2020 New York Redistricting

[EastAnglianLefty]

Now that DRA has CVAP data, I took a look at what a commission State Senate map might look at, assuming nobody put their thumbs on the scales too much. Outside of NYC, I prioritised minimising county and municipality splits, with the exception of Rochester and Buffalo (as keeping them whole is effectively a Republican gerrymander and I suspect that Democrats will be able to successfully argue to the commission that enough areas outside those cities are functionally part of them that a split is better for CoIs.)

Inside NYC, I worked out groups of boroughs first, then prioritised the VRA first, then tried to keep the lines as clean as possible. I didn't pay much attention to partisan data, but as I was trying to put areas with similar interests together a few packs did naturally form themselves. I took no account of incumbents' residences, so obviously this might be dead on arrival in Albany.

The link is here: https://davesredistricting.org/join/f0b86bc6-7100-4dbc-8cc4-585a5da79375

Statewide:

NYC:

In partisan terms, 45 districts have a Democratic PVI and 18 have a Republican one (though the latter will probably go up once 2012 drops out of the averages.) 34 districts are D+10 or above and 37 are D+5 or above.

On Long Island, the 8th district is majority-minority by total population but white-majority by CVAP. It's possibly a black-opportunity district, but a bit of a reach. Most of the other districts were drawn based on what combinations of towns were within the acceptable population deviation.

Queens and Manhattan are paired, because the former can't stand alone and whilst the latter theoretically could, it's got the population for 5.24 districts which in practice is too hard to balance whilst remaining within the 5% limit. There are two black-majority districts in Queens and one performing black-plurality district in Harlem (49.6% by total population, 46.4% by CVAP). The 14th in Corona and the 32nd in northern Manhattan are Hispanic-majority by CVAP, whilst the 16th has a substantial Hispanic plurality by total population but a narrow white plurality (39.9% vs. 39.1%) by CVAP. In NE Queens, the 12th is drawn as a performing Asian VRA district (57.5% by total population, 47% by CVAP) and the 13th is narrowly plurality Asian by population (37.3% vs. 36.1% white) but has a substantially white plurality by CVAP (though John Liu would hold it without difficulty.)

In Brooklyn, there's still the population to draw 4 black VRA districts, so I don't think there's any legal justification not to. The snag is that this means they all have to take long-awkward shapes, so they can grab enough areas in the east of the borough to keep a black-majority and pair them with whiter areas in the west to avoid them becoming packs.

I couldn't draw a clean majority Hispanic district by CVAP, but my 17th will probably perform (47.9% Hispanic by total population, 40.9% H 30.7% W 20.8% B by CVAP.) The cut of Williamsburg is ugly, but if you take the entirety of it in then the Hispanic CVAP falls precipitously and if you take none of it you impinge on the black VRA districts instead.

In south-western Brooklyn, there's enough of an Asian population that an opportunity district should be drawn (I'm assuming that they form a politically and ethnically coherent group, so correct me if this is incorrect.) That's my 22nd, which is narrowly plurality Asian by CVAP (38.5% versus 37.5% white) and has a much stronger plurality by total population.

Once you factor all those in, the non-VRA districts more or less have to fill in the gaps. The 19th ends up as a Republican district in south Brooklyn, but Borough Park gets sunk into a black VRA seat. Unless you claim that Orthodox Jews should also be encompassed by the VRA, this would always be likely to happen. The only way to preserve the current vote-sink Felder holds whilst still drawing the Asian opportunity seat would be to stretch one of the VRA districts along to Coney Island, which would hardly be compact.

In the Bronx, you've got 4 Hispanic-majority districts by CVAP, and one black-majority one which also takes in Mount Vernon and parts of Yonkers.

Upstate, Democrats can rely upon the Albany and Syracuse seats and two seats in both Monroe and Erie, and under most circumstances also the seats based on Ithaca and Ulster County.

I would very much expect this not to be what gets passed, given incumbent preferences, but I thought it might serve as a good baseline of what to expect. Thoughts?

[EastAnglianLefty]

Clean-ish bipartisan 26 district map: https://davesredistricting.org/join/064e8710-ae0f-4867-b163-d11d9066e620

I figure we're not going to see a full-on incumbent protection map, because most of the Democratic delegation isn't vulnerable in the general election full stop, and plenty of others have only vulnerable if a Republican gerrymander combines with a wave year. In contrast, there aren't really any districts in the states Democrats couldn't theoretically win in a good year against a weak incumbent. So Katko isn't going to get a safe seat, for example, because Democrats have no reason to agree to that.

That said, Democrats do have lines they want in particular places that they would need the Republicans on the commission to agree to, so they do have an incentive to make concessions in places. This is a guess at what they might go for if both sides wanted a reasonably clean-looking map and Democrats looked to maximise their options rather than trying anything too ambitious.







NY-1: R+5
NY-2: R+3
NY-3: D+2
NY-4: D+8
NY-5: D+30, black VRA district
NY-6: D+20, Asian opportunity district
NY-7: D+37, Hispanic opportunity district
NY-8: D+35, black VRA district
NY-9: D+27, black VRA district,
NY-10: D+37
NY-11: D+4
NY-12: D+29
NY-13: D+43, Hispanic VRA district
NY-14: D+34, black/Hispanic opportunity district
NY-15: D+43, Hispanic VRA district
NY-16: D+14
NY-17: D+0
NY-18: D+2
NY-19: D+6
NY-20: R+2
NY-21: D+4
NY-22: R+9
NY-23: R+3
NY-24: D+7
NY-25: R+12
NY-26: D+9

This map assumes that all incumbents win re-election this year bar Brindisi (and that the retirements are succeeded by someone from their party.) Even if Brindisi wins, I'm not sure there's that much incentive to protect him as a) he's proved he doesn't need that much protection; b) he's not got any seniority to speak of, nor close allies in Albany and c) protecting him means throwing him in with Katko, which Republicans aren't going to agree to and Democrats shouldn't need, when they could just get a competent Syracuse-based candidate instead.

All incumbents have somewhere to run, with the short stick being held by Lowey's replacement - although if she's succeeded by a Westchester-based Democrat, they might have a shot at NY-16. That district reaches into the NW Bronx because I assume that's where Engel lives, but doesn't include Mount Vernon to try to protect him from a primary challenge from Bowman. If Bowman does beat him, NY-16 would probably become entirely Westchester-based, although what happens in the Bronx is anybody's guess.

Meeks, Jeffries and Clarke each get a district that is 47% black by both total population and CVAP, which ought to be safe enough (especially as the white population in their districts is heavily Republican.)

Kathleen Rice gets a district that is narrowly minority-majority by total population, though I could see her being eager to give some heavily-Democratic minority precincts to Suozzi to ensure she's protected from primaries if her pre-Congress career becomes more controversial in the coming months.

Meng's district has a strong Asian plurality by total population, but only a narrow plurality by CVAP. Substitute Asian for Hispanic and you can say the same thing for Velazquez. Meng shouldn't mind either way and whilst Velazquez might, she's in no position to complain - she's at much less risk of a primary than Jeffries or Clarke, so she's just going to have to take her share on NW Brooklyn hipsters.

Rose gets Park Slope, which doesn't quite make him safe but certainly makes him favoured. Getting Republicans to agree to this might be a stretch, but if they do kick up a fuss over that then Democrats are likely to argue for combining Ithaca and Syracuse, or for redrawing Long Island in a more aggressive fashion.

Espaillat's district becomes narrowly Hispanic by CVAP, whilst AOC's district is approximately one-third black and one-third Hispanic by CVAP. In the event Bowman does win, you could theoretically make AOC's district majority-Hispanic and put Bowman and Espaillat into a plurality-black district, thought that would be ugly and it seems a little unlikely a commission would chuck Espaillat under the bus like that when such a district would be unlikely to pass the Gingles test.

Upstate, Maloney and Delgado get districts that aren't radically shifted from their present partisanship, whilst Stefanik has little to complain about. Brindisi's district gets a few points redder as it swaps Binghamton for Jefferson County, whilst Reed might be a little worried about having to deal with a district containing both Binghamton and Ithaca. If this is not to your taste, you could shift Binghamton into Brindisi's district, Wayne, Cayuga and Seneca counties into Reed's and have Katko's district head north along Lake Erie, though that's probably a little Republican-friendly for Democratic commissioners to agree to it.

[EastAnglianLefty]

As it's still possible, I took a look at what a 25 district map might look like, working from the same basic assumptions as in the 26 district map I posted earlier.

I didn't come in with any particular assumptions about which seats were going to get axed and I was a little surprised to find that the two that disappeared were Delgado's and Ocasio-Cortez's.

Maps:



Link: https://davesredistricting.org/join/bb77aac8-9803-4811-9899-466381aa2b9a

PVIs:
NY-1 - R+4
NY-2 - R+2
NY-3 - D+4
NY-4 - D+5
NY-5 - D+30
NY-6 - D+21
NY-7 - D+34
NY-8 - D+36
NY-9 - D+26
NY-10 - D+37
NY-11 - D+5
NY-12 - D+30
NY-13 - D+44
NY-14 - D+44
NY-15 - D+24
NY-16 - D+8
NY-17 - R+2
NY-18 - D+5
NY-19 - R+4
NY-20 - R+2
NY-21 - D+3
NY-22 - R+3
NY-23 - D+6
NY-24 - R+13
NY-25 - D+9

Explanation:
On Long Island, the town of Hempsted is almost exactly the right size for a congressional district in a 25 seat map and whilst I don't expect those drawing the lines to care too much about respecting municipal boundaries, it creates a fairly sensible seat for Rice. If you follow that arrangement, Suffolk plus the southern part of Oyster Bay is the right size for two seats and a north-south orientation probably reflects COIs best.

Suozzi then has to take in about 300k from NYC and if you're going to do that whilst still keeping NY-06 as an Asian opportunity seat, that means heading up into the East Bronx.

The higher population required with a 25 district map means it's harder to keep Velazquez's district Hispanic-plurality if you maintain its present orientation. However, if you loop it round Meng's and into Jackson Heights and Corona, it becomes comfortably Hispanic-majority by total population and there's a very clear plurality by CVAP too. I'm not sure Velazquez would be overjoyed, as this might be the best district for AOC to run in.

Meeks, Clarke and Jeffries then divide up southern Brooklyn between them, with each having a district that's just under 47% black by CVAP and which should maintain their base. Sunset Park and Park Slope should make NY-11 safe D and Maloney and Jeffries divide up the remainder of Brooklyn between them - this split could be eliminated, but I think the map looks nicer this way.

In the Bronx, Espaillat, whoever wins tonight and Bowman/Engel each get a district that's pretty close to what they have now and Lowey's district also has a recognisable successor.

Upstate, things get a bit trickier. NY-17 is clearly more Maloney's than Delgado's, though it's a little redder than the current NY-18. Tonko has the best claim on this NY-18, although I just now note I've drawn him out and there must be a decent chance he'll retire anyway, in which case Delgado might have a shot at this one.

NY-19 is perhaps winnable by Brindisi, although there's a good chance he'd be beaten by an Ulster County Democrat in the primary anyway. Stefanik ought to be fine in the North Country seat.

You're going to have 5 seats for the bits of the state west of Syracuse/Binghamton and this is far from the only solution. You're going to get a Rochester seat and a Buffalo seat, but everything else can be varied. Here NY-21 is designed to contain areas in the orbit of Syracuse, but it'd be easy enough to remove Oswego, for example, and to chuck in Ithaca instead. Similarly, for the two rural seats I went with an east-west arrangement, but north-south is equally feasible and turns NY-22 from a potential swing seat into a Republican banker.

[EastAnglianLefty]

Slight tweak to my 26 district map in light of Engel's defeat: https://davesredistricting.org/join/064e8710-ae0f-4867-b163-d11d9066e620

[EastAnglianLefty]

Fairly similar to my Upstate map, I think. Though I suspect a bipartisan map wouldn't draw out both Stefanik and Tonko from their districts, so a couple of tweaks to prevent that would be likely.

[EastAnglianLefty]

If New York loses only one district, I think NY-27 (Chris Jacobs) will be the district eliminated.

If New York loses two districts, I'd expect three targets. Jacobs and Jamal Bowman would probably get their seats eliminated, with AOC also being drawn out of her district (though she'd still have a district to represent, if she wants it, it just likely would be territory that might not be winnable for her).

I don't think the legislature would bother with Bowman or AOC if New York only loses one seat (though they might target AOC in the sense that they'll draw her out of her district, but still give her a seat to run in, that's unwinnable for her)

This is definitely wrong. There are still too many people west of Syracuse to squeeze them into 3 districts, so Jacobs will have a district to run in.

With one lost district, the state's geography forces NY-16 to be divided, with the likeliest option being that Bowman's base in the NE Bronx gets thrown into a district with AOC.

With two lost districts, Reed is the likeliest victim Upstate. Around NYC preserving Bowman's seat is pretty easy and AOC's seat is the obvious one to split up.

[EastAnglianLefty]

a) It's not going to a court, if the commission don't draw a map that the Democrats are happy with they can just pass their own through the State Assembly and Senate
b) Of course it's possible to draw three performing Hispanic districts. It's not even that complex - just adding Corona and Elmhurst to Velazquez's district would do it, although that's using a sledgehammer to crack a nut.

[EastAnglianLefty]

Im surprised people think the legislature will go after AOC and Bowman. It doesnt look at all likely. Not only has the legislature seen large growth in left wing over the course of these 4 years, but the two are also in very strategically important areas that are hard to dismantle.

AOC's seat has rapidly seen a demographic shift, to the point that it could considered a VRA Hispanic seat. By eliminating her seat, the surrounding seats have to take in very Hispanic territory, which would dilute many of the VRA required seats in the area. Meng cant take them in, and neither can Espaillat nor (probably) Torres. There are no good ways to divide up AOC's seat, and, in fact, it would probably better serve the map for her seat to just be given more Hispanics from other neighboring territories.

Bowman's seat also has a roughly similar problem, as cutting his seat means the largely Hispanic VRA seats have to take in African American territory, but besides that, his seat is rather geographically secure. Its very difficult, not impossible but difficult, to draw a map that doesnt include a seat that crosses between Westchester and the Bronx. Its definitely possible, but it would require a lot of contortions and twists that would likely endanger many other seats.

Also, sidenote, they're likely not going to eliminate a Hispanic seat. The Hispanic population in NYC is growing, and it wouldnt look good to either the courts nor other Democrats to crack one of their seats. There's also the problem of dividing up the Hispanic seat's population, which would lead to problems similar to the ones outlined with eliminating AOC's seat. If NYC+LI are losing a seat, its probably going to be a white seat.

Whatever happens, there's got to be at least some disruption to NY-16, as there's no other way to get from Upstate to NYC and you aren't going to get an integer number of seats north and south to it, and probably nowhere near. It's possible to minimise the disruption if you loop a seat round the eastern edge of the seat, but that only makes sense if those drawing the lines particularly want to protect Bowman - he can still get badly affected even if there's no deliberate effort to hurt him. But in general he'll do better if two seats are lost rather than one, because if one seat goes in NYC and one Upstate, Bowman's seat gets moved around less.

AOC is probably more secure (both due to location and because nobody wants to face her in a primary), but she's still going to be protected less than her near neighbours - Meng needs to have an Asian-opportunity seat, Velazquez needs a performing Hispanic district, Maloney wants to be shored up in a primary (although that may mean she wants less of Queens in her seat.)

With one seat, the numbers actually work reasonably well for keeping AOC's seat similar to it's current lines, but shifting north to grab most of the black-majority areas of NY-16. It would be a comfortable Hispanic plurality and could be considered Hispanic-access at worst. Alternatively you could shift it further into Torres' seat and try to make it Hispanic-majority, but if you do that then the black-majority areas probably end up with Espaillat, which he isn't going to want.

With two seats, it's pretty difficult to keep Velazquez's district as a performing Hispanic district under something resembling the present lines (although she herself probably wouldn't be in difficulty), because whichever direction you go you take in lots of high-turnout white liberals. It depends how high a Hispanic percentage she wants, but if she wants a clear plurality then the easiest way to do that is to take most of the Queens portions of AOC's district. In addition, Suozzi's district will need to gain more of Queens (Rice is probably blocked in by Meeks) and as he can only go so far into Meng's district, he'll probably also have to reach into the Bronx. That could leave NY-14 divided multiple ways.

[EastAnglianLefty]

Anyone guessing what NY D's might do if Balter wins, that would probably cause a crapfest especially if NY loses 2 seats.

If Balter wins and Brindisi loses, it’s easy.  Just cut up Tenney’s district and give Utica to Balter.

Why would you move Utica? It's not that Democratic, if Brindisi loses there are better places to shore up the Syracuse district.

[EastAnglianLefty]

In previous cycles the numbers had been increased from 61 to 62 and from 62 to 63 to assist Republicans in shorting up marginal seats. Very much a case of bolting the stable door after the horse has bolted, though.

[EastAnglianLefty]

If you're worried about NY-3 eating into the Asian parts of NY-6, it's perfectly feasible to skirt round it and top the seat up with parts of the Bronx. This will actually tend to allow NY-6 to become more Asian.

If you're considering extending NY-5 further into Long Island, bear in mind that Kathleen Rice lives in Garden City. So it's feasible to make the minority communities on the border with Queens, but it's difficult to add in Hempstead et al. without drawing Rice out of her district.

[EastAnglianLefty]

Dems have supermajorities in the Assembly and State Senate, so if they remain united they can overrule the Commission. However, out of idle curiousity I thought I'd take a look at the Assembly to see what sort of map they might be overruling (if they needed to.)

I started by trying to assign whole numbers of seats to county groupings, then within them as much as possible tried to avoid dividing municipalities (and where I had to, I tried to do it along natural lines and to keep areas of continuous development in the same seat.) I also tried to maximise the number of VRA seats and where that wasn't possible to create plausible opportunity districts. It's much more of a good government methodology than the Commission is likely to produce, but it was easier and more interesting to work to those strictures than to work out what a bipartisan incumbent-mander would look like.

The map is here: https://davesredistricting.org/join/62046ddf-178d-4951-9dde-623fea8d9aa8

On 2016 PVIs, 100 lean Democratic and 50 lean Republican. There are 18 black-majority seats, two black-plurality seats and another three where I'm reasonably confident they're a majority or plurality in the Democratic primary. For Hispanics, the figures are 13, 5 and 2 and for Asians the figures are 2, 3 and 2. I don't know NYC well enough to know when I've sliced together eg African-American and Caribbean-American areas, but if anybody wants to point out things like that I'd be interested to hear it.

If I get round to it, I'll summarise what I did in different areas of the map later on this week.

[EastAnglianLefty]

I'm going to start off at the end, with the western NY districts, which I'm here defining as Syracuse, Binghamton and points west. There are thirty districts here, which I assigned to the following groupings:

• Oswego/Onondaga/Madison: 5
• Broom/Tioga/Cortland/Tompkins: 4
• Cayuga/Seneca/Schuyler: 1
• Chemung/Steuben/Yates/Ontario/Wayne: 3
• Monroe/Livingston/Wyoming/Genesee: 7
• Orleans/Niagara: 2
• Erie: 7
• Allegany/Cattaraugus: 1
• Chautauqua: 1

These aren't the only possible groupings and I did consider some alternatives. In particular, Ontario/Yates; Wayne/Seneca and Cayuga/Cortland all just about fall within 5% of the acceptable population deviation, but the issue is that they're all at the bottom end of the range and it makes it really tricky to get the districts in the Southern Tier small enough without splitting counties and municipalities all over the place. Chemung/Tioga is also a pairing that works, but only if you slice up Tompkins County in a distinctly ugly fashion.

I think most of the groupings are fairly easily to justify, except for the Wayne-Chemung group. However, whilst that's ugly taken as a whole, I think each of the districts works reasonably well on its own as representing a distinct community (Rochester exurbs/Finger Lakes/Elmira-Corning) so I'm happy enough with it.

Syracuse area



AD-118: Oswego County and parts of Cicero Township in Onondaga to the north of State Highway 31. Splitting a township wasn't ideal, but all the ones that border Oswego County are too large to go in in their entirety. R+5.4.

AD-119: Northern suburbs of Syracuse, plus a tiny slice of the city proper (which is slightly too large to remain intact.) D+3.8.

AD-120: The rest of Syracuse. Only 50% white by total population, but nearly 60% white by CVAP so can't realistically be called a minority opportunity seat. D+29.5.

AD-121: The western portions of Onondaga County. The rural parts are moderately-red, the Syracuse suburbs are light-blue and they average out to produce a toss-up seat at a presidential level, though in practice Onondaga is still pretty Republican down-ballot. R+0.8.

AD-122: Eastern Onondaga and Madison. Similar to AD-121, except the suburban portions are slightly more diverse and Democratic, and that's enough to give it a mildly Dem PVI. D+1.0.

Binghamton and Ithaca

AD-126: Tioga County, most of Cortland County and the rural portions of Broome County. R+11.0.

AD-127: Binghamtom and the surrounding townships containing part of its urban area (Dickinson, Vestal and Union.) D+4.6.

AD-128: This is deliberately drawn as a college district, so as well as Ithaca it also takes in SUNY Cortland. D+16.8.

Finger Lakes

AD-129: By splitting Steuben north-south, you can get a decent Elmira-Corning district, which you can't do if you keep Steuben whole. R+8.7.

AD-130: Yates, northern Steuben and the bulk of Ontario County. R+8.3.

AD-131: Cayuga, Seneca and Schuyler counties. Voted for Obama fairly comfortably in 2012, but Trump won by about 7000 votes this year. R+3.1.

AD-132: Wayne County and the northern tier of Ontario County townships. The latter are a little less red than the former, though not enough to make it remotely competitive. R+9.1.

Rochester area



AD-133: Perinton, Penfield and Webster townships in eastern Monroe. D+0.03 and the first two townships actually swung towards Clinton in 2016, so probably a little more Democratic-inclined by now than that indicates.

AD-134: Townships south of Rochester and east of the Genesee river. Most of this territory is reasonably marginal, but Brighton township makes this safely Democratic. D+9.6.

AD-135: An inverted-C district, made up of Irondequoit and the whiter bits of Rochester. D+17.8.

AD-136: Western, central and inner-northern Rochester, taking in most of the black and Hispanic residents of the city (by total population, whites are only the third largest group here.) Black-majority. D+38.5.

AD-137: Stretches from Greece, on Rochester's north-western boundary, to the Orleans County boundary. The territory covers goes from being Republican-leaning to staunchly Republican (and this is the bit of the county that swung hardest to Trump in 2016). R+5.7.

AD-138: South-western Monroe and northern Livingston. The former is one of the more conservative bits of its county, the latter is the most liberal portion of its county, which means both parts vote more or less the same way. R+3.7.

AD-139: Genesee and Wyoming Counties and the bulk of Livingston. If it wasn't for southern Staten Island and Borough Park, this would be the most conservative district in the state. R+17.6.

Orleans and Niagara

AD-140: The entirety of Orleans County and around three quarters of the land area of Niagara County. R+13.3.

AD-141: Niagara Falls, North Tonawanda and their hinterland. Not quite the most Democratic district you could draw in Niagara County (technically you could draw a Lockport-Niagara Falls district that Biden narrowly won) but it's pretty close. R+2.0.

Erie area



AD-142: Grand Island, Tonawanda and a little bit of north-east Buffalo. Technically you can squeeze Buffalo into two whole districts, but this does a better job of keeping other municipalities in the county whole. It also burnishes the Democratic margin slightly, though even without it this would still lean D. D+2.4.

AD-143: Technically Amherst is large enough to stand alone, but if you do that it causes knock-on problems elsewhere. So this district grabs around 10k electors from Clarence to make up the numbers, which also makes it a little more marginal. D+1.1.

AD-144: The Erie black-majority district. You could increase the black percentage a lot more, but it's not necessary so I prioritise compactness. D+37.6.

AD-145: Lackawanna, South Buffalo and the Lakeshore. D+21.6.

AD-146: Cheektowaga and West Seneca. A classic rust-belt swing district, which went Obama-Trump-Biden. D+0.2.

AD-147: Eastern Erie County. Securely conservative. R+13.1.

AD-148: Cattaraugus and Allegheny Counties. R+15.6.

AD-149: South-western Erie County. The northern townships here are swingy, but the southern ones are Republican enough to keep this from becoming competitive. R+7.0.

AD-150: Chautauqua County is the only county that's exactly the right size for a single district. R+9.7.

[EastAnglianLefty]

Alternative arrangement for Erie County, which only splits Buffalo once and Amherst once. Biden won all but one district and the south Buffalo district will probably be minority-majority by 2030, so Democrats would find much to like here:

https://davesredistricting.org/join/e6c051e1-357f-47f4-a766-e55a4b6d19b7

[EastAnglianLefty]

Next up, northern NY, the Mohawk Valley and the Albany area. 16 districts in this group, in the following combinations:

• Delaware/Greene/Schoharie: 1
• Columbia/Rensselaer/Albany: 4
• Schenectady/Montgomery/Fulton: 2
• Washington/Saratoga/Warren/Hamilton/Essex: 3
• Clinton/Franklin: 1
• St Lawrence/Jefferson/Lewis: 2
• Chenango/Otsego/Herkimer/Oneida: 3

Alternate county combinations are possible, but the only one that I think merits further investigation is combining Albany and Saratoga and throwing the rest of their respective groups together, so that you don't have any districts crossing the Hudson.

Mohawk Valley

AD-125: Chenango County, Otsego County and most of Herkimer County. A bit of a leftovers district, if I'm honest. R+7.9.

AD-124: Utica, some of its suburbs, and towns along the Mohawk as far as Herkimer. I would have preferred to keep it solely in Oneida County, but doing that makes AD-123 look too weird. Somewhat marginal. R+3.0.

AD-123: The rest of Oneida County. R+12.8.

The North Country

AD-117: Lewis County and the bulk of Jefferson County. Watertown is the largest settlement here and is somewhat marginal, but the rest of this district was firmly Republican even before 2016. R+9.2.

AD-116: St. Lawrence County and north of Jefferson County. Voted strongly for Obama in 2012, but swung hard towards Trump in 2016 and that looks to have continued in 2020. R+1.15 but these days significantly redder than that implies.

AD-115: Franklin and Clinton Counties. Went for Trump in 2016, but this district seems to have trended back to the Democrats in 2020. D+4.9, though given Republican strength downballot it's probably more of a toss-up in the Assembly.

AD-114: Essex County, Warren County, Hamilton County and northern Saratoga County. Essex County leans Democratic, Warren County is finely balanced and Hamilton and northern Saratoga will push this into the Republican column in most years. R+3.2.

AD-113: The entirety of Washington County, plus Saratoga Springs and points east. In 2020, the latter narrowly outvoted the former leading to a Biden victory. R+1.2.

AD-112: Southern and western Saratoga County. Seems to be trending somewhat Democratic and Biden won this by about 4000 votes, but still fundamentally a marginal seat. I suspect a Democratic gerrymander would try to throw Saratoga Springs in here to guarantee one seat in the area. R+2.7.

Capital Region



AD-111: Montgomery and Fulton Counties and the town of Glenville from Schenectady County, using the Mohawk as the southern boundary. R+8.7.

AD-110: The rest of Schenectady County. D+6.

AD-109: This seat combines Troy and northern Rensselaer County with Watervliet, Cohoes, Green Island and about half of the town of Colonie from Albany County. Whilst it is a trans-Hudson seat, the two portions of the district have good links and are fairly similar to one another. D+4.4.

AD-108: Most of the land area of Albany County. D+4.8.

AD-107: Albany and the town of Bethlehem. D+25.4.

AD-106: Columbia County and southern Rensselaer. D+1.3 and would appear to be trending more Democratic.

AD-105: Delaware, Schoharie and Greene Counties. R+11.9.

[EastAnglianLefty]

Moving on to the Hudson Valley, which gets 18 districts, in the following groups:

• Sullivan/Ulster: 2
• Dutchess/Putnam: 3
• Orange: 3
• Westchester/Rockland: 10

In terms of groupings, Westchester definitely needs a partner. Conceivably you could put it with Putnam instead, then group Orange, Sullivan and Rockland, but you can't pair Ulster and Dutchess for a whole number of seats so you'd have to combine both with areas to the north. I'm still not wild about a district covering both sides of the Tappan Zee, but I think the alternative is messier.

Hudson Valley

AD-104: Sullivan County and southern Ulster County. The 2020 township results for Ulster County aren't up yet, but I suspect Trump held on here. R+2.2.

AD-103: The rest of Ulster County. D+9.9.

AD-102: This district stretches along the Hudson from Poughkeepsie to Rhinebeck. Ended up as something of a Democratic pack, which was a consequence of most of the Dutchess County's major roads running north-south and the populations of its various townships. D+7.8.

AD-101: The eastern end of Dutchess County, with the awkward edition of outer suburbs of Poughkeepsie. R+7.9.

AD-100: Putnam County, plus Beacon and Fishkill. There aren't enough Democratic votes in the latter to outvote the former. R+5.2.

AD-99: Orange County is the right size for three whole districts, and this is the one covering Newburgh. D+1.5.

AD-98: And this is the Middletown district. R+1.5.

AD-97: And this one takes in Kiryas Joel and the south of the county. R+4.7.

Westchester and Rockland



AD-96: This district is co-extensive with the town of Ramapo. D+1.4.

AD-95: The towns of Haverstraw, Stony Point and enough of Clarkstown to make the numbers work. D+2.2.

AD-94: North-west Westchester. There are enough Democratic votes in Peekskill to make this fairly secure, even without trends. D+3.9.

AD-93: North-east Westchester, Mount Kisco, Ossinning and the Pleasantville area from Mount Pleasant. Looks uglier than it actually is if you account for where the population is. D+11.0.

AD-92: Eastern Westchester, along the Connecticut border. 25% Hispanic by total population, but only 16% by CVAP. D+8.2.

AD-91: White Plains, Scarsdale and eastern portions of the town of Greenburgh. Fairly close to being minority-majority by total population. D+19.9.

AD-90: This is the district that crosses the Tappan Zee, and it contains portions of three separate towns (plus another in its entirety.) That's got more to do with the large size of Westchester towns, though - on the ground its borders do follow settlement boundaries reasonably well. D+8.7.

AD-89: Technically you could draw two seats just using Yonkers and Mount Vernon, but the most likely outcome is that you'd produce two white-plurality seats, so I opted to group them with New Rochelle and some smaller municipalities. This is the western Yonkers seat. By CVAP it's 35% white, 37% Hispanic and 23% black. That might not quite be enough to perform as of today, but by total population it's 46% Hispanic so it should do by the end of the decade. D+23.7.

AD-88: Eastern Yonkers and northern New Rochelle. 72% white by CVAP. D+4.0.

AD-87: Mount Vernon and southern New Rochelle. 46% black by total population and CVAP, which is comfortably enough to make this perform for VRA purposes. D+29.3.

[EastAnglianLefty]

I managed to find a new arrangement for Upstate for the Assembly which minimises municipality splits and eliminates some of the more awkward districts I'd drawn: https://davesredistricting.org/join/9e483f55-2dd7-419c-aad2-9fce0ebf24c1

The downsides are that Syracuse is split 50:50 between districts (though that could theoretically be eliminated by swapping territory between 123 and 124 if you wanted); Oneida County is split four ways; and that 95 is technically not quite contiguous by road (because the on-ramp for the Bear Mountain Bridge is in Orange County) but those are all pretty minor issues. Overall I'm much happier with this version.

[EastAnglianLefty]

For NYC and Long Island, it's worth looking at the state assembly quotas by county/borough:

• Bronx: 11.00 districts
• Manhattan: 12.48
• Staten Island: 3.63
• Brooklyn: 19.89
• Queens: 17.57
• Nassau: 10.37
• Suffolk: 11.38

Obviously, the Bronx can stand alone, but elsewhere there are separate options.

Staten Island needs a partner. Numerically Manhattan works, but a connection via the ferry isn't ideal and whilst you can get 2 districts out of the Financial District and the North Shore, this tends to make it difficult to draw a a minority-majority district on the North Shore. Alternative, you can connect to Brooklyn via the bridge, but that gives you a quota of 23.52 which either means 23 small seats or 24 large ones. You could add in Manhattan to fix that problem, but then you have an extra cross-borough district.

What's more, if you group those three then you either have to squeeze 18 seats out of Queens when there's only the population for 17 and a half, or you have to pair Queens with Long Island. If, on the other hand, you don't put Manhattan in with SI or Brooklyn, then Queens and Manhattan pair nicely.

Finally, both the Long Island counties can theoretically stand alone for 10 and 11 districts respectively, but that forces all the districts to be large and makes it hard to respect town boundaries. You can credibly argue that Long Island towns are mostly too large for this to matter, but that's a rather different methodology to the one I've followed elsewhere. Alternatively, you could pair them for 22 seats (a little smaller than average but not enough to cause many problems) or you could put them (or conceivably just Nassau) in with Queens.

From a Democratic perspective, I think the optimal solution probably gives 10 seats to Nassau, 11 seats to Suffolk, pairs Staten Island and Brooklyn for 24 seats and pairs Manhattan and Queens for 30 seats. The Republican perspective would probably favour assigning 22 seats to Long Island, pairing Staten Island and Brooklyn for 23 seats and pairing Manhattan and Queens. I think the latter is slightly easier to justify, so I've gone with that. YMMV.

So, with that settled, let's turn to Long Island:

Suffolk County



AD-1: The five eastern towns of Suffolk County make a certain amount of sense as a district, but they're just slightly too large for a single district, and given that links between the North and South Forks aren't great, I thought it was easier just to combine those towns with Brookhaven for five districts. This district covers the Hamptons and Moriches. R+0.8, though you could easily flip that by swapping parts of Brookhaven for Shelter Island.

AD-2: This district is made up of the North Fork and areas along the north coast as far west as Sound Beach. R+6.8.

AD-3: A compact district centred on Port Jefferson. R+2.4.

AD-4: This district takes in parts of Brookhaven north of the Long Island Railroad. R+4.5.

AD-5: Another south coast district, stretching from Patchogue to Mastic Beach. R+1.1.

AD-6: Smithtown is slightly too small for a district of its own, but it pairs nicely enough with Islip (other options are available.) This district pairs Islip itself with the east of the town, in order to allow AD-8 to become a minority-majority district. R+10.1.

AD-7: And this is the Smithtown district, with the bits of Islip is grabs serving the same purpose as in AD-6. R+12.9.

AD-8: Hispanic-plurality seat centred on Brentwood and Central Islip. 58.9% Hispanic by population, 46.8% by CVAP. Whites are 25.6% CVAP, blacks are 23.9% CVAP so this ought to perform without trouble. D+28.7.

AD-9: This district combines parts of the towns of Babylon and Islip, running from Bay Shore to Lindenhurst. R+6.9.

AD-10: I wouldn't claim this L-shaped district is lovely, but it is possibly a viable coalition district.  Whites are 53.4% by CVAP (and the whitest parts of this district are pretty red), blacks are 24.7% and Hispanics are 17.3%. D+10.0.

AD-11: Huntington then pairs with Oyster Bay, North Hempsted and Glen Cove for 6 districts. This district contains the north and east of the town, including parts of Huntington itself. R+4.0.

AD-12: And this is the cross-county district, which combines minority-heavy bits of Huntington with Plainview in Nassau County. D+3.7.

Nassau County



AD-13: Southern Oyster Bay - the Peter King district, if you will. R+13.2, the most Republican district in Long Island.

AD-14: The remaining three districts are all orientated north-south, although now that I think about it arguably an east-west orientation would be better, both in terms of compactness and producing minority-influence seats - I will probably revise this later. This district stretches from Hicksville to Glen Cove. R+2.1.

AD-15: Westbury and the Cow Neck Pensinsula. Only 67% white CVAP, but the minority population is split equally between blacks, Asians and Hispanics. D+6.6.

AD-16: Mineola, New Hyde Park and Great Neck. D+1.3.

AD-17: Hempstead and the City of Long Beach are covered by six largeish districts. This one is designed to absorb heavily-white areas near the LIR and is accordingly relatively Republican - R+5.7.

AD-18: Levittown and the southern shore of the town as far as Merrick. R+8.1.

AD-19: This Freeport-Baldwin-Rockville district is rapidly diversifying, but for now it's still white-majority. D+11.2.

AD-20: Whereas this Hempstead-Uniondale district is black-majority by CVAP, though by total population it's 39% Hispanic and less than 45% black. D+29.8.

AD-21: This Elmont-Valley Stream district is designed as a black-opportunity district and is plurality black by total population. CVAP is 39% white, 30% black with growing Hispanic and Asian populations. To make a stronger district you'd either have to cross the NYC boundary or send a tendril towards Hempstead. D+15.0.

AD-22: Long Beach and the Five Towns. Internal connectivity could be better, but this was a leftovers district and by that metric it's good enough. R+3.5, which is entirely down to the Orthodox communities in Cedarhurst, Woodsmere and Lawrence having voted Republican by margins that wouldn't be out of place in Borough Park.

[EastAnglianLefty]

Here's an amended version for Huntington and northern Nassau:



The affected districts are 11, 12, 14, 15 and 16. 14 and 15 become North Shore-focused districts; 16 fits neatly between the LIR and the Long Island Expressway (and becomes minority-majority by total population); 12 is now a reasonably compact Hicksville-Plainsview district.

New PVIs:

11: R+4.3
12: D+2.9
14: R+0.3
15: D+3.6
16: D+3.6

I also played around with an alternative scheme for Hempstead in which 21 took in the Five Towns rather than Franklin Square, since white voters in the former are more Republican and this means it's more likely the district might perform. In the rejig this forced, 19 became minority-majority but at the cost of 20 becoming white-plurality. Which I could see working as a method of maximising Democratic performance, but I'm not sure it really improves minority representation so I rejected it.

[EastAnglianLefty]

We now come to NYC, and I should probably say a word about my methods here. Whereas elsewhere in the state I prioritised compactness, that was primarily a consequence of those areas being overwhelmingly white (and where there was a large enough minority community to control a district, they're generally geographically concentrated anyway.) Here, on the other hand, the VRA is almost constantly in play.

As I was aiming to maximise minority representation and to avoid packing (even where the pack has very compact lines), a certain amount of salamanders are unavoidable. I tried not to do this more than necessary and to avoid tendrils narrower than a couple of blocks, but the resulting districts are still much less compact than elsewhere in the state. Additionally, isolated white neighbourhoods and minority communities which aren't large enough to form a majority in a district of their own often got the short end of the stick. Some of this could have been averted if I'd been more willing to draw districts that were only 45%+ CVAP, but I thought it better to err on the safe side even if it meant uglier districts.

I also didn't distinguish between different Asian, Hispanic or black communities. It may therefore be that these arrangements unfairly privilege Chinese communities against South Indian communities or African-American communities against Caribbean communities, to pick two examples at random. There I can only plead ignorance.

South-East Queens



AD-23: The Rockaway Peninsula is exactly the right size for a district. This is black plurality by population, but very narrowly white plurality by CVAP (38.4% vs. 37.6%.) Given that the white vote here is pretty Republican, I strongly suspect this will elect a black candidate. On that basis, if you aren't combining Queens and Nassau then there's pretty much no reason not to draw this seat. D+18.2.

AD-24: Contains JFK airport and various communities surrounding it. 54% black by CVAP. Howard Beach is added in to minimise packing and to make it more like AD-25 will perform. D+33.7.

AD-25: Ozone Park, Woodhaven and Richmond Hill. Total population is 17% W, 49% H, 27% A, but CVAP is 24% W, 44% H, 25% A. Would probably elect the candidate of choice of Hispanic voters, but I wouldn't stake money on it. D+27.9.

AD-26: An ugly snake stretching from Laurelton to the outskirts of Ozone Park. Ideally you'd want 26 and 27 to be divided east-west rather than north-south, but doing so eliminates a black-majority district. D+43.5.

AD-27: St. Albans-Morris Park. 64% black by CVAP, making it the blackest district on this map outside Brooklyn.

AD-28: Cambria Heights to the Grand Central Parkway. Surprisingly neat-looking all things considered. D+40.8.

North-East Queens



AD-29: Drawing three Asian-majority districts by total population in NE Queens isn't hard, but it usually means that all are white plurality by CVAP. So in order to shore up 31 and 32, this district is Asian-plurality by total population (47% vs. 35% white) but white plurality by CVAP (42% vs. 40%.) D+16.6.

AD-30: Takes in the whiter bits of NE Queens, plus College Point. The most Republican district in Queens, but still D+6.0.

AD-31: Asian-majority district in Flushing (65% by total population, 52% by CVAP.) D+19.5.

AD-32: Not quite Asian-majority, but a very strong plurality (48% vs. 21% white by CVAP) which is good enough. D+26.6.

AD-33: Forest Hills/Kew Gardens/Kew Gardens Hills. White-majority by CVAP. D+17.0.

North-West Queens



AD-34: This district, stretching from Forest Hills to Maspeth, is designed as a white-sink in order to try and make 35 into a Hispanic-opportunity district. D+7.3.

AD-35: A district centred on plurality-Hispanic areas near the Brooklyn border. Unfortunately, you run out of those quickly, which means this district is white-plurality by CVAP (47% vs. 41%.) D+23.8.

AD-36: Asian-majority district extending from Elmhurst to Jackson Heights via Woodside. D+28.1.

AD-37: Hispanic-majority Corona/Jackson Heights district. D+35.4.

AD-38: Hispanic-majority East Elmhurst/Ditmars Steinway district. D+33.8.

AD-39: A compact Astoria-based district. D+32.3.

AD-40: 60% of this district's population lives in Queens, around Long Island City and Sunnyside. The remaining 40% is in the Upper East Side and given differential turnout rates in primaries it probably controls the district. D+30.6.

[EastAnglianLefty]

As a thought experiment, I tried to create a commission map for the State Senate that might be basically acceptable to the Democratic caucus. It's not as implausible as I'd thought.

Essentially I followed neutral redistricting principles (grouping counties and minimising splits; not splitting towns unless unavoidable or they're very large; paying attention to COIs and transport links) but with a definite thumb on the scale in terms of which areas got grouped. The sort of thing you might get from a commission with a pronounced Democratic lean.

Here's what I got:

https://davesredistricting.org/join/63ff1e92-4815-48fd-a0c7-bb3cfe859e06

Long Island

9 seats here, as at present. This is probably the bit of the map I'm least happy about, because with 2016 PVIs and incumbent residences it's quite hard to give all Democratic incumbents a seat they'd like without really squiggly lines. So there are a couple of losers here.

NY-1: Northern Brookhaven and the town of Riverhead. Republican pack, R+5.5.

NY-2: Southern Brookhaven, the South Fork and Southold. Drawn as a competitive seat. R+0.7.

NY-3: This Brentwood-based seat is entirely in Islip, which limits how Democratic you can make it. Majority-minority by total population but not by VAP. You can accomplish the latter by grabbing Wyandanch in the Town of Babylon, but it wouldn't be a VRA seat so I'm doubtful a commission would go for it. This should still be Democratic more years than not. D+1.5.

NY-4: The entirety of the towns of Huntington and Smithtown. Very neat, but does put a Democratic incumbent in a seat he can't win. R+6.

NY-5: Babylon, the rest of Islip and southern Oyster Bay. Another unhappy Democratic incumbent. R+3.8.

NY-6: Central Oyster Bay and the east of the town of Hempstead, designed to suck up Republican votes. R+4.3.

NY-7: North Hempstead, Glen Cove and northern Oyster Bay. Not radically dissimilar to the current seat. I don't think Kaplan would be overjoyed by this, but it's fine. D+1.9.

NY-8: Majority-minority seat in central Hempstead, taking in the home of 6th district senator Kevin Thomas. If you weren't bothered about majority minority seats you could easily boost several neighbouring seats, but realistically I think Nassau County needs one. D+12.5.

NY-9: Long Beach and eastern Hempstead. Not quite majority-minority but should be by 2030. D+4.5.

That equates to one super-safe Democratic seat, one reliably Democratic seat, a couple that lean Democratic, a swing seat and four seats that at present are reliably Republican. Stronger Democratic maps can of course be drawn, but it's no worse than at present and it's a swingy area.

Queens

Currently there are seven districts entirely in the borough, but it now needs part of an eighth. The natural partner is Brooklyn.

NY-10: Black majority seat on the Rockaway Peninsula and around JFK. D+32.

NY-11: Asian majority seat in Northern Queens. D+12.8.

NY-12: Western Queens seat covering the entire length of the East River. Majority-minority but the primary would be controlled by white liberals. D+32.

NY-13: Corona/Jackson Heights seat. Hispanic majority with a significant Asian minority. D+35.

NY-14: Another black majority seat in SE Queens. D+43.

NY-15: The cross-border seat, taking in Ozone Park, Cypress Hills, Bushwick and Ridgewood. Hispanic majority. Addabbo lives here but he'd likely shift to NY-17 to avoid a primary with Salazar. D+36.

NY-16: The other Asian majority seat, taking in areas of Eastern Queens. Stavisky doesn't live here but must be on retirement watch anyway. D+22.

NY-17: Central Queens seat. Majority-minority but white plurality. Addabbo could run here, Stavisky could too at a push. D+14.3.

All eight of those seats are safely Democratic in all circumstances and disruption to incumbents is minimised as much as possible.

Brooklyn and Staten Island

Brooklyn currently has seven seats plus one shared with Manhattan and one with Staten Island. It keeps the existing seats and adds another seat shared with Queens. Staten Island sees little change.

NY-18: Felder's seat is renumbered but not massively altered, because it's quite an effective pack. R+9.3, not that presidential voting particularly matters here.

NY-19: The first of four black-majority seats in Brooklyn, stretching from East New York to Gerritsen Beach along the edge of Jamaica Bay. D+30.

NY-20: The second, a compact Crown Heights-Flatbush-Prospect Heights seat. D+44.

NY-21: The third, shadowing NY-19 further inland. It's ugly, but you have to break up eastern Brooklyn to get 4 black-majority seats and even so it's still about 70% black. D+36.

NY-22: Narrowly Asian majority, although I'm not sure this is VRA protected as the various Asian communities here doesn't necessarily vote cohesively. Doesn't include Gounardes' Bay Ridge base, so he'd probably run in the 26th instead. D+9.4.

NY-23: The North Shore, Coney Island and enough territory to connect them. Felder's district and the South Shore aside, this is the most marginal NYC district on the map. D+7.3.

NY-24: Most of Staten Island. R+12.5.

NY-25: The fourth black-majority district, taking in Bed-Stuy and Clinton Hill. D+46.

NY-26: White majority uber-liberal district, running from Bay Ridge to the Brooklyn Bridge and reaching inland as far as Windsor Terrace. I assume Gounardes would run here? It'd be an open seat if not. D+34.

NY-27: Based in Williamsburg, this also extends over the river to grab the Lower East Side. Brian Kavanagh lives here. D+37.

The 18th will vote for whoever can best mobilise rabbinic support and the 24th is safely Republican, but nothing else is in any danger of being competitive.

Manhattan

Currently has four seats to itself plus one shared with the Bronx and one with Brooklyn. By drawing small seats, I was able to give Manhattan five of its own seats plus one shared with Brooklyn.

NY-28: Lower Manhattan district curving round the 27th district. D+34.

NY-29: The Upper East side. D+25.

NY-30: The Upper West side. D+35.

NY-31: The Harlem/East Harlem district. You'd need super-squiggly lines to get this black-majority (and even then it probably wouldn't be by 2030) but this is a strong enough plurality to perform throughout the decade. D+44.

NY-32: Hispanic majority northern Manhattan district. D+42.

Nothing to see there, unsurprisingly.

Bronx

Currently the borough has three districts to itself, one shared with Manhattan and two with Westchester. You can simplify this to four districts in the Bronx and one shared with Mount Vernon (and no other part of Westchester.)

NY-33: South Bronx district. Less than 2.5% white. D+45.

NY-34: Not much to say here. D+42.

NY-35: There's no need to preserve the seat drawn for Klein, so this is just a compact district in the North West Bronx. D+37.

NY-36: Compact East Bronx district. Only very narrowly Hispanic majority, as that's how the lines naturally draw themselves and I thought it might protect Biaggi from a primary challenge. D+33.

NY-37: Comfortably black-majority seat centred on Mount Vernon and Eastchester. D+41.

Even less to see there.

Westchester

Right now the county is split six ways, but this map has three districts wholly in the county and one shared with the Bronx.

NY-38: Yonkers and Greenburgh are the right size for a (majority minority) district. Two State Senators currently reside in Yonkers, but Stewart-Cousins definitely has first dibs (and Mayer has a district to go to.) D+17.

NY-39: South-eastern Westchester. The town of Harrison is split, because somewhere had to be and it's not an actual settlement. Mayer would probably run here. D+12.5.

NY-40: Northern Westchester. Harckham lives here and would likely enjoy having a safer seat. D+6.8.

The Hudson Valley

NY-41: All of Rockland County bar the town of Stony Point. D+2.7.

NY-42: Stony Point, eastern Orange County and south-western Dutchess. R+1.4, which is not as safe as I'd want for a Democratic incumbent.

NY-43: All of Putnam and most of Dutchess. R+2.3. A more aggressive map would probably removed western Dutchess and make this into a pack along the Connecticut border.

NY-44: Ulster County and western Orange County. You could make this safer, but at the cost of probably conceding NY-42. Hinchey would run here. D+3.2.

Northern NY and the Capital Region

NY-45: Washington, Rensselaer, Columbia and Greene. R+1.4, although it'd be quite easy to flip this if you swapped Greene for part of Albany County.

NY-46: The entirety of Albany County. There are wasted votes here, but if you're doing a good government map then you probably do have to give the only county that's the right size for a district it's own seat. D+13.2.

NY-47: Schenectady, Fulton, Montgomery and south-western Saratoga. R+1.6, again flippable if you split Albany.

NY-48: Most of Saratoga County, from where it stretches north to the outskirts of Plattsburgh. R+1.8, so yet another seat that leans Republican but is swingy.

NY-49: Jefferson, St. Lawrence, Franklin and most of Clinton. R+0.3.

NY-50: The entirety of Oneida, Lewis and Franklin counties. R+9.3.

NY-51: Six counties west of Albany and east of Binghamton. R+6.6.

The Democrats can rely on the three seats wholly in Westchester county and the Albany seat, and another two have D PVIs. That's one down from the seven seats in the region they hold, but there are another five seats with PVIs below R+2 so Republicans would need to focus on defence in these areas and only James Skoufis amongst the incumbents has a seat with an R PVI.

Western NY

NY-52: Broome, Tioga and Chemung counties. R+4.

NY-53: Oswego county and northern Onondaga county, including about half of Syracuse. D+3.4. Whilst this is an obvious example of putting a thumb on the scales, it's effective enough and a much cleaner split of the city than is currently the case.

NY-54: The rest of Onondaga County and the entirety of Cayuga. D+5.3.

NY-55: Eastern Rochester and southern Monroe County. I think Rochester is large enough that you can argue keeping it whole is a pack, so this doesn't feel too force. D+12.3.

NY-56: The other Rochester seat. I had hoped to keep this east of the Genesee, but I had to cross in order to keep it safely Democratic. D+5.1.

NY-57: Bits of the southern tier plus Livingston County and about half of Wyoming County. R+12.5, just slightly less Republican than NY-24.

NY-58: Cortlandt, Tompkins, Schuyler, Seneca, Yates and Steuben counties. D+0.4. Alternatively, you could combine Broome, Cortland and most of Tompkins in a D+4 seat, but that's an extra county split and it's not necessary to secure a majority.

NY-59: Ontario and Wayne counties plus western Monroe. R+3.7.

NY-60: Northern Buffalo and its suburbs. D+5.8.

NY-61: Southern and western Erie County. R+9.3.

NY-62: Niagara, Orleans and Genesee counties, plus about half of Wyoming. R+10.1.

NY-63: Southern Buffalo, Lackawanna and Cheektowaga. Minority majority by total population, white majority by VAP but probably has a decent chance of electing a black candidate after Tim Kennedy moves on. D+22.

This gives every Democratic incumbent a decent seat to run in, all but one above D+5 and also creates an opportunity for an Ithaca Democrat.


Overall, such a map would create 40 districts with a Democratic PVI, 33 with a PVI above D+5 and a whole host of marginal Republican seats (11 at R+5 or below) that would force them to play defence in most cycles. That's not as good as you can get with an aggressive gerrymander, but it's secure and means a two-thirds majority would usually be in reach.

[EastAnglianLefty]

It was done from 2003 to 2013, but it was pretty unreasonable then, as it jumped over the Tappan Zee rather than using the bridge.