goes with either the Thumb counties or with Pontiac and northern Oakland (I think the latter is slightly better, as Pontiac is similar enough even if northern Oakland isn't, but reasonable minds can disagree on this.)

The Thumb then goes with northern and eastern Macomb and the 9th and 11th are neatened from their current iterations but not wildly different. The two Detroit districts can be characterised as one covering the East side, Downtown and Downriver, and one covering the West side and inner suburbs to the north and west.

Livingston, Washtenaw and Monroe aren't exactly a natural community, but it is at least a coherent district based on counties on the edge of the metro area, and I think the Kalamazoo-Battle Creek-Jackson district works rather nicely.

That then forces Ottawa into the 6th and the 3rd compacts into an essentially rectangular shape. I opted to split the rest of the Lower Peninsula on roughly east-west lines, but some might prefer a north-south split.

So the Flint and Saginaw districts are both a bit awkward, but everything else works out surprisingly neatly. Trump won 7 of those districts, but in the case of the 5th his margin was only 0.9% and it practice it would probably return seven Democrats most years.

COIs yes, but I don't see why adjusting for the geographic disadvantage is a necessary part of the commission's responsibilities.  I'm sure some on the commission would think so, but not all. 

The instructions for the commission say partisan fairness is one of many criteria included.  Certainly different commissioners will have different ideas about what a fair map would actually look like, but folks here suggesting partisan fairness automatically means drawing lines with an intent to help out Democrats are kidding themselves and don't understand the first thing about partisan fairness.  There is no specific formula or criteria included in the law to define partisan fairness, for good reason.  The intent of the law wasn't to create a map that elects a delegation that exactly matches the statewide popular vote, it was to ensure neither side draws a map that was clearly biased, even it it follows the other criteria.  The current map is like that, it favors the republican party and it's clear that was the rationale behind every part of the map.  The commission has Republicans, Independents, and Democrats, people from each side have to agree.  Democrats could try to make the case that a fair map means drawing lines with intent to help them, but I doubt it would be very convincing to the republicans who need to agree on the map.  The map needs to be fair, sensible, and bipartisan.  If the commission works like it's supposed to, you'll see a map that is't obviously drawn to favor one party.  Here is a good starting point for a fair map.  Ik it has 14 districts, but it could be worked with https://projects.fivethirtyeight.com/redistricting-maps/michigan/#Compact
https://davesredistricting.org/join/7a9f5f27-c686-4b38-b432-85ae3efe10e4

I really like this map, and this does seem realistic. The only place I would try to change for fairness is the Bay City area; if it's possible to put the split in a rural county instead, I think that's better. (I generally prefer splitting rural counties to urban ones because you're affecting fewer people that way.)

Going into Bay was mostly a legacy effect - initially I wanted to get Bay City in (on the basis that southern Bay is fairly distinct from much more rural northern Bay) but then I found that based off 2018 population estimates it's not quite big enough. That will probably still be the case on the census figures, but it might be possible to take most of Midland instead (just leaving off the rural fringes.)

While your map doesn't really favor one party,  you split Oakland 5 ways heavily diluting their voice.  It is possible to only split it once, with one district entirely within it (probably a Dem seat) and exurban areas going elsewhere.  Also some dems might be nervous about all those Clinton+5 or less seats.  They are all trending R.

https://imgur.com/a/pQIkR45 (https://imgur.com/a/pQIkR45)
https://imgur.com/7d7Ddl7 (https://imgur.com/7d7Ddl7)
6-1-6 map
4 county breaks

that looks more like 7 splits. also your number of splits definitely has any significance with regard to redistricting

I only counted a split when a county that could in theory be kept whole wasn't.  Wayne, Oakland, and Macomb will be split no matter what, unles Baker v Carr is overturned.  But regardless let's not get stuck in the weeds.

()
https://davesredistricting.org/join/b89d1b54-490e-4440-b3b6-29d27d0081dd (https://davesredistricting.org/join/b89d1b54-490e-4440-b3b6-29d27d0081dd)
1 - R+11
2 - R+7
3 - R+7
4 - R+4
5 - R+5
6 - R+5
7 - R+4
8 - R+5
9 - R+6
10 - R+6
11 - D+18
12 - D+28
13 - D+32
The Michigan GOP would like to introduce you to this totally fair map which definitely has no ulterior motives.

how do u directly upload a pic like that?
but it's amazing u could draw a map that lopsided that clean looking.  The only obvious give away is arm reaching into Pontiac.

I copy and paste into discord and use that link here.

And the arm reaching to Pontiac can be justified under the VRA without too much effort.
And it's a reminder that just because a map that looks clean doesn't mean it's at all fair. That's why Independent Commissions take into account partisan fairness.

The arm into Pontiac isn't needed to fufill the vra tho, 2 roughly half black districts can be created with limited intrusion into Wayne.  It can be legally justified (but there is an argument it is using race as a predominate factor which isn't legal)  but courts have not struck down a district similar to that so who knows.  My point was that the arm is the one part of your map that looks gerrymandered to the average voter.  How did you upload the photo?

a) Oakland isn't a single thing - it's a huge county ranging from core bits of the Detroit urban area in the SE, to suburbs in the middle to rural areas in the north. 9 and 11 split the suburban bits and you can reasonably argue that it's better to unit them and then have one district solely in Macomb, but otherwise the map matches up to the different COIs in the county reasonable well.

b) Those seats aren't trending R. The northern bits of 4 are trending R, but the southern bits are trending D and there's more population in the south of the seat (and the south is growing and the north isn't). Also, Saginaw County has a significant non-white population, so there are limits to how much further it can swing unless it goes full Kentucky (in which case Democrats are screwed whatever happens.)

In 8, Washtenaw has about half the electorate and is trending D (although they may be close to tapped out in the east of the county and it's not clear whether the same will happen in the west too.) Livingston is also trending D, so it's only Monroe which is the issue and that's only about a fifth of the electorate.

In 9, most of the Macomb bits are swinging R, but the minority population in south Macomb is growing rapidly, which ought to take the edge off that. In Oakland, the trend is towards Democrats. That said, if you swapped St. Clair Shores for Mt. Clemens and part of Clinton Township (which looks cleaner on a map but splits more townships) then it probably would be an issue.

ok, you can keep kidding yourself.  Sure maybe the massive trends from 2012 to 2016 will stay frozen while d trending areas stay trending D :D For now those Clinton+4 seats might be somewhat safe but in a Dem midterm your map collapses.  It is clear you spread dems incredibly thin so they still have a chance at winning 7 seats, but your map is too partisan for republicans, funky for independants, and risky for dems.  This is a better alternative in every way: https://imgur.com/a/pQIkR45

You seem to have the idea that everybody else is biased, whereas you are fair. The former may be true, the latter is frankly delusional. You're just much better at seeing other people's biases than admitting your own.

I would agree it's not necessarily a plan that is likely to be drawn, but that wasn't the point. It was just an experiment to see what happens if you draw a Lansing-Saginaw district. Answer: the Flint and Lansing districts are a bit odd, everything else makes sense in isolation but you split a few more counties than you necessarily need to.

A lot of the decisions on the map seem to come down to how Flint, Saginaw and the Thumb get treated. If you put Flint with the Thumb, you're drawing a Republican gerrymander. If you put Flint and Saginaw in the same district and extend it to the Huron coast, you're going to produce a strong Democratic map around Detroit (as you soak up heavily Republican bits of northern Macomb and/or Oakland, allowing Democratic strongholds in the south of those counties to outvote lean-Republican areas further north.) If you put Flint and Saginaw in the same district but let the Thumb district head north via Bay, it's more mixed but tends to benefit Republicans as they're going to be favoured to win the Macomb district.

I never claimed not to have biases, but on this map https://imgur.com/a/pQIkR45 I turned off partisan data and based it on 538's map keeping counties whole as much as possible.  Not every map I've drawn is fair but that one is as the data shows https://imgur.com/7d7Ddl7 6D, 6R, 1 even.    Also I don't think you can get a Flint-Saginaw district all the way to northern Macomb unless you make a weird tendril which the commission isn't doing.  You can get a Flint district to the Huron coast, only if you drop Saginaw.  I think the most fair and straightforward way to divide up the northern suburbs is a solid R district that includes Livingston county, northern and western Oakland, and northern Macomb.  Then a solid D district in the remainder of Oakland, and a tossup in the remainder of Macomb.  The one potential alteration is having a black seat go into southern Oakland to get more black voters, but that doesn't really change the partisan makeup because the Oakland seat still solidly votes Clinton due to Pontiac and places like Royal Oak.  1D-1T-1R is a fair breakdown of the Macomb-Oakland-Livingston area.  Those counties combined voted Trump narrowly and Obama narrowly. 

could you link me the map plz?

https://davesredistricting.org/join/bf7fe969-7260-4079-9fba-713e5eb41629
I looked at presidential results and down-ballot tendencies and it would be like this:
4 safe R (6, 9, 10, 11)
2 lean R (7, 13)
2 tossups (1, 8)
1 lean D (12)
4 safe D (2, 3, 4, 5)
Based on the '12+'16 composite it's 6R-1T-6D
In addition to keeping counties whole, I tried my best to keep metros whole too.  MI has a lot of mid sized metros like Flint, Lansing, Grand Rapids, Kalamazoo, ect which can all have their own districts.

I quite honestly love your map, my only minor complaint relates to the 1st and 9th and the fact the border is not a straight line, but that's not really a major thing by any stretch.

Sorry, I didn't put that clearly - I'm not talking about the Flint-Saginaw district going into Macomb, but about the knock-on consequences. The Thumb has the electorate for half a congressional district and if you don't get the other half from Flint or from the north, then you either have to take it from Oakland or Macomb or both, which creates a Republican pack and means that you're more likely to get Democratic seats in the rest of Oakland and Macomb.

I have  mentioned before that if I personally tend to think thee thumb belongs with the upper chunk of the state, connected through Bay City of course. Now, I didn't always hold this view, I was converted by talking with various residents and reading the discussions I alluded to in the opener. In general, only really St. Clair is seen as potentially being a good partner to various Detroit Metro seats. The rest are seen as culturally more aligned with those on the other side of the Tri-Cities, itself a recognized region of Michigan by most groups which require regional divisions.

Is it possible to better compactness by putting Oscoda and Crawford counties in the 1st in return for Osceola joining the 9th?

https://davesredistricting.org/join/435dcad3-d0f2-435b-abe8-c5eee7aa8575
In this map we have:
a 1st district more along the lines of the 2000s iteration of the district
a Flint-Thumb district, with Saginaw and Midland thrown in a Lansing district
a nice, neat, compact Grand Rapids seat
1 seat nested entirely within Macomb, another nested solely within Oakland, and 2 nested solely with Wayne, and then 2 seats around those, enveloping them
Kalamazoo and Battle Creek are kept in the same seat
the Macomb CD and Flint CD are the tipping point districts

I don't really see the point of Oakland getting it's own district.   It seems...meaningless.  Oakland is huge and very diverse,  both with income and demographics.    It's as though it's making a district simply to follow county lines and literally nothing else. 

It makes the most sense to have southern Oakland cross into either Wayne or Washtenaw,  depending on what communities you want to put together. 

I could understand a Macomb-exclusive district though.   That makes way more sense since it's much more White Working Class and generally is it's own community.

An all-Oakland seat is logical especially because 8 Mile Road is better not crossed if one can help it and/or its not absolutely essential to one's plans elsewhere. The main benefit is not necessarily in a homogenous CoI but better districts elsewhere. There is much reason and much elegance in two exurban districts wrapping around the more urban metro Detroit districts.

The 8 mile rd thing is more for the Macomb-Wayne border than Oakland,  since it's what separates Black Detroit from White Macomb (the difference really is pretty stark).   In Oakland you have Oak Park and Southfield areas to the north of Detroit which are both pretty black, and the difference between the two is minimal further west.

Most of Detroit itself is suburban sprawl, there's extremely little true "Metro" in the sense of what you see in Los Angeles or New York in Detroit.

But you don't have to reach into Oakland for an adequate black district, and the more areas you take from Southern Oakland the more you need to eat into exurban metro Detroit, which is a CoI worth keeping together.
So crossing 8 Mile is still undesirable in most cases.

I like your map, it does split counties more than mine but I'd say respects COIs a bit better.  I see you made the Lansing district more dem but the Flint district is more competitive.  Some might criticize your map due to the tipping point districts being red leaning in 2016, but down ballot Flint and Macomb are more Dem than they are presidentially so those districts would be closer to true tossups.  I like how you fixed my long western MI strip and made more compact districts.  My only criticism of your map is the UP-Bay City district.  I'd trade Traverse City and Bay City between their respective districts.  Other than that, I would support this map as a commissioner. 

Oakland isn't as cohesive as other counties but I don't get how splitting it up better reflects COIs.  The main divide in Oakland is exurban vs suburban, and both my map and the speaker's map keep those areas together.  I do see potential merit in having some black areas in Oakland go to Wayne because they would probably be better represented by Lawrence than Levin; and such a move might be necessary due to population loss in Wayne.  But Dems would probably prefer keeping the county border unbroken, more blacks in Levin's district will make it safer but pushing it out further into the exurbs could endanger him.  I know incumbent interests aren't taken into account, but partisan fairness is, and a safe Dem district in Oakland is fair. 

for what its worth you could shift Bay, Arenac, Iosco, Omegaw, Roscommon, and Gladwin into the Muskegon seat and Grand Traverse, Wexford, Mason, Manistee, Benzie, and Leelanau into the UP district. But this overall arrangement hurts compactness a bit. UP is 58 and Muskegon is 43 - in my original its 59 and 88, respectively.
And in all honesty I'm not sure if UP is necessarily a better fit with any specific area of the LP besides the counties immediately to its south in an around the Mackinac bridge. At least here we have Western Michigan less divided.

Even then,  Detroit is in Northeastern Wayne, and southeastern Oakland is more "Metro" and urban than southern Wayne.    There's exurban metro Detroit in western and southern Wayne, and suburban Detroit in Oakland.  

Anyway you put it the county borders don't really mean much in Detroit,  at least in respect to Oakland and Wayne.  

The current MI-11 was probably drawn as a Suburban district surrounding the AA districts (to the benefit of Republicans at the time), which puts all the communities of interests together, and the current MI-9 is pretty close to the Macomb working class district.

The current MI-09/11 were drawn to to maximize GOP potential in the Metro region. The 9th packs in the near suburbs which at the time were the solid obama parts of the region. The 11th is designed to squiggle around and collect the 'further' suburbs which at the time were more GOP friendly. As people moved further out, as the exurbs got pushed further back, and as attitudes changed, this no longer was a viable dichotomy that made much sense. Instead, the defining divide is now between the well understood 2016 style coalitions. In fact, one could say those coalitions are even more stark here because this is Detroit, the epicenter of the financial crash where it all began. Looking to these districts for guides on....anything is a horrible choice.

County borders aren't perfect but at least ensure relative compactness.  Arms going this way or that in Detroit are really easy to gerrymander with going either way, at least compact sensibly shaped districts limit gerrymandering.  I have seen some really funky Detroit maps here favoring both sides, I think the county based division is the most fair and generally avoids ripping up communities.

I could see a Stevens vs Levin primary happening, since those districts are getting reconfigured big time.  Unless Levin runs in a swing Macomb district, but I think he'll want to stay in Oakland.

Thanks for the history lesson, but that completely misses the point.  

MI-11 was drawn to maximize GOP influence in the district, by incorporating certain groups of people in it, and MI-9 was drawn as a vote sink with other groups of people.  

The districts have changed their voting patterns, but the people are still there generally.  Northwest Wayne has way more in common (today) with parts of southern Oakland than it does with other parts of Wayne, the vice versa is still true today.

Anyway,  I'll drop the point,  doesn't seem anyone here really agrees.

Well, I agree with you on the  topic of wayne parts vs other wayne  parts, I just think the pop and the commission rules favor the later. It also allows the west suburbs to go potentially with Ann Arbor and Washtenaw, which they also share common traits with. Or they could go with other wayne towns like Dearborn and bits of detroit part of a Arab access seat.

While we are  on this topic though, I'm surprised there are not many maps that try to get a non-detroit seat purely in Wayne, something like what occurs below. I guess it's because the numbers are less favorable this cycle when compared to the last when it comes to wayne vs the suburbs. Such a seat would have cascading effects like requiring both minority seats to head north and then push the suburban seats even further north. Livingston couldn't be paired fully with Oakland. Maybe I will play around with it since it will Allow Livonia & Co. to be paired with the near side of Oakland, even though it would be in a AA seat.

()

I would think Romulus and Inkster should be in the AA districts to get their numbers up. 

They likely will be, that's what my map did.  With population decline in Detroit it's possible the black district will need Pontiac too, I'm not sure though.  But with 2010 demographics it isn't necessary.

If you compare the 2010 precincts on DRA with the 2016 ones, it looks like that doesn't really apply to Macomb-Wayne any more. Eastpointe is on the path to becoming black-plurality and the bottom few rows of precincts in Warren have pretty high black populations these days. There's still a divide between 90% black precincts and 40% black ones, but on the west side of Macomb it doesn't become super-white until 9 Mile or 10 Mile Road.


Has anybody successfully made 2 VRA districts solely in Wayne in this thread? It was certainly doable in 2010, I'm not convinced it is any longer.

I'm also not convinced it would be that controversial on a partisan level. You have to put a lot of SE Oakland into a Detroit district before a district entirely in Oakland starts looking competitive.

This depends entirely upon how we define AA seats in this context. The 2010 GOP believed in packing said seats to the brim with as many AA voters would be allowed before the map ran became an example of racial packing. In this regard, they all passed 50%. Now, it is still possible to get two seats out of the entire metro above 50% if you use the 2016 data and ALL areas of AA concentration. There is even enough leeway in the 2016 data for it to still work in 2020 after four more years of pop decline.

However, it is unlikely the commission takes this approach since it would require the destruction of almost every other Metro COI in favor of the AA community. Instead, we should perhaps look to CA, where we got Asian access seats like CA39 and two AA access seats in west LA. All three seats have the relevant demographic far below 50%.

So then we should lower our criteria. Is it AA plurality? Can it be lower, to something like AA greater than 40%? Do the other voter groups matter? Is a white republican who will not vote in the AA dominated dem primary equal to one who will? How one defines an AA seat will  decide how much of near Oakland, if any at all, is required to make the seats pass the minimum criteria.

I'd argue it would need to be high 40s, which is possible without weird shaped districts.  In the 2012 election Gary Peters was able to win a primary with a plurality against multiple non white candidates.  Even with a black majority seat there is not a guarantee it will elect the black candidate (Peters likely won due to the Oakland County portion of the district which isn't as black, due to the odd nature of the race it's likely Peters did not win a majority of the black vote).   Due to this fact the vra districts should try to get as close to 50% aa as possible with compact districts.  Since it has been demonstrated a 58% black district in Detroit can "fail" under the right circumstances, a 40% black district would be FAR more likely to not elect a black candidate. 

Some thoughts on incumbents:

-John Moolenaar and Dan Kildee live not too far from each other. They could potentially be in a district together.
-Elissa Slotkin wil almost certainly have her hometown of Holly be put into a Solidly R district. She’d probably have to run to Lansing to be viable.
-Andy Levin is also in deep trouble. Assuming he gets a district entirely in Macomb County, he’ll be a heavy underdog. His opponent will most likely be the next MI-10 rep.
-All of the major MI-3 candidates are from the Grand Rapids area which means it’s clear what district they’ll be running in in 2022.
-Fred Upton will almost certainly retire in 2022 if he doesn’t do so this year.

https://davesredistricting.org/join/f9f2fff2-6ca3-4bd6-a202-511f71a26636
This is my take on this concept. Two black seats forced north into Oakland > rest of Oakland+Macomb CD makes Bay City-Thumb district inevitable > Flint+Saginaw seat formed and whole county district formed in SE MI east of Wayne > Livingston threw in with Lansing metro and Calhoun added for partisan balance reasons and Grand Rapids-Muskegon CD also created for partisan balance reasons > the rest of map draws itself.
(ignore the district numbers - assume they'd be numbered in a continuity-driven matter)

https://davesredistricting.org/join/b99a88d3-23fe-4d86-a4e3-8a237a3424da
another map, the aim here was (unsuccessfully) a nice, compact, whole county CD with Washtenaw in it, and for the all-of-southern Wayne district to move south if so necessary. I compromised and then the rest of the map basically fell into place into place. On a sidenote: this makes the Muskegon-Grand Rapids more natural looking on the map, due to Ottawa having to pair with Michiana as opposed to areas north and east of Grand Rapids. Overall the Washtenaw CD becomes more marginal, going down to D+2. The Oakland district's partisanship is barely affected despite its losing its tiny share of Macomb, as Southern Oakland is paired with southern Macomb. The Thumb-Bay City district loses Midland as it is pushed further south, forcing Saginaw to separate from Flint due to the need to avoid Midland, Saginaw, and Bay City all ending up in different districts. Flint's district is forced south into Metro Lansing, and the Lansing CD switches out Battle Creek for Kalamazoo. Saginaw and Midland meanwhile are thrown in with a wide area of rural Michigan.
This map is a bit more D-favorable and an argument could be made it is D-leaning.

https://davesredistricting.org/join/945fa102-1e26-4ee0-859a-1fbe56ec6af2 I made this based on your map, but united Ottawa and Kent instead.  The shifts is causes are particularly noticeable in Detroit, including a swing seat nearly all in Wayne, which I didn't know was possible.  I'd say this is more fair, but dems have big potential if they can keep improving in the suburbs since there are 3 suburban swing seats.

Interestingly this only further highlights the rotation involved.
Calhoun and Branch taken from Washtenaw CD forces it to take in the rest of Washtenaw and most of Monroe > southern Wayne CD takes in Livonia and other areas > NW Detroit CD takes in SE Oakland > almost-all-Macomb CD is created > Bay-Thumb CD takes in Saginaw > Midland CD takes in Muskegon > Grand Rapids CD has to take in most of Ottawa County.
I would like to assume that the double splitting of both Jackson and Hillsdale counties was a mistake - that's not something permitted under Michigan law with the sole exception of VRA compliance (something that MI Rs used in 2011).

https://davesredistricting.org/join/dc4df638-b30d-48ba-bbf2-4fdc6fd78031
competitive district map.  8/13 districts are competitive ('16 margin 10 points or less for either candidate) with 5 being highly competitive.  Of course if the state continues to trend red like it did last decade this turns into an R gerrymander.  I like how I made a tossup entirely within Wayne and 2 tossups in central MI. 7 Trump 6 Clinton

Two versions of possible maps for the Michigan State Senate, based on 2018 population estimates:

https://davesredistricting.org/join/04f91558-5287-438a-8efe-aa9246cbef8a

https://davesredistricting.org/join/17d7f5ff-be35-4006-95c0-9cf51d2c0e52

They're pretty similar, with the differences limited to the Tri-Cities area where I couldn't quite decide which option was superior. As I'm depending on estimates, I tried to avoid having seats right at the top or bottom of the population deviation range, unless it was unavoidable. Although the 2016 block groups I drew it with don't always respect township boundaries, in actual fact the only municipalities I had to split were Detroit, Dearborn Heights (avoidable if you're willing to use touchpoint-contiguity) and Sterling Heights (I couldn't see a way to avoid it, but it may exist.)

Highlights worth pointing out:

• On 2020 numbers, Wayne County could just about have 7 districts but in practice it isn't really feasible now and it won't be at all by 2020
• Wayne and Macomb combined are entitled to almost exactly 10 Senate districts, which is a better fit than you can get with Wayne and any other one county. I acknowledge the problems with crossing 8 Mile Road, but it makes it a lot easier to maintain 5 VRA districts in Detroit.
• I think the districts I drew in Oakland are rather neat, except for the somewhat ugly pairing of Troy and Pontiac (which I think was just about avoiding splitting municipalities.) Nevertheless, they are pretty decent lines for Democrats and would undoubtedly be controversial, especially if the final census numbers do allow alternate configurations. District 11 is drawn as a black-opportunity district.
• Livingston and Washtenaw are now two large to share two senate districts, but Washtenaw plus Monroe works perfectly. Yes, this has a definite partisan effect, but I don't see a whole-county alternative. Swapping Ypsilanti for rural Washtenaw would probably make Republicans happy, but would look ugly (though maybe not worse than my district 22?)
• I don't love my district 19, so if you're looking to avoid a Washtenaw-Monroe combination, hopefully the solution would improve that too.
• I tried to draw a majority-minority Grand Rapids district, but the numbers aren't there for it.
• It's not very competitive right now, but back in 2012 district 34 would have been great for Democrats. Incidentally, does anybody know why Lake County was so Democratic up until 2016? It really shows up on a map, but I have no idea what prompts it

In a neutral year, I suspect this would be something along the lines of 22-16 or 21-17 Republican, but a Democratic majority looks like a very hard ask - it would probably entail running the table in the Detroit metro, locking down the Saginaw-Bay district, winning the Muskegon district and somehow getting over the line in either the Jackson district or the UP district.

If you're trying to draw a map that doesn't guarantee a GOP majority, you should be able to create a Lean R (and trending D) seat and a Safe R seat (instead of two Safe R seats) out of districts 14 and 15 by concentrating one seat in the suburbs (Novi to the areas west of Pontiac) and the other in the exurbs.

Anyway here's the results of my investigation into the Wayne non-AA seat style map. Like the others upthread discovered, this leads to the rotation of districts around that state. Now I didn't intend for it to become a D-Gerry/D-Favoring map at the start. However, once I saw how the Detroit Metro seats (first drawn) basically favored the Democrats to an unfair degree, I kinda made that unfairness part of the map. I do not endorse this plan, and I think it should show why an anchored Wayne  non-AA seat is probably asking the data to provide something that isn't available.

()

()


Districts 11, 12, 13 are Safe D (The Wayne Seat is right there is a D gift)

District 9 is probably Likely D (Clinton+10, D+4 CPVI, Sunk Livingston is growing just like Ann Arbor)

Districts 3 and 10 are marginal Clinton with D+1 CPVI's, but they are both moving towards the Dems

District 5 is marginal Clinton with D+4 CPVI, but it is moving towards the GOP (yep Clinton won a majority of seats...)

District 7 is marginal Trump with a R+1 CPVI, but moving towards the Dems

Districts 1, 2, 4, 6, and 8 are safe GOP, all with CPVI's above R+8 and an average Trump win of 58% - 37% (one seat a bit above that trump win, one seat a bit below that).

An all Wayne non AA seat doesn't have to cause the map to favor dems tho.  Having the Ann Arbor district go into southern Oakland instead of red exurban/rurals could be done. 

how do you put photos directly in?  When I put them in it on ImgBB just becomes a link

Have you tried this?
- right-click on image
- hit the 'open image in new tab' button
- there'll be a new link, use that

THANKS

()
()
()
This proves a non AA Wayne seat doesn't have to necessarily help Dems.  Besides there's an argument Ann Arbor does better paired with educated parts of Oakland rather than western Wayne.

()
https://davesredistricting.org/join/ee18ca89-11d7-4635-9684-3dfb273592b1 (https://davesredistricting.org/join/ee18ca89-11d7-4635-9684-3dfb273592b1)
1 - R+11
2 - R+12
3 - D+5
4 - D+6
5 - R+15
6 - D+6
7 - D+5
8 - R+14
9 - D+6
10 - D+6
11 - D+6
12 - D+14 (47% Black)
13 - D+11 (47% Black)

Michigan's House of Representatives map is much less gerrymandered than its congressional or State Senate maps, because at this scale minimising splits of counties and municipalities means there's much less room to put a thumb on the scales. There are areas where it's clearly drawn to favour Republicans, but equally there are plenty of areas where you can more or less maintain the present alignments, once you've accounted for population shifts.

Once again, I worked with 2018 population estimates and tried to make sure that as much as possible seats were not right at the upper or lower limits of the allowable range. I tried to keep similar areas together, which probably promotes more safe seats, and I tried to draw as many seats winnable by minority candidates as possible, which definitely promotes safe seats. However, there are still a fair few marginals and I think it's clear that the House is a much easier target for Michigan Democrats than the State Senate.

In a few places (primarily Macomb, Grand Rapids, Lansing and Jackson) there looked to be two viable solutions, so I drew two maps. This is the first:

https://davesredistricting.org/join/90a2f6be-8fcf-4293-9b92-1a6b6a9c3d4b

• This map has 10 black majority districts (8 in Detroit, one in Southfield, one in Flint); 1 Hispanic plurality district in Detroit (although that's probably more likely to elect a black candidate as it stands); a black-plurality district based on Pontiac; a majority-minority district in Grand Rapids; two districts where the Democratic primary is probably plurality-black (the Inkster-Romulus district and the Saginaw City district) and several other districts where the white percentage is below 60% and which might be viable coalition districts either now or by 2030. Except for the Inkster-Romulus district, these are all pretty clean. Sadly, the latest update to DRA seems to have screwed up the display of electorate data and is now only showing white and Hispanic populations, so you may just have to take me at my word.
• Wayne and Monroe combine for 21 districts, which is one fewer than they presently have. It's mathematically possible to give Wayne 19 large districts on its own, but then Monroe lacks a convenient partner and there's nothing terribly wrong with a cross-county district there so I left it be.
• In Macomb, I went for a least-change plan as much as possible. Warren and Clinton need to be chopped due to size and there's an extra chop in Sterling Heights. Macomb Township is now large enough for its own district.
• In Oakland, you can mostly leave things be in the south-east. I tried to create a black-opportunity district based on Oak Park, but the numbers aren't quite there. Rochester Hills is now large enough for its own district, but Troy and Clawson are too large so Troy gets a chop. The county could theoretically stand alone for 14 seats, but Livingston is too large for two so I stuck in a cross-county seat.
• In St. Clair, I undid the current gerrymander to create one rural district shared with Sanilac and one district focused on the St. Clair river towns. This might have been a swing district in 2008, but isn't now.
• At the moment, Jackson County is sliced and diced to try to dilute the votes of Jackson itself. This map undoes that and creates a swing seat. Similarly, I've simplified the lines in Calhoun County. On the flipside, Shiawassee County gets carved in two.
• Kent County is cracked quite hard at the moment. I've undone and given it seven seats of its own, which ought to usually split 4-3 Republican, with the Wyoming district possibly on the path to becoming competitive.
  
• I tried a lot of options, but I don't think there are any plausible solutions in the UP that let you keep Marquette whole. That makes it even harder for Democrats to hold down a seat there than it already is.

Overall, Clinton won 42 districts to Trump's 68, but several of Trump's victories were extremely narrow and in several more Democrats tend to overperform downballot, so it would be competitive in good Democratic years.

I will post the second map either later tonight or tomorrow.

Wow, interesting how a fair house map favors Rs so much.  Rs should be thankful for the rules which limited them last time.

() (https://imgbb.com/)
https://davesredistricting.org/join/1a8fff1c-a5dc-4bf1-82aa-cd22367f05c1
my take on a "Washtenaw paired with Southern Oakland" arrangement paired with a Wayne non-AA seat. I tried to make it balanced but Rs still have a verrry slight advantage on this map.
Two black seats were possible even when only one of them crossed into southern Oakland AND Macomb has a CD to itself starting north from 8 Mile Road.
Clinton won the Washtenaw-southern Oakland seat by 24 points in 2016.
Dem-leaning seat can be created from northern Oakland and Genesee, but it's not safe.
Multiple whole county CDs can be formed.
Even with the inclusion of Saginaw, the Bay-Thumb district is very hard for Dems, measuring up to be R+9.
Only 4 Clinton districts. Ouch. The Washtenaw-southern Oakland seat reallly hurts here.
Tipping point seats are Lansing (R+1.71) and Macomb (R+1.93). Lansing is most likely to be marginal here due to Livingston having to end up somewhere. Lansing is the most natural place for it to go.

()
9-4 map with Trump winning all 9 by at least 10.  Amazing how it doesn't take an egregious map.  I can get a better map out of MI than TX and it looks better.  TX is 8-9 pts to the right of MI.

This is the second map I drew: https://davesredistricting.org/join/a2f88c52-cda9-4396-bc78-2c16f33c5600

Changes from the first map:

• The Inkster district (16) is significantly more compact and there are two fewer municipality cuts as a result. This drops the black percentage from 40% to 33%. However, that's still probably an improvement from what it is at present and it's represented by a black candidate, so it's probably functioning. The partisan ramifications are pretty limited.
• There's a totally different arrangement in Macomb County. Whereas in the first map I left the 18th district unchanged as a combination of Eastpointe and St. Clair Shores, in this one I combined Eastpointe with southern Warren in an attempt to create a black opportunity district. It's only 34% black at the moment, but diversifying fast and probably already black-majority in the primary.
• The knock-on consequences mean more compact lines in northern Macomb and one less split to Sterling Heights, but Shelby township gets cut and there's also a small nick in to Clinton Township (although that may be avoidable, depending on population shifts.) I probably prefer this option, but it is more disruptive.
• Rather than pairing Eaton and Clinton for two districts, I stuck Eaton in with Shiawassee and Genesee and Clinton in with the Lansing group. This flips district 71 and arguably better respects community identities round Lansing, but it's uglier in Jackson.
  
• There's a slight re-jig in Kent in an attempt to create a purely suburban district wrapping round Grand Rapids to the north and east. Only partially successful, I think it's fair to say.
• Slightly re-jigged lines in Ottawa, switching from a East/South/West orientation to an East/South-West/North-West one. Also minor tweaks to Muskegon, no great partisan effect in either case.
• A slight re-jig to the lines in Saginaw, making district 95 slightly more effective as a black opportunity district.

()

Here's a Detroit metro arrangement I like. Two districts over 50% AA and each of Wayne, Oakland, and Macomb gets its own non-VRA district. Of course, it'd result in an Ann Arbor/Rural pairing and no Flint to Saginaw Bay district like the current CD 5.

What does the rest of the map look like? Presumably the numbers would work for putting Livingston in with Washtenaw, which is a better fit than more rural counties.

looks good

()
I made a map based on yours.  Not ideal for either party but favors Rs a bit.  Livingston county is challenging, pairing it with Flint or Lansing is definitely unfavorable to Dems.  I'd say pairing Livingston with exurban Oakland is the most sensible (and fair) way to draw the map, but then you might not get a 100% Oakland district.   

()

My latest version, which solves the Livingston problem by pairing with Washtenaw and Oakland in a coherent D+1.66 district that is quickly trending more D.  Oakland, Macomb and Wayne each have a district entirely within them.  The-Oakland only district is D+0.76 but Clinton won by 5.  The white parts of Wayne are reasonably paired with Monroe; it's basically a tossup district (DRA says R+0.61).  Flint/Saginaw and Lansing/Isabella districts are D+3 and D+0.15 but Trump very narrowly won both. 

I'd say good map but Rs will hate all those suburban swing seats and Dems will be sweating since  Clinton won only 4 seats!  Of course it is a citizen's commission.  If they prioritize competitiveness we could see something like this.  But I predict they'll focus more on keeping boundaries unbroken and COIs.

Well given the political parties have literally no input into the commission it doesn't really matter what they think. It'll be whatever 13 random citizens with no political links who are probably lawyers think best fits the criteria they follow.
And it should be noted that respecting COIs is completely different from keeping Counties intact. And the commission is bound by law to prioritise respecting COIs and competitive elections over county and municipality integrity.

But in the absence of other COIs, a county works just fine. In the past, true commissions were limited to the west, where huge, unnatural, counties are destined to be cut. Michigan on the other hand has more rational county breakdowns. At the congressional level with only 13 seats, it's very likely COI discussions will focus on which counties should be paired with which other counties (Saginaw+Flint for instance) - except in several obvious situations usually within one of the metro areas. On lower level maps these cross-county COIs will be more imperative.

This will probably be my final take on it -

()

()

https://davesredistricting.org/join/86c3575b-47c5-455b-bdef-ba4557967108

6 Clinton seats, 7 Trump seats.   Seat in Macomb and Oakland, 2 AA seats.  Keeps most communities together.   Overall a good fit.

MI-5 (Flint/Saginaw) won by Clinton but trending R.

MI-3 (Grand Rapids) won by Trump but trending D.

I'd expect those two to flip by 2022 or maybe 2024.

The Oakland district was barely won by Clinton but it's trending D too.

Other than that the rest are pretty much all safe for either party.

Oakland to Newaygo and Wayne to St. Joseph just aren't good districts. 

I particularly like my latest map because the district shapes are incredibly compact -- 3 of them are basically perfect rectangles, and the Flint, Washtenaw and Lansing districts are really good shapes as well (in addition to being decent COIs).  I might swap Newaygo for Barry in the Grand Rapids district; that's the only real annoyance I have. 

The Oakland to Newaygo district is actually a really good collection of all the Metro's Exurban areas around them, from Grand Rapids to Detroit.   They actually do have a lot in common.   It isn't compact sure, but if CoI > Compactness then it's a very solid district.

()
There's a lot of ways you could draw the state outside of the metro; here I wanted to get the whole Lansing area in one district. You could swap territory between 7 and 8 and get a Livingston/Washtenaw/Monroe district that doesn't look too bad. Not great from a COI standpoint either IMO but it's a hard area to deal with.

I do agree with Idaho Conservative that the best fit for Livingston is probably pairing it with exurban Oakland, but that requires a different arrangement for most of the metro. I've also been playing with Washtenaw/Ingham, though that does sort of tilt the rest of the map in Republicans' favor.

While Livingston being paired with Lansing isn't ideal, your map is better than most I've seen here due to compact.  The three counties which make up Lansing are a COI and that's good they were kept together.  I would also say Washtenaw is a better match for Wayne but again, I get that changes stuff elsewhere.

()
Competitive map, I guess;

12 is 50.3% black, 13 is 49.8% black. Both, along with 7 (Clinton +8.1) and 8 (Clinton +13.3%) are Safe D.

9 is Clinton +7.2, so safe for now, but will probably shift closer through the decade. Likely D.

10 is Clinton +1. Tossup/Lean D.

3 is Trump +4.4 and 6 is Trump +6. Lean R. 6 was probably won by Whitmer; 3 definitely.

The rest (1,2,4,5,11) are safe R, with the closest being 11 at Trump +19.

Seems like it would be 7R, 6D but with the realistic chances of going 8R-5D or 8D-5R.

I just drew a new version of a MI map, and I'm extremely satisfied with it. The goal was to keep very close to the rule of minimizing county and municipal splits while keeping COIs together, and I think this map does an excellent job of both while also coming out with a very fair partisan balance of 6-6-1. It also has two seats that are majority or nearly majority black (and you could fiddle around with the edges to get them both to majority black, such as by switching Hamtramck). The UP-Traverse City district has no county splits at all (!!!), and the Lansing and Flint-Saginaw districts share a de minimis county split but have no splits with any other districts. The Detroit metro has two majority/near-majority black Detroit+ districts, one district entirely in Wayne, one district entirely in Oakland, one district covering the exurban parts of Oakland and Macomb and one district containing the western/southern exurbs+Ann Arbor, which I think is the best possible COI arrangement in the Detroit area. The three counties of the Lansing metro are also kept together without combining them with any other major metro (no combination with Flint, Saginaw, Livingston County, Kalamazoo, etc.)

Without further ado, the map:

()
()

MI-01 (Upper Peninsula, Traverse City): 37-59 Trump, Safe R
MI-02 (Holland, Muskegon): 37-57 Trump, Safe R
MI-03 (Grand Rapids): 41-52 Trump, Likely R
MI-04 (Bay City, Port Huron): 32-63 Trump, Safe R
MI-05 (Kalamazoo, Benton Harbor): 42-52 Trump, Likely R
MI-06 (Lansing, Battle Creek, Jackson): 48-46 Clinton, Lean D
MI-07 (Flint, Saginaw, Midland): 48-47 Clinton, Toss-up
MI-08 (Ann Arbor, Monroe, Howell): 50-45 Clinton, Likely D
MI-09 (Pontiac, Royal Oak, Novi): 53-42 Clinton, Safe D
MI-10 (Oxford, Sterling Heights, New Baltimore): 36-59 Trump, Safe R
MI-11 (Dearborn, Livonia, Romulus): 51-44 Clinton, Likely D
MI-12 (Detroit West, Southfield, Inkster): 78-19 Clinton, Safe D (51% black)
MI-13 (Detroit East, Warren, Mount Clemens): 69-28 Clinton, Safe D (49% black)

Total: 6R (4 Safe, 2 Likely), 6D (3 Safe, 2 Likely, 1 Lean), 1 Toss-up

https://davesredistricting.org/join/28069751-857c-44d0-8e44-8abde1dc2f25

Edit: I realized I had an unintentional extra municipal split in Macomb County. I've fixed that on the DRA version, and it actually increased the black percentage in MI-13 just a hair (nearly to the point of rounding up to 50%!).

()
Optimal map for metro Detroit
Yellow: Clinton+56  (45% black VRA seat)
Blue: Clinton+42   (47% black VRA seat)
Red: Clinton+26
Pink: Clinton+13
Green: Trump+20
Purple: Trump+16

Districts are compact, follow VRA, communities of interest are largely kept together.  Also a 4D-2R breakdown is fair for the metro with 2 districts having potential to be competitive in a wave election (purple and pink).

Should the tri-cities area be considered a COI?

I think that's probably the best effort I've seen yet in this thread.

I particularly like that MI-2 - it's very hard to make a doughnut seat that doesn't look incredibly awkward, but you pulled it off and it's a good Grand Rapids exurbs seat.

Thanks! I will say it's definitely not necessary to draw the Grand Rapids area that way, though I personally think this is the best COI version because it keeps the Grand Rapids urban+suburban areas together with also keeping rural/small city western Michigan together. I often like donut maps because they allow you to keep a clear separation of urban/suburban areas from rural areas, which have very different interests.

I also like most of the map, though I personally think Donuts don't really work on the congressional level. That said, another advantage of the map is CD's 2 and 3 are in a 'closed ecosystem' so if you have different goals for the seat you can adjust as needed. If you want another cooperative (GOP leaning) seat to counteract the weaker Clinton seats in the east, you could do Muskegon+Newaygo +Grand  Rapids part of Kent. If you just don't like Donuts you could put all of Ottawa or Montcalm+Ionia+Gratiot and then cut away the redder parts of Kent to compensate. It's a great map and template to base future work off of.

I had drawn Muskegon+Newaygo+Mecosta+City of Grand Rapids as an alternative (definitely favorable to the Democrats) map. That version actually voted for Clinton by a few tenths of a percent. But, agreed, really it's just a matter of fiddling around with those two districts if you don't like the donut.

()

Slight tweak to get that MI-13 to 49.9% black: strip out St. Clair Shores and the remaining bit of Harrison Township, add in all of Clinton Township bar five precincts in the north-west. It makes the map look a little uglier, but it splits fewer municipalities and it doesn't really look worse on a map than the Dearborn/Dearborn Heights split does.

I got it to 50% by cutting into Clinton Township along the west instead of the northwest (as the NW corner actually has a fairly high black population).

As an aside, it's shocking how Republican St Clair Shores is. In some precincts, the white vote has to be nearly 80% Trump given the black population, which is almost unheard of for a suburb outside of the Deep South.

yeah weird doughnut districts should be avoided.  Livingston should be paired with North and western Oakland, not Ann Arbor.  Also there should be more of an attempt to make a tri cities district.  This means Flint can be paired with the Thumb or Lansing.  I'd say the Thumb makes more sense overall but pairing it with Lansing created a D district.  Otherwise there won't be a Dem seat out of Detroit, just tossups.

()

I tried my best to satisfy the important COI considerations that have been discussed.

1: UP and northern LP
2: Northern Lake Michigan coast and interior
3: Grand Rapids and relevant Ottawa County suburbs
4: Lansing metro area and surrounding rural areas
5: Genesee County and the tri-cities
6: Southwestern MI, with Kalamazoo and Battle Creek
7: South Central MI and right-trending Wayne suburbs
8: Livingston and Oakland exurbs/outer suburbs
9: Oakland and Macomb inner/middle suburbs
10: Thumb and exurban Macomb
11: Most of Detroit and nearby suburbs
12: Washtenaw and upscale, left-trending western Wayne
13: Western Detroit and blackest parts of Oakland

Districts 4, 5, and 9 would be highly competitive.

The 7th is sort of the "leftovers" district, but I like the arrangement it allows for the 12th and it's pretty uniformly right-trending areas, at least. You could alternatively put the remainder of Wayne with Monroe (almost exactly a full district) and get another competitive district, but that means putting Ann Arbor with rural areas and that's something I wanted to avoid.

And this my friends is what you get if you try to preserve the gerrymandered lines as best as possible. I applaud your efforts to update them, but they are still tainted.

()
()
2012/2016 composite data

1 Detroit Exurbs, Livingston+outer Oakland
2 WWC suburbs, Macomb
3 Highly Educated suburbs, southeast Oakland+northwest Wayne
4 VRA Urban Detroit
5 VRA Inner suburbs
6 Lansing, Mount Pleasant, and rural central MI
7 UP and northern mitten, Traverse City
8 Southwest Coast, Muskegon and Holland
9 Ann Arbor, southern Wayne, Monroe County
10 Grand Rapids and eastern Ottowa County
11 Flint and thumb
12 Tri Cities area, rural central and western MI
13 Kalamazoo, Battle Creek, Jackson. rural southern MI

I like how I keep most COIs together throughout the state. Flint gets its own district without competing with another city, Tri Cities get their own district, Lansing metro is kept whole and the Northern suburbs of Detroit don't get ripped apart by being put with urban or rural areas.  I get that the thin strip in district 8 is ugly, but after the 2010 census it will be thicker, taking up the western half of Ottawa County.  I've seen too many maps doing weird stuff like doughnuts, half of Macomb with Detroit, cutting Oakland county 10 ways 'til Sunday, pairing Flint with the Detroit suburbs, an Oakland county to western MI seat YUCK!.  I see way more concern for the exact partisan makeup of the seats and less for MI communities.  

"Flint gets its own district" is a particularly euphemistic way of describing a fairly obvious attempt at cracking.

why is half this thread just idaho conservative posting gerrymanders/praising himself for “keeping COIs together”

Imagine seriously defending the idea that Saginaw has more of a COI with the west coast of Michigan than with Flint.

I think more than anything what this thread really needs is a definition of what exactly the "CoI's" in Michigan really are supposed to be. 

Because looking over the maps here they seem to be all over the place.

Michigan has changed so much in the last decade that a 2010 gerrymander looks nothing like a 2020 one. The biggest similarity I see here- the 5/9/10 arrangement- was advantageous for the GOP last decade but is likely now Dem-favoring. The Flint district is drawn about as favorably for Dems as it can be (it could easily be split into two double digit Trump districts via the Thumb), and doing so limits the ways 9 and 10 can be compactly drawn, so that's why that portion ends up looking similar. Besides that, many of these districts have a significantly different character than the current ones.

Overall, I think it ends up being pretty fair from a partisan standpoint, at least relative to a reasonable gerrymander for either side.

The constitution prioritizes the criteria:

(1) Equal Population.
(2) Contiguity
(3) COI " Districts shall reflect the state's diverse population and communities of interest. Communities of interest may include, but shall not be limited to, populations that share cultural or historical characteristics or economic interests."
(4) Political fairness.
(5) Not favoring/disfavoring incumbent or candidate.
(6) Reflect consideration of county, city, and township boundaries.
(7) Reasonably compact.

I would read that as requiring political boundaries to be ignored if necessary to reflect communities of interest.

COI and fairness are contradictory. If it is a community of interest, it will have political coherence.

Here's a rough attempt at defining COIs. First group are county groupings that should never be split apart because they form a clear combined COI. There are only four of these, standing alone:

Detroit Metro (Wayne, Monroe, Macomb, Oakland, Washtenaw, Livingston): There should be six seats entirely contained in these six counties with no overlap with the rest of the state (other than needing to pull in just a few thousand voters from somewhere else on the 2016 numbers)

Lansing Metro (Ingham, Eaton, Clinton): These counties should all be in the same district and treated as one county.

The Thumb (Saint Clair, Lapeer, Tuscola, Sanilac, Huron): These counties should all be in the same district and treated as one county.

The Upper Peninsula: These counties should all be in the same district and treated as one county (this maybe goes without saying, haven't seen anyone propose splitting it up).

After those groupings, it becomes a bit harder because you're talking about metro areas that are mostly self-contained in one county so could go a few different directions. But there are still obvious groups, like Genessee with Saginaw, Calhoun with Jackson (which forms almost exactly one district with the Lansing metro described above so is ideal; also see my map that shows that this grouping with Genessee/Saginaw/Shiawassee/Midland is a perfect pair of two districts, population-wise, needing only one de minimis split), Kent probably best paired with eastern Ottawa to keep Grand Rapids suburbs with the city, etc.

Not with western MI, with the tri-cities area.  Saginaw-Flint breaks up the tri-cities area.

No, cracking would be splitting Gennesee county.  Saginaw and Flint being in different districts is not cracking, they are not a metro or COI.  The tri cities are much more of a community than Flint and Saginaw are.

What's so special about that grouping in particular? They don't have anything more in common with each other than Saginaw has with Flint. Less, generally, as there are significant demographic differences. For example, Bay City and Midland are nearly 100% white while Saginaw has a large black population. Saginaw is also historically an industrial city while Bay City is a shipping center and Midland is more high tech and services.

because you don't

https://en.wikipedia.org/wiki/Saginaw,_Midland,_and_Bay_City_metropolitan_area
Tri cities are a region of MI, Flint and Saginaw are two different cities that both happen to have black people.  Race doesn't make a COI, but it matters for VRA purposes in Detroit.

It's explicitly not a region based on that link. Midland, Saginaw and Bay City are all separate MSAs. The "Central Michigan" region as defined by the state of Michigan also includes Flint as well as Mount Pleasant and some rural counties.

The recognized tri-cities community is Saginaw-Midland-Bay City, but most groupers also throw in Flint because of the regions economic ties. They are all oriented along route 85, and all are at least somewhat postindustrial. They all are distinct from the rural thumb, whose most similar cousins are across the bay in upper Michigan (all residents I have chat with want something like this), all distinct from the universities to their west, and distinct from the Detroit suburbs to their south.

The point of keeping counties whole is that counties are the default COI. If there is no better or clearer COI, the the county level is best observed. If there is a better COI, than it comes before the county. If there are lots of counties like in Michigan, you get cross-county COIs that deserve the same respect as inter-county ones. We have cases here where those outside of Wayne have clear cross-county COIs: rural Thumb+Upland, the route 85 tri-cities corridor, the central universities, and the Wayne exurbs.

Well it's not a metro but a region which is a CSA.  Not all CSAs can be kept together but this can.  Tri cities are more a region than Flint and Saginaw.  The tri cities have economic ties and share an airport.

Well the tri cities and Flint can't be together.  You can do Tri Cities or Flint-Saginaw.  If Flint isn't paired with Saginaw it can go with Lansing or the Thumb.

Let's look at the actual criteria for COI:


The existence of minority communities can fairly clearly be argued to be a cultural characteristic (and to some extent also reflects historic economic patterns.) Historically, Genesee and the bulk of Saginaw's population have been in the same congressional district since 1992 and Saginaw and Midland haven't been in the same district since at least 1972. And yes, post-industrial cities rapidly losing population have more in common with each economically other than they do with rural areas with no industrial heritage whatsoever.

I would also note that it's not even an either/or whether Saginaw goes with Genesee or with Bay and Midland. Based on 2018 numbers, that four county group has about 2% too many people, which can easily be dealt with by cutting out western Midland or northern Bay, neither or which are obviously out of place in a northern Michigan district. Your argument actually depends upon the notion that the Thumb can't go with exurban Macomb and/or Oakland, which is not a proposition you've actually made coherently yet.

Most Clear COIs:
6 county Detroit area
3 county Lansing area
UP
Thumb
Flint
Grand Rapids and suburbs in eastern Ottawa county

Other COIs that should be respected if possible:
Macomb
Livingston+exurban Oakland
Tri-Cities Area
Ann Arbor+college educated communities in western Wayne
Arab communities in Wayne
Kalamazoo-Battle Creek
Huron Coast

And Lansing and the thumb have better partners than Flint. Remember how I said that the rural, Lakeshore oriented, thumb is best paired with the upstate. How it is something every Michigan resident I have consulted with agrees to? Hell, Even
 Dave Wasserman in a hypothetical map linked the two. Well, we cannot link the two via water across the bay. Therefore, going through Bay City is the easiest solution. Guess what? Flint + Saginaw + Midland is a viable cd, with a bit more tacked on of course.

I thought it had been settled metro Detroit was a COI.  Exurban Detroit with the Thumb makes no sense.  The thumb would go better with Flint, tri cities, or the rest of the coast.  Detroit is its own thing.

Detroit is its own thing. The Detroit metropolitan area is a thing. I see no evidence that's the same thing as the northern townships of Oakland and Macomb, which aren't urbanised to any significant degree. The fact that Ray township is in the same county as Warren doesn't mean it is actually meaningfully like Warren.

FWIW, I think the Thumb probably goes best with northern Michigan via Bay (I might even be tempted to cut out Bay City into the Flint/Tri-Cities district and jump the Saginaw River, though I doubt a commission would go for that.) But I see no reason to get precious about breaching the St. Clair-Macomb boundary or the Oakland-Lapeer boundary if you're not going to reach down into the actual urban area itself.

()
()
2012/2016 composite

Well I listened to complaints and made adjustments, now there's a Flint-Saginaw-Midland district which votes Dem in the last 2 elections, albeit narrowly in 2016.  Now the Thumb is part of a Huron Coast district.  I still prefer the previous map, but this one would be more passable being 7-6. 

For what it's worth, you appear to be using 2010 census figures (thus the larger Grand Rapids-based district than my map, e.g., and the bigger split of Lenawee), which may be skewing the map in other ways. 2016 estimates are not the same as what will be reported in the 2020 census, of course, but they should be closer to 2020 than 2010 figures would be.

Hell, I'm not even sure why you are using the 2010 module for Michigan - the 2016 one uniquely has 2016 election data.

()
my map adjusted for 2016 population estimates.  GR area looks better.

Michigan does not have party registration.

Why are suggesting looking at Arizona?

I don't understand the part about lawyers and additional degrees.

Explain?

To get to 6 districts in SE Michigan you have a choice between adding Genesee or Washtenaw.

Ypsilanti is tied to Detroit as much as it is Ann Arbor. Going north you are in clearly exurban low populated areas before you get to Flint.

Why can't you go across Saginaw Bay?

"Districts shall be geographically contiguous. Island areas are considered to be contiguous by land to the county of which they are a part."

Doesn't this say that counties encompass all offshore waters, including any islands.

Districts should be contiguous unless they physically can't be (like the UP).  I think the vast majority of people would agree.

In the 19th Century, the two northern Michigan districts were Huron+Superior shoreline and Michigan shoreline, both UP and LP.

I agree in theory, and I don't think this is a completely crazy idea as it does keep genuinely rural areas together and separate from cities, but I think it would meet with a fair amount of resistance.

Genesee+Tri-Cities counties is indeed only 16k over pop on 2018 estimates. And for comparison they were 57k over in 2010 and 35k over in 2016. Given the rate of decline it looks very likely that Genesee+Saginaw+Bay+Midland Counties will be extremely close to perfect population in the 2020 census. Which only makes the Flint+Tri-Cities pairing more appealing.

Clinton won such a a district by about 2,000 votes. If you do that, you all but assure a thumb+northern Macomb district.  This is similar to the current map, but there will definitely be differences.  I predict Lasing will be kept whole and an exurban Oakland+Livingston district.  MI-4 is likely the seat eliminated.

A Thumb+N Macomb/Oakland district seems like the best way to balance COIs. A Thumb+N Michigan would be better but then you either split the Tri Cities or force Flint into a horrid pairing with the Detroit exurbs or combined with Lansing in a clear gerrymander. The Thumb doesn't really have anywhere good to go that doesn't mess with other COIs, but combining it with the Detroit exurbs and rural fringe seems like the pairing that causes the least harm. And I don't think anyone debates that the 4th will be the seat that's eliminated. Hard to see how the 4th could possibly survive, really.

()
Here is a map with a Flint+Tri Cities district.  Overall a pretty fair map and I could see the commission passing something like this.  Likely a 7R-6D map but either side could pick off another seat or 2 in a good year.

()
Somewhere between your map and the above map looks likely to result from the commission.

It will be interesting to see what the dynamic of the commission will be.

Justin Leavitt made a presentation to the the selection panel in California, in which he said one of the main skills that commissioners should have is the ability to question their lawyers and demographers and other experts.

In Michigan, the commissioners are going to be drawn by lottery, with almost zero screening. Michigan does not have partisan registration, and party selection in primary elections is secret. Yet the commissioners are expected to declare a party affiliation and be selected on that basis.

The SOS who is in charge of the lottery added a couple of optional questions, letting an applicant explain why they considered themselves affiliated with a party, and why they wanted to serve on the commission. Each of the four legislative leaders may make 5 strikes  from a randomly selected pool of 200 (60D, 60R, and 80I).

Remember they won't be choosing commissioners, they will be knocking potential commissioners. If you are a Republican leader who do you go after? Some independents who you think might be biased? Some Democrats who you think might be forceful leaders. No doubt they will try to do some background checks, but even if you work with the other leader of your party, you can only take out 10 of 200.

The commission of 13 total strangers of varying competence will be expected to choose a lawyer and mapping specialists, and arrange hearings, etc. The SOS is designated as the secretary of the commission. Will they be susceptible to being led?

I mentioned this in my original writeup. White there is some degree of self-selection, you need to respond to the SoS's mailed invitation, and said self-selection will trend towards 'professionals' with the time and knowledge to commit to their potential undertaking, it will lead to a ore random and fragmented selection than in CA. This could very well lead to cliques of councilors forming, a handful of individuals dominating the commission, or potential 'guidance' from the SoS. However, it's most likely to result in individuals committed to their preconceived COIs from their region of the state, and therefore will only approve maps that conform to those guidelines, along with the most vocal of public input.

So after everyone discussed COIs a while back, I decided to explore a map that was mainly based on COIs. It ended up unusual. The 'guiding' districts in this case were CD5 which has all of the tri-cities and Flint, CD4 which crosses the Saginaw river to link the Thump and the Upstate, CD1 which actually gets all of the non-urbanized west coast, and CD7 which puts all the notable central MI college towns together. CD9 gets the Grosse Pointe's because their local lines cross the border of Wayne and Macomb. One of the AA seats has all the arabs, as I tend to prefer when possible. The main victim of the mid-state getting their COI's is CD8, but it isn't affected too much as far as pop distribution is concerned.

Trump won 7 seats when he won by less than 1%. When whitmer won by 10% she no only got the 6 Clinton seats and the swingy Macomb seat, she also got the Grand Rapids seat by <3K votes.

https://davesredistricting.org/join/c7f30e42-14ef-444d-a14b-2e7b25639720

()

()

Is it possible to put Shiawassee with Lansing? I think that would look neater and be a bit better from a COI perspective.

The rest of the map is decent, but this is on the 2010 figures. If you pull up the 2016 map, you get 2016 presidential results with updated Census estimates.

Regardless of whether or not DRA has data on the 2016 map, I always draw my lines there. Then I redraw the map in the 2010 data, since the 2010 map has more information on the precinct level and the census tracts often used on the 2016 level are sometimes weird. Michigan just is nice enough to have 2016 data in the 2016 module. Every image/map I have ever posted here is in the 2010 module, transcribed from the 2016 one. So, this is the 2016 map.

As it happens, I just finished drawing a similar map based off a district jumping over the Saginaw River:

https://davesredistricting.org/join/38383e2f-7132-4b5e-89b0-7b8c8bdcd477

The major differences to Oryxslayer's map are that Shiawassee goes in the Lansing district and Jackson mostly in the Kalamazoo district and that Grand Rapids district goes west instead of east.

Clinton wins six districts in this map, but two of them only narrowly. Obama carried ten of them in 2008 (though not the Grand Rapids district, which these days is probably the seventh best Democratic prospect.)

I think this map has some positives to it, but overall I think it shows the problem of a Thumb/Huron Shore district, which is that the two areas aren't really big enough for a congressional district. To get the necessary population, you either need to reach into exurban Macomb, or the Tri-Cities area, or well into the interior of northern Michigan, or some combination thereof. And if you're going to do that, why not just go the whole hog and tack the two areas on to separate districts?

I think the Macomb district is still the Democrats' 7th potential seat on this map (despite being stronger for Trump), but both the Grand Rapids seat and the Kalamazoo-Battle Creek-Jackson seat have some potential. Long-term, I can't see Macomb doing anything but boomeranging back towards the Democrats, demographically, as the black population is booming in southern Macomb, and fundamentally Macomb is still a suburban county.

Anyway, here's a quick alternative to Shiawassee in the 8th. Shiawassee in general is rather small when compared to her neighbors, and lacks a clear COI partner since it is kinda the 'empty' space between Flint, the Tri-Cities, Lansing, and The Detroit Exurbs. If you stick Shiawaasee in the Lansing seat on my map, you trade it for bits out of the south, something I wished to avoid. If you however put it with the 5th and the cities seat, then you get a less obtuse map, though you so add a cut. It also opens discussion of whether those southern towns in Gennessee are part of the Detroit exurbs or the Flint region.

()

Here's what it looks like overall.

()

EastAnglianLefty does have a good question about the thumb+upland seat, but I tend to think that the CD1 on this map is a viable answer to said question. The seat puts two COI's together: the upstate and the non-urban west Coast, allowing the fourth to work.

I think that's a significant improvement, though it's hard to tell how closely Bay City is being cut. I was hoping by putting Shiawassee in the Lansing seat you could pull the Lansing seat south from Mount Pleasant, though, which is a long northern appendage now (though I suppose I see the reasoning behind putting a smaller college town with Lansing). Ultimately Shiawassee is fine with either Flint or Lansing, though. I don't love it with Detroit exurbs, though, because Shiawassee has an interstate connection to both Lansing and Flint (and so has some exurbs of each on the edges) but doesn't have any highway connection to Livingston County.

It looks like quite a tight cut on Bay City. Given that there isn't a bridge over the Saginaw north of there and the 4th hence can't be contiguous by road whatever you do, you may as well do a looser cut.

Regarding Mt Pleasant, I still think the solution is to shift Jackson into the sixth (making it more compact), send the Lansing district into Montcalm and Ionia (as they're not that closely linked to GR), stick eastern Ottawa into the 3rd (as that is closely linked to GR) then re-align the boundary between the 2nd and the 6th (which should allow you to keep most of Michiana together.)

Yes the point of this 7th was the college town COI. The idea was to pair Lansing (a clear college city) with Mt. Pleasant and Albion which are also obvious college towns in mid-michigan. So it's naturally going to be perpendicular, like I said the maps goals made it strange.

()

Also it pays to keep an eye on the Bay county towns if the county is getting cut. In this map it's all of frankenlurst, Bay City, and Midland, plus a bit for pop equity. On the other map it adds monitor, Auburn, and Williams, along with a precinct for pop.

The GR district should go into Ottawa, not rural areas to the east.  Also, D-8 can be kept within Livingston and Oakland.

Your original statement says that the SOS will "select" the 200 persons in the final pool.

This is not true. The 200 finalists will be chosen to match state demographics (e.g. 36 from Wayne, 24 from Oakland, 17 from Macomb, 12 from Kent, etc. Perhaps below this level, the selection will be by region. It will likely be sex- and race-, and ethnicity-stratified. It may be age-stratified. While applications are available in Arabic, there is no way to indicate this on the application (if someone applies in Arabic can they really participate on the commission - will there have to be translation?).

It is unknown how political balancing will be done. Logically, a larger share of the Democratic pool will be Blacks from Detroit.

Half of the 200 will be chosen from those who applied on their own. This may lead to more professional applicants. The other 100 will be chosen from those who responded to the mailout to 250,000 random Michigan voters. That should be 45,500 in Wayne. If 1% respond, that is 2500 persons. There is going to be a lot of non-professionals.

In California, there were 21,004 original applicants. This was reduced to 17,090 after an initial cursory screening. 2,206 completed a supplemental application that included essay questions, 2,003 also provided letters of recommendation. 90% of initial applicants went away after it started looking like a job application.

In Michigan there is nothing like the supplemental application. The hardest part is that the application must be notarized, but 100s of notaries have volunteered to do this for free. The only barrier is that some people have never been before a notary. The pool in Michigan is going to be a lot more like that initial 21,000 who applied, rather than the 2000 that bothered to fill out a full application with essay questions and recommendations.

The applicant pool in California was reduced from 2003 to 685 in November based on reading the applications by the 3 auditors and their staffs of the Auditor Review Panel, who on average recommended 367 applicants by each panelist. A second screening reduced the 685 to 342, with each panelist recommending an average of 230 individuals.

The current process is to reduce the pools to 40 Democrats, 40 Republicans, and 40 other, who will be called in for interviews. The ARP is meeting this week to do that. This latest screening includes background checks and requires a statement of financial interests.

These 120 will be reduced to 60 after the interviews. The four legislative leaders may each strike 2 applicants from each subpool of 20, potentially reducing them to 12 each. 8 members (3R, 3D, and 2O) will be drawn randomly from the three pools. The other 6 members will be chosen by this initial 8. Austin used a similar process. Their random drawing produced an almost entirely Hispanic panel, which had to pick Anglos to balance the commission.

In Michigan, the SOS will randomly draw, though stratified, pools of 60R, 60D, and 80O. These will in turn be split in two subpools of 30 from voluntary applicants and 30 from the 250,000 solicited voters.

The legislative leaders will be able to strike 5 each from the total pool of 200. They will be doing this largely blind other than what they can garner from investigation by the political parties.

The final drawing of 13. While a naive assumption would be that there would be one commissioner from the region of each potential congressional district, this is quite unlikely.

In Michigan you have the additional difficulty that there is no partisan registration, and party selection in primaries is anonymous. Are Sanders supporters necessarily Democrats? Are Trump supporters necessarily Republican?

It is entirely wishful thinking that the Michigan commissioners will be particularly competent or representative of the state as a whole.

How much population is north of Muskegon-Kent-Clinton-Midland-Bay? Can you get two districts, even coming further south without touching those urban counties.

Assuming that you treat Gratiot and Ionia as the rest of the boundary, just under 1.2m people according to the 2016 estimates. So not that much more than a district and a half.

()

My latest iteration (this time with 2016 populations) -- you can have a reasonable pan-Huron district if you just keep all of Bay.  The Flint district takes Midland and was won by Clinton.  The Lansing-Mt. Pleasant district is fairly coherent, Trump won here by less than a point.  Muskegon down to Benton Harbor/St. Joseph is coherent, Kalamazoo-Battle Creek-Jackson works well, and Monroe is paired with southern Wayne rather than counties to its west.

Monroe + Southern Wayne is a pairing that seems very natural to me and I was wondering why it was so rare in the maps that were posted so far.

The challenge I can see is to draw a Monroe + Southern Wayne district without splitting the Middle Eastern communities in e.g. Dearborn, Livonia and other areas.

What areas of Wayne besides Dearborn have the highest concentrations of Arab-Americans? 

IIRC there are fair few Yemenis in Hamtramck.

Monroe is funny. The population along the border with Wayne is quite low. Most of the population is in the southern part of the county and consists of suburbs/exurbs of Toledo (over the border in Ohio), plus the small city of Monroe itself. There is some outer Detroit metro spillover in the north of the county, but it doesn't pair naturally with inner suburbs deep in Wayne. It pairs much better with Washtenaw, which is also an area connected to but peripheral to the Detroit metro.

I'm not sure Hamtramck is Muslim Majority, Wikipedia metions that it's city council is. Even though the Bengali and Pakistani Muslims are different from their Arab cousins, I'm sure they would prefer to be together rather than cracked between the AA seat. This is why I always reach an arm in there for my MI-12/13s. Heading west you got some in Livonia and the other townships in that western line (Redford, Plymouth, etc), though they are more of an Arab Christian Descent.

()
This seems like a useful image.

So other than Dearborn and Dearborn Heights (and maybe Hamtramck if you can manage it) it doesn't seem to be worth it to try to connect any other particular town for the sake of the Arab Americans?

Monroe doesn't naturally belong with Dearborn but the outer portions of Wayne would seem to be just fine; better that than rural counties to the west.  I don't see what it really has to do with a college town like Ann Arbor other than "somewhat outside Detroit". 

I was wondering whether it is even possible to draw two districts, where the largest L.P. city is Traverse City (but I see even that would exclude Mt.Pleasant). But if we wanted to make those two the largest cities are we still short?

If so, then there may be three acceptable options:

(1) Take Muskegon (the UP district then goes down the Huron shore)
(2) Take Midland and ... (accept that are combining COI)
(3) Sneak across to take the thumb (but I'm not too keen on going down to St.Clair) which will be a significant chunk of the population).

Not-acceptable.
(4) Into Kent
(5) Into Clinton

Historically, Monroe has stronger ties to Toledo than it does Detroit. Toledo is on the Ohio line, while Detroit is on the northern Wayne line or has grown out to it. You can't go directly south from the center of Detroit without going into Canada. Detroit has generally grown more to the north than to the west, or southwest.

Practically, if you are trying to keep whole counties, it may be easier to exclude Monroe, St.Clair and Livingston if you want equal population districts (that is, you start with Wayne, Oakland, and Macomb) and determine the whole number of districts, and add in others to complete the last district.

Monroe is either flexible, or not significant a COI to be treated as other than X people.

Since we are on the topic of Arab communities, I decided to go download the latest ACS data and see for myself what the present lay of the ground is. Here is a map of all people claiming Mid-Eastern descent in the metro - not just Arabs but Turks, Iranians, and others that would prefer being in the arab seat to anywhere else. Compared to the more limited view shown above, the Larger Arab pockets in Dearborn, Dearborn heights and Hamtramk have expanded. This is mainly because of the turnover since then, Arabs moving in and the older residents moving out. The map though also captures the Arab Christians that I mentioned earlier to the west of Dearborn - Assyrians, Lebanese, and others who may be missed by a more limited scope. If we are solely confining the Arab or Arab+AA seat to Wayne, and preventing it from tendrilling into the suburbs, then those western towns are the next best additions. There is this idea of doing parallel cuts into Oakland, so that the Arab seat can grab more Arabs and the AA seat grabs Pontiac, however such things appear to be banned by the commission.

()

Option 1 is probably not feasible - if you add Muskegon, Ionia and Gratiot to the northern group then you're still short about 50k from the necessary population for two seats, so you need to reach into the fringes of Ottawa/Kent/Clinton/Midland/Bay or some combination thereof to get the numbers up. And if the 1st district is going right down to Muskegon, then there's only room for the LP part of it to be one county wide (and even then you need to lose 10k people.)

Option 3 is also cutting a COI, because whilst you can draw a map putting Bay City in with Flint and the rest of Tri-Cities, realistically it's probably not going to happen because of the road contiguity issue.

You could go into Barry County from Ionia. That's messy and dipping pretty far south, but Barry County is enough to make up the difference and is pretty rural.

That said, I'm not sure why Muskegon would be preferable to Midland and/or Bay. Muskegon is a bigger metro than either.

Midland has been in districts with at least part of northern Michigan since at least the 1970s redistricting, so presumably there wouldn't be that much resistance locally.

West Macomb is a surprise, must be Christians.  Arabs are very divided by religion.  Arab Christians and Muslims wouldn't make a COI, about as much if a COI as jews and muslms lol.

Sterling Heights apparently has a large Chaldean population (Iraqi Christians).

Map: The Geography of Arab Detroit (https://detroitography.com/2018/06/13/map-the-geography-of-arab-detroit/)

Map in above article shows concentrations of population.

Hamtramck is interesting since at one time it was about 90% Polish, while Highland Park was almost 100% Black.

Here is my take on a 13-district Michigan:
https://davesredistricting.org/join/cf3b2fa2-7fcd-4b0d-adeb-6347e4ef6b63 (https://davesredistricting.org/join/cf3b2fa2-7fcd-4b0d-adeb-6347e4ef6b63)

Tries to preserve COIs, but the breakdown is as follows:

MI-01: Northern Michigan, Trump +21, Safe R
MI-02: Muskegon and the remnant of current MI-04, Trump +18, Safe R
MI-03: Grand Rapids, Ionia/Barry/Montcalm, Trump +9, Lean R
MI-04: Lansing, Saginaw, and rural areas in between, Clinton +5, Lean D
MI-05: Flint, Livingston, Western Oakland, Trump +9, Tossup
MI-06: Kalamazoo, SW Michigan, inland Ottawa County, Trump +14, Likely R
MI-07: Monroe County, Indiana border, Trump +23, Safe R
MI-08: Pontiac, Troy, West Bloomfield, southern Oakland, Clinton +12, Likely D
MI-09: southern Macomb County, Trump +9, Tossup
MI-10: northern Macomb, NE Oakland, the Thumb, and Bay City, Trump +29, Safe R
MI-11: Washtenaw and western Wayne, Clinton +21, Safe D
MI-12: Wyandotte, Romulus, central Wayne County, 45.6% black, Clinton +46, Safe D
MI-13: downtown Detroit, 56.3% black, Clinton +67, Safe D

Overall, I think this map has an ample amount of competitive districts. Districts 1, 2, 3, 6, 7, and 10 should likely go to the GOP. Districts 4, 5, and 9 are competitive but should be winnable for the right Dem (Levin, Kildee, the Macomb executive, etc). Districts 8, 11, 12, and 13 are all solid D holds. Overall, this leads to a 4D-3C-6R spread, which would roughly average out to 6D-7R or 7D-6R. Trends as well will make the 4/5/9 combo, as well as MI-3 and to a lesser extent MI-6 competitive.

Finally, a Dem who draws fair districts.

These are my population projections for 2020. Two annual change estimates were made. One was based on the 2010 Census (adj base) to the 2018 estimate (8.25 years), and the other from the 2016 estimate to the 2018 estimate (2 years). Linear growth was assumed in both cases. The two estimates were averaged, and projected forward from the 2018 estimate for 1.75 additional years. Given the slow growth of the state, they can be considered to be largely the 2018 estimate with an additional tweak.

()

()

Michigan has 14 Regional Council of Governments, covering the entire state except for Barry. Some have an obvious focus (Detroit, Lansing, Grand Rapids, Kalamazoo), others tend to be simple groupings of nearby counties.

Region II (Lenawee, Hillsdale, and Jackson) apparently never came up with a name for their region. GLS is based on the initial letter of Genesee, Lapeer, and Shiawassee, with the latter two in the group because they are adjacent to Genesee and don't fit with any other group. Barry may have been in the West Michigan RCOG, but finds itself between Grand Rapids, Lansing, and Kalamazoo.

The numbering of the West Michigan Shore RCOG suggests that it broke off from the West Michigan RCOG, perhaps concerned about Grand Rapids dominance.

The RCOG's have a scale somewhat similar to congressional districts, and reflect a COI which is recognized at least at the local government level (both the federal and state governments provide incentives for regional planning.

It happens that they can be grouped in super regions with populations equivalent to a congressional districts.

North: 3 RCOG in the UP (WUP, CUP, EUP), and two in LP (NW and NE) 0.959 districts.

West: West Michigan and West Michigan shore: 2.030 districts. One Grand Rapids based, and the other the remainder.

Southwest: SW, SC, and Barry: 1.101 (Barry placed in this group because it had the lowest excess over a whole district).

Central: Tri-County (Lansing) and Region II (Jackson, etc.): 1.022.

East: East (Tri-Cities, Thumb, etc.) 0.973.

Southeast: SE (Detroit) and GLS (Flint) 6.913 7 districts.

The three eastern areas: North 0.959, East 0.973, and Southeast 6.913, are underpopulated, while the three western areas: West 2.030, Southwest 1.101, and Central are overpopulated. Counties will need to be shifted. It may be possible to get the districts within acceptable limits by shifting counties.

The Michigan Constitution requires that congressional "[d]istricts shall be of equal population as mandated by the United States constitution." The mandate is as interpreted by the SCOTUS, most recently in Tennant v Jefferson County.

Very very interesting. To start off with Flint+Tri-Cities whole counties is 1.01100 quotas, or about only 7k excess population.

This shows grouping of RCOG's into areas equivalent to a whole number of CD's.

()

To get the regions closer to an whole number of representatives we need to shift some counties. The greater southeastern region (Detroit, Flint, and Ann Arbor) is short about 0.087 of the population needed for 7 districts.

There are three plausible options.
(1) Move Lenawee into the region.
(2) Move Sanilac and Huron into the region.
(3) Move Sanilac, Huron, and Tuscola into the region, and Shiawassee out.

This map explores the first option

()

Lenawee in the past has been included in the Ann Arbor metropolitan area, so it has some community of interest.

Elsewhere Barry is added to the Lansing-based central district. Barry is a free-agent not belonging to any RCOG. The district is not ideally compact, but forms an area between Detroit, Grand Rapids, and Kalamazoo.

The southwest area consisting of two RCOG's is unchanged and quite compact. It places Kalamazoo and Battle Creek in the same district.

Mason and Alcona were shifted for population balance. Otherwise the northern district consists of five RCOG's and the eastern district of one RCOG.

The western two RCOG's were divided into two CD's. Kent was placed with Allegan to avoid dividing counties. Any Grand Rapids district is going to be an imperfect representation of the city and its surroundings. You could go east from Kent adding in more rural areas, but Grand Rapids is distinctly in the western part of the county and somewhat southerly. Or you could go into suburban areas in Ottawa, but that divides the county and requires a slender connection.

In the greater southeastern area, St.Clair is added to a Flint-based district. St.Clair, particularly Port Huron is distinct from the Detroit core, and is needed for population balance.

On the west, Livingston, Lenawee, and Monroe are added to Washtenaw. Livingston and Lenawee have in the past been part of the Ann Arbor metropolitan area.

This leaves the inner three MOW counties with about 5 districts. Counties were always going to be split in this area.

Overall population equality is good (-1.6%, -1.4%, -2.2%, +1.5% +2.1%, (+4.0%/7 = 0.6%). If further refinement is needed by dividing counties, only round 75,000 persons need to be victimized.

This alternative shifts Sanilac and Huron to the southeastern region rather than Lenawee

()

The central district (Lansing-Jackson) no longer needs adjustment so Barry is added to the western Michigan area.

Taking of Sanilac and Huron forces the eastern district (Tri-Cities) westward. This results in Barry and Ionia being added to the Grand Rapids (Kent) district, since they would be cut off otherwise.

In the southeast, the Ann Arbor district will need to pull in some population from far western or southern Wayne.

Number of interested applicants has risen to 6K.

https://twitter.com/mcpli/status/1233773797377007616?s=19

Correct me if I'm wrong but it looks like you put Moolenar, Kildee, and Slotkin in the same district(5)

A 6th Circuit panel just unanimously upheld Michigan's redistricting commission against a federal constitutional challenge.  The panel was 1 Trump appointee and 2 Clinton appointees.  The Trump appointee wrote a concurrence that seemed to grant even stronger deference to Michigan and cited Roberts' comments from Rucho that seemed to bless state commissions.

I think the commission is safely in place going forward.

https://www.michigan.gov/documents/sos/ICRC_Whos_applying_682238_7.pdf

just interesting stats
So the commission applicants are a decent bit whiter, they are also 60% male, and also much older than the average Michigan voter, but however 35% are Democrat to 15% Republican, doesn't really matter , just really hillarious its the most Republican Demographic but most partisan applicants are Democrats.

So it's going to be a commission of former Woodstock attendees?

Michigan does not have partisan registration. Primary ballots have all parties on them. The voter selects a party anonymously and votes for candidates of that party.

An applicant can say whatever they want. The restrictions are so tight, on political involvement anyone with any sort of overt political activity might be excluded.

Anyhow when the random pools are selected, they will be weighted.

IIUC, a "Republican" will be weighted by 30/15, a Democrat by 30/35, and an unaffiliated by 40/50.

Weighting will also be done for sex, race, age, and geographical area.

The age weightings are not based on the age categories shown, but are off by 10 years. So the 55+ group will tend to be heavy in the 65+1, the 35-55 group by those in the 45-55 group, and the 35-and-under by those in their late 20s and 30s.

The geographic areas are Wayne; Southeast (which excludes Wayne, but includes Lansing and Jackson; Western; East Central: Flint, Thumb, and Tri-Cities; Northern L.P.; and U.P.
There are more applicants for Oakland alone than Wayne, and partisan hacks from Lansing might be drawn to represent the Detroit suburbs.

The commission has all the makings of a train wreck.

With a Democratic governor, will the map be less hostile to Democrats?

With the commission it's actually the Democratic SoS that matters more, at least that's how it looks now.

Here's a link to a possible map; too lazy rn to screenshot, my apologies

https://davesredistricting.org/join/1f84b5fd-62f4-4f75-8320-484b0288b314

GR should stay whole, and Ann Arbor goes with Wayne better than the Exurbs.

Not really, Ann Arbor works ok with the 3 Western cities in Wayne but anything not part of the 2 VRA seats should be put with Monroe.

Between 3 and 4 fix that unnecesary double cross.(just split 1 county)
Then I'd say put battle creek in the Kalamazoo district and  put all of Jackson with the Lansing district.  Dont really like the flint district but not sure how to fix it within your map.(would require a very large rotation)

I don't care for the Ann Arbor-Livingston thing either, but my hand was kind of forced by drawing two majority (not plurality) Black districts, which will probably be done by the commission since it's still possible. That means the Dingell district has to take more of Wayne, and that's too much to still include Ann Arbor. I'd be happy to hear your suggestions, but I didn't want to f[inks] up the balance I got in the other part of the state by drawing Ann Arbor into 6th or whatever.

 I'd be happy to see a recommendation which has two majority Black districts and does nicer things with Ann Arbor though :)


Grand Rapids is whole..

Tbf Ann Arbor doesn't really go well anywhere, the closest county in a cultural term is probably Ingham which splits 2 metroes and is also a GOP gerrymander.

It does have a certain Detroit metro aspect to it though, no? That said the closest cultural companions to Ann Arbor are Ferndale and co., who aren't exactly easy to put in a Washtenaw district either...

the metro

()

Opinion of this Oakland Macomb split?
Its not really partisan(both districts are Clinton +7 and Clinton +5. Making it more compact would just switch those numbers but I wanted a white collar vs blue collar split.

Its very ugly and would scream gerrymandered to idiots but I think it better represents the COI's of the north Detroit suburbs.)Grosse point is white collar but thats way too far away to "gerrymander" inside.

I don't think its really worth it either just because of how ugly it is but was just gathering some ideas.

I avoided royal oak. the cities to the south of it are more working class. (also I consider Pontiac working class of course)

Uh, the idea of crude class-based redistricting sounds ugly.

 I mean different classes clearly have different interests? And its still within the same metro area etc. Again my version was too ugly but Im trying to cook up a reasonable version.

()

Obviously violates the VRA but this is my non partisan map of the detroit metro that fits almost perfectly
Creates a nice downriver WWC+arab seat thats Clinton +4 but trending R. Lean D
Ann Arbor College town and upscale Wayne county suburbs. Clinton +25(Safe D)
Titanium D Clinton +90 Detroit seat.
Livingston + exurban oakland and Macomb(Trump +24) Safe R
Inner Oakland(Clinton +25) Safe D
Most of Macomb +7 Trump = lean to Likely R and trending right.

()
My take on a fair map.  For communities of interest, I mostly used metro areas and municipal/county borders.  2 black VRA seats+a black opportunity seat in the suburbs, 25% black.  But Levin would likely represent it.  In terms of partisan outcome, my map is decent if you take into account 2018 and 2012, but if 2016 trends continue, it would probably end up with 8 Republicans.  But who knows, Whitmer won the Flint/Thumb district so Kildee could survive there awhile. 

Did you split the tri county lansing metro Idaho?

Also for your northern 2 districts. Divide them east west. Don't put a cross lake district besides for the UP district.

https://davesredistricting.org/join/00ba0430-eccf-4ee4-9cb5-82090bd38e07
no it's kept together
and It's actually really hard to do it east/west, because of Muskegon and the road connection to the UP. 

()

Turns out we can make copies of maps now on DRA :)
Anyway made 2 edits to your map. Kept Kalamazoo whole and tried the East West thing. Not sure I like the split of Saginaw Bay and Flint into 3 separate districts. Its not a mega partisan gerrymander as you merely made Lansing Lean to Likely D from tossup and then made Flint Likely R from tossup. Whitmer lost the pink district 82 votes btw !. You did go for a light GOP gerrymander in Macomb with the black opportunity district but its not the worst idea as its still fairly compact.

I didn't think of connecting the UP to the east coast.  Honestly, I prefer my cut aesthetically but to each his own.  Keeping Kalamazoo whole is probably a good idea tho.  As for Macomb, yeah my cut slightly benefits the GOP, but Dems are already getting a safer suburban sea than they currently have while keeping Dingell safe, so if I cut northern Macomb rather than southern, basically every cut in the metro area would favor dems.  As for the tri cities, keeping them whole would complicate the rest of the map.  Either the thumb would need to go into Detroit suburbs or up way north, and also messes up sw MI. 

This is much nicer (of course except the Detroit sink, which one can recognize isn't quite right without even knowing what the VRA is)

If you combine the Downriver and Detroit districts, what's the black population? Just wondering if there's a viable east-west split there that keeps them both as VRA districts.

I decided to give this a try with a fair map and uh, here is what I got. Not really a fair map in terms of partisan outcome but the COIs seem reasonable?

()

https://davesredistricting.org/join/9ed5b836-c557-4f51-ad5e-6c9199cf8098

MI-01: Trump+22, R+10 (Safe R)
MI-02: Trump+22, R+10 (Safe R)
MI-03: Trump+9, R+7 (Likely R)
MI-04: Trump+19, R+7 (Safe R)
MI-05: Trump+11, R+3 (Likely R)
MI-06: Trump+6, R+2 (Lean R)
MI-07: Clinton+8, D+4 (Likely D)
MI-08: Trump+4, R+2 (Tossup)
MI-09: Clinton+3, EVEN (Tossup)
MI-10: Trump+19, R+7 (Safe R)
MI-11: Trump+0, D+2 (Tossup)
MI-12: Clinton+58, D+29, 51% Black CVAP (Safe D)
MI-13: Clinton+52, D+28, 51% Black CVAP (Safe D)

4 Safe R
2 Likely R
1 Lean R
3 Tossup
1 Likely D
2 Safe D

For the sake of completeness, I will say that in the Governor race Whitmer gets all the Clinton districts plus the 5th (Flint), the 6th (Kalamazoo), the 8th (Lansing) and the 11th (Southern Detroit); getting in fact a majority of districts in the process (8/13)

However, Stabenow does not get a majority of districts as she only flipped the 11th and the 8th (for a total of 6/13)

The Black population in that yellow seat is essentially negligible outside of River Rouge, Ecorse, Inkster, and Romulus, and none of those are a huge percentage of the population. So yeah, I'd imagine it may even be majority white? The Oakland county district might actually be Blacker since it has Southfield and Pontiac.

sending both VRA seats into Oakland would be interesting.  But they can both be kept within Wayne, I think that's the best configuration

Agree I prefer to avoid that cross(I might put southfield in a vra seat and grosse point in the macomb seat.)

I just question to what extent one of two VRA seats which just stays in Wayne actually works as a performing seat. I'd be a little concerned about a legal challenge if the Black % in both seats isn't as high as reasonably possible; even under the current district Tlaib managed to win a primary in 2018 where she was not the Black candidate of choice IIRC.

As far as the Pointes go, I don't see anything wrong with putting them in a Black-majority seat.. They aren't exactly a great fit for working-class and increasingly diversifying Macomb either, and you avoid a county split.

I mean, I don't think that the VRA has anything to say about very crowded Democratic primaries where the winner barely gets 30% of the vote.

ok i tried my best at a fair COI-based map with partisan data off. think i'm pretty satisfied with it since it's compact, only splits 10 counties, and keeps the grand rapids, lansing and tri-cities areas together. the sacrifices made are the detroit exurbs (shared among the two suburban detroit districts and lansing) and wayne's WWC neighborhoods (necessary for VRA districts)

it ends up being 7-6 obama, 7-6 trump, 8-5 whitmer. tipping point seat is CD8 which is consistently 2-5 points to the right of the state, so there's a GOP tilt

()
MI-01: Obama +1 | Romney +9 | Trump +23
MI-02: McCain +6 | Romney +14 | Trump +9
MI-03: Obama +2 | Romney +6 | Trump +24
MI-04: Obama +6 | Romney +5 | Trump +21
MI-05: Obama +24 | Obama +18 | Clinton +1
MI-06: Obama +7 | Romney +2 | Trump +8
MI-07: Obama +20 | Obama +14 | Clinton +16
MI-08: Obama +10 | Obama +3 | Trump +5
MI-09: Obama +16 | Obama +15 | Clinton +6
MI-10: McCain +1 | Romney +10 | Trump +31
MI-11: Obama +8 | Obama +1 | Clinton +2
MI-12: Obama +61 | Obama +62 | Clinton +54
MI-13: Obama +64 | Obama +63 | Clinton +48

Could you post the DRA map?
I want to make a change or 2.

my posted map is with extrapolated 2020 populations but here's  with 2018 numbers (https://davesredistricting.org/join/7be2549f-5a4a-43f4-a8eb-c7eef59d7076)

()

CVparty?
Opinion of these metro changes based on your map?

The median district is still Trump +5 but its now the Macomb seat. The lansing seat moves to Trump +2.4. The Flint seat is Clinton +0.6.

Also changed the cut of Flint county I took but didn't really make a major difference.

()
A good government R map.  Keeps cities and metro areas/COIs together, but would probably be 9-4.  Each red district is at least Trump+10.  In 2010 Steven Wolf would love this tho LOL. 

()
Michigan map premised on competitive districts.
10, 11, 12, and 13 are all safe Dem; 1, 2, 4 and 6 are safe GOP; and the rest of the districts are at least fairly evenly divided, if not very evenly divided. The map does have a slight GOP lean but I wouldn't make much of it.

https://davesredistricting.org/join/67c4b87b-568c-4733-93eb-06bf4571758d

()
This was done without looking at partisan data. Emphasis was on county integrity and having an all-Oakland CD. Unlike the previous map it has 2 black-plurality seats, not just one.
Only 6 county splits.
7, 11, 12, and 13 are Democratic. 1, 2, 3, and 4 are Republican. Not all of these 8 seats are safe. All other districts are too competitive to consider leaning too much to either party.
Ironically, while not aiming specifically for more competitive seats, I got a more competitive map than last time.
https://davesredistricting.org/join/530ccacc-b579-42c9-99fb-785785196b7d

()
Here's a similar map with 5 Clinton districts instead of 4, but fewer competitive districts.
According to the 2012-2018 composite, 6, 9, 11, 12, and 13 are solidly Republican, 5 and 10 lean Republican, 7 is highly competitive, 8 leans Democrat, and 1, 2, 3, and 4 are solidly Democratic.
https://davesredistricting.org/join/00ba0430-eccf-4ee4-9cb5-82090bd38e07

https://twitter.com/rbeggin/status/1295417327849177093?s=20

Commissioners have been chosen.

Here's another shot at Michigan--IMO this is my favorite map I've made yet. Here's the DRA link. (https://davesredistricting.org/join/bcbb861e-3e13-4b8b-a6f3-eb82d5679aa8)

()
()
()

The only town or township split is Clinton Charter Township, in Macomb County, which has such humongous municipalities that it makes not splitting something rather difficult.

MI-11 and MI-12 are both majority-Black.

Let me know if there any issues with the image displays.

That is a very nice map.  About the only thing I'd quibble with is the 8th's long western tail. 

Yeah, not my favorite. Though it's kind of forced if you want to put Battle Creek in with Kalamazoo. Forumites, are those two particularly connected?

You may be able to keep Battle Creek if you give most/all of Allegan to the 2nd; that way Holland isn't split. 

How about this? (https://davesredistricting.org/join/c2cffd5c-1631-491d-88ce-9a2b1ce2e27e)

()

Still had to split Calhoun, but I do agree that the 8th is nicer when conciser.

I think Calhoun is destined to be split.  If you want to make things even more compact you could give Barry to the Lansing district, split Allegan between 2 and 5, and then lop off the top of the Lansing district and put it back into 2. 

It seems the most important decision is regarding Macomb and the Thumb. Keeping a district fully with Macomb all but forces Levin and Stevens together while also putting Flint with the thumb which should be Likely GOP.

However keeping something like the current MI 10th will help D's a decent bit even if the Detroit districts expand as Washentaw can take all the red territory and still remain Safe D.

Are these just random people, like a jury trial, or do they have certain qualifications? Pretty cool if it's the former.

()

What is partisan breakdown of that map? Looks like 1, 2, 3, 5, 7, 8 are Republican, 6, 9, 10, 12, 13 are Democratic, and 4 and 11 are swingy?

3 is lean Dem I suspect, and 5 could go swingy if the Dutch belt keeps moving to the Dems. The only really pretty safe CD's for the Pubs are 1 and 7. At the moment 4 is probably tilt Pub, but may not be that way for long. The one really swing CD is 11, and it will go tilt Dem if trends continue. Trump 2020 probably got beaten pretty badly there, but he was a horrible fit for that CD this year.

The way the trends have played out, there are a lot more competitive CD's in MI, which is a beautiful thing. :)

I think 3 would lean Republican and 4 and 11 would be the swing districts.

Well you can peruse the numbers yourself and make up your own mind. It is harder than it used to be to predict these things with the complexion of the partisan coalitions in such a state of flux.

https://davesredistricting.org/join/d5fbe65c-1976-45b4-a98f-ab09f694fad4

Well that does not seem to work. Sorry.

()

Honestly I would give Bay and the county to the north to the 1st and give Leelanau and Grand Traverse to the 2nd.

Your wish is my command.

()

For fun decided to try and make a fair State senate map:

()

Detroit area inset:
()

https://davesredistricting.org/join/e7e3428a-47d3-490a-88b2-f6bde8043a81

Per DRA, there should be 18 Republican districts, 11 Democratic districts and 8 competitive districts.

The tipping point districts would be:

Dem majority: District 34 (Muskegon area): Trump+10, R+3
Tie: District 13 (Southeast Macomb County); Trump+9, R+2
Rep majority: District 22 (Saginaw area): Trump+7, R+2

So basically this map would make the state senate likely R, but not 100% safe

Just to see if it's possible given how easy a 9-4 GOP gerry is, a solid attempt at a 9-4 Dem gerry. Every Dem incumbent resides in their own seat except for Slotkin who really lives in the worst place possible, though I assume she'd move to the 7th. The 3rd is marginal but trending blue while the 4th and 11th are both zooming right.
Overall it turned out a lot better than I was expecting.

https://davesredistricting.org/join/9f0a01dd-f779-4b11-82ce-ed008b8a1ce9 (https://davesredistricting.org/join/9f0a01dd-f779-4b11-82ce-ed008b8a1ce9)
()

()

()

()

My first attempt at a fair Michigan map. I tried to minimize county and city splits, but in doing so I mistakenly created a mild GOP gerrymander. Detroit is kept whole, but the result is a Dem pack, but I think Detroit should be kept whole if it can be. The data is President 2016. Biden likely flipped the Lansing CD and possibly flipped the Grand Rapids CD, but the likely breakdown was 4-9 in 2016 and 5-8 in 2020. I tried to make a thumb seat but that involved incorporating Flint, which cost Dems a seat. Not making a seat that went across Macomb and Oakland probably cost another seat. The map could've been 9-4 with a pro-Dem drawn while still being fair.

Not VRA compliant.

MI-10 is 76.65% Black?  Not happening.

Do I need to make a second AA seat in Detroit?

Bit of a bummer, I was surprised that Detroit could be fit into a nice, neat district. I suppose it can be rearranged without really changing the outcome of the map though.

What qualifies as VRA compliant? I was able to make both Detroit CDs around 45% black. Should I shoot for 50?

()

The east Detroit is now down to 50% Black and the West Wayne is now 42% Black. DRA has minority rating at 98.

A compact Detroit CD is excessive packing of black voters. That being said - you don't have to touch any CDs on the map other than 11 and 10. Just rearrange them.

Rearranged above.

Ok, that's decent.

You can definitely get over 50% pretty easily and compactly. It does require dipping into Southfield and Pontiac, and you do have to do a little careful work to avoid splitting municipalities, but it's very doable.

Pontiac shouldn't really be reached for, its way too far away and messes up the rest of the Oakland district.Southfield is ok though.

This is an example of why the VRA is stupid and goes against the whole idea of fair redistricting which is to put communities of interest together. From a fair standpoint it makes sense to make Detroit one district because it fits nicely and a city is a perfect COI but no we can't do that because it packs blacks so instead we have to spread them out through the suburbs to make sure we can get two black districts.

Does going from 42 to 50% warrant another split of Oakland? I saw all of the Black population in South Oakland but I have an existing district with basically all of incorporated Oakland in it. Counties are pretty arbitrary political boundaries, but I don't like how Wayne/Washtenaw/Macomb/Oakland are all carved up in weird ways in the actual House map.

I do think 42% Black is enough to elect a representative of their choice consistently. They make up over 60% of the Dem electorate in each.

Imo, I think it can go either way. Going into Oakland or not going into it are both justifiable choices.

Yeah, Pontiac doesn't belong in there. Period. If you reach into Oakland, take Southfield. That's it.

You have to draw two Black districts in Detroit, suck it up.

The VRA does not require some grotesque arm leading up through Oakland all the way to Pontiac. Just no.

Given you have 2 black districts you could go for 2 districts at 46% black as well instead of one at 42 and one at 50.

I also agree with TPH that in this particular case the COI based map would keep Detroit in a single district. But over all 50 states the effects of the VRA probably cancel out.

Yeah, the Supreme Court has ruled that you can get a black performing district at like 42-44%.