Dave's Redistricting App

[jimrtex]

This wouldn't even be a horrible looking map if gray and green weren't so stringy.

Use microthreads to connect areas.  They could be a millimeter wide and run along the edges of precincts.  You can define them as having no persons living in them.

[jimrtex]

This app just made me realize that NC-06 and NC-13 rely on a rather "unique" definition of continguous.

If they would just adopt my idea of microthreads they would not need to resort to that sort of deception.

[jimrtex]

Not really as even those would have to cross each other. The current setup is an obvious attempt to keep a chunk of Republican exurbia out of Brad Miller's district. And NC-12's long stringy shape makes things a bit more difficult. NC Democrats kind of drew themselves into a corner really though I'm sure they'll find a way to preserve all the Democratic-held seats when redistricting comes around. But it's kind of tricky.

Not really.  The map on the upper left is equivalent to the case in North Carolina.  The map on the upper right is the same, but the two districts no longer have point contiguity.  The connecting sections can be made extremely narrow.

The bottom map is a bit more complex with 4 discontiguous sections in the green section.  But they can be connected.  The map in the lower right is topologically the same as the 3 concentric circles.  I think it is similar to making balloon animals.

[jimrtex]

Nice maps sbane. Too bad if the redistricting initiative passes (I can't imagine that it will not), it all will be totally illegal.   Plus, even if it does not pass, Governor Whitman would veto it.

The initiative that created the redistricting commission for legislative districts, also established the same parameters for congressional districts (respecting political boundaries, not bypassing communities, etc.), while leaving the legislature as the body to perform the redistricting.   The only difference is that the legislature is permitted to make political considerations and take into account the residence of incumbents and potential candidates in drawing congressional districts.

The Legislature is also obligated to "provide access to redistricting data and software, and otherwise ensure full public participation in the redistricting process."  Failure to provide internet access to the data and software precludes full public participation.

So even if the initiative to entrust congressional redistricting to the redistricting commission fails; and Governor Brown Jr. signs the Sbane plan, the California Supreme Court will overturn it on grounds that it violates the California Constitution.  The special master that draws a legal plan for the court will use the members of the redistricting commission to draw the congressional district plan and direct that they not take into account the residence of incumbents and potential candidates, or make other political considerations.

[jimrtex]

Nice maps sbane. Too bad if the redistricting initiative passes (I can't imagine that it will not), it all will be totally illegal.   Plus, even if it does not pass, Governor Whitman would veto it.

Hmm, I don't see what's so wrong with my gerrymandering at all, besides not being visually pleasing enough for Xahar. I mean a district that includes both Long Beach and Compton makes a lot of sense doesn't it? Of course it also contains Seal Beach and Cypress. Or the district that joins Sunny hills in Fullerton to Huntington park. There's a lot of epic gerrymandering in there. I'm particularly fond of the central valley districts. I didn't know 3 Hispanic districts could be created there.

California Constitution requires that congressional districts comply with the following requirements:

(b) The population of all congressional districts shall be reasonably equal.
(2) Districts shall comply with the federal Voting Rights Act (42 U.S.C. Sec. 1971 and following).
(3) Districts shall be geographically contiguous.
(4) The geographic integrity of any city, county, city and county, neighborhood, or community of interest shall be respected to the extent possible without violating the requirements of any of the preceding subdivisions. Communities of interest shall not include relationships with political parties, incumbents, or political candidates.
(5) To the extent practicable, and where this does not conflict with the criteria above, districts shall be drawn to encourage geographical compactness such that nearby areas of population are not bypassed for more distant population.

I doubt that your plan complies with (4) and (5).  I doubt that you can show that your plan was necessary in order to ensure contiguity.  So you are stuck with showing that your intent was to comply with the VRA, and it was not practicable to do so without violating (4) and (5).

[jimrtex]

I think that a GOP gerrymander would play it safe and keep OH-11 majority Black. The best way to do that is to run it down to Akron. However, that would keep OH-10 on the west side of Cleveland.

In 2000, it was possible to include all of Cleveland in a majority-black district, though the included suburbs had to be very selective.  In 2010, more of the eastern and southeastern suburbs will be majority black, and presumably the black share of Cleveland itself will be greater (based on the 2006-2008 ACS) the district is now 59% black vs. 56% at the time of the 2000 census, this is because whites have been leaving at a greater rate than blacks (the district has lost 80,000 since the census and the loss of 2 seats is going to require a huge pick up of population.

The black population in Akron is concentrated in the SW part of the city, and Akron itself is only 30% black, so you would need a long isthmus running through the northern part of Summit County and into Akron, which also results in eastern and western Summit County being split, or having to wrap around CD-11.

I don't think you need to string cities together to pick up enough black residents if you can create a compact district in Cuyahoga county that is very high plurality black.

[jimrtex]

I think that a GOP gerrymander would play it safe and keep OH-11 majority Black. The best way to do that is to run it down to Akron. However, that would keep OH-10 on the west side of Cleveland.

In 2000, it was possible to include all of Cleveland in a majority-black district, though the included suburbs had to be very selective.  In 2010, more of the eastern and southeastern suburbs will be majority black, and presumably the black share of Cleveland itself will be greater (based on the 2006-2008 ACS) the district is now 59% black vs. 56% at the time of the 2000 census, this is because whites have been leaving at a greater rate than blacks (the district has lost 80,000 since the census and the loss of 2 seats is going to require a huge pick up of population.

The black population in Akron is concentrated in the SW part of the city, and Akron itself is only 30% black, so you would need a long isthmus running through the northern part of Summit County and into Akron, which also results in eastern and western Summit County being split, or having to wrap around CD-11.

I don't think you need to string cities together to pick up enough black residents if you can create a compact district in Cuyahoga county that is very high plurality black.

The Bartlett decision might imply otherwise. If there has been a pattern of racial bloc voting in NE OH, and there is a potential election district with 50% plus black voting age population, then there could be a valid section 2 claim against a map that did not create a majority-black district. The decision rejected plurality districts that relied on crossover white votes or on a multiple-minority coalition.

According to the ACS 2006-8 CD-11 is now about 59% black, because whites are leaving faster than blacks (its population is now 550,000, down from 630,000)

The majority-black census tracts in Akron have 35,000 blacks in a total population of 50,000 (about 70%),.  Add in enough census tracts to make the area contiguous (most of the black population is in SW Akron, but there are also areas in the SE and near north side), plus provide a connection to Cuyahoga County, and you might be able to get 45,000 black out of 75,000 from Summit County.

Assume 720,000 for a CD.  So a majority black district would need 360,000.  If a Summit portion were 45/75 black, the Cuyahoga portion would be 315/645 black (48.8%).

Let's assume that we instead added in areas in Cuyahoga county that were only 30% black.  That would end up being 46.9% overall.  So your choice is between a totally ungainly district that runs 20 miles along the Cuyahoga River for strictly race reasons to get up to 50%, vs. a compact district in Cuyahoga County that might be 47% black.

[jimrtex]

The one I was speaking of, before SCOTUS tossed it out I think in 2001, makes what you put up a perfect circle.

North Carolina 1990s redistricting plans.

Maps of CD 6 and CD 12 from the 1992 redistricting are particularly interesting.

Also note that the 1998 plan was used in 1998, and the 1997 plan in 2000.

[jimrtex]

The map below uses the Cleveland-Akron connection to make a new CD-11 that is 56% black (402 K / 718 K). I don't think the shape is any worse than many acceptable VRA districts.

 

How is the population projected?

Haven't you split Akron among 4 districts?

[jimrtex]

The map below uses the Cleveland-Akron connection to make a new CD-11 that is 56% black (402 K / 718 K). I don't think the shape is any worse than many acceptable VRA districts.

 

How is the population projected?

As I understand it, Dave's Red. App uses the 2008 estimates for the county, including the estimates for each racial and language group tracked by the Census. The population changes for the county in the estimate are applied within each group uniformly throughout the county. Variations in population growth within a county are not part of his app. Similarly, new communities in transition this decade between racial groups are not captured by his approach.

I think that is where it is going to be tough.  Places like Cleveland, Cleveland Heights, and East Cleveland are losing population even faster than the county as a whole.

The areas of greatest growth, or at least slowest decline, are the outer tier of townships on the south, west, and east.  North Royallton and Olmsted Falls are the big winners with about 2.5% growth each.  While you might find blacks moving into areas like Solon, you might be picking up cities that are 20% black.

In Summit County, you actually have growth areas like Richland, but Akron is declining.  And I'm sure you are going to find the black population more dispersed.

[jimrtex]

The map below uses the Cleveland-Akron connection to make a new CD-11 that is 56% black (402 K / 718 K). I don't think the shape is any worse than many acceptable VRA districts.

 

How is the population projected?

As I understand it, Dave's Red. App uses the 2008 estimates for the county, including the estimates for each racial and language group tracked by the Census. The population changes for the county in the estimate are applied within each group uniformly throughout the county. Variations in population growth within a county are not part of his app. Similarly, new communities in transition this decade between racial groups are not captured by his approach.

I think that is where it is going to be tough.  Places like Cleveland, Cleveland Heights, and East Cleveland are losing population even faster than the county as a whole.

The areas of greatest growth, or at least slowest decline, are the outer tier of townships on the south, west, and east.  North Royallton and Olmsted Falls are the big winners with about 2.5% growth each.  While you might find blacks moving into areas like Solon, you might be picking up cities that are 20% black.

In Summit County, you actually have growth areas like Richland, but Akron is declining.  And I'm sure you are going to find the black population more dispersed.

At 56% I think the district I drew will hold up even with accurate numbers. I'm anticipating better tools soon. That should give us all a better look at the possibilities.

If you can get 56% you are capturing 90% of blacks in the two-county area, and your proposed district will be short of 720,000 because it is based on countywide estimates.  To get it back up, you are going to have to include adjacent areas that have a substantial black population, but nowhere near a majority.

In Summit County, the countywide estimate is for no-change, but Akron is declining.  In Cuyahoga County, the 2008 estimate is for an 8.7% decline, but there are heavier declines in Cleveland, East Cleveland, Cleveland Heights, etc.

So you are going to end up closer to 50% than 60% and you are connecting two cities 20 miles apart with a national park and splitting both cities.

In 2000, a district with 720.000 persons and a bare majority black population could be drawn in Cuyahoga County using whole cities, including Cleveland.  It will have lost 10% of its population, but will be a bit blacker.  Adding in 70,000 in the eastern part of the county may drop the percentage just below 50%.

So can you rationalize a non-compact district that may end up just a couple of percentage points blacker, than a single-county district made up of whole towns?

[jimrtex]

The rationalization would be based on avoiding a potential VRA challenge. The premise of the map is that the GOP is in charge of the pen and wants to maximize seats. Presumably the Dems would want to challenge the map, and a successful VRA challenge could throw their entire map away. Based on the Bartlett decision, I think that this map survives, but a GOP map that has only a plurality black district when a majority is possible would lose.

I agree that the population is probably less than indicated, but the map only needs 50% VAP to meet the goal I set for the map. I think that with 56% in this estimate, it is likely that 50% will be achievable. Actually, being able to expand into additional D areas of Akron to add population to CD-11 would help the GOP improve the PVIs of neighboring districts.

If the GOP is in charge of the pen, they will want to get rid of Kucinich.  They can put all of Cleveland in the black district, and put the western suburbs in a district extending into Lorain county, and the southern suburbs in a district extending into Medina, and the far eastern suburbs into a district extending into Geuga.

They argue that the Cleveland-Akron connection traverses a near wilderness, and effectively creates a Shaw I district solely in the pursuit of isolating blacks in a racially gerrymandered district, as traditional redistricting criteria such as counties and towns are ignored.

[jimrtex]

Looks like it's not possible draw a McCain State House seat in Philly after all.

You give up too easily. This district is 52% McCain.

You have 3 enclaves, two along the western border, and one in the northeast that are cutoff, and the area that you wrap around might not have enough population for a district of its own.

[jimrtex]

You have 3 enclaves, two along the western border, and one in the northeast that are cutoff, and the area that you wrap around might not have enough population for a district of its own.

Always a complainer. Perhaps this is more to your liking then? Only 51% McCain, but no enclaves.

What is the population of the center area?  Maybe you can create a donut?

[jimrtex]

What is the population of the center area?  Maybe you can create a donut?

I was curious about that as well, so I tested it. The center area would form most of a district, but there would have to be a little bit on the outside. Some of the precincts are rather inconveniently shaped; I can't get population equality for the center district without splitting a precinct or forcing a neighboring district to go outside of Philly. Even if the outer district remains entirely within Philly, it would have to be extended along the northern and western edges of the McCain district and include a small area to the south as well.

I just did it quickly, so could have missed one of them, but the population seemed short by about 6,000.

Using estimates, a 203-seat plan, and keeping districts entirely in Philadelphia, this is what Northeast Philly would probably look like with the McCain district. The largest deviation from the ideal population is -68, in the green district.



I'm not sure that the green district is really contiguous, but if you zoom in close enough, it looks like it is. If it isn't, the district will have to go into Montgomery County, but existing districts in the area also go into MontCo, including a district that is somewhat similar to the green district.

It is contiguous.  The beak of the green VTD actually extends about a block south of the southern border of the blue VTD.  So while the southern tip of the green VTD is on city limits, there is an east-west connection of about a block between the two green VTDs.

Tomlinson Road is the southern boundary of the blue VTD and takes a jog before entering Montgomery County.  It is the jog that forms the border between the green VTDs.

[jimrtex]

Here is a map of California using the redistricting commission's guidelines with 53 districts.

Will more respect for county lines be required?  You have 4 districts crossing the Los Angeles San Bernadino county line.

[jimrtex]

How does one know how Hispanics actually voted?  Exit polls?  Or is the game just one involving what surname seems to get elected? Has SCOTUS ruled on this notion, that if 50% +1 Hispanic VAP seems to still elect Anglos, and in particular Anglo Republicans, that further inquiry is needed?

You take election results and correlate precinct results with the portion of the voters who are Hispanic-surnamed.  Depending on what you are trying to prove, you choose your expert witnesses carefully, and they choose the elections that prove the point you want to prove.  You roll in some 10-year old census data, and make adjustments for Hispanics who are not Hispanic surnamed, and non-Hispanics who have a Hispanic surname.

[jimrtex]

Tom DeLay already tried that loophole with heavily Republican districts with non-voting Hispanics, and it got shot down by a federal court, hence why Ciro Rodriguez is back in Congress. Safe to say the Obama DOJ won't be more lenient.

Really it would take a rather vile gerrymander not to draw at least one new district in South Texas that is supermajority Hispanic.

You can only do that if you draw another fajita strip district.

[jimrtex]

The new seat I drew was just Cameron County plus a strip into Hidalgo County.

Which means you cut Nueces County loose which will need counties to the NE and will be a chance for a Republican pickup, since you will still need counties like Willacy, Kleberg, and Jim Wells for the border districts.

[jimrtex]

Here is a map of California using the redistricting commission's guidelines with 53 districts.

Will more respect for county lines be required?  You have 4 districts crossing the Los Angeles San Bernadino county line.

CD 26 is all along I-210 and makes sense as a district, notwithstanding the different counties it is in. It's really just one metro area, the county lines don't matter much. CD 25 connects high desert semi-exurban communities that have commonalities. CD 38 joins West Covina/Pomona with Ontario/Chino in San Bernardino county creating a majority Hispanic district. These areas are more inland and could be considered a community of interest. CD 40 is a district that takes in more Anglo/Asian areas in LA county like Diamond Bar and Walnut as well as wealthy Chino Hills in San Bernardino county. This is joined with more Anglo/Asian areas like Fullerton and Buena park in OC. These areas are all close by and are quite similar to each other.

You have explained why you drew the districts that you did.  You did not explain why your districts complied with the California Constitution.

[jimrtex]

Nope. The 27th kept Nueces County and is still 64% Hispanic.

LULAC v Perry specifically rejected the McAllen to Austin district since it was merely connecting disparate areas of the state based on skin color.  Some of the remedies suggested drawing districts that went from McAllen to almost Austin, but were rejected by the district court, which implied that Laredo-San Antonio and McAllen-San Antonio districts were suspect.  In the past, these could be justified on the basis of insufficient population along the border.    But in 2011, you'll be able to draw 4 districts along the border.   (1) El Paso; (2) Laredo-Mid Valley-Trans Pecos;  Cameron-Hidalgo;  Hidalgo.  Include the Brush Country as necessary.

[jimrtex]

As long as there is a good reason to cross a county line, it should be OK.

You must be logged in to read this quote.

The previous subdivisions deal with population equality, VRA compliance, and contiguity.  It may be necessary to span county boundaries for purposes of population equality or VRA, but in general it should be possible to do this without more than one district spanning a county boundary. 

The findings section of Proposition 11 explicitly noted the splitting of cities, and I think that a reasonable interpretation of legislative intent was to avoid splitting counties and cities.  "counties shall be respected to the extent possible" should be read as "counties shall be disrespected only to the extent necessary to comply with population equality and the VRA"

The proposition that would put congressional redistricting under the redistricting commission adds language to that section:

"The geographic integrity of any city, county, city and county, local neighborhood, or local community of interest shall be respected in a manner that minimizes their divsion to the extent possible without violating the requirements of any of the preceding subdivisions."

It also further refines the definition of "local community of interest".  Perhaps you could argue that districts spanning county boundaries were necessary to avoid splitting a "community of interest".  But unless, the extent of these local community of interests were defined in advance, and consistently respected it gives the appearance of mere rationalization.  BTW, defining these "local community of interest" will give the commission something to do while waiting for census data.

[jimrtex]

New Hampshire? Why?

There are 10 townships in New Hampshire that are covered, including one that has no population.

The counties in New York and California, and the townships in New Hampshire were caught under a provision that if there was a literacy test and less than 50% of the CVAP voted in the 1968 presidential election they were covered.

New Hampshire's literacy test had not been enforced, but was still on the books.  Since elections in New Hampshire are based on towns, the percentage test was also applied to each town.

The 4 counties in California were all tied to military bases (some of which have been BRAC'ed.  Military were counted in the population, though relatively few voted locally.  In one county, the census bureau may have miscalculated the CVAP, by projecting all adult population growth from a census to have been citizens, and the county barely missed 50%.

In California, the state Supreme Court had already declared the literacy test to be unconstitutional, and it was officially repealed a month after the specified trigger date.

The two townships in Michigan were caught under the language provisions, both due to a large Spanish-speaking population coupled with low voting participation.

[jimrtex]

The New Hampshire townships have a total population of about 16,000.

The two in Michigan includes one that's about 30% Hispanic (kind of odd, it's in the middle of nowhere and the county population is only about 5% Hispanic), the other is just outside of Saginaw and is less than 10% Hispanic but majority black.

The one near Saginaw is kind of odd, because it is covered under the language provisions, but is majority black, and people think it is covered because of the black population.  In recent elections, participation by blacks was higher than among whites.

Clyde Township I assume must be migrant farm workers of some sort.  It is about half way between Muskegon and Benton Harbor on Lake Michigan and well away from Grand Rapids and Kalamazoo.  The census shows an extremely large share of the Hispanics were born in Mexico.

[jimrtex]

New Hampshire? Why?

New Hampshire hired a redistricting consultant who wouldn't believe that parts of New Hampshire were covered by the VRA.