House breakdown if...

[jimrtex]

P-4 = zero. Only Putnam and Richmand are smaller than a CD, and both stayed intact. It does create the only geographically strange district as Putnam is linked to the northern Bronx along the eastern edge of towns  in Westchester. Richmond is connected to Kings by the Verazzano Bridge.

I'm wondering if the split county rule could be stated a different way.  Something like: if a county has population equivalent to N+F districts, where N is an integer (including 0) and F is fraction between 0 and 1.  Then the county should not be split between more than N+1 districts.

That would maintain the rule of not splitting small counties, but would permit splitting large counties in more natural ways.  I'm assuming that there is a way that would split Westchester in two parts, the northern placed with Putnam, and the southern with the Bronx, as opposed to one whole district in Westchester, and the remnant linking Putnam and the Bronx.

On the other hand, I think that the whole district rule works OK in places like Erie and Monroe Counties.  Trying to express it in terms of linking, might not work (eg the remnant of Erie links Niagara and Chautauqua).

A contiguity rule could probably be stated in a form that takes into account surface transportation routes, such that for example, Staten Island is connected to Brooklyn at the end points of the Verazzano Bridge and with Manhattan via the ferry landings.

Islands within a legal jurisdiction would be contiguous at their nearest point.  This definition might be extended to enclaves as well.

[minionofmidas]

I'd like a link to data on these NY hoods, please.

[muon2]

I'd like a link to data on these NY hoods, please.

I used data from the New York City Department of Planning. They have compiled data on each of their community districts.

[Gustaf]

How's it going Muon? Can we hope for a non-gerry mandered house map before the next midterms to see what the house division *should* be?

I would help with what I guess is a LOT of work if I knew anything about this. But I don't.

[minionofmidas]

I'd like a link to data on these NY hoods, please.

I used data from the New York City Department of Planning. They have compiled data on each of their community districts.

Very very interesting. Thanks a lot.

[muon2]

I just calculated to asign the seats to the NYC are.  It should be possibe to give Bronx two seats and Manhattan-Staten Island three seats. The rest of the counties is are obviously more difficult.  The districts over Long Island Sound are difficult to make decent.


We just wait the political analysis. Not many GOP districts ?

Not likely, but I can't really say at this point. Dave doesn't have town information for NY yet. Also it will be hard to assess how strong some of the districts are without a breakdown within the city.

[socaldem]

I've completed Ohio using the usual rules.



My comments:

Only one city, Columbus, is larger than the population of a district. Columbus also has a number of unincorporated pockets inside, and I'm only guessing that I can make a district entirely within the City limits, and provide the necessary links between the pockets and the adjacent districts. With that assumption, rules P-1,2,3 are all zero.

OH has a number of counties with large urban centers, and that forces some county splits.  I split  Montgomery and Stark that were less than one district size, and split  Stark into three pieces to avoid
splitting a third small county.  Rule P-4 is two.

All three large counties had a district wholly within them, and Cuyahoga has two wholly within. Rule P-5 is zero. All three also only have two partial districts. Rule P-6  is 11 partial districts, and rule P-7 is three maximum in any county.

I've completed Ohio using the usual rules.



My comments:

Only one city, Columbus, is larger than the population of a district. Columbus also has a number of unincorporated pockets inside, and I'm only guessing that I can make a district entirely within the City limits, and provide the necessary links between the pockets and the adjacent districts. With that assumption, rules P-1,2,3 are all zero.

OH has a number of counties with large urban centers, and that forces some county splits.  I split  Montgomery and Stark that were less than one district size, and split  Stark into three pieces to avoid
splitting a third small county.  Rule P-4 is two.

All three large counties had a district wholly within them, and Cuyahoga has two wholly within. Rule P-5 is zero. All three also only have two partial districts. Rule P-6  is 11 partial districts, and rule P-7 is three maximum in any county.

I hear Ohio will now have an independent commission drawing their maps... I think they should hire Muon as an advisor...

What would happen if a plan like this were really adopted?

I know others have done a partisan analysis already, but i'll insert some political figures...

OH-01, Steve Chabot might not even try to compete in the blue Cincinnatti-based district as he may prefer to go to more friendly suburbs... but I would predict that either way the district would probably elect the loser in the current race for Cincinnatti Mayor (either Pepper or Mallory) or former mayor Quills in a heartbeat...

OH-02, the Cleremont-county-based red district would be home to Rep. Jean Schmitt, though she may face a primary challenge again...

OH-03 and 04 the green Cincinnatti suburban district and the neighboring red district with which it splits Montgomery county and apparently dayton would probably still be good bets for Republicans.  I am unsure which incumbents would run in what district, though, as Boehner, Turner, and Chabot might all be interested and its really unclear which Turner or Boehner would prefer to run in...a united Montgomery county, though, it would seem to me would be a better district (and one which would offer dems a better chance!)...

OH-05, the blue South Ohio district would not actually be home to any of the incumbants but its traditionally Democratic territory that has long been supportive of Democrats like Ted Strickland and the counties that were added from OH-02 went overwhelmingly for Paul Hackett.  It includes much of the territory now in Bob Ney's district as well, though I don't think Ney lives there... Ney's likely opponent, Chillicothe Mayor Joe Sulzer would likely run here and, I think he'd win even though the district would have gone for Bush...

OH-06, the yellow Eastern Ohio district combines areas currently held by Strickland with those held by Ney and Ralph Regula, who would proably run in this district because he's based in Stark County, even though that's his weakest performing county...  The Democratic candidate state senator Charlie Wilson whom I believe is favored to replace strickland would probably be favored here if no incumbant ran and would have a good chance against any Republican comers...especially since this district very closely fits his current state senate seat...

moving up the oh/pa border...

OH-07 the Mahoning-based green district would easily elect Tim Ryan even with the addition of republican voters...

OH-08 the yellow district above it would also easily elect a Democrat in a district that is now held by Rep. LaTourette by putting Geouga county in Tim Ryan's district, leaving this district for any number of Democrats including either of their Lake County state reps or state sen. Mark Dann...

OH-09 and OH-10 will be the Cleveland-based districts where Democrats should win easily...with Stephanie Tubbs Jones easily keeping Cleveland proper... Kucinich, however, might actually have difficulty getting reelected if his district were changed too much...

OH-11, the blue summit county seat, which is essentially a recreation of Tom Sawyer's old seat would pretty clearly favor a Democrat.  Perhaps even the former congressman...

OH-12, the yellow Eastern Cleveland district would be a toss-up race in an open seat, but I don't see why Rep. Sherrod Brown would lose there, though he might have a few close contests...

OH-13, the red Northwesternmost district would go solidly for Rep. Marcy Kaptur

OH 14, the green district in the northwest of the state and OH-15 the red district in the center of the state are both very republican and since each represents areas represented by a number of incumbents its hard to tell who would run where...

OH-16 and OH-17 would be Columbus suburbs with the western district likely retained by Deborah Pryce and the Eastern district a combo of Dave Hobson and Pat Tibieri's district.

OH-18, Columbus proper would be a solidly Democratic district that could easily elect Mayor Coleman or any number of city councilmen...

So, if this is what the redistricting plan leads to, I think it would be a wonderful map, with dems having very good chances at winning eleven seats and possibly a 12th if the Dayton race were competitive.... that would be a gain of 5 seats! 

[jimrtex]

I've completed Ohio using the usual rules.

I hear Ohio will now have an independent commission drawing their
maps... I think they should hire Muon as an advisor...

What would happen if a plan like this were really adopted?

There is an initiative on the November 2005 ballot that would establish an independent redistricting commission.  The commission would have fairly limited line drawing powers.  Instead it would evaluate plans submitted by "private individuals" on the basis of how competitive they were. 

Competitiveness is determined on the basis of an objective scoring system.  A partisan index is determined on the basis of the three closest statewide elections over the previous 8 years (4 elections).  For a district, the partisan index is the difference in the average of the percentage of support between the Republican and Democrat candidates.

• Districts where the margin is less than 5% are considered competitive.

 
A competitive district is worth 1 point.  But if there are two competitive districts that favor opposite parties, they are considered balanced and together are worth 4 points.   If one district has a 0.1% Democrat index and another a 4.9% Republican index, they are considered balanced.  If the one district flipped to 0.1% Republican they are no longer balanced.

• Districts where the margin is greater than 15% are considered uncompetitive.  If there are two uncompetitive districts that favor opposite parties, and the absolute value of their partisan index is within 5% of each other, they are considered to be be balanced uncompetitive districts.

An unbalanced competitive district scores a minus two points.   Balanced uncompetitive districts score zero points.

• Districts where the margins is between 5% and 15% are neither competitive or non-competitive, and score zero points.
  
  muon2 calculated a competitiveness score for his plan.  The score could be improved by a small amount.  For example, adjusting the partisan index so that a district was barely within the 5% threshold, or so that there were more pairs of balanced concompetitive pairs.
  
  But to get a really high score, you need to maximize the number of balanced competitive districts.  In Ohio, you can create balanced competive pairs in the Cincinnati and Columbus areas, but for the others you need to combine the Democratic areas of NE Ohio (Akron, Canton, Youngstown) with the Rural and small town Republican areas of western, central, and southern Ohio.  Such districts will need to be very thin and long.
  

[WMS]

I've completed Ohio using the usual rules.

OH-12, the yellow Eastern Cleveland district would be a toss-up race in an open seat, but I don't see why Rep. Sherrod Brown would lose there, though he might have a few close contests...

You meant Western Cleveland, right?

[muon2]

Just a note on Cleveland. The map I drew had all of the city of Cleveland in a single district, with a handful of nearby suburbs to bring it to the required population. The specific suburbs added were: Bratenahl, Cleveland Heights, East Cleveland, Lyndhurst, Mayfield Heights, North Randall, South Euclid, and University Heights. This district is 81.3% D, and I agree would stay with Rep Jones.

The rest of Cuyahoga, except for Euclid, Highland Heights, Mayfield, and Richmond Heights is in a single district. It includes none of the city of Cleveland. The district is strongly D at 58.3%, but I don't know if Kucinich could hold it in a primary or against a moderate R. That eastern bit of Cuyahoga with Lake, Ashtabula, and Trumbull should also elect a Democrat, based on a 55.9% D rating.

Where I would differ from socaldem's analysis is in some of the other seats. In predicting 11 or 12 seats there is an implicit assumption that conservative Dems with run in conservative districts, but liberal Rep aren't competing in leberal districts. I tried to avoid those assumptions, but look at the leanings of the districts based on their 2004 performance. I get 8 R, 8D and two competitive districts with less than 2% margin. Those include the east central district and north central district, both in yellow.

[minionofmidas]

The rest of Cuyahoga, except for Euclid, Highland Heights, Mayfield, and Richmond Heights is in a single district. It includes none of the city of Cleveland. The district is strongly D at 58.3%, but I don't know if Kucinich could hold it in a primary or against a moderate R.

He seems to be quite popular with his constituents though.

[muon2]

The rest of Cuyahoga, except for Euclid, Highland Heights, Mayfield, and Richmond Heights is in a single district. It includes none of the city of Cleveland. The district is strongly D at 58.3%, but I don't know if Kucinich could hold it in a primary or against a moderate R.

He seems to be quite popular with his constituents though.

I know he was the mayor of Cleveland, but I don't know how his popularity is in the suburbs in western Cuyahoga. He represents them now, so he might well have built a strong base.

[muon2]

Since precinct data is now available in NY, I can start filling in some of the NYC/LI districts.



NY 1 (green) Brookhaven - 50.5% R (competitive)
NY 2 (blue) Islip - 51.2% D (competitive)
NY 3 (red)  Hicksville -  51.4% D (competitive)

I'm looking for on-line precinct maps of Nassau and NYC to get the remaining districts analyzed.