Dave's Redistricting App

[jimrtex]

10 districts for Mass on 2000 population.   Boston, 3 inner suburban, 3 outer suburban-exurban, Worcester-Springfield, Western Mass-Springfield, and Fall River-New Bedford-Cape Cod-Islands

[jimrtex]

...Good Lord. I think even Tom DeLay would consider that a bit excessive.

Lest we forget the true evil of Martin Frost, ex-congressman from Texas.

The red district is the part of Delay's district in West Houston, the green district is Mike Andrews-Ken Bentsen-Chris Bell's former district.  The yellow district is part of Craig Washington-Sheila Jackson_Lee's district.  The blue is Bill Archer's district.  This was a totally legal political gerrymander.



This is further east in SE Houston.  Delay's district is the red district at the bottom.  The magenta district is Gene Green's district.  This of course is an illegal racial gerrymander.



This is in north & northwest Houston.  Besides all the weird peninsulas in Green's district, you have another district wrapping around it.  The neck of the magenta district in the lower right is about 3/5 of a mile across.  To get between the yellow district on one side to the other is about 60 miles.

In the Ohio Redistricting contest the measure of compactness is area divided by perimeter squared.  A perfect circle would be 0.0796.  A square 0.0625.  The median district in the contest was about 0.029.  The worst district, which wrapped completely around Cleveland plus almost got chopped off in Parma was about 0.007.  Districts in Texas were .00115, 0.00075, 0.00056, and 0.0189.



This is Lubbock County.  Again a legal political gerrymander.   Guess where blacks and Hispanics are concentrated?



The districts meet in Midland which is east of the SE corner of New Mexico.  Dalhart is in the extreme NE corner of the panhandle.  Decatur and Granbury are exurbs of Fort Worth.  Round Rock is north of Austin and New Braunfels is NE of San Antonio.



This is zoomed out quite a bit.  The area in the previous map is in the far NW corner of this one.   The orange district is Austin.  It is kind of a donut hole district with one district stretching west to Midland, and the magenta district stretching SE to Galveston.  The Democrat congressman from the magenta district switched parties and then was beaten by Ron Paul in the primary.  The yellow district is central San Antonio.



This is San Antonio and the area to the NE around New Braunfels.  The red district stretches to El Paso, Midland, and Laredo.  The area on the north side of San Antonio was added at the request of the Democrat congressman who was building a house there.  After being caught in the Post Office Banking scandal, he was beaten by Henry Bonilla.



This is Fort Worth with Dallas on the eastern part of the map.  The link in the red district in the NE part of Tarrant County is by the surface of Eagle Mountain Lake.  That is why the edges are a bit fuzzier.  The yellow district is former congressman Martin Frost's district which links parts of Arlington to parts of Dallas via Waxahachie.   The red district was a legal political gerrymander.



The blue district is Eddie Bernice-Johnson's.



This is a detail of the northern part of the district, with a bit extending into Collin County.  In a deposition, she claimed that she was just trying to include her friends in her district (she was a state senator at the time the district lines were drawn, and she was elected from the new district).  Parts of this area was also thrown out as a racial gerrymander.



The whole state.

[jimrtex]

Send an e-mail to Dave Bradlee.  He appears to be receptive to assistance to make his program better.  Since block groups nest within census tracts, it should be relatively easy to adjust the data for your estimates, especially if you are willing to do the grunt work.

I took your advice, but he never responded.

When I wrote him, he said that he had a new job and a lot less time to work on it.  Looking at his response it was pretty non-committal, though he said he was working on making it open source.

[jimrtex]

10 districts for Mass on 2000 population.   Boston, 3 inner suburban, 3 outer suburban-exurban, Worcester-Springfield, Western Mass-Springfield, and Fall River-New Bedford-Cape Cod-Islands

A New England Republican might even win the red district.

It is probably the district that goes away with a 9-district plan.  You should be able to get all of the Connecticut Valley in to the western district, which makes the green district more pure Worcester.  And then the gray and light blue would slide westward taking up the rest of the red district.

[jimrtex]

Yeah, the combination of "let's create districts for white Democrats and minority Democrats" plus "ooh, computers" was ugly. Weren't the squiggly inner-city districts in Houston and Dallas overturned by the courts and redrawn?

Bush v Vera, a 5-4 decision by the Supreme Court found that race was the predominating factor in three of the districts 18, 29, and 30.

Justice Stevens in his dissent noted that this district was just as bad as the others, and it wasn't based on race, so this made the others OK.  The district starts in Ellis County, loops around Fort Worth and ends up in Johnson County.  Yellow are areas that were in the district before and after Bush v Vera, red areas were removed, and green areas were added.  Many of the changes appear to have been made for population balance.

13 (of 30) districts were modified to some extent and were contested as special elections concurrent with the November 1996 general election.  The results of the primaries were thrown out, and any candidate could run regardless of party, with a majority required for election.  In several cases you ended up with a rerun of a primary.



This is Martin Frost's district.



This is Eddie Bernice Johnson's district (30).  As state senator she had originally proposed a relative compact district black majority district, but it would have placed several incumbents in the district.  One of the defenses offered for the plan was that her original plan had been adjusted to protect Democrat incumbents, including Martin Frost and John Bryant, and that some Democratic areas that just happened to be black remained with their districts, so that she had to seek blacks further afield.  Since she had a lot of influence over drawing her district, it was claimed that she along with Gene Green and Ben Reyes were de facto incumbents, and it was a legislative prerogative to protect incumbents.

The majority opinion noted that the plan did not follow traditional districting values since the northern tip of the district went into Collin County and a tiny bit went into Tarrant County (the very western nob of the red area on the southern part of the western extension.  Justice O'Connor seemed to suggest that 18 and 29 were somewhat better because they were entirely in Harris County.



This is Tom Delay's district in Harris County.  The areas that were chopped off appear to make up for an area in Fort Bend County that was added.



This is Mike Andrews-Ken Bentsen-Chris Bell district, with a lot of cleanup in the middle.  The green area to the east is the Houston Ship Channel, with very few people, but providing a connection to some Hispanic areas in Baytown (the green tongue at the far east end).  This district is the main reason for the gerrymander in Houston, since they didn't want to force the district any further south into Republican areas.  Now that it has been converted to a majority black district, it is fairly compact.



This is CD 18 which was Craig Washington's district (and before that Barbara Jordan and Mickey Leland, and now Sheila Jackson Lee.  Major concentrations of blacks are in NW, NE, and S Houston, while Hispanics are in the E and SE, N and WNW.  One reason that the districts are weird is that you can only connect one of the districts through downtown.

The original plan had CD 18 wrapping around the Hispanic district to the far west, and then connecting the NE and NW black areas at the top (the airport - IAH) is at the far north so it isn't that heavily populated and neither is the extreme NE.

The revised plan connected the black areas through downtown.  This is still the current configuration.



This is CD 29, that was drawn (somewhat) as a Hispanic district, but has always elected Gene Green, who was instrumental in drawing the map.   A lot of the bizarre fingers were to include apartment complexes, which have population, but less than their share of voters.



This is DFW after the federal district court had adjusted the boundaries.



And this is Harris County afterwards.

[jimrtex]

Whichever court redrew the lines in 1994 seem to have cleaned up the lines a lot better in Houston than they did in Dallas-Fort Worth, from a worse starting point.

<i>Bush v. Vera</i> was delivered in June 1996.  Originally, 24 districts were challenged, but the district court ruled that only 3 districts, two in Houston, and one in Dallas had race been the primary motivation.  In other cases the intent was merely to pack Republicans so that Democrat incumbents could hang on to their seats.  BTW, the Oyez project has a recording of the oral arguments in <i>Bush v. Vera</i>.

In the case of Houston, two districts were interlocked, one black and one Hispanic.  There were hundreds of little twitches to take in a block one way or the other based on race.  So once you defined the main areas you wanted to connect it was relatively easy to provide a connection.  And there was only one other Democratic district that they were trying to preserve,  But if you notice, Tom Delay's district has lot of little indentations and was left alone except to balance the population.  The boundaries drawn by the district court were not as balanced as perfectly as possible, and they expected the legislature to finish the job up and perhaps straighten out some other boundaries.  But that didn't happen until the district court redrew the boundaries after the 2000 census.

The Supreme Court decision was after the primaries had been held, and rather than wait until 1998, the district court ordered the elections for the new districts be held as a special election in November 1996.  In Texas, a special congressional election is held without party primaries or nominations, and requires a majority election.  Even with the minor tweaks to some of the districts, 13 districts had special elections.  In some cases, the runoff was a rerun of the party primary in the spring. 

Ironically, the only material effect of the changes was that Nick Lampson was elected, defeating Steve Stockman.  Stockman had defeated Jack Brooks, a 42-year incumbent in 1992.  Stockman had led the November vote, but had not received a majority.  The runoff was in December, and turnout was much higher in Beaumont and Port Arthur, than in Houston where the election got little coverage.  The only changes to the district were quite small, and if there had been an actual effort to avoid a special election by not making the changes, they probably could have been avoided.

[jimrtex]

10 districts for Mass on 2000 population.   Boston, 3 inner suburban, 3 outer suburban-exurban, Worcester-Springfield, Western Mass-Springfield, and Fall River-New Bedford-Cape Cod-Islands

This is the updated version based on whole towns.  The smallest deviation is to place Milton and Winthrop with Boston.  This shifted Chelsea to the NE, and then the 3 districts north of Boston were rotated counterclockwise, producing more of a coastal Cape Anne district, and interior district that includes both Lawrence and Lowell in the same district, and a district concentrated in the densely population towns immediately north of the Charles.  In the west, I tried to balance the population from the Springfield and Worcester areas, which meant that Chicopee, and all of Hampden County east of the Connecticut river is now in the district.  This resulted in more of an L-shaped district.  Some other small changes were made to better balance population.

Maximum deviation is 0.22%, with a mean absolute deviation of 0.11% or just short of 700.

In a 9-district plan, the Central Massachusetts district between Worcester and Boston disappears.  The western districts should take in Fitchburg and Leominster, and hopefully let the western district include Springfield, while the next district will be mostly Worcester County.

The SE district should take in Taunton, while the next district takes in Quincy.  Other districts may take in Brookline and Newton, which will result in major westward shift in the Norfolk district to take in the southern parts of the Central Mass districts.

Proposed names for districts:

Western Massachusetts
Springfield and Worcester
Central Massachusetts
Lawrence and Lowell
Cape Anne
SE Middlesex
Boston
Norfolk
Brockton, Taunton and Attleboro
Fall River, New Bedford, Cape Cod, and Islands

[jimrtex]

10 districts for Mass on 2000 population.   Boston, 3 inner suburban, 3 outer suburban-exurban, Worcester-Springfield, Western Mass-Springfield, and Fall River-New Bedford-Cape Cod-Islands.

Plus a modified version based on 2007 estimates.  10 towns were shifted:

(1) + Ashburnham
(2) no change.
(3) - Ashburnham, - Ayer.
(4) + Ayer, + Merrimac, + West Newbury, - Wilmington
(5) + Melrose, - Merrimac, -Wakefield, - West Newbury
(6) - Melrose, +Wakefield, + Wilmington
(7) no change
(8 ) + Plainville
(9) + Freetown, - Kingston, - Plainville
(10) - Freetown, + Kingston

For the new plan maximum absolute deviation 0.16% (1015); mean absolute deviation 0.05% (329)

2000

2007

[jimrtex]

2000 10 districts:

2007 9 districts:

Mean absolute deviation 0.1% or about 694.  As expected, the Central Mass district disappears.  It is possible to get most of the more populous Connecticut Valley into the West Mass district.  The plan pretty much follows county boundaries with Franklin County minus one town with Worcester; and all of Hampden county, and all of Hampshire County minus 3 towns with Berkshire County.  Fitchburg and Leominster were shifted to the west.  The city of Worcester is close to the edge of its district, but I'm not sure that that there are better alternatives.

Quincy and Milton are the best fit with Boston (0.003%) which meant that the suburban districts ended up being shifted clockwise, to account for the shift in Quincy.

Given the slow growth of the State, something like this plan could be stable until Massachusetts loses its ninth district, though the Boston district could have major shifts in order to maintain population equality and not splitting towns.

[jimrtex]

Massachusetts with districts of approximately 100,000 persons each.  If there were 3 members per district, the 189 members would approximately comply with the cube root rule (185 for Massachusetts population).

The average deviation is 0.96%, 38 districts are within 1% of the ideal population, 55 within 2%, and all 63 are withing 3%.

Split cities/towns are Springfield, Worcester, Lowell, Newton, Somerville, Melrose, Lynn, Boston, Quincy, Fall River, and New Bedford.









1. Pittsfield & The Berkshires
2. Northern Connecticut Valley
3. Amherst, Northampton, & Central Connecticut Valley
4. Westfield, Agawam, & Southern Connecticut Valley
5. Holyoke. West Springfield, South Hadley, & Easthampton
6. Chicopee, Ludlow, Belchertown, & Palmer
7. Springfield Central
8. Springfield East & Greater Springfield
9. Worcester Southwest
10. Fitchburg, Gardner, & Worcester North
11. Middlesex Northwest & Worcester Northeast
12. Leominster & Worcester Central
13. Worcester City North
14. Worcester City South, Auburn, & Millbury
15. Shrewsbury, Worcester East, & Middlesex Southwest
16. Worcester Southwest
17. Chelmsford, Dracut, Westford, & Middlesex North
18. Lowell
19. Billerica, Tewksbury, Burlington, & Middlesex,Northeast
20. Wakefield, Reading, Stoneham, & Wilmington
21. Malden, Saugus, & Melrose
22. Medford & Arlington
23. Woburn, Lexington, Winchester, & Bedford
24. Waltham, Concord, Sudbury, & Lincoln
25. Marlborough, Acton, & Middlesex West
26. Framingham & Natick
27. Needham, Wellesley, Dedham, & MIddlesex South Central
28. Brookline & Newton South
29. Newton North, Waterton, & Belmont
30. Cambridge
31. Somerville & Everett
32. Lawrence & Andover
33. Metheun, North Andover, & Essex West
34. Haverhill & Essex North
35. Gloucester & Cape Anne
36. Salem, Beverly, & Marblehead
37. Peabody, Danvers, & Essex South
38. Lynn, Swampscott, & Nahant
39. Revere, Chelsea, & Winthrop
40. Boston (East Boston, Central Boston, & Charlestown) & Somerville South
41. Boston (Allston, Brighton, Back Bay, & Fenway)
42. Boston (Dorchester & South Boston)
43. Boston (Roxbury & South End)
44. Boston (Mattapan & South Dorchester)
45. Boston (West Roxbury, Jamaica Plain, Roslindale, & Hyde Park)
46. Franklin, Milford, Norfolk West, & Middlesex South
47. Mansfield, Easton, Norfolk Southwest, & Norton
48. Stoughton, Walpole, & Norfolk West Central
49. Randolph, Norwood, Canton, & Norfolk Central
50. Quincy & Milton
51. Weymouth, Braintree, & Quincy South
52. Attleboro, North Attleborough, & Bristol Northwest
53. Taunton, Bridgewater, Raynham, & Halifax
54. Dartmouth & Bristol Central
55. Fall River & Westport
56. New Bedford & Acushnet
57. Brockton & West Bridgewater
58. Plymouth Central
59. Marshfield & Plymouth North
60. Plymouth Town & Plymouth South
61. Falmouth, Wareham, Buzzards Bay, & Martha's Vineyard
62. Barnstable Town, Sandwich, Bourne, & Mashpee
63. Yarmouth, Cape Cod, & Nantucket

[jimrtex]

I can't believe you did that.

I've never seen MV and Nantucket split before, but the ferry routes do argue for your division, and it gets the islands double representation.

I would have had to split Barnstable town if I had left Nantucket out of the Cape Cod district.  And then if I had gone from Martha's Vineyard to Wood's Hole and then northward towards the Cape Cod Canal, I would have had to get two districts where there was just one line of towns.  And Plymouth has a fairly substantial population so it might have needed to be split.

So this worked out pretty good population-wise.  I probably could have tried to include the ferry landings in New Bedford, but it ended up that just a small bit of that city, so I took an area adjacent to Fairhaven.

[jimrtex]

The Boston City Council with 21 districts




District number- White%-Black%-Hispanic%-Asian%-Oth%

01- 79%-2%-6%-12%-1% (blue, SW Allston/Brighton)
02- 62%-6%-14%-16%-1% (green, Central Allston/Brighton)
03- 66%-4%-10%-18%-1% (purple, East Allston/Brighton, Boston University)
04- 77%-4%-12%-7%-1% (red, Charlestown)
05- 41%-3%-51%-4%-2% (yellow, East Boston)
06- 53%-10%-10%-26%-1% (teal, Chinatown and surrounding areas)
07- 87%-2%-4%-6%-1% (gray, Back Bay)
08- 87%-1%-9%-2%-0% (blueish, East Boston)
09- 68%-10%-10%-11%-1% (skyblue, SE of 3, NW of 16)
10- 26%-22%-47%-4%-1% (fuchsia, east of 20)
11- 90%-2%-4%-3%-1% (Green, SW Boston)
12- 42%-23%-20%-12%-2% (skyblue, East, with an extension into central)
13- 12%-53%-23%-8%-4% (pinkish, North of 14)
14- 4%-68%-21%-3%-4% (gold, NW of 17)
15- 39%-32%-26%-2%-1% (orange, S/SE of 10)
16- 32%-35%-25%-5%-3% (green, SE of 9)
17- 48%-25%-9%-16%-2% (blueish, NE of 18)
18- 18%-66%-10%-3%-2% (yellow, south, east of 19)
19- 16%-69%-12%-1%-1% (olive, south)
20- 63%-12%-15%-9%-1% (soft pink, west of 10)
21- 54%-25%-16%-3%-1% (maroon, SW corner)

Any thoughts

The census bureau data for voting precincts is organized by wards, but in Boston it appears that the wards are no longer used for political purposes, since they have quite variable populations and are reasonably compact.  In other cities in Massachusetts, it appears the wards are still used for city council elections.  In many cases there will be some fingers coming out from the city centers, trying to maintain the basic district alignment (as opposed to completely eliminating a district in one area, and creating a replacement elsewhere).

I started by trying to get enough population for a whole number of districts.  Boston is just short of 6 districts.  Most of the towns around Boston have significant population which makes it difficult to find combinations that would work together.  Cambridge and the other three Suffolk County towns (Chelsea, Revere, and Winthrop) are almost perfect for a district each.  It would be nice to use Brookline since Boston surrounds about 80% of it, but it doesn't get anywhere near the target population.  So that is how Boston ended up being matches up with Somerville and Everett.  The connection from Boston is OK since Somerville is adjacent to Charlestown and gives some more weight to Charlestown which is rather small.  I don't think you can actually drive from Somerville to Everett so those two are somewhat iffy.  And to avoid separating those two, the part of Somerville placed with Boston is a long finger along the Cambridge line.

Starting with East Boston and Charlestown, the only adjacent areas are in the central area including North End, Beacon Hill, etc.) to make up enough population.    Allston/Brighton is about 3/4 of a district and by the time you include Fenway about the only choice is to include Back Bay.  If you head south you're just wandering around to pick up enough population with no focus.

I had thought the Black population was generally moving southwest, perhaps because I mentally connect West Roxbury and Roxbury.  But the growth is more to the south into Mattapan and parts of Dorchester.

I was surprised that there were two Hispanic majority districts in Boston.  One in East Boston and the other in Roxbury.  Are these mostly Puerto Rican?  Or Dominican or something else.  Are Azoreans counted as Hispanic, and are there signficant concentrations in Boston, moving up from New Bedford, Fall River, and Providence?

So I was thinking that I had drawn two White districts in from South Boston through Mattapan, and a Black district and a mixed district from Roxbury to West Roxbury and Hyde Park.

I had played around a bit with putting South Boston and the South End together.  Note that #8 on your map is South Boston not East Boston, though it does include Logan Airport.  This is a weird precinct that also includes the Old Harbor between South Boston and Quincy, and likely the islands in the Harbor.  I'm not sure where the people actually live.

If I had included South Boston with the South End then I could also have included the more northern and eastern parts of Dorchester.  And then the Roxbury district could have gone somewhat more  southward.  This would have rotated the Forest Green, Brown, and Slate districts about 1/3 or a district counter-clockwise.

[jimrtex]

That will be an interesting quandary in MA if splitting the islands keeps a town intact.

According to my atlas, there is not year-round ferry service from Martha's Vineyard to Nantucket, nor from Falmouth to Martha's Vineyard.

Instead year-round service is from Yarmouth (Hyannis Port) to Nantucket; and from New Bedford to Martha's Vineyard.  So my districts actually comply with the transportation links.

[jimrtex]

Who does this map pair? It looks like Halvorson and Biggert, would that be accurate?

Where do Quigley and Gutierrez fall in the two earmuff successor districts?

I did some research and saw that Melissa Bean now has a Cook County district and Bill Foster is in the district to the west.

I think the northern earmuff is the Puerto Rican district, and the red district by the stockyards is the Mexican district, so I'd guess Gutierrez has the northern district.   The question then does the Mexican district have enough US adult citizens to be an effective minority majority district.

[jimrtex]

My take on CO tried to minimize county fragments while bring population deviations down to 100 or less. Presumably, adjustments at the block level could get perfect equality without significant change to the map. While I only split 4 counties while maintaining some compactness, I also note that the result is remarkably similar to the current map. The major difference is that Aurora and Arapahoe County move to CD 7 from 6 and 1, and Lakewood moves into CD 6 from 1.

This is a good map.  I'd drop the split of Lake County, you can't get from Lake to Pitkin in the winter.  Go ahead and take a few 1000 from southern El Paso instead.

[jimrtex]

This is a good map.  I'd drop the split of Lake County, you can't get from Lake to Pitkin in the winter.  Go ahead and take a few 1000 from southern El Paso instead.

I've driven Independence Pass (12,093 ft) between Piktin and Lake so I know what you are saying about the connection. Actually, my preference was to split Custer, but the voting districts didn't provide fine control to get to my target of a 100 person deviation. With block-level projections, that's probably where I'd go.

My quality measure for a county split would be the relative size of the smaller fragment to the population of the county.  I'd still go for a couple thousand from El Paso rather than splitting either Lake or Custer counties.

The election officials in larger counties are going to be able to handle multiple legislative districts, and probably have an integrated GIS system to draw updated precincts.  A smaller county might have hand drawn maps, and the only reason they have precincts is to match up with the comissioner districts.

[jimrtex]

I know we were asked not to post every swing state diary map here when people can see them there on their own, but this "New York 28-0" takes the cake for treating redistricting as an abstraction. I'm actually offended that someone submitted it.

http://www.swingstateproject.com/diary/6166/#108516

That is a work of art.  Major kudos to the creator for putting parts of Buffalo into Jerry Nadler's district.

What is up with that weird Rochester district though?

He also didn't take advantage of the Queens-Richmond and Westchester-Nassau contiguity.

BTW, NY at one time had a Kings-Richmond-Rockland district.

[jimrtex]

or http://i49.tinypic.com/qyu33r.jpg

Total Population: 720, 584 (-1412 from ideal w/ new population estimates)
White Population: 334, 050 (46%)
Black Population: 161, 746 (22%)
Native Population: 1, 920 (0%)
Asian Population: 68, 915 (10%)
Hispanic Population: 140, 107 (19%)
Other Population: 13, 846 (2%)

Combines all of Cambridge, Somerville, Everett, Chelsea with parts Boston.

The three parts of Boston that are included: Brighton, East Boston, and Roxbury and points south are only connected through the other towns, and it appears that you have split the rest of Boston into 7 separate areas: Back Bay, West Roxbury-Jamaica Plain, 3 areas along the southeastern border, South Boston-Downtown-North End-Charlestown, and part of East Boston near the airport.

So can you slide your district SE to match the border, and then link to East Boston through the harbor.   This will give you two segments for the rest of Boston, and you can simply treat Brookline as an inclusion.  IIRC, the area between Quincy and South Boston just to the east of the main Boston part of your district is actually in Dorchester Harbor and actually iincludes the harbor islands and is part of the East Boston ward (I think the people actually live on the island south of Winthrop).

[jimrtex]

If the courts draw the lines in Mass, because in 2010 the Pubbies get more than one third of the seats in one of the houses of the legislature, and a GOP governor is elected, causing a deadlock, what would a non partisan map look like that comports with the law, and would that cause any seat to have a GOP lean, or close to it?  Just curious.

This is based on 2007 town estimates, but 2010 should not be too much different - though the key is which towns fit best with Boston.  On the 2007 estimates, Quincy and Milton were almost a perfect match, but in another year it could be towns on the other side of Boston which could mean shifting about 100,000 people around as the suburban districts rotate.

The federal court (in my better world) will rule that it should not be picking winners and losers as Massachusetts loses a representative, and so incumbents are totally ignored.  Three districts outside the immediate Boston area are created, and the districts are compact and equipopulous.

The Republicans would be hopeful about the light blue district in the western suburbs, and fantasize about winning the yellow western district (Worcester) and the southern and the other three outer suburban districts.

[jimrtex]

What does including those precincts from Dorchester do to the demographics?

I'd guess that it is about 20,000 population total?  So even if it were all white it would be around a 3% shift.  But the direction of growth in the black population is to the south and southeast.  The black population of Quincy has about doubled between the 2000 census and the 2006-2008 ACS (still under 10%) but indicative of what is happening on the other side of the line.

[jimrtex]

I can now apply the election data from the 2006 Governor's race, using just the R and D numbers, without the other two candidates. Both MA-4 and MA-5 in the map above could be interesting to the GOP. The 2006 two-party split in MA-4 favors the Dems 52.4 - 47.6, and in MA-5 it would favor the Dems 52.8 - 47.2.

Do you think you can calculate how many districts in this map Scott Brown wins?

Maybe 5.

NYT election results

Coakley would win the 3 inner districts around Boston (6, 7, and 9), and 1.

You might be able to get it up to 6, if you split 7&9 more on north/south basis and really pack the eastern district, and then split CD2 and 5 into 3 districts and crack the remnant of 7&9.

You might be able to bring CD4 north, to get Quincy and Brockton, and shift CD3 eastward.  The northern part of CD3 could then be used to bulk up the new districts.  CD6 could probably be adjusted a bit to take in some suburbs in the eastern part of CD7.

In a more conventional election this might go back9-0 Democrat.  

Some pretty big differential turnouts vs 2008 presidential election:

Metheun 67%, Lawrence 50%

Westfield 71%, Springfield 55%.

Boston 67%, Dedham 84%, Chelsea 52%.

New Bedford 59%, Fall River 58%, Dartmouth 82%.

Norfolk 88%, Wrentham 87%, Walpole 83% (only 3 towns in Norfolk carried by McCain).

[jimrtex]

How high of an Asian percentage can you get in NYC, assuming you can use the East River to connect areas in Manhattan, Brooklyn, and Queens.

[jimrtex]

These Asian-majority districts become interesting as well if CA passes the initiative to include congressional redistricting with their commission. The commission will probably be swayed by linking communities of interest, and I would not be surprised to see a pan-Asian COI promoted.

The law requires that the legislature apply the same standards when drawing congressional districts, as the commission does when drawing legislative lines, except the legislature is not restricted from considering the residence of incumbents or political candidates, and may draw districts for the purpose of favoring or discriminating against an incumbent, political candidate, or political party.

A geography-based "community of interest" would seem to be pretty ill-defined, so I would stick to cities and counties since they do have a clear definition, and then consider neighborhoods and maybe CDP (California may has alternatives defined for planning purposes) when the others had to be divided to reach population relative equality.

Incidentally, there are now 12,000+ applicants to the commission with two weeks to go.  About 10,000 appear to qualify based on legal eligibility.

The next step is for the auditors to reduce that to 60 persons (20 Democrats, 20 Republicans, and 20 other) "on the basis of relevant analytical skills, ability to be impartial, and appreciation for California’s diverse demographics and geography." 

The Phase I applicants will now be given a supplemental application which will include

You must be logged in to read this quote.

Based on the supplemental applications, 120 will be brought to Sacramento for personal interviews, from which 60 will be chosen.

It will be interesting to see how many persons follow through with the more extensive application.

I would be concerned that commission would be too narrowly qualified, as opposed to the sort of variety you would get if a more random process, such as that used for jury selection were used.

An interesting demographic about current applicants is that the median age appears to be about 57, which is older than I would expect - but perhaps younger persons were simply more realistic about the potential job interruption if they were selected.

[jimrtex]

Here's an attempt at Iowa:



Not sure what algorithm they use exactly, but it's pretty nicely drawn I think and would fit the general criteria. Interestingly this map could easily elect an all Dem delegation.

It might not qualify.  Iowa actually demands an extreme level of population equality.  A lot more than you would intuitively expect to be possible with whole counties.  In the 2000 redistricting the legislature rejected the first plan saying it had too great population equality, and it was way below 1000 deviation.

Iowa also uses a somewhat odd definition of compactness, comparing the east-west extent of the district with its north-south extent.   You can see this in the western district.   Visually, it looks to be basically a parallelogram that is somewhat narrower east-west than north-south.  But the east-west extent is measured from the westmost point in the NW corner to the eastmost point in the SE corner.  The district get some bonus width simply due to the fact that Council Bluffs is east of Sioux City.

It also means that Iowa tends to get districts that are somewhat L-shaped and toothy.  Think of a 4 x 4 block of counties that is almost square, and place a 3x3 block inside one corner of the district, so that you now have an L-shaped district along two edges of the 4 x 4.  Iowa thinks that the L-shaped district is as compact as the original square.   You can also move counties back and forth along the interior boundary, without any effect on the compactness.

If you look at a current map, you can see this effect, especially with 1, 2, and 4.   For 4 to follow about 2/3 of the Iowa-Minnesota border also requires it to drop south of Des Moines.

I suspect that a compact 4-district map would maintain the western district and then have districts in NE and SE corners that have some additional north-south extent by overlapping.  The 4th district will include Polk and an almost arbitrary set of counties.  If you add a county to the east, you have to also add one to the north or south.

[jimrtex]

I gerrymandered Maryland to a 5-4 GOP advantage, and a 6-3 GOP in a good year.

Maryland has only 8 U.S. House seats now, and I haven't heard that it might gain a seat.  Is it within the realm of the reasonably possible?

Only if DC is retroceded.   So reasonable yes.  Possible no.