GOP should stop gerrymandering for its own good

Yes it is time, way past time really. Otherwise, the GOP  runs the risk of winning the battle, while losing the war. Focusing on CD districts, it seems pretty clear to me, that the net effect of gerrymanders currently in place nets the GOP a max of maybe 10 CD’s nationwide. I suspect it’s less now, as the the GOP loses cosmopolitan voters in big metro areas, while cutting into the white working class elsewhere in more exurban and rural areas, and smaller industrial towns. Meanwhile the Dems are launching a jihad to gerrymander CD’s in their favor, on the premise that the share of the pie in legislatures, should be more reflective of the share of the pie in statewide popular votes. That has some appeal, and the GOP is more vulnerable to the extent it plays its own gerrymandering game.

A case in point is PA. Just look at the graphics. This is a map I drew based on projected 2020 census numbers, in other words, a map that one might draw in 2021. Without a gerrymander, net the GOP cedes the CD that PA losses after the next census. Sure, the PVI numbers are based on the 2016 election alone, where Trump cut into the white working class in a way that might not hold for the GOP in general going forward. But that is not the point. The point is that the GOP abandoning its hideous gerrymander in PA, does not cost it much, while giving it the high ground, the better to the deflect the coming Dem jihad on this issue. It’s kind of ironic, that the GOP Dem pack of PA-17, is now having the effect of saving the seat for the Dems, that otherwise might well go down the drain. One would think, the GOP would go about the unpacking business. Yes, some seats are somewhat marginal, that could with a hideous gerrymander, be made safe GOP, but I think more marginal seats is just what the public square needs. It tends to encourage folks to run, actually interested in trying to find common ground, rather than divisive slash and burn policies, that in the end tend to be public policy failures. Just the opinion of this old man, who has been running around the track since rocks cooled.

So says this former Pub, now a Dem, but really not a very partisan person. I am more of a data based good government person, more of a technocrat that anything else. What do you think? Yes, I know it might be something of a pipe dream. In my opinion, both parties are well, not very helpful when it comes to intelligent public policy. No, they tend to be more into high school towel snapping, but I digress.


Below is a slightly revised PA map that gets a slightly higher score, using totally objective criteria. I won’t bore you as to the details as to why, unless someone is interested. But as you can see, it looks a bit better to the eye. Avoid chops, either by county or county subdivision, and  minimize erosity, and keep metro areas intact, as much as one can, to get the max score. A computer does it all. Mere humans need not apply. When it comes to redistricting, humans are just too biased. Even God may be too biased to do it. One needs to set up tight rules in advance, and let a computer do the implementation. This works guys. It really does. 

Addendum 2:

Somebody below mentioned Missouri, the “Show Me” state, expressing doubt that the rules as outlined by me, would effect anything other than aesthetic excrescence of a map for the state. So I drew the map, based on 202o census projections to find out. What do you guys think? Is it map malpractice, or a museum quality object d’art?  I feel the force is with the creator of this algorithm, yes I do.  And you guys are getting an advance preview. Isn’t that exciting? 🙂


Addendum 3:

Addendum 4; a higher scoring Michigan map

Addendum 5: New York map based on the March 2017 county population estimates projected forward to the 2020 April, 1, 2020 census date. 

Addendum 6:  New map for PA based on the latest county  census estimates for the 2020 census redistricting.

Previous Post Next Post


  • shamlet February 5, 2017 at 4:57 pm

    As someone who initially got into politics and became a partisan Republican almost exclusively as a reaction to my state’s horrible Dem gerrymander, I give this a hearty Amen. The Trump coalition is actually abnormally well geographically distributed, and fair maps really aren’t going to lose the GOP much.

    R, MD-7. Put not your trust in princes. Process is more important than outcome.

    • Torie February 19, 2017 at 12:02 pm

      Michigan: The urban-rural divide writ large

      In the opening post, I added Michigan to the pile, with a more detailed partisan stats chart because it tells an interesting political story.

      As some of you may know by now, the redistricting rules that I follow include a component that rewards drawing districts that are either nested within metro areas, or outside them, rather than including both. For larger metro areas, that metric tends to drive where the lines on the map go.

      So, in drawing up Michigan after the 2020 census, putting aside exorcising gerrymandering, the CD’s tend to hew more closely to following the urban-rural divide than they do now. And consequently, the partisan stats of the map below highlight well the urban cosmopolitans going one way, and the rest of the nation going another, at least when it comes to white voters.

      The PVI chart below is organized based on the swing to Trump from Romney. Putting aside the Macomb County anomaly (big metro area white voters swing to the same degree as their rural and smaller city compatriots), one can see that the swings to Trump (or away from him), are within the zone of the big three metro areas (Detroit, Grand Rapids, and Lansing). Everything outside that zone, swung massively to Trump (plus Macomb County).

      As to the partisan effect of this map, as compared to the existing GOP gerrymander, it depends whether you think Trump is the future, Trump is half the future, or Trump is an anomaly, and in due course, his impact on politics will disappear as if he never existed.

      If Trump is the future, then the Dems take the hit for the seat Michigan loses (the old MI-09, which disappeared, and is now the number for the old MI-14), while the GOP drops MI-11 to the Dems, but picks up from them MI-05 (who knew that a Flint-Saginaw based CD would now be a GOP CD?!).

      If the future is Trump lite, with the PVI figures using the Cook method of average the PVI’s for the last two election cycles, then each party shares the loss of Michigan’s CD, each losing half a seat. MI-04 and 11 go swing from safe GOP (for a net loss of one seat), but for the GOP MI-05 goes swing from Dem, leaving each party with a half seat loss (the Dems lose MI-09, with MI-05 going swing, but MI-11 and 4 go swing to them in exchange, also netting out to half a seat loss for the Dems).

      If Trump is an anomaly, and mere vagrant on the waters of the public square, then the map does more what would be expected: the GOP loses two seats (MI-04 as the Lansing seat is created for the Dems, rather than being gerrymandered away, plus MI-11 as that gerrymandering is also tossed out), for a net gain of 2 Dem seats, less the loss of the old MI-09 that disappears from the map.

      You choose as to what the future may hold. My guess, is Trump lite, lite, when it comes to more upscale precincts that used to have a GOP lean, and Trump lite for the balance, leaving the GOP with a true swing seat with MI-11, with MI-04 tilt Dem, but within striking distance, and MI-05 lean Dem (with the incumbent needing to work hard, and have political skill to avoid becoming vulnerable in what is potentially a highly volatile CD now).

      65 - NY-19 (D)

      • Jon February 19, 2017 at 3:59 pm

        On the Michigan districts on the map: as drawn 11 & 12 are illegal districts for Michigan. Absolutely no double crossing of counties between the same two districts allowed.

        45, M, MO-02

        • Greyhound February 19, 2017 at 4:27 pm

          The current map does something similar, as I think the VRA takes precedence. The 14th and 11th are both Oakland/Wayne districts.

          R, 27, CA-18. Anti-Anti-Trump

      • Torie February 19, 2017 at 4:45 pm

        Correct. In any event, the idea is to get a law passed requiring the use of the algorithms used here, in lieu of the currently toothless Michigan law. Ditto for the other 49 states. 🙂

        65 - NY-19 (D)

  • rdelbov February 5, 2017 at 5:09 pm

    The CA and AZ maps have left me a bitter taste in my mouth for what “fair” means.

    Even this “fair” map made partisan decisions. May split Butler county as opposed to Beaver county?

    I would suggest that fair maps all across the nation would not cost the GOP very many seats and the only reason I say this IMO maps in MA, AZ, IL, CO, OR and CA cost us nearly as many seats as our favorable maps win for us.

    The problem is I don’t Mathis or some CA professor telling me what is fair.

    • Torie February 5, 2017 at 5:17 pm

      Without getting into more, about nice rectangles, and what counties are in statistical metro areas, and so forth, spitting Beaver rather than Butler, would have almost a nil partisan effect. In other words, focus on the big picture, not this rounding error harzarai, is what I suggest. Yeah, the CA map, was a Dem gerrymander” lite,” at the time, but now, it makes zero difference, and in fact may be saving a couple of GOP seats. The GOP marketing posture nationwide, is totally toxic in CA, outside a few rather unpopulated zones. I suspect more GOP seats in CA will be going down the drain soon.

      65 - NY-19 (D)

      • shamlet February 5, 2017 at 5:24 pm

        Can we put this stupid canard to rest? The CA map was overall a model of fairness. For every Dem-friendly decision like CA-9 there’s a GOP-friendly draw like CA-21. The real issue is that the GOP baseline has eroded so dramatically there over the last 5 years. It’s coalition change, not the maps. Ironically we would have been better off with a Dem gerrymander a la what’s happened with IL-6. Even the most aggressive Dem map would have likely GOP-packed CA-49 for example.

        AZ OTOH is definitely a Dem gerrymander in disguise. IMO we should focus on getting truly non-partisan CA-style commissions where buy-in from both sides is required (instead of an AZ/NJ type winner-take-all arrangement that can easily be gamed) should be our focus.

        R, MD-7. Put not your trust in princes. Process is more important than outcome.

        • twinpines February 5, 2017 at 5:29 pm

          I strongly disagree regarding California and there are many articles out there about how the dems gamed the process in CA. Coalitions have changed, yes. But based on 2012 numbers its hard to give the dems that many more districts without extreme gerrymandering. CA-21 is by no means GOP friendly. Its actually quite easy to make a district that is super hispanic and also very GOP heavy in the southern central valley. CA-21 is basically a district gerrymandered for the dems which they keep losing because the state dems turned the water off to the central valley.

          MI SD-38, Conservative Independent, not associated with either party.

          • shamlet February 5, 2017 at 5:38 pm

            Not having the seat go into Fresno makes it winnable for the GOP.

            But there are plenty of others where a geographically reasonable decision could have made the map better for Dems. CA-36 could have been given Imperial and made Hispanic-majority. CA-25 could have been given a chunk of the San Fernando Valley. Santa Barbara and SLO Counties could have been split instead of put together to make two safe D seats. CA-1 could have been given the North Coast. CA-7 could have been given the southern suburbs of Sacramento at the expense of some of the more Republican northeastern suburbs.

            R, MD-7. Put not your trust in princes. Process is more important than outcome.

            • twinpines February 5, 2017 at 6:12 pm

              Could you make the CA-21 even more gerrymandered? Yes but that does not make it GOP friendly by basically saying it doesn’t take the seat to the absolute dem limit. CA is a strong dem gerrymander that splits counties needlessly to help maximize dem seats. It should be noted after this non-partisan map came out, we lost 4 seats in the next election and only avoided a 5th by a fluke in the top 2. Yes, an even more extreme gerrymander was possible just like a more extreme gerrymander could be done in IL. That doesn’t make the IL or the CA maps fair. As others have stated, the CA map was supposed to minimize city and county splits. Yet it completely ignored the rules it was given so it could maximize dem seats.

              MI SD-38, Conservative Independent, not associated with either party.

            • twinpines February 5, 2017 at 6:14 pm

              Due to dem self-packing its actually possible to give almost half of the congressional seats to the GOP if one wanted to and the map would not look that much worse then the current map if any worse at all.

              MI SD-38, Conservative Independent, not associated with either party.

              • shamlet February 5, 2017 at 6:20 pm

                While respecting the VRA, including S5? Show me.

                R, MD-7. Put not your trust in princes. Process is more important than outcome.

                • twinpines February 5, 2017 at 6:24 pm

                  Respecting section 5 is subjective based on the court.

                  MI SD-38, Conservative Independent, not associated with either party.

                  • shamlet February 5, 2017 at 7:31 pm

                    You can use a reasonable interpretation of the mean jurisprudence c.2011, which is what commissions and legislatures did anyway.

                    R, MD-7. Put not your trust in princes. Process is more important than outcome.

            • twinpines February 5, 2017 at 6:21 pm

              Not to belabor the point but actually going into Fresno would have hurt CA-16 to the north which has its dem base in Fresno. Any votes given to CA-21 would have come at the expense of CA-16. The southern central valley is strongly GOP which is why its hard to do what the dem “non-partisan” commission tried to do. Draw a 2dem-2rep map in strongly GOP southern central valley.

              MI SD-38, Conservative Independent, not associated with either party.

              • shamlet February 5, 2017 at 7:32 pm

                Ok. Quick and dirty version. Blue seat was 55% Obama in 2008 (and 57% HVAP) and Red seat was 58% Obama 2008 (and 51% HVAP). I believe Valadao’s actual seat was something like 51% and Costa’s actual seat was in the 57% range.

                R, MD-7. Put not your trust in princes. Process is more important than outcome.

        • Torie February 5, 2017 at 5:36 pm

          What we really need are objective rules, that minimize chops and erosity, and respect metro areas and thus minimize subjectivity. This PA map gets close to that, although the expert on this issue, a brain in the state legislature in Illinois, who will be coming out with a white paper in due course, after his metrics have been tested by computers doing monte carlo runs, would no doubt play with the lines.

          The problem with a non partisan commission without tight objective rules, is that they tend to be gamed by those advocates using subjective criteria, pushed by partisan hacks in disguise, although sure, that process is better than just “going for it,” like in the states with vomit inducing gerrymanders. And you know what? Even the vomit inducing gerrymanders don’t net much. That is what happens when the partisan divide is so geographically segregated. I know this is largely wrong, but it is as if, we hate living near folks who disagree with this politically and culturally. Perhaps many of us need to go into primal scream therapy, if only to chill out. In the meantime, I plan to run and win for an office in my 80% Dem city. That’s the beauty of a small town. It’s one on one., and all this garbage can be transcended with hard work.

          65 - NY-19 (D)

          • shamlet February 5, 2017 at 5:43 pm

            No argument there, though in my mind the best argument for something like CA’s is that you also need buy-in from the hacks on both sides. You’re never going to get a 100% objective measure of district fairness and there will always be more than one way to skin this cat. What you need is both a bipartisan tug-of-war but then objective criteria to clean up the map so you avoid the constant cycle of incumbent-protection gerrymanders that places like WA produce.

            R, MD-7. Put not your trust in princes. Process is more important than outcome.

            • rdelbov February 5, 2017 at 5:57 pm

              IMO the hallmark of gerrymandering is nasty twisted lines. One reason that refuse to call NC congressional map a gerrymander–yup yup the map favors the GOP but like Michigan’s map it has clean lines. Except within counties to equalize population.

              CA congressional map has so many double crossing of lines that I can’t list them all. I note as well the amendment was passed with the idea that counties and cities would be kept intact as a priority. Oh my goodness the map is unreal as to how cities and counties are sliced up.

              If does not count Presidential numbers, and most GOP congressional run ahead of Presidential numbers, 40 or even 42% is the generic R number–year in and year out for CA.

              18 to 22 seats should be the range for the CA GOP. Even in off years those numbers are out of reach for the GOP

              I might add that CA democrats are much more self packed then GOP voters are so 22 seems more reasonable then 18.

              • krazen1211 February 5, 2017 at 6:37 pm

                I think there is some merit to avoiding tiny chops and daisy chaining precincts if only to avoid ballot initiatives and other silliness that puts a damper on future gerrymandering.

                I take the 2002 FL map, with that fractal pattern district drawn for Clay Shaw. It’s a hypothetical, but maybe if Jeb and company don’t go overboard, we don’t have this disastrous FDF as applied by these old Chiles judges.

                Some of these things like digging out certain office buildings in Columbus seem unnecessary.

            • Torie February 5, 2017 at 6:13 pm

              Yes, but the folks on the commission in CA are picked for diversity, and thus the almost totally white GOP party in CA, had only one white GOP member, and the rest were either false flag, or “gameable,” and thus the GOP members were patsies, with one exception, a half native American GOP member with whom I had extensive discussions. As I said, now the map is more of a GOP friendly map than anything else, but that was not what it appeared then. But the system is better than the partisan spoils system definitely, I agree. But we can do better. This is too technical for laypersons to be game proofed. That is my experience.

              As an aside, I applied to be on the commission in CA, and did not get to first base. Apparently, I was not sufficiently a community activist, or too WASP or something (even as a Pub, as I was then), even though I had been drawing maps for decades, and knew this stuff cold. So it goes.

              65 - NY-19 (D)

              • rdelbov February 5, 2017 at 11:00 pm

                Lets just say that CA Ds have 10 to 13 seats where they have natural massive self packing margins. So 1/5 of the state is massively D (+80% in SF, Berkley, San Jose, LA, downtown San Diego and several college areas. Yet in the other 80% of the state where the Ds have a 52-48 margin the Rs win 1/3 of the seats.

                Just call me skeptical.

                Here are my simple rule to avoid gerrymanders.

                Respect city and county lines as much as possible. The High court is leaning towards race based neutral redistricting so if you claim to be non-partisan, like CA, avoid Maryland like nasty districts.

                I for one am for partisan redistricting but I have no problem with Michigan rules. Look at the map above for PA. Making innocent decisions(or were they partisan?) affects partisan results. Philly is a bit more then two seats–Which district gets the leftover? Bucks and Chester does not make a whole district what county fills how their CD? Why does Chester county get split three ways or even two ways?

                To mean its an easy decision. We have a party based system let the parties decide. You want nasty looking maps like commissions in NJ, AZ and CA do it.

        • rdelbov February 5, 2017 at 5:37 pm

          I just have to humble disagree between drawing a map as ugly as CA congressional map gives a nasty taste to the concept of non-partisan. PVI for CA, outside of president elections in is 40% and that translate to 21 seats in a fair map. The map IMO is bizarro in concept. Imperial valley to downtown San Diego –Ventura county split 4 ways–Long Beach to center of Orange county and so forth

      • twinpines February 5, 2017 at 5:26 pm

        What I would like seen nationwide is that elected legislatures, not so-called non-partisan bureaucrats draw the lines. However those elected legislatures should be required to meet national standards such as no double crossing county lines, minimizing city splits, etc. This would keep things in the hands of elected officials while at the same time minimizing extreme gerrymanders. Basically Michigan style laws would be my preference. MI is probably the most dem packed state in the country so I doubt it would benefit us that much nationally but due to self-packing, fair maps rules would give our party a slight benefit at the moment. In the future, of course that is always subject to change. But fair national standards would be great and since we have a trifecta, this is a good time to do this.

        MI SD-38, Conservative Independent, not associated with either party.

        • Torie February 5, 2017 at 6:03 pm

          The MI rules are toothless, because you can pick where the chops are, and bridge chops are just fine. So it does not slow down much gerrymanders. In fact, the MI map generates more GOP seats net than any other map anywhere in the nation for either party, or at least did, prior to the Trump thing upsetting the apple cart. Still it netted the GOP at least 2 seats in the Trump era, if not necessarily about 3.5 seats as in the pre-Trump era.

          65 - NY-19 (D)

          • twinpines February 5, 2017 at 6:07 pm

            Any rules are going to be at least somewhat toothless. Gerrymandering will never cease. What should be done is put limits how extreme. MI does prevent the baconmandering we see in chicago and other places.

            MI SD-38, Conservative Independent, not associated with either party.

    • Torie February 19, 2017 at 4:42 pm

      Correct. In any event, the idea is to get a law passed requiring the use of the algorithms used here, in lieu of the currently toothless Michigan law. Ditto for the other 49 states. 🙂

      65 - NY-19 (D)

  • Son_of_the_South February 5, 2017 at 6:02 pm

    Washington State’s system is the only one that’s truly fair and useful in every state. All there proposals and systems that I’ve ever seen are either inapplicable elsewhere (IA), gameable systematically (AZ, NJ), or gameable in the cases of individual districts (CA).

    24, R, TN-09
    Classical liberals are a minority. Fusionism is the answer.

    • shamlet February 5, 2017 at 6:08 pm

      Problem with the WA system is that you can easily get an endless cycle of incumbent-protection gerrymanders. WA-6 for example.

      I do wonder what the CA system would look like elsewhere. I think a big part of the problem is that CA is so huge (and thus has so many lines to settle) and also has huge subdivisions that don’t provide much guidance for lines. Following county lines for example does zero good for half the districts or more. I think the same system would be far less gameable in a smaller state with smaller divisions.

      R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • Son_of_the_South February 5, 2017 at 6:41 pm

        Yeah, but even if WA-06 was drawn neutrally, it would still be fairly safe for the Dems, at least for a while. I’ve tried it and the real problem is the populated section of Jefferson County on the sound. Even if you get rid of Tacoma, that seat would elect a Democrat in most normal circumstances for the foreseeable future. Furthermore, even though WA-08 was gerrymandered somewhat to help Reichert, coalition shifts mean that his seat will likely be competitive in the future. You could say the same for a Tacoma-less WA-06, but more shifts on the peninsula could get you there anyway.

        24, R, TN-09
        Classical liberals are a minority. Fusionism is the answer.

        • shamlet February 5, 2017 at 9:54 pm

          Just because it’s not ugly or maximally effective doesn’t mean it’s the best plan from a COI standpoint. The whole pacific coast should be in WA-6 for example, and as we litigated before the WA-1/2 split doesn’t make a lot of sense.

          This is why I like the CA system. You need to triangulate off of three poles: Dem partisans, GOP partisans, and nonpartisan objective criteria, and find a map that passes muster with all three.

          R, MD-7. Put not your trust in princes. Process is more important than outcome.

          • Greyhound February 5, 2017 at 10:46 pm

            We should try to do some test maps. I mean, the RRH community is like the PERFECT set of people to try this out on a bunch of other states too.

            R, 27, CA-18. Anti-Anti-Trump

    • Jon February 5, 2017 at 6:25 pm

      For congressional districts, I’m thinking Michigan style rules (including absolutely no recrossing) combined with a rule to prefer contiguous whole county(ies) adding up to within 1% deviation over random rural counties having to be split.

      For state senate & house districts, several states already have strict county integrity requirements.

      I’m also a fan of those states which which a county is at or above ideal population for 5+ state house seats rounding the seats the county gets to the nearest whole number and having them all stay within the county.

      45, M, MO-02

      • rdelbov February 5, 2017 at 11:07 pm

        I tend to agree with that Jon. I add that I prefer partisans to decide which counties to split and what counties not to split. I hate that an indie got to decide to split Pinal and Pima three ways as opposed to two way split for Pima and not splitting Pinal at all. These are real communities and attaching Pima county to Northern AZ was a shame.

        My biggest complaints with the CA maps were hideous number of splits and double cross cities and counties.

        • rdelbov February 6, 2017 at 7:42 am

          Two last thoughts

          1st I prefer a Michigan style map for states like KY-IN-TN-MO. I prefer to lock in seats for the GOP and not to bacon strip the large urban counties. I think I have made myself clear over the years on that. Under certain circumstances I can see a three way of Marion and Davidson counties in IN and TN but probably would prefer not to.

          2nd other Iowa and its computer the concept of non-partisan or bi-partisan or fair redistricting is a total myth. You can set standards and rules and let the computers whirl for other states but whoever writes the standards will be setting in biases that will benefit either party. For instance in CO IMO it will help the GOP if make it a rule that no county can be split more then once. If you make a rule that any county can be sliced and diced by your computer as many times as needed to equalize population then it might benefit the Ds. I am just laying those without knowing but whatever standard you set will inevitably help one party or the other. I note again the most hideous county splitting and city splitting lines of 2012 cycle IMO were done by commissions in CA and AZ. I found it that the poster suggests that the GOP get out of it but comments not on D maps in MD-IL-MA as well our commission maps.

          • rdelbov February 6, 2017 at 10:16 am

            I know its probably pointless to respond to my own posts but this is a subject that I am passionate about. It is why I was at SSP and now here.

            1st talk about your alternate facts-calling any map you don’t like a gerrymander is an abuse of historical facts. The term gerrymander was used to describe weirdly distortly draw seats. The next use of the term was when states created seats of unequal size. Yes yes until 1960s congressional and legislative seats were grossly under or over populated in many states.

            2nd maps that produce disparate results like CD maps for Michigan and NC are called Gerrymanders but yet the lines are clean (from a historic basis ) and they are equally populated. States like OR, MA, CA, KS, AR, WV, TN, NJ among others produce disproportional results for a particular party and yet are these maps Gerrymanders?

            Call your 1984 references –Ms Mathis from AZ gets kudos for her fair legislative maps but the variations in population and are full of weird jagged lines.

            I note for years we saw no county splits in AR congressional maps (1981-1991-2001) but the GOP was packed into one seat The other GOP areas were exactly split into the other seats to make it hard for any R to win. Was it a gerrymandered map-no it was just drawn to benefit the Ds. The seats were about compact as can be and likely an Iowa type computer plan might have drawn seats just like it. For that matter the GOP hopes for extra seats in Iowa rests on one variable in the CD map. If the Western boundary of the state is placed in one CD (ergo the GOP is packed into CD4 or whatever is named) you will see the GOP struggle to win 2 or 3 seats in a 4 district map. The next map could be a disaster for the GOP and its all about whether western Iowa is packed into one CD

    • rdelbov February 19, 2017 at 6:02 pm

      WA system is 4 guys(all very political) trying to agree on a plan that gets 3 votes with a very liberal court as a fallback

      no Thanks

  • Republican Michigander February 6, 2017 at 12:27 pm

    The questions is this.
    1. Define gerrymandering.
    2. What’s the solution.

    A lot of people here like the APOL standards in Michigan, but if you ask just about any democrat in this state, it’s going to get called a major league gerrymander. To them, anything that is not 55% D favorable is a gerrymander.

    I oppose any “commission” for two reason. 1. It’s not nonpartisan, no matter what they say. 2. There’s no accountability to the commission. If I don’t like the map my legislator supports, I can at least toss that person out (theoretically).

    If VRA is modified a bit, we could see less ugly districts like MI-11 and MI-14. I won’t defend those two. MI-11 is shaped that way because of MI-14.

    The main rules I would like to see are as follows.
    In order:
    1. Minimal county breaks.
    2. Minimal municipal breaks (city, township, unincorporated area) – that’s going to happen with exact vote splits.
    3. No bacon-stripping.

    Reps are supposed to represent districts and communities. If a party is self-packed in certain areas, well that just the way it is. I can’t help it that in Michigan, democrats are self-packed into a few regions – Detroit, Southern Oakland east of Novi Road, Pontiac, Southern Macomb, Flint, Lansing/East Lansing, Eastern Washtenaw, Saginaw, Muskegon, Grand Rapids, Kalamazoo. I can’t help it that my base R county is next door to some heave D areas. It’s the way it is. Baconstripping part of Ann Arbor to my area to create 51% D districts is gerrymandering as well and not fair to either of our communities.

    MI-08 - Michigan is a red state again. We need a 50 state strategy and an 83 county strategy.

    • rdelbov February 6, 2017 at 1:58 pm

      R Michigan–I agree.

      I chortle when folks call particular maps gerrymanders. NJ is a good example. Until 2016 the delegation was 6R-6D while OR is 4D-1R. I have had people argue with me over and over that OR is not a gerrymandered map for the Ds but both states are clearly blue but one is 50-50 while the other is 80-20. So what is the standard for calling one a gerrymander and the other a model of fairness?

      I also chortle at the idea that the Ds are responding with maps in IL or MD or MA because of what the GOP did in MI or OH or NC.

      The entire South for generations was drawn (at least as far as I know from 1961 to 1991) to benefit the Ds. The 1982 CA congressional and legislative maps changed that state forever!! The idea that the GOP needs to play fair or the Ds will do retribution flies in the face of history. The democrats rigged most of the maps in the states from 1932 to 1992 and rarely if ever showed any mercy to the GOP.

      I do not buy that any national standard will be truly fair.

      I might add that when Obama talked about redistricting reform last fall and all of the MSM went on about his next national crusade–his effort to fix the system. That chatter stopped when the press found out that his idea of reform was putting democrats in charge of redistricting in as many states as possible why taking away GOP control. Oh yes reform–Chicago style.

  • Republican Michigander February 6, 2017 at 12:38 pm

    Interestingly, I just got a call from a Durham, North Carolina number about “bipartisan redistricting reform” pitch in Michigan.

    MI-08 - Michigan is a red state again. We need a 50 state strategy and an 83 county strategy.

  • Izengabe February 6, 2017 at 12:49 pm

    Unilaterally disarming or enacting redistricting reform on a state by state basis is insane. If we want to end gerrymandering the proper way to do this is via a constitutional amendment which establishes national standards or even some sort of national commission for redrawing lines.

    Follow me on Twitter: @Izengabe_

    • Jon February 6, 2017 at 7:22 pm

      Missouri is one of those states that could be done before hand without costing us anything.

      (Basically it would result in the unneeded double cross between St Charles & Jefferson counties between 2 & 3 and also #5’s bypassing part of Jackson county to take in an exurban and two rural counties beyond it.)
      The state legislative lines here are already drawn by commissions when they can come to an agreement and the courts when they can’t.

      45, M, MO-02

      • rdelbov February 6, 2017 at 9:34 pm

        MO congressional lines(especially MO5) seem over down now. I would likely just suggest doing Jackson plus 100K from Clay. I would make sure that MO6 got only the reddish of the Red counties north of the Missouri river but it should be fine in that form. I think the Trump number would still be near 59 or 60%?

        MO3 would not be hurt if lost part of St Chalres-and MO3-MO2-MO8 swapped some area.

        The only way IMO to get a third CD for the Ds in MO, other then having the GOP nominate really bad candidates in an open seat situation, is to combine MO1 and MO2 together. Attach downtown St louis district to either Jefferson county or St Charles county while slicing and dicing st louis county.

        • Jon February 6, 2017 at 10:16 pm

          Attaching St Charles County with St Louis City is non-viable politically; the idea would be hated by those in St Louis City, whatever portion of North county is used to connect them, and St Louis City.

          It’s likely also a VRA violation to split the AA portion of North County in a different district than the adjoining AA portion Northern St Louis City.

          It would also overpopulate a CD to include the entirely of all those areas in a single CD.

          The D controlled St Louis City leadership though would be ok with reattaching Southern St Louis City to Jefferson County using South County to connect as long as Northern St Louis City restored to a separate district. (They think they’d have more influence if the city was included in two CDs instead of just one.)
          By chain reaction, this would mean the other district would be Northern St Louis City and North county and indeed the remainder of the population of St Louis County not placed in the above district. In fact it would be underpopulated by 2020; to avoid the St Charles county problem it would instead cross into Franklin County.
          The main problem for Democrats though is this was pretty much the “Save Carnahan” redistricting proposal from D groups in which it’s known a Carnahan would have lost in 2010 under. Looks winnable for Ds in 2012, but not in 2014. The same trends impacting Democrats in neighboring IL-12 across the river though impact this as well for 2014, 2016, and likely going forward.

          45, M, MO-02

          • rdelbov February 6, 2017 at 10:24 pm

            That is why no commission or fair plan would ever do it. After Obama there is no enough rural D votes IMO for a 3rd D seat. I note Clinton did worse in rural MO then Obama did

  • FreedomJim February 6, 2017 at 8:00 pm

    What if a commission consisted primarily of statewide elected officials? I think that was a backup plan in Texas at one point.

    • Jon February 6, 2017 at 8:04 pm

      Prior to this current election, it would have turned over control of redistricting in Missouri to Democrats.

      45, M, MO-02

  • Torie February 7, 2017 at 4:08 pm

    I added a slightly revised PA map in my opening post, and pontificated a bit more. On this one, I really believe in what I am saying, after years of work on the topic, with another brain involved, considerably smarter than I, but yes, I did have some input into the process. I hope within a year, a white paper will come out, after the monte carlo runs have been done. First stop with the white paper in all probability is the courts, who are getting more antsy about redistricting (the political sphere seems to be getting rather hopeless on this issue, among others). Offensive gerrymanders will be dead within a decade. You can write that prediction down for future reference. Yeah, I might be dead, before one knows one way or the other (yes, I am that old), but that is OK. Cheers.

    Oh, in case one is interested, Toomey carried PA-17 as I drew it by about a margin of about 1% (matching his statewide margin more or less). So it is at once an uber volatile yet swing CD.

    65 - NY-19 (D)

    • rdelbov February 7, 2017 at 5:12 pm

      I guess it depends on what you mean by “offensive gerrymanders” . I happen to believe that even with a court ruling against “political gerrymandering” we will still see “offensive gerrymandering”. It will just be court ordered. As I have commented before court maps in OR and CO are ugly and nasty-they do not respect COI or county or city lines. I am very skeptical based on what maps that I have seen courts create that “offensive gerrymander”

      For that matter look at VA new congressional map that was drawn by the court. The lines for the four affected districts are worst then the previous ones. CD2 is worthy of Maryland CD. So I laugh when say that courts are going to end offensive redistricting. They draw nasty maps–look at CO congressional map.

      I look forward to seeing a Monte Carlo runs or a three card monte report on “redistricting reform”. Just call me skeptical.

      IMO your new PA still has Chester County split up too many ways.The Pittsburg seat could be more compact. Doing it like you did seems as it you had political intent.

      • Torie February 7, 2017 at 5:27 pm

        Yes, it is unfortunate that Chester County is split three ways, but it has to be split three ways. Granted, the split could be smaller, if PA-06 was just Berks and Chester, with PA-16 having a chop into Chester of about 42,000 people to take the balance, but that creates a mess of a map. Play with it yourself. I don’t understand your point about Pittsburg. It’s as compact as it could be, both as to the city, and to the metro area at large. Yes, there might be a map out there with a higher score, but this one has a relatively high score, and was drawn in good faith, following objective rules, no matter where it might lead, from a partisan standpoint. If someone has a better approach, show me with maps and objective rules, that don’t use subjective criteria, such as communities of interest, not recognized by the census bureau, that are gamed to death by partisans. That is my challenge to you. Draw your own PA map and show me. I would be happy to email you the census data, that I used to draw the map, using the 2015 estimates by county, and projecting it forward to 4-1-2020.

        Anyway, what you found offensive, I doubt follows any objective rules at all.

        65 - NY-19 (D)

        • rdelbov February 7, 2017 at 6:24 pm

          I am unable to draw a “fair” PA map as the concept is IMO is lucridous. Every decision I would make would have political ramifications. For instance if I decide to attach the left over portion of Philly to a Delaware county seat you get 4 seats that will be D in the Philly with only the Buck county seat being a likely or possible R seat. You attach the leftover portion of Philly to Bucks county and Delware you get 3 D seats with maybe two R seats. So what is fair?

          There is no rule that anyone can design for any program that will decide whether the left over Philly population goes to Delaware county or Montco or Bucks? Flip a coin?

          Why should the Erie county seat head southward instead of Eastward? PA10 is a huge geographic seat and the standard deviation in size seems all wrong. Of course it makes PA10 a vote sink for the GOP so is fair that seat goes south or east? I am not terribly up on the political subdivisions around Pittsburgh but something tells me that a Pittsburgh seat that moves eastward as opposed to Southward or Northward is more of a D vote sink. I think those far southern suburbs lean R while all the eastern ones are D. So what is fair? A compact seat or one that randomly centers the seat in the north or south or in the eastern or western portions of the county? I don’t know? Some party will benefit. What magic formula you come up will benefit one party or the other.

          As a side note IMO the election of Trump puts off “redistricting reform by judicial fiat” until perhaps the 2031 cycle. If then. Oh I guess Kennedy could decide to retire in the summer of 2021 and a Kagan clone could get confirmed and we see some sort of “reform”

          Until then I suspect we have our status quo.

          • Torie February 7, 2017 at 6:45 pm

            I would love to see your map. Chat is cheap, drawing maps that can withstand the slings and arrows, and are based on objective criteria, much harder. Oh as an aside, the suburbs of Pittsburg to the south in Allegheny county, don’t vote much different from those to the north, particularly when it comes to Trump. Yes, with Trump, the northern suburbs did trend Dem (all the way into southern Butler County), because that is where the bourgeoise lives. But no matter how you draw the lines, everything these days outside of the city of Pittsburg based CD itself, is safely Pub. That is just not where the action is. The action is in the 3 CD’s in the NE corner of the state, PA-08, 15 and 17. The rest is noise, after you unravel the offensive GOP gerrymander, that created a GOP CD out of PA-07, buttressed by a talented GOP incumbent, that cannot stand under any metric. The loss of the GOP CD in the west, the old PA-05, is due to population drains in the region, rather catastrophic ones, sort of like a rural Detroit.

            We shall see what the courts do. In the meantime, gerrymanders are being cut back, bit by bit, including in NY, which state alas does not preclude bipartisan gerrymanders, which is disgusting, but NY is not so much about ideology, and much more about pork, and feathering nests. It’s very cynical in my neck of the woods. But it’s slowly changing, bit by bit, and I am proud to say, that I am part of the change, at the local level anyway.

            65 - NY-19 (D)

            • rdelbov February 7, 2017 at 7:39 pm

              I don’t have a map per say because I am content with the current map. I don’t buy into the concept that partisan decisions are gerrymanders. Your PA map could easily be made into a 13-4 map for the GOP. Adjust Delaware, Philly, Montco and Chester a bit and it would be a likely GOP seat. As likely as PA8 and PA7 or PA6 are GOP seats are now. All three can be had by Ds.

              If you give Lackawana, Pike and Wayne county from your PA17 to PA10 and whatever you need to balance out put into PA17. Then move some voters between PA6 and PA15 you would get 4 seats in that area that the GOP could win. You don’t even need to split counties more then what Michigan style map would require to get close to 13-4 in PA. If you do not attach Lackawana county to Luzerene county you can get 4 seats there. One D seat in Pittsburgh and three Philly area D seats. That IMO is not a gerrymander but rather just dividing the counties out to benefit the GOP.

              It is simple. There are two simple decisions to make. Where do you attach the leftover Philly voters. Do you seperate Luzerne and Lackawanna counties. You get two extra D seats if you attach Philly leftovers to Delaware and the two LL counties together.

            • FreedomJim February 8, 2017 at 10:47 pm

              I have tried to draw a fair PA map and assumed Allegheny would have one full seat and two partial seats. I am impressed that you drew a clean map with only one partial seat. Kelly or Murphy might not live in your 3 and 5, respectively, but would be likely favorites there.
              The Dems may concede all but one western PA seat, but I do not think that is inevitable. If North Allegheny were with Pittsburgh, the rest of Allegheny could probably anchor a D-leaning seat. Erie might swing back to the Dems, but the nearby counties will probably still keep the district red unless it is gerrymandered.

  • CTIronman February 7, 2017 at 7:20 pm

    This map works as a good start point. Some possible tweaks w/reasons

    A) PA 7/8/13. PA 7 hasn’t had any part of Philly since 1990. PA8 could –like now–0nly be Bucks/Montco. The PA13 incumbent lives in Philly. Since Philly>PA1&2 seems you could get 1 less cut of Philly putting balance in 13 with less of Montco; the balance could go back–in a more normal fashion–to PA7.

    B) Lackawanna & Luzerne were in separate districts before 2000. Split them between 10 & 17 w/17 becoming a NE corner seat & 10 a Wilkes/Barre & Williamsport seat.

    C) Putting Penn Hills & Plum into 14 makes it a more viable minority opportunity seat.

    Anyway; this is just my effort to take historical events & cobble them to a “quiet” map

    • Torie February 7, 2017 at 7:49 pm

      Philly is growing faster than the state, so now it has more than 2 CD’s within it, or will, if population trends hold. The very small cuts into Philly by PA-08 and PA-07, reduce county cuts elsewhere, and make no partisan difference. In fact, the Philly part of PA-08, per the Trump numbers, is very slightly more GOP than the CD as a whole. But then, objective criteria, ignore the partisan impacts (except maybe as a tie breaker, and it rarely is a tie, with tight constraints). So if one starts focusing on those, and it all goes down the drain into the pit of despair and hell.

      65 - NY-19 (D)

    • rdelbov February 7, 2017 at 7:53 pm

      There you go out of your month and into a map CT!

    • GOPTarHeel February 7, 2017 at 8:09 pm

      Works well as a good starting pony for a GOP 2020 map too. Almost all of the ugliness from the current map comes from trying to keep 3 suburban SEPA Rs instead of 2.

      R/NC. Waiting for a non-ossified establishment or sane populists. Not optimistic.

    • Torie February 8, 2017 at 5:17 am

      Luzerne and Lackawanna are part of one metro area per the census bureau, so you get penalty points for splitting them up into two difference CD’s. Putting that aside, if you do split them up, you get two parallel elongated northern CD’s, that look horrible on the map (yes, I drew that map just for kicks, and really, really winced). That would be a GOP gerrymander par excellance, which is precisely what the rules are designed to shut down.

      65 - NY-19 (D)

      • rdelbov February 8, 2017 at 8:29 am

        Well I note Wyoming county is also in that MSA. I also note that Chester County is part of Philly MSA but it gets sliced and diced between several MSAs. I note that just using MSA as a COI is a decision unto itself.

        I go back to my original point. Why is slicing Philly 4 ways (instead of three) a model of fair redistricting but attaching it to Montco a GOP gerrymander? Why is slicing Chester county up three ways fair but chopping up Lehigh county two ways a gerrymander? Why is attaching Monroe county to PA17 “fair” while attaching Columbia and Montour counties to it a GOP gerrymander?

        The critical aspect of your map is that PA10 is a GOP vote sink and I don’t see how that is fair?

        • Torie February 8, 2017 at 1:16 pm

          Lots of good questions. First MSA was short hand, It is more complicated than that. Counties deemed too rural are excluded based on the following:

          “These are the “final” delineations of Urban County Clusters, using the 25K/40% threshold, where the urbanized area population of a county must be 25,000+ or 40% of the total population.”

          So Wyoming is too rural. There are not 25K persons within urban clusters. If it were part of the urban cluster, then my map gets a penalty point. So the map would include Wyoming in PA-17, and the chop of Susquehanna would become more substantial. One might quibble with the rule. That is fine. Using MSA’s including rural counties is better than nothing, but does cause the MSA to spread out so much, that it makes map drawing more difficult, and can cause distortions, for no good reason.

          Sure, you could have one chop into Philly, rather than two, but then some other county gets chopped, as in Montco. Maybe in fact, even though it makes the map look less pretty, it might generate a higher score in fact, because the map gets a penalty point for chopping ward 40. If PA-08 could go deeper into Philly, taking another ward without a chop, that map would get a higher score. Or the other way, if PA-07 could take two wards without a chop. Neither is likely to be true. (A 0.5% population variance is allowed between CD’s as allowed by the court decisions by the way.) The chop you are talking about is small, and makes no real difference in any event. Regarding chopping urban clusters, we have a pack and a cover rule. My map gets a pack penalty point (I take credit for the pack penalty idea, over the skepticism of the brain, but he finally acceded), because two CD’s go into the Philly metro area that are partially outside it (PH-06 and PA-16), when it could of course have been but one. Avoiding the pack penalty, would make a much more erose map, which itself would be penalized. One most choose. I chose. There is no cover penalty, because at least 7 cd’s must impinge into the urban cluster. If the Philly urban cluster were 42,000 projected persons smaller, then it would get a cover penalty, and I would revise the map. A two point penalty is too much.

          Attaching Columbia and Montour to PA-17 and excising its portion of Monroe, makes PA-17 more GOP (not by that much, but some), but that is not the point. The point is how it affects the score. It would have a negative impact on the score, because twisting the map counterclockwise there, would make it more erose. Erosity is measured by highway cuts, which are a good proxy for what looks erose to the eye (unless its looks erose due to physical barriers like rivers with no bridges, or mountains with no state highways going over them). It might also entail a county chop that is more than 5.0% of the CD quota (about 38,000 people), and I was careful to keep the chop into Monroe below 38,000 people, which took some work while still avoiding subunit chops.

          Yeah, you may quibble with the rules, although I find they make a lot of sense and generate good maps. The point is to have a set of rules, such that the maps can be scored. We have a system, were the maps that make it to the legislature to consider are on the pareto optimality frontier, meaning that there will almost always be at least two maps on such a frontier, one with the best erosity score, and one with the best chop score. One can allow more maps in second or third place to be considered as well, as long as there is some buy in by both parties. The goal of course is to get subjectivity out of the process, by having a set of rules in place long before any map is drawn, and before one knows who will be benefited. And sometimes small changes in whatever is the final census populations that are used, can have a substantial impact on the map (e.g. as noted above, if the Philly urban cluster ends up having 42,000 less persons than anticipated).

          Aren’t you sorry that you asked all of these questions? 🙂 It is complicated. And there are many other rules, that I did not mention, because you would lose patience with me (as in what happens with scoring in macro chopped counties, population inequalities, and as a tie breaker, and only a tie breaker (I emphasize that because what the Dems want, is for that to become the major map drawing metric), how closely do the partisan spoils match the political split of the state). The idea is to get a court to buy into it. As soon as courts know there is an objective way to get this done, that ends up being reasonably fair, and generates good maps, they are far more likely to outlaw gerrymanders. It’s been the subjectivity that has given them pause. Yes, I know, you might not like it, and that SCOTUS may well end up doing it. Such is life. Urge Trump to appoint judges that love gerrymanders, if you want to slow this all down. 🙂

          Thanks for your interest. You may seem a bit confrontational, but I don’t mind. I am a lawyer, and have a thick skin. Cheers.

          65 - NY-19 (D)

          • CTIronman February 8, 2017 at 1:45 pm

            Lemme suggest another SEPA tweak
            Add the surplus Philly pop to Bucks in PA08. Avoid adding Montco Twps
            Montco then is all of PA13 & a sliver of either PA06 or PA07. (probably the Merions) Adjust PA06/PA07 line accordingly.

            • Torie February 8, 2017 at 2:00 pm

              PA-08 would then need to chop into another county (the chop into Montco entails more people than the PA-07 chop into Philly, and alas the only counties available for Bucks to chop into are Lehigh and Northampton, which are in their own urban cluster, and chopping into it, is a penalty point, causing the map to have one more penalty point than otherwise, knocking it off the Pareto Optimal frontier (assuming it is otherwise on it), unless the move reduces erosity (it doesn’t), or saves a penalty point somewhere else on the map, which is highly unlikely, but perhaps not impossible.

              In short, to just shove PA-07 out of Philly still leave PA-08 with a smaller chunk of Montco, so Montco now gets a triple split. If you then force PA-08 entirely out of Montco, then the replacement chop is in the Allentown urban cluster area, which is a double penalty, with the chops of Monroe and Carbon County getting smaller, as the map is in this final step pushed counterclockwise. It’s tricky, because in step one, the map twists clockwise, and in step two, move counterclockwise. I got a headache just thinking it through. So in step one, there either there is no benefit (and that assumes that as PA-08 digs deeper into Philly, and PA-07 in turn digs into Montco (if that generates two subunit chops, in lieu of the existing one (the chopped subunit entailed by PA-07’s cut into Philly) then you have an extra subunit chop, which generates an erosity penalty point), and in step 2 you add an extra penalty point, so step two is a no-go from the get-go as it were. 🙂

              65 - NY-19 (D)

              • rdelbov February 8, 2017 at 3:58 pm

                These comments and your other comments show that some of “fair” redistricting is in the eye of the beholder. It also shows that are dozens of individual decisions made by Torie that are just that. Personal but rational decisions to draw lines a certain way. I am not attacking Torie for his decisions but rather going by MSA is a personal decision, putting part of one county in a MSA in another CD -dividing Philly 4 ways is a personal decision while splitting Montco two ways is another while splitting Chester county three ways is yet another.

                All rational decisions but as noted we do a Michigan style map with very few split counties and just as rational as Torie.

                Here is my way to avoid extra splits.

                8th All of Bucks plus the balance north of Montco (northern R part of county)
                13th is Montco plus some in Philly
                1st is all Philly
                2nd is Philly and rest comes out of Delaware
                7th is the rest of Delaware plus what you need in Chester–the rest of Chester goes to CD16

                Okay now there are no splits–everything is easy smooth lines with no double crosses.

                To me it is beyond fair as no extra cuts are made.

                • Torie February 8, 2017 at 4:13 pm

                  The main thing is to come up with detailed rules, so that maps can be scored in detail, so in the end in general, you have but two pareto optimal maps. Sure it is possible that perhaps some maps will be tied in both erosity and chops, but that is unlikely. Under the Michigan rules, there are zillions of maps that are tied, leaving tons of room for mischief, as the GOP in its gerrymander there so brilliantly demonstrated, as it chose the location of chops which maximized partisan advantage. With the rules I live by, each chop is tested for erosity, preservation of integrity of urban clusters, and macro chops and so forth, so that all chops are not equal, and in general will not be equal at all. And that is what takes out the subjectivity – you don’t end up with tons of equal maps, with huge differences in partisan impact. Make sense?

                  I might add, that most of the gerrymandering mischief is achieved with macro chops, and where they are drawn, and how they are drawn. Macro chops cause the subunits to be treated like counties, so each subunit chop counts as a chop, and each subunit that generated road cuts with another subunit that crosses a CD line, generates an erosity penalty point, so you end up with tons of erosity penalty points, if you do something like the erose GOP macrochops into Oakland County, MI. Man, the number of erosity penalty points generated by what the GOP did in Oakland County, would sink the Titanic. That nasty bridge chop through Saginaw County is also something of a disaster, and we penalize bridge chops too I might add. Bridge chops are done by folks up to no good, almost always.

                  Persuaded now? No, of course not! 🙂 That’s OK, the point is to persuade the courts. Persuading you guys would just be a nice little bonus, but it is not the main course.

                  65 - NY-19 (D)

                  • Jon February 8, 2017 at 7:45 pm

                    This whole MSA thing doesn’t make sense applied to Missouri.

                    First at the state senate level here it’s totally unneeded, there was only one or two legal ways to draw the subset of districts not self contained within a single county.
                    Second is that some MSAs/ CMSAs here are still growing geographically and the definitions lag as it’s typically years ending in “4” that census bureau updates them.

                    Our state house seats have been drawn by the courts every decade since the commissions were formed (every house commission so far has deadlocked; in large part over the St Louis county subset)

                    At the congressional level, it would also have resulted in one heck of an ugly map in the western side of the state; #6 starting in the NW corner would have been required to detour around the northern counties of the KC metro area which have almost as much in common with the rest of the Platte county purchase counties as they do KC.

                    45, M, MO-02

                    • Torie February 10, 2017 at 2:16 pm

                      Regarding Missouri, I drew a map for you, for your viewing pleasure in the OP. Does it change your mind at all? This map is probably one that a computer would draw. I tend to doubt there is another map out there with a higher score. It just “works” too well in all particulars, vis a vis the scoring system. It is a very carefully thought out scoring system, that has been worked on, with hundreds of hypos tested, for over two years now.

                      65 - NY-19 (D)

                    • rdelbov February 10, 2017 at 3:11 pm

                      Not to be overly negative about every map you draw, because that is not my intention. Yet dividing Jackson county when one does not need to be seems to be redistricting 101 for a “fair map”. One would think county integrity would be a desired goal?

                      I note as well that there is enough of Kansas City Missouri in Clay County, just north of the river. Cass county is your CD5 does have some suburban voters but nearly half the county’s population is rural or smallish towns that are no focused on KC. By any measure the rest of Jackson county and the rest of Kansas city Missouri should be in CD5 as opposed to rural Missouri area.

                      I see this pattern both in your PA and MO map–very simple decisions that help the D party seem to come naturally . The decision to place Cass county in CD5 instead of Clay county population makes CD6 more democratic. Still likely to be R but still more D then it should be. Ditto for St louis area. While the northern part of Jefferson county has a st louis commuter feel. The bulk of the county stands on its own. St Charles county has a more suburban feel to it then Jefferson county. For that matter Franklin county has IMO more a connection to SE St louis county then Jefferson. It makes CD2 a bit more D but still likely R but its funny how little decisions get made that way.

                    • Torie February 10, 2017 at 4:15 pm

                      Sometimes the rules help one party, sometimes another party. These maps are drawn without any concern for party impact, except as a tie breaker, and there are rarely ties. The CD’s are nested in the KC urban cluster. Cass is party of the urban cluster. You are discussing communities of interest, and that instantly becomes a subjective exercise. In any event, some county in the urban cluster needs to be chopped, or you will have a pack and a cover penalty point for having 3 CD’s in the urban cluster, or a county outside the urban cluster will be chopped, with the other CD making a ring around Jackson, that would look terrible and have a huge number of erosity points. Try it yourself.

                      As I said, the main thing is to have objective rules, so a computer can draw the map, If you don’t have that, then, well, the discussion will devolve into the type that we are having, which frankly goes absolutely nowhere. What would be more productive, is suggesting other rules, if you don’t like these, that would also end up with just a few winning maps, that a computer would select. I warn again, that if you don’t have objective rules as I describe, or some variation thereof, the courts will come up with another objective rule – the map that comes closest to the partisan balance in the state wins, which would help Democrats in most states (not all, e.g. Mass). That is what Eric Holder is pushing for. Blow me off if you want, but hopefully some of you will listen to me. And it is going to happen one way or the other, and sooner rather than later. You shall see.

                      It’s odd that you speak of partisan issues when it comes to MO. It is still a safe 6-2 GOP map. And trying to go for 7-1, would probably end up being a dummy-mander.

                      Anyway, I obviously am not going to persuade you of anything, and visa versa, and that is OK. The goal here is to persuade the courts, and/or legislatures. Thank heavens for powerful computers. That is what makes this all possible. Doing it manually, as I did, for the PA and MO maps is time consuming, and in general will not generate the highest scoring maps, although MO turned out to be so fortuitously simple, that I think it probably is.

                      One final thing, that I assume that you are probably inferring. is that I think I am the GOP’s best friend on this one, trying to save it from itself as it were. It’s getting greedy, and you know what they say: pigs get fat, and hogs get slaughtered.

                      Thanks for you interest, and no, I am not offended at all. I get many of the same types of comments from partisan Democrats. If both sides are aiming the slings and arrows at me, that gives me a good feeling that I am probably on the right track! 🙂

                      65 - NY-19 (D)

                    • Jon February 10, 2017 at 6:59 pm

                      Boundary between #1 & #2 in St Louis County : 40 (I-64) would normally be the least favorite major highway to split along (it’s cut both Mid County & West County in half)
                      But if St Louis City + St Louis County + Jefferson county are that precisely two districts, it’s the best way to draw them.
                      No complaints about #3,#7,#8

                      But as to the KC area : First Jackson should contain all of #5 rather than cutting out the NE corner. Then since I assume that Jackson county by itself isn’t quite enough for a CD it should go east if you are looking to avoid county splits; (the ones in the East are much smaller)

                      As drawn the ratings would be:
                      #1 Safe D
                      #2 Either R- replacement for Wagner after she runs for Senate or OPEN : Likely R (on the border with safe)
                      #3 Either R-replacement for Wagner after she runs for Senate or OPEN : Safe R
                      #4 OPEN : Safe R
                      #5 The current incumbents of #4 (R) and #5 (D) would be doublebunked: Currently Likely D but could drift to Lean D
                      #6 Safe R
                      #7 Safe R
                      #8 The current incumbents of #3 (R) and #8 (R) would be dunkbunked: Safe R

                      45, M, MO-02

                    • Torie February 11, 2017 at 12:43 am

                      Thanks for your comments. I drew MO-05 the way I did, because if MO-05 strays out of the KC urban cluster, you get a pack and a cover penalty (2 penalty points), because now two CD’s would be both within the urban cluster, and outside it, rather than just one (pack penalty), and three CD’s are partially within the CD rather than just two (cover penalty), with now Cass County chopped into by MO-04, assuming that you don’t propose that somehow MO-06 snakes around to take in Cass County (if it did, and were possible, you would have a ton of erosity points, as the CD lines cut the state highways that connect the county seats between adjacent counties). Thus under the rules, it is not a very high scoring map. The idea is to try to separate urban cluster CD’s from other CD’s.

                      65 - NY-19 (D)

                    • Jon February 11, 2017 at 1:59 am


                      The KC MSA from the Census Bureau officially includes both those two counties east of KC.

                      So I’m seeing it as drawn 1 district fully within the KC metro area and #4 & #6 partly in and partly out plus a county integrity violation.

                      In any case I’d give several more penalty points to a district with county integrity violations than crossing MSA lines; but at the moment here are the other MSA violations from our cities being in natural places to split districts:

                      St Louis MSA : In addition to including 2 full districts and aprox 1/2 of the population of district 3, the two northern most counties in #8 are within the St Louis MSA; if you go by CMSA rather than MSA the Farmington MSA fully contained in #8 is also within the St Louis CMSA for a total of 4 counties in #8.

                      Jefferson City MSA: While most of the MSA is within #4, Callaway county in #3 is also in the Jefferson City MSA and in fact has stronger ties than most of Cole counties other neighbors (multiple bedroom communities just across the county line). Also note that it’s only a matter of time before the census bureau creates the Columbia-Jefferson City CMSA.

                      Columbia MSA: While most of the MSA is within #4, Extends north into county in #3.

                      Springfield MSA: It’s now composed of seven counties; so as drawn it’s spliced and diced between #4, 7, and 8.

                      Hannibal MSA: Currently split between #3 & #6. (This one is fairly significant; the smaller of these two counties contains bedroom communities for the larger one)

                      Here is the flow on how I’d fix these:

                      #5 All of Jackson county plus one or both counties east (whichever combo is closest to the ideal district size) ; it will still have #4 & #6 both partly in and out of the KC MSA, but the county splits would be eliminated.

                      #6 First the county north of the Missouri river in #4 gets moved to #6. Next priority county to move into #6 if needed for population would be the one in #3 near Hanibal. If additional population still needed continue to eat into #3. This fixes having a district cross the Missouri river in two awkward spots and reunites Hanibal MSA.

                      #4 Add Fulton; as needed take away from counties in #4 near Springfield; giving them to #7. This reunites the Jefferson City MSA.

                      #3 Shift South as needed

                      #8 Shift West as needed, start with SE most county in #7

                      45, M, MO-02

                    • Torie February 11, 2017 at 8:00 am

                      That’s fine Jon. As I noted above, counties that are too rural are spliced off from the urban cluster zone, because otherwise in general the flexibility to draw a map that minimizes chops and so forth is constrained for no good policy reason, since the counties are really more rural than urban. There was an attempt to have a policy reason for each and every rule we set up. And for large subdivisions that sometimes need to be chopped (e.g., any city over 750,000 or so depending on the state), we set up zones within the city as sub-sub units, based on what is out there (in Phoenix they have community planning districts, in others, wards work pretty well, as in Philadelphia, but not in NYC where the council districts are gerrymandered).

                      Anyway it all depends on the rules that one sets up. If you strictly follow the census bureau zones, rural counties and all, then of course one draws a different map. The important thing is to have the rules. The only problem is that if the rules vary from state to state, then parties might figure out what set of rules favor them, and cherry pick. Thus the desirability for one set of rules that applies nationwide. But reasonable people can differ as to precisely what the rules should be, and there have been a lot of arguments about that. Hopefully after thousands of monte carlo runs are done, we will have a better idea if indeed reasonable maps are generated nationwide. I doubt most reasonable people would claim the maps generated above, are unreasonable, even as one quibbles about some details. They are good enough. They are certainly better than what we have now. Reasonable is good enough. Make sense?

                      65 - NY-19 (D)

                    • rdelbov February 11, 2017 at 9:06 pm

                      Jon your comments about St Louis county is spot on. The North South delineation is clear.

                      I made most of my comments above but let me add that have I heard many people discuss non-partisan computer programs and formulas with fair outcomes. Call me skeptical. Its funny as the PA and MO maps both made little moves here and there that benefit the Ds-Oh I see some that benefit the Rs. It is called packing them into safely R seats while others are marginalized.

                      As noted you divide up MSA areas all you want but you can’t keep counties and cities intact the underlying scheme or plan or software bug is pro-democratic. The secret sauce with attaching Cass county to CD5 while not attaching Clay to it suggests a pro-democratic bias. It makes CD5 less compact while splitting city and county lines.

                      As noted as well in the PA map there are 6 or 7 ways to arrange Bucks-Montgomery-Delaware-Philly and Chester county. The computer picked the one that would bring the maximum advantage to the D party. They lock 4 seats and have a great chance at a #5. You can roll the dice and at least come up with a way to make PA8 come with a R edge. You do the map several ways and have the GOP as the favorite in two seats. To have the map with 4 safe D seats and have the Rs at even in the 5th the only way to do is with this map. Using a Michigan style splitting as few counties as possible.

                      So let me suggest that if I can see this D flaw in this system many others will also see it.

                    • Jon February 11, 2017 at 11:26 pm


                      An interesting thing though on St Louis County is it wouldn’t change the PVI indexes much to have had #1 fully take in all of Mid County; the section of West County he grabbed by using 40 (I-64) is the only part of West County that contains some D majority areas.

                      A glance at the PA map near Phillie shows the flaw is it doesn’t penalize extra county splits enough. It is trivial to make small tweaks that would reduce the number of county splits there. It looks a lot like the map started in the wrong part of the state without sufficient second passes being made to eliminate the extra splits.

                      45, M, MO-02

                    • Torie February 12, 2017 at 10:58 am

                      As to the Philly area CD’s, I honestly don’t think it is possible to reduce the chops given how the populations of the respective counties plays out, along with the Philly urban cluster, as I described in detail above (e.g., if you eliminate the double chop of Philly, you get a chop popping up somewhere else). But if it were possible, then of course that might well be a higher scoring map, and it would be revised accordingly.

                      65 - NY-19 (D)

  • CTIronman February 7, 2017 at 8:15 pm

    Adding Chester to PA7 makes it blue enough. Even a small sliver of W Philly makes the seat uncompetitive. And how many incumbents are we double-bunking? Boyle has no district now; Tim Murphy needs to move already (& this can’t be helped given Allegheny pop loss) & Thompson/Shuster & Barletta/Cartwright are already matched up on this map.

    • Torie February 7, 2017 at 8:50 pm

      Excising a portion of the 40th ward in Philly that is appended to PA-07 moves it from a Dem PVI of about 11% to a 10% Dem PVI. Yawn. You guys just can’t get the partisan juices out of you system apparently. That will have consequences, if that is the GOP attitude across the Fruited Plain. You shall see, since I tend to agree, that partisanship in this arena has no miracle pharma drug to exorcise it. And all over a handful of seats, to try to win this battle while losing the war. Sad.

      Thanks for listening. I think we have beat this drum to death. Unless someone wants to draw their own non partisan map using objective criteria. We shall see.

      65 - NY-19 (D)

      • GOPTarHeel February 7, 2017 at 9:00 pm

        I think your map is mostly fine and I’d take a system like this in a heartbeat nationwide.

        R/NC. Waiting for a non-ossified establishment or sane populists. Not optimistic.

        • rdelbov February 7, 2017 at 10:19 pm

          GT if I could understand what sort of “system” Torie used to draw this map I might could be convinced. Might could be convinced–Maybe.

          I don’t any real logic or scheme in this map of PA. If I was trying to be “fairer” to the Ds then the GOP is right now this would be a D alternative. The map keeps Montco intact but slices Chester–what sort of plan with repeatable logic can be used with this? I would not call it a “system” unless it is repeatable.

          Lets see this same “system” in maps from Florida-TX-IL-MD-KS-CO-OR. Not to be cynical but the above map ensures that the lost seat is a GOP one plus CD7 and CD8 are less republican then they are now. I would love to know what sort of system this is and whether it is repeatable.

  • krazen1211 February 7, 2017 at 9:52 pm

    The PA-17 argument is a stretch, even if you believe that Republicans would have won it this year. I find that prospect unlikely as all Democrats running for row offices would have won that district. A better district for that line of discussion is IL-06.

    More importantly, there is limited marginal value of an additional House seat. Going from 241 to 242 House seats does little for a political party. Conceding a single seat for the purposes of safeguarding multiple seats and allowing for allocation of resources elsewhere is a greater gain. If the goal is to set a high floor of close to 218 seats, and accept a low ceiling, that is a reasonable tactical decision, and a party with a natural advantage has incentive to pursue high floor schemes.

    All of this is of course arguing over small pieces of a bronze medal in our current elected political system. You can enact policy without the House of Representatives. See the last administration.

    • rdelbov February 7, 2017 at 10:10 pm

      I guess it depends on Gov race 2018. If the GOP loses I would go to Boss Brady and toss a 12-5 map at him in a heartbeat. I would let him draw the 5 D seats.

      If the GOP had the trifecta you could modify Torie’s map and a bit then do 13-4. Or nearly 13-4. You could get a chance at PA17. You would have to swap PA7 for PA17 but IMO it could be cleanly done. You can’t stretch out three Philly surburban GOP seats

  • Jon February 7, 2017 at 10:12 pm

    On PA, with control of redistricting, my target for elimination would be #17; which as a side effect should greatly reduce the mini-baconator of NE PA to Central PA.

    But I’m doubting we’ll have full control of redistricting anyway, and so perhaps #7 would be the seat to abolish. It looks like three mini-districts held together by thin strings instead of a single district. (I wouldn’t be surprised if among all R held districts in the country it scores the worst in compactness scores.)

    45, M, MO-02

    • rdelbov February 8, 2017 at 10:27 am

      Without Trifecta it is let’s make a deal time. Both parties IMO will lock up safe seats with the GOP taking the loss on #6 or #7—name it whatever you will. One less suburban Philly seat.

  • FreedomJim February 11, 2017 at 10:50 pm

    I noticed that, with 2010 census numbers, a group of counties in SEPA was very close to the population for exactly seven districts. I thought a clean map of them would be something like:
    District A: Lancaster and part of Chester
    B: Parts of Chester and Delaware
    C: Parts of Delaware and Philly
    D: Berks and the northern parts of Bucks and Montgomery
    E: Most of Montgomery
    F: Philly
    G: Parts of Philly and Bucks.

  • rdelbov February 12, 2017 at 12:51 pm

    I am not by nature a gadfly but let puncture in detail the whole concept that the Rs used Gerrymandering to win and to keep control of the US house. I appreciate Torie giving me this forum and let me fully explain what I mean in a brief way. If there was “fair” redistricting as opposed to the partisan or non-partisan redistricting of 2011 here is where IMO we changes. FL and VA IMO have already been done. I can argue that Florida map tilts D while others wail about VA10 but hard to argue that the maps are not fair as both sides complain about parts of them. With fair maps from Maine to Hawaii I see the GOP and Democrats getting chances to win additional seats in the following states.

    GOP gains chances to win 16 seats with fair maps in the following states: MA (2). CT(1), MD (2) IL (3), AZ(1), OR(1.5), WA (.5) and CA (5)

    The Democrats gains chances to 17 seats with fair maps in the following states: PA (2), NC(2), SC (1), GA(2), AL(1), OH(2), MI(2), WI(1), LA (1) and TX(3)

    I tried to be thoughtful. Think of Indiana. You do a fair map of the state and come very close to what we have. Torie’s fair map of MO gives the Ds better chances in a few districts but IMO still would only elect two Ds most elections. I note that the GOP might win two GOP congressman in MA in a fair map but we have no chance IMO under the current map. Ditto for CT which is a compromise map but we compromised poorly. The gist of many GOP maps in 2011 was not so much creating new seats for the GOP but locking them up for us. In TN we did a 7-2 map and IMO a fair map would do the same. Oh you could do a map in rural west TN and exclude any Memphis suburbs but it still IMO lean to the Rs. It would take in Montgomery county and few more rural counties but it would still lean R. I note we locked up seats in LA, AL and SC. Redoing the maps and being “fair” giving the Ds a chance to win seats in those states would still lock up others more tightly for us.

    So IMO overall redistricting 2011-2012 only favored the GOP as we created seats more seats that we had PVI of 8 or 10 but a fairer map with PVI of 5 or 6 would still likely see us win those seats in most years. Sure we locked the Ds out of 20 or so (18 by count) seats with maps. Well either the Ds or commissions or judges (OR) did the same thing to us. I should add that TX courts likely cause us a seat as well but that’s another story for another day

  • rdelbov February 13, 2017 at 7:39 pm

    Here are the listing of the GA6 8

    Ossoff will be the bold progressive in the race with access to the Sanders grass roots money machine. Slotin was a D state senator from around the area when this area elected white democrats. He lost in a D primary to Cynathia McKinney in 1996 and has not run for office since. There is rumoured to be one D gal running but she has not filed yet. No news from Karen Handel yet–Moody(who filed today) represented much of her power base in Fulton county in the state senate. Hill is from Cobb county and likely can count on alot of votes from that area.

    Stay tuned

  • Upstater22 February 14, 2017 at 10:56 am

    Late to the party here, but wanted to get some comments in.

    1) You don’t do a very good job of proving your thesis, that the GOP needs to abandon gerrymandering for “its own good”. You seem to insinuate that they should do it because Democrats want to gerrymander seats to be proportional to the statewide vote. That doesn’t hold much water, IMO. Unilaterally disarming in the redistricting battle seems like a terrible idea. If you are going to make such a declaration, you need to back it up better.

    2) It’s BS to say that humans are biased but computer algorithms are not. Humans write the algorithms and their biases are built into the algorithm. One guy came along on SSP with a similar proposal. He wrote an algorithm to draw districts that were visually appealing. It was beautiful: a lot of evenly-sized districts that were smoothly drawn. Until you realized that in order to do that, he had to split cities to a ridiculous level. One man’s beautifully-drawn map is another man’s gerrymander.

    The bias in your algorithm is your decision to keep MSAs together (such as Lackwanna and Luzerne). This undoubtedly helps Democrats, as it allows you to draw all-rural Republican vote sinks like your PA-10. Instead of MSAs, you could have chosen to use media markets, which makes a whole lot more sense. Your PA-10 crosses 5 media markets. Tioga, Bradford, Wayne, Susquehanna, etc are all in the Scranton media market. When people in this counties go to “the city”, they mean Scranton. Of course, drawing two sensible districts, one based in Scranton and one based in Wilkes-Barre that both take in rural territory would be horrible for Democrats, so your algorithm avoids doing it.

    3. I agree with rdel. There will always be political decisions to be made when drawing maps. I think the NY map is very well drawn. NY-20, especially. It makes a heck of a lot of sense to draw Albany, Troy, Schnectady, and Saratoga together. But no doubt that this would not be allowed by some archaic “We must keep county splits to a minimum” rule.

    I would rather know the affiliations of the people drawing the maps than to have it done by some shady software designed by who knows who with who knows what agenda and biases. Everyone drawing maps has biases.

    Conservative, because facts are more important than feelings

    • rdelbov February 14, 2017 at 11:44 am

      UPState–well said. I said that the same computer program keeps several PA counties together in the same MSA but chops up counties in the MO map. I note that MSA have changed over the years and some counties have begged to be in them for economic or political reasons but others fight the census folks to stay out of them. Not sure folks want to be in the Memphis MSA but in Nashville area its a popular matter to be associated with that city metro area.

      Not to beat a dead horse about Philly and its suburbs. To me the most salient reason leftover Philly city needs to go the Montco seat is that Chestnut hill and Roxborough plus a few surrounding areas are so close in economics and politics to most Montco county. If you give NE Philly (mostly blue collar and some lower income white voters) to Bucks (mostly white collar/Suburban) you mix and match social economic areas. Ditto for giving leftover South Philly to the Delaware county seat. The only logical transfer of similar area is Chestnut hill and surrounding area to Montco seat. Yes it helps the GOP but it makes the most sense. I might add that no computer program can ever have that variable plotted in. That’s a human decision and IMO a fair map maker would make the same decision.

    • FreedomJim February 15, 2017 at 7:25 pm

      I wholeheartedly agree about media markets. Unifying media markets would reduce the number of people forced to see ads for candidates they can not vote for.

      • Jon February 15, 2017 at 7:58 pm

        For most states that would work well for Congressional districts. (In the case of state senate / state house; county integrity rules will usually suffice)

        One place where it could result in a geographically unpleasing shapes though is the Nashville area districts; this rule might force a district that comes close to completely surrounding TN-5.

        45, M, MO-02

        • FreedomJim February 15, 2017 at 10:47 pm

          I agree. If the Nashville (or any) market is too big for two districts, I would consider any map that keeps it in three to be compliant. Regardless of partisan implications, I would be okay with putting one suburban county in TN-5 if that would make the neighboring districts neater.

          • rdelbov February 16, 2017 at 11:30 am

            Davidson county will need to have surrounding area added to it for population standards. That is a given. In fact the GOP in 2011 was the 1st map drawn in years that did not divide Davidison county by attaching GOP areas in Belle Meade and Brentwood to CD6 or CD7.

            • Jon February 16, 2017 at 6:47 pm

              At a rough guess; I think for 2020 Davidson + Robertson counties would come closest to ideal population if only full counties used.

              Using media markets this would appear to cause the district that starts in Sumner county to not only take in Wilson, but also Rutherford and Williamson. Note that the same problem occurs in this case using MSA.

              45, M, MO-02

              • rdelbov February 16, 2017 at 8:21 pm

                My sense of things are that TN GOP will not baconstrip Nashville into two or three seats in 2021.

                Just because I support partisan drawing of seats that does not mean I would slice and dice up my home state or any other in a MD or IL type map.

                I am happy with a MI or NC style map with minimum number of chopped up counties.

                Of course I note even the judges in Florida chopped up Orange county in weird ones for a political outcome. Likewise Federal judge in VA chopped up that state. That is a major reason I support partisans doing the chopping as opposed to unelected comissions or judges.

                • Jon February 17, 2017 at 12:33 am

                  My best guess on how TN will actually draw the lines in 2020:

                  1. First TN-9 will be bumped back up to the right population level. The first part they’ll add will probably be the unincorporated area South of Collierville and Eads / Fisherville. If it still needs population after that, next they’ll grab the sparsely populated NE portion of the county that’s East of Millington and north of all three of Bartlett, Lakeland, and Arlington. Everyone is Bartlett and Germantown is going to hope that’s enough population to get #9 up to the right level; otherwise a big internal fight develops in the legislature over which of Bartlett & Germantown needs to have part of their city sacrificed to #9.

                  2. TN-8 brought up to correct population basically by grabbing everything else within the West TN grand division.

                  3. Whatever minor changes need done to TN-1 will be made.

                  4. TN-5 brought up to correct population; I think they’ll add more of Chentum County if needed to bring it up to proper population rather than replacing the western counties with Robertson even if the later would avoid a county split.

                  5. How they’ll divide up the suburban counties of Nashville will depend upon who’s holding the seats, but we can be positive that all of #4, 6, and 7 will all contain at least one county that adjoins Davidson and will also all contain at least county that’s outside the Nashville media market.

                  6. I’m more hopeful that they’ll swap counties between #2, #3, and #4 to make #2 more Knoxville centric and #3 more Chatt. centric as these moves can be done without hurting any incumbent that matters.

                  45, M, MO-02

    • krazen1211 February 17, 2017 at 9:43 pm

      In light of this, some clown at Bloomberg wrote this article. He gets everything basically wrong of course.

      Republicans’ Gerrymandering Could End Up Helping Democrats

      • Republican Michigander February 17, 2017 at 11:17 pm

        That guy is an idiot. He had nothing to back up what he said outside of his mouth.

        “”””Here’s where the “Republican gerrymander” could help Democrats. In a big enough anti-Trump wave, the R+4 districts intended to favor Republicans could tip toward Democrats.”””

        Well, no shit, Sherlock. In a big enough R wave in certain communities, those 53% D seats can go R, too like Bill LaVoy’s seat which was gerrymandered for D’s to protect Pat Somerville ‘s downriver and the Western Monroe County seat.

        Since this guy cited Michigan, I’ll ask this. Who do the dems have that can take MI-6, MI-7 (not sure if that’s R+3 anymore), MI-8, or MI-11 assuming incumbents run again. Even in a wave year unless you are in a strong D district winning based on crossover, you can’t beat someone with nothing (or crap). Bill LaVoy didn’t lose to crap. He lost to Joseph Bellino who won with hard work and door knocking.

        MI-06 – Fred Upton refused to get caught looking. He took his last race seriously.
        MI-07 – What else can be thrown at Walberg that hasn’t been thrown at him already? He beat the two best dems already in Mark Schauer (lost and won in rematch) and Doug Spade (moderate radio talk show host and later 3 term rep – Walberg’s last state rep win). Pam Byrnes and Gretchen Driskell were at least B team candidates as well.
        MI-08 – UAW went all in to stop Bishop. D’s couldn’t break 40% last time in a seat Trump ran around generic R. Gretchen Whitmer in MI-8 is running for governor. Tom Cochrane would worry me IF he got past a D primary, but that’s about it right now. Dems do have a good bench in Ingham County.
        MI-11 – Canton’s getting bluer, but Trott has money. In addition, there is not a strong farm club for the dems here. This is the quintessential “Romney voter” district outside of Canton (Bush won twice, Obama won BIG twice, Hillary won) and Livonia (narrowly for Romney, but liked Trump).

        MI-08 - Michigan is a red state again. We need a 50 state strategy and an 83 county strategy.

      • rdelbov February 18, 2017 at 10:40 pm

        I get offended by the casual use of the term gerrymander but that is another story. I note the article did not spent any time on states that had maps drawn by democrats.

        His other assumptions are chuckles.

  • roguemapper February 18, 2017 at 2:22 pm

    I retitled this diary because it’s a pain for me to scroll through the extra lines over and over again in the directory logs. Sorry! Great diary btw!

    Dem NC-11

    • Torie February 19, 2017 at 12:28 pm

      No problem. Michigan turned out to be a quite fascinating enterprise, and the GOP got lucky, because the population figures caused the old MI-14 to remain a Dem vote sink, rather than Washtenaw County being appended to another CD (i.e., MI-07), which would cause it to go safe Dem. If the Detroit metro area plus Washtenaw did not have just about exactly 6 CD’s, that is probably what would have happened.

      65 - NY-19 (D)

  • Torie February 21, 2017 at 7:31 am

    Yes, another Michigan map is added to the pile at the bottom of the opening post. It’s a useful exercise to display the possibilities for map drawing, and how to approach the matter.

    A poster on Leips suggested a revised approach for Michigan map that turned out to be clearly a winner. It loses two chops from my effort, including a macro chop in Wayne County, so it also has a much better erosity score (which it probably does even without the lost macro-chop). I still take a pack penalty in the Grand Rapids urban cluster, to avoid a macro-chop. I learned from the guru of the system that in most cases, one takes a pack penalty to lose a macro-chop (a county chop that involves 5% or more of the CD quota (38,364 residents in this case), because it tanks the erosity score (each county subdivision in a macro-chopped county is then treated as itself a county, what we call “zooming”) . There may be a higher scoring map that the uber computer can find, but I tend to doubt it. The population array worked to make this map a winner. When I did that rectangle in Oakland County for MI-11, and the population fit perfectly, I was just amazed. It was like a winning lottery ticket. I was blinded by the perfect fit for Washtenaw to round out to a whole number the Detroit metro area CD quota, so I did not adequately consider the alternative approach.

    I also like the chaos that it creates from a partisan standpoint, including a host of potentially marginal CD’s, but I digress. It also shows that while the Trump coalition may be disaster for the Congressional GOP in CA, the reverse is true in Michigan. With this map, the Dems are down to but two safe seats – the two black seats. Everywhere else, the Dems will have to sweat. But then, then there are seats the GOP will need to sweat in too – like in four of the CD’s potentially. And that is the way it should be. And oh yes, the incumbent politicians will hate this map. Good! ☺

    Yes, I know, some of you hate this approach, and the evil computer, and the whole idea of having metro area integrity rules (which may hurt the GOP slightly, but really not by much). Krazen gets that this is the best defense to the GOP continuing to hold the high ground in Congressional political battles, and as a vaccine to ever more restive courts when it comes to gerrymandering. Follow his lead I say. He may be an uber-partisan hack, but he’s a darn smart one. ☺

    65 - NY-19 (D)

    • rdelbov February 21, 2017 at 10:40 am

      My only concern whatever a computer program driven is that inevitably the secret sauce in the program favors the Ds. Your PA and MO maps confirm my thoughts on that. In the PA counties gets sliced and diced in a way to favor the Ds all in the name of non-partisan. Then in MO we see several counties chopped up and in one case they don’t need to be but then again it all favoring the Ds.

      I don’t see the point of trading partisan redistricting done by pols for partisan redistricting done by computer

      I might as well that I dispute the underlying concept in this diary the GOP is anymore at fault in this process then the Ds. I notice that not a single map has been discussed from Maryland or IL or MA or AZ or CA where the Ds clearly have a huge edge. The GOP had total control in only about 1/3 of the CDs that are currently being used in the USA. I don’t see any huge edge for the GOP in the current house. Rather the natural overall edge for the GOP in redistricting is in play in the USA. IN other words the Ds self pack and overall nation is GOP by house seat numbers.

      • rdelbov February 21, 2017 at 10:44 am

        I might add that Trump’s election in 2016 pretty much assures that we will have status quo/2011 redistricting for 2021-2022 cycle. Talking about restive courts is so 2016–hoping and dreaming of a Hillary win.

        • Torie February 21, 2017 at 10:56 am

          The Dems are going to hate this computer even more than you do I bet. That is the beauty of it all. If both sides hate it, it must be the cat’s meow. Hey, putting aside that mini chop into Jackson County, that got so many worked up (over nothing really), my MO map was a chop-less wonder, a veritable work of art. Even Krazen loved it. Have you no taste? 🙂

          Moving right along to PA, putting aside the urban cluster integrity concept that you dislike so much, if you split Luzerne and Lackawana into two separate CD’s, you get these two hideous bacon strip CD’s running right across the northern part of the state. Next to nobody lives up there outside those two counties.

          65 - NY-19 (D)

          • rdelbov February 21, 2017 at 12:33 pm

            I have no evidence that the Ds would hate this program more then the Rs so far I don’t see enough evidence of that idea. Honestly using MSA as a focus does not account for how the 30 seats in LA, Orange, San Diego, San Bernindino and Riverside counties are chopped up. All of the seats are in the same MSA ditto for the 1st 18 seats in NY. If you don’t set standards for preserving city and county lines (hard standards not like FL and CA where courts and commissions abused them) then MSA means very little. Of course I note in your maps MSA are the standard until they are not the standard.

            I note that courts in the 4th circuit told legislators to draw maps in NC and VA. Even in WI dispute the ball is back in the legislators hands.

            Thankfully IMO the courts will continue to, even more so now with Trump in the whitehouse, place redistricting in the hands of the legislators.

            • Torie February 21, 2017 at 1:54 pm

              Within large jurisdictions everything is macro-chopped, so you treat each town or township and city as a separate county, with the same chop and erosity rules, and in large subdivisions, they are further divided by planning district, or ward, or city defined neighborhood district (which NYC has) or whatever, whatever makes the most sense. NYC has five boroughs to boot, so you get penalized for chopping those as well. The computer would give the Pubs a seat in south Brooklyn for sure that the court cheated the Pubs out of, but I digress. Most of NYC is driven by the VRA, with the 4 white CD’s picking up what is left after the VRA does its thing.

              So the computer knows how to do all of this. Be happy!

              65 - NY-19 (D)

              • rdelbov February 21, 2017 at 2:47 pm

                I am not worried about the computer its the programmers!

                Basically you have programmers setting the rules for redistricting as opposed to either elected pols or commissions. As Jon mentioned the same program that slices up NY city or LA county can’t be used to slice up St Louis county with its 92 cities.

                Someone at some point has to program the perimeters and decisions has to be made in setting those standards. I went through this with RM last year. You can set the very same standards for NY state but if you start you map in Buffalo or Albany or at Montauk you get very different maps.

                If you lay out the 1000 decisions that you plug into any algorithmic program and start picking and choosing what decisions that your program will make maps will change.

                One can change prioritize county or city integrity within an MSA or not. You can classify a county within an MSA as outlier like Wyoming county is.

                I don’t worry about the computer its the software and who designs it. I might add that whatever standard any person is just that an arbitrary decision. If you give 1st priority to MSA composition or 1st priority to county integrity or city integrity or you put an emphasis on overall compactness or neighborhoods or race or whatever. Those are decisions. I don’t see why we should trust programmers and scholars to make those decisions more the pols.

                We have maps above where cities and counties are sliced and diced. Philly 4 times instead of three-Chester county three times instead of one or two. Jackson county is split–St louis county is split in a strange sort of way-to me at least. I note that AZ and CA commissions produced among the most contorted lines in history of redistricting so I refuse to cede the moral high ground of redistricting to either commissions or programmers.

                I for one would happy with a MI or NC style congressional map for every state. With the pols in charge. Just these two maps show that decisions by pols can be as rational if not more rational then those made by some programmer.

                • Jon February 21, 2017 at 7:18 pm

                  New York is actually one of those states where there’s only two logical places to start redistricting at:

                  Eastern tip of Long Island (from which they are renumbered to follow, and the other is western edge of the state near Buffalo. (Long Island somewhat preferred) There’s a significant risk of painting yourself into a corner starting significantly away from those two.

                  TN is another example of these; either start in the Tri cities or Memphis.

                  Missouri maps pretty much have to start in St Louis City to avoid painting into corners.

                  On the other side there are states like Iowa: You can pretty much start anywhere in that state without running into painting issues later. (Here is a place where their rules are extremely easy to follow; there are several possible configurations for their congressional districts that would have reasonable population variance. If I recall correctly, in addition to them requiring there be no split counties for congressional districts they aren’t allowed to split a state house seat [under its new lines] either. There is still a human element, what “looks most compact” among the choices with low population variance.)

                  45, M, MO-02

                  • rdelbov February 21, 2017 at 8:51 pm

                    When you start in Erie county you get one CD and some leftovers. So do you go north or south? Compact or not how can you split the Southern Tier? Decisions decisions? I might add IMO NY MSAs are gerrymandered. Counties that are just next to Monroe county but not really economically connected to it are put in that MSA.

  • rdelbov March 3, 2017 at 12:41 pm

    Hogan calls for redistricting reform

    I would love to see our KOS friends and others who have all sorts of non-partisan mathematical ways to draw congressional seats to support reform in this state

    Looking at MSAs and “fair” redistricting in MD it really falls together fairly neatly

    Seats 1 is Montgomery
    Seat 2 is PG
    Seat 3 is rest of Montgomery, Rest of PG, Charles plus some 1/2 of Howard
    Seat 4 is Garret to Carroll plus adjustments
    Seat 5 is Balt city plus Balt county
    Seat 6 is Balt county plus adjustment
    Seat 7 Anne Arendel-Calvert-St Marys plus a minor adjustmnet
    Seat 8 is Eastern Shore plus some Harford

    GOP should have a chance to win 4 seats under a fair map. I hope our D friends stop doing this currernt gerrymandering for their own good.

    • Jon March 4, 2017 at 10:58 am

      Too me that looks like a 2 – 6 map in a normal year; which is about the best you can do without a realignment.
      The problem is compounded by the the closest to even places being in areas where Democrats move into.
      I do note that a clean 1 – 7 map is easy to draw for this state.
      The difference between these is what you attach to the westernmost Maryland district. (Baltimore western exurbs or DC western exurbs)

      45, M, MO-02

      • rdelbov March 4, 2017 at 9:41 pm

        My seat#4 & #8 are gimme GOP seats

        Trump won #7 by 10K before any votes from Howard county. Trump as opposed to say Hogan ran poorly in MD. If he wins CD7 then its a GOP lean seat.

        I fully admit that my Baltimore county seat is an unknown. If you give the inner strong D suburbs to Balt City seat then clearly my #6 seat which will need voters from Harford county will be tossup to lean R

        • rdelbov March 5, 2017 at 9:26 am

          I should further explain that there are several easy ways to attach –If attach the AA suburbs to NW of the city then the Baltimore county seat would be overall R leaning. If you go southward or eastward it would be less so. Of course the seat would still need population from elsewhere which would tilt that seat to the right.

  • Torie April 1, 2017 at 5:15 pm

    The county census estimates came out about 10 days ago, and I have updated my NY map for the next redistricting cycle based on extrapolating forward the population trends to the April 1, 2020 census date. The maps have been added to the pile at the bottom of the opening post. NYC growth has slowed down, which causes given the funnel shape of the state, an amplified impact on the CD lines in the Hudson Valley. It will be interesting how close the NY legislature gets to this map, under NY’s new redistricting law. Again, all of this is done rather slavishly hewing to an algorithm. There is some judgement exchanging chops for a better erosity score, but now much. Enjoy or complain or ignore as you wish. 🙂

    Oh, NY-07 is a creature of the VRA, and it is an interesting question whether borough lines need to be cross chopped to get the HVAP up a couple of percentage points (that is about all that is available) from its existing 50.8%. I doubt it, so I didn’t do it. But the erose shape does unite contiguous Hispanic areas, and thus will very probably be required.

    65 - NY-19 (D)

    • GOPTarHeel April 1, 2017 at 6:43 pm

      Do you have 2016 results numbers for those seats?

      R/NC. Waiting for a non-ossified establishment or sane populists. Not optimistic.

    • Jon April 1, 2017 at 10:37 pm

      There’s an interesting note in the NYC area regarding VRA implications that doesn’t show up well on DRA.
      Caribbean-AA currently has a separate congressional district adjoining another AA district ; in DRA these groups are added together, but they are considered separate enough groups there that combining the two wouldn’t count as a cohesive group. Basically if drawing a realistic map, don’t change the boundary between those two districts from what they are currently.

      For #7, is that a portion where the Hispanics are mostly from PR (and thus already citizens)? Or from another area? (That’s another thing that doesn’t show up on DRA)

      Minor nitpick: With the boundaries you’ve chosen, I’m positive that what is labeled as #3 and #4 would be flipped to follow New York’s renumbering all districts; there’s probably a few others that would also get renumbered under these boundaries.

      45, M, MO-02

      • shamlet April 1, 2017 at 10:45 pm

        Is 8 or 9 the mostly-Carribean seat? I used to know and now I’ve forgot.

        R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • Republican Michigander April 2, 2017 at 10:55 am

        Are the “black Carribeans” considered “black” or “Hispanic” in the case of Dominicans.

        BTW – I think that one question in and of itself describes how friggin stupid the government, bureaucrats, and especially academia are on the issue of race in this country. I didn’t mention “interest groups” because they have self-interest in keeping the policy. $$$

        MI-08 - Michigan is a red state again. We need a 50 state strategy and an 83 county strategy.

  • Torie April 2, 2017 at 10:25 am

    Clinton Trump Margin
    E Suffolk 1 43.81% 56.19% -12.39%
    W Suffolk 2 49.10% 50.90% -1.81%
    S Nassau 3 55.85% 44.15% 11.71%
    N Nassau 4 53.61% 46.39% 7.22%
    Jersey border 17 49.46% 50.54% -1.09%
    E Hudson Valley 18 54.05% 45.95% 8.09%
    Albany 19 53.58% 46.42% 7.15%
    Syracuse 23 50.55% 49.45% 1.11%

    The above are the 2016 figures, although the Long Island CD’s are estimates, but should be pretty close. My NY-04 might be the most inaccurate, because I estimated the results in Queens from the McCain numbers, adding a bit more of a pad for Clinton.

    I suspect the black Carribeans are along the Queens-Kings line. There is not too much flexibility as to how to draw if you want two BVAP CD’s in Kings that are more than 50% BVAP, or close to it, plus an Hispanic CD. The VRA does not take cognizance of the existence of Hispanic blacks, or various subsets of Hispanics, as a separate ethnic group that triggers section 2. But the NYS legislature may well do so.

    65 - NY-19 (D)

    • krazen1211 April 2, 2017 at 12:00 pm

      Splitting Suffolk East-West and Nassau North-South is, well, interesting in its own way. I wonder if splitting Nassau East-West would change the numbers for the 3rd district.

      • Torie April 2, 2017 at 1:38 pm

        It’s driven by the way the town lines run, and minimizing the size of chops, and avoiding macro-chops like the plague where possible. One needs to track the town lines, and as you know, some of the towns in Suffolk and Nassau are huge. Hempstead is larger than a Congressional district.

        65 - NY-19 (D)

        • krazen1211 April 2, 2017 at 4:22 pm

          Oh, sure. Here’s a problem I see. You have created a rather lousy set of districts for the GOP, at least on Long Island. The current set of districts that were created by the judge are somewhat favorable to the GOP as they have 2 districts that Trump won by about 10 points. Trump won Nassau + Suffolk by about 1.6 points. In this map, 3 out of 4 of districts are moved in favor of the Democrats by sizable margins compared to the existing.

          I could see the GOP trying to get a better map via the commission. A NY-01 that adds the remainder of Smithtown + maybe portions of Huntington, and a NY-02 that is Islip + Babylon + larger portions of heavy Trump Oyster Bay (places around Bethpage) would be good too.

          Alternatively, if one went back to the 2002 configuration and split NY-02/NY-03/NY-04 vertically, a NY-02 that is Islip + Babylon + Huntington and NY-03 that contains all of Oyster Bay + Glen Cove + portions of Hempstead would work as well.

          The old ‘each side loses 1’ rule doesn’t look like it will be in place in 2021 as the GOP is taking the hit of the lost district.

          • Torie April 3, 2017 at 8:23 am

            The GOP is actually getting a pretty good deal. The GOP takes the hit for the lost CD (Faso loses to Maloney), but it still has a pretty favorable second CD on Long Island in my NY-02 that is competitive, so say a loss on paper of 1.5 seats on the liability side, but in reality, -1.0 if King is willing to move a couple of miles to the east and run in NY-02 (now very hostile NY-03 is another matter for him). Balanced against that, it picks up the south Brooklyn CD that it was cheated out of by the court, and my NY-17 should be pretty favorable to it, particularly as a open seat, without the popular incumbent Maloney to contend with. So in the end, it is pretty much a partisan wash to maybe a gain of a seat, with the lagniappe for the GOP based on upstate population stagnation, that the CD’s up there move to the GOP in varying degrees as the CD’s expand out from their urban cores. Heck even the Albany CD could be competitive with the right GOP candidate if the seat opened up, and Katco will be safe in the more GOP Syracuse CD. The ying and the yang.

            The map you want would send the erosity score through the roof as the towns on Long Island are macro-chopped. The algorithm punishes macro-chops pretty severely. That is where the gerrymandering game is mostly placed, in generating and then creatively placing, macro-chops.

            65 - NY-19 (D)

            • krazen1211 April 3, 2017 at 4:55 pm

              Well, let’s look at all the districts, then.

              I agree that the upstate configuration you have is rather favorable, but the court and Roanne Mann already gave us a semi favorable map in 2011 to form as the basis of the status quo, right? More importantly, since an R district is being dissolved, that is 700k theoretically R leaning votes that have to be dispersed among other districts.

              Here are the points I see in critical districts (I think you flipped NY-03 and NY-04 compared to the current map)

              NY-02 moves to the left by about 9 points. Currently Trump won by 9. I cannot characterize this as favorable.
              NY-03 moves to the left by about a point. Currently Trump lost by about 6.
              NY-04 moves to the left by about 2 points. Currently Trump lost by about 10.
              NY-11….where is this? Trump won the current district by 10. We all know that you can flip this district by running it into the ‘wrong’ places in Brooklyn or Manhattan.
              NY-10….I will take this and reconsider pending the last question. But I am unconvinced that we can get 2 districts in NYC even if we get 20% of the vote in NYC. Usually when I have tried to get a S Brooklyn district it squeezes the SI district.
              NY-17 (Maloney?) you have as Trump +1…..but Trump won Maloney’s existing district by 2 points.
              NY-23 (Katko) certainly gains.

              So we are taking the hit for the lost district, and not really gaining ground in most of the R districts. In fact losing ground in many of them. And this is not compared to a gerrymander….the existing map is a court map! So I do not think it makes the case for the thread title too well.

          • rdelbov April 3, 2017 at 10:10 am

            I have seen this map after map from Torie–Every so called non-partisan arrangement of the lines in his maps favor the democrats!!

            If you just rolled the dice or flip cards somewhere around 50% of the line moves would favor the Rs. Every map above has too lines that favor the Ds.

            Too comical.

            • krazen1211 April 3, 2017 at 5:14 pm

              I do not mind chopping Long Island 2-2, just adding 30k each to NY-01 and NY-02, and calling it a day…..even as Long Island is Trump territory. A nice gerrymander might give us 3 districts by flipping NY-04, but its hard due to the location of the non-white population right in the center, and the need to pour in 100k additional 2-1 Queens voters into some district crossing into NYC.

              But certainly 1-3 the other way is not good at all. King is certainly safe in any district there, but I’ve learned not to draw for certain incumbents when you can avoid it, and the current map does.

    • Jon April 2, 2017 at 1:09 pm

      The key phrase within the VRA is “sufficiently cohesive” ; when both groups were placed together in the past, they frequently opposed each other in primaries. This was amplified in one of the SCOTUS decisions involving redistricting in Texas post 2000 specifying stating that “coalition” seats aren’t protected. (Involving a seat in Texas in which Hispanics and AAs combined for more than 50% held by a White Democrat)

      This same phrase is one of the reasons that IL was able to get away with post 2010 not drawing a second Hispanic VAP district near Chicago (which by VAP standard would have both been trivial and a lot cleaner); the second one would have mixed different Hispanic groups even if it could be gotten above 50% by CVAP. (Plaintiffs and Defendants disagreed on which side of the 50% CVAP line the plaintiffs preferred map would have been ; citizenship status wasn’t on the census.)

      45, M, MO-02

  • Greyhound April 3, 2017 at 5:04 pm

    This might be utterly off-topic, but has anyone figure out how to get DRA to work since all the major web browsers discontinued MS Silverlight support? I mean besides using the older versions of said browsers that are going to get discontinued next year.

    R, 27, CA-18. Anti-Anti-Trump

    • shamlet April 3, 2017 at 5:05 pm

      Internet Explorer still seems to work.

      R, MD-7. Put not your trust in princes. Process is more important than outcome.

      • Jon April 3, 2017 at 7:05 pm

        IE’s drop of support is more a matter of don’t call us if you have any problems with out browser with Silverlight rather than a change to stop allowing the silverlight plugin to work ; in large part it’s because Microsoft is pushing Edge and not really supporting changes to IE. (IE itself is going to become unsupported in the near future ; Windows will still run it but MS answer to any browser crash questions will be upgrade to Edge.

        45, M, MO-02

    • roguemapper April 3, 2017 at 5:21 pm

      Firefox ESR will support Silverlight until at least March 2018.

      Dem NC-11

  • rdelbov April 7, 2017 at 9:03 am

    With Gorsuch being sworn in soon I believe we will seen any major changes in court actions til the 2031 cycle at best.

    One can’t take anything for granted but our partisan algorithm map makers can take a pause for awhile.

  • Torie July 3, 2017 at 9:56 am

    A new map for PA is up at the bottom of the introductory post. Yes, in case you are wondering, Trump carried PA-06 as drawn, by a grand total of 53 votes. And Toomey carried PA-10 by 89 votes, with the small chop into Pike County pushing him over the top. Trump of course carried PA-10 easily.

    65 - NY-19 (D)

  • Leave a Reply

    Do NOT follow this link or you will be banned from the site!