Partisan Population Distribution

[Sol]

A discussion in the 1,000 district series thread prompted this:

What states population distribution naturally favor one party over another? What states are "naturally gerrymandered"?

It seems like most states favor the Republicans, due to the Rural-Urban divide. Maryland comes to mind as an extremely democratic state that nonetheless favors the GOP in redistricting.

New England States seem to favor the Dems. In Massachusetts, drawing a McCain district is barely possible.

North Carolina and Colorado seem to be roughly neutral.

What are your thoughts on this?

[traininthedistance]

A discussion in the 1,000 district series thread prompted this:

What states population distribution naturally favor one party over another? What states are "naturally gerrymandered"?

It seems like most states favor the Republicans, due to the Rural-Urban divide. Maryland comes to mind as an extremely democratic state that nonetheless favors the GOP in redistricting.

New England States seem to favor the Dems. In Massachusetts, drawing a McCain district is barely possible.

North Carolina and Colorado seem to be roughly neutral.

What are your thoughts on this?

Maryland is not actually the worst, at least with CDs as large as they currently are.  The horrible lines are more a matter of accommodating the residences of Ruppersburger and Sarbanes than anything else.  A clean, sensible 6-2 map is trivially easy there.

The absolute worst states for this are in the Midwest, I think.  I'd say that relative to its overall partisan lean, Illinois is probably the single most skewed state in the nation.  A state that has returned 10-plus point Democratic victories in the past six Presidential elections should not naturally lend itself to maps that are liable to elect more Republicans than Democrats.  Ohio and Michigan are also particularly bad. 

[krazen1211]

The states that are best naturally gerrymandered are generally those where the Democratic party is most reliant on blacks in order to achieve 50%.

[traininthedistance]

The states that are best naturally gerrymandered are generally those where the Democratic party is most reliant on blacks in order to achieve 50%.

Wrong as usual.

North Carolina is relatively good for Dems, and they most certainly rely on black voters there.

Replace "black" with "urban voters" (not the same thing by any stretch) and your statement would be more accurate- with the caveat that packing Dems in urban centers is sometimes not that bad in heavily-Republican states, letting them get a couple seats they might not have with a more even distribution.  The best example of this is probably Utah- a fair map would give the Dems one seat in Salt Lake City and environs, despite the state being so strongly Republican in general.

[Miles]

NE and KS are pretty natural Democratic gerrymanders as the intensity of the Republican's support lessens as you go eastward.

[krazen1211]

The states that are best naturally gerrymandered are generally those where the Democratic party is most reliant on blacks in order to achieve 50%.

Wrong as usual.

North Carolina is relatively good for Dems, and they most certainly rely on black voters there.

Replace "black" with "urban voters" (not the same thing by any stretch) and your statement would be more accurate- with the caveat that packing Dems in urban centers is sometimes not that bad in heavily-Republican states, letting them get a couple seats they might not have with a more even distribution.  The best example of this is probably Utah- a fair map would give the Dems one seat in Salt Lake City and environs, despite the state being so strongly Republican in general.

North Carolina is 'relatively' good for the Democrats for Congressional districting if you make convenient choices that favor the Democratic party.

It is of course a not 'relatively good' for Senate/House legislative seats where excellent performance was achieved by the Republicans despite barriers to vicious gerrymandering. As proof one can merely look at the delegations from Mecklenburg County.

[krazen1211]

It seems like most states favor the Republicans, due to the Rural-Urban divide. Maryland comes to mind as an extremely democratic state that nonetheless favors the GOP in redistricting.

Texas has a huge Urban population but a poor layout for the GOP. Reason being that the urban Hispanic districts, despite being effective % packs, pack in fewer actual votes and leave them with heavily overpacked suburban districts and counties.

[Miles]

Lets be nice...

By your own admission, its hard to draw a clean map that favors the Rs; I happen to agree. Most clean maps I've seen of NC favor Democrats.

As someone who's lived in Mecklenburg, I don't really see your point. Yeah, the Republicans have a few swingy Assembly districts, but the downballot fundamentals of those districts are more Republican than you'd think.

[krazen1211]

Lets be nice...

By your own admission, its hard to draw a clean map that favors the Rs; I happen to agree. Most clean maps I've seen of NC favor Democrats.

As someone who's lived in Mecklenburg, I don't really see your point. Yeah, the Republicans have a few swingy Assembly districts, but the downballot fundamentals of those districts are more Republican than you'd think.

Yes, as I said then, the issue comes with the spread of the population and that is due to the size of the districts.

Smaller districts neutralize that spread factor. If you do not like Mecklenburg (which is 6-6 in the assembly and 3-2 in the Senate), you can look at Wake County.

Even those clean congressional maps rely on over performance by local Democrats.

[Sol]

The states that are best naturally gerrymandered are generally those where the Democratic party is most reliant on blacks in order to achieve 50%.

Wrong as usual.

North Carolina is relatively good for Dems, and they most certainly rely on black voters there.

Replace "black" with "urban voters" (not the same thing by any stretch) and your statement would be more accurate- with the caveat that packing Dems in urban centers is sometimes not that bad in heavily-Republican states, letting them get a couple seats they might not have with a more even distribution.  The best example of this is probably Utah- a fair map would give the Dems one seat in Salt Lake City and environs, despite the state being so strongly Republican in general.

North Carolina is 'relatively' good for the Democrats for Congressional districting if you make convenient choices that favor the Democratic party.

It is of course a not 'relatively good' for Senate/House legislative seats where excellent performance was achieved by the Republicans despite barriers to vicious gerrymandering. As proof one can merely look at the delegations from Mecklenburg County.

Try thinking next time.

The Republicans tend to overperform a decent amount in both Wake and Mecklenburg county. Wake however has a lot precincts that are 48% Democrat, 52% Obama, so there still is a decent chance for a Dem to win those places.

NC really does favor Democrats if you follow communities of interest, like a commission would. Some of this is due to incumbent strength- Shuler's seat would lean Republican in all but the most extreme Democratic gerrys. But all the others would all still lean towards the Democrats with MCIntyre's being a tossup.

Of course, a fair map would also do things that the Democrats wouldn't like, like shoring up Ellmers and Foxx.

[TJ in Oregon]

Wisconsin is naturally resistant to gerrymanders (at least on the level of congressional districts though it favors Republicans with smaller districts). With 8 seats, you basically end up with 2 safe seats for each party (Madison and Milwaukee for the Dems and the Milwaukee suburbs and the Sheboygan/Fon-du-Lac/Manitowoc area for the Republicans) and 4 swingy seats unless you do some extreme gerrymandering. Vast swaths of the state vote close enough that the rest are difficult to make safe, so you can have Obama carry 7 districts in 2008 and Republicans win 5 in 2010.

[traininthedistance]

Wisconsin is naturally resistant to gerrymanders (at least on the level of congressional districts though it favors Republicans with smaller districts). With 8 seats, you basically end up with 2 safe seats for each party (Madison and Milwaukee for the Dems and the Milwaukee suburbs and the Sheboygan/Fon-du-Lac/Manitowoc area for the Republicans) and 4 swingy seats unless you do some extreme gerrymandering. Vast swaths of the state vote close enough that the rest are difficult to make safe, so you can have Obama carry 7 districts in 2008 and Republicans win 5 in 2010.

But since Wisconsin is Lean D overall (it hasn't voted Republican for President since 1998), 2D/2R/4 Tossup is naturally gerrymandered for the Republicans, albeit not quite as badly as some others.

[they don't love you like i love you]

The worst state for the Democrats is probably Indiana. Even with the 2008 numbers, it's hard to draw a State Senate that'd be under 60% R even if we assume every Obama won seat went to a Democrat. The 1000 districts map I drew for it was even worse, the Democrats are too spread too thinly in the state outside of NW Indiana and Indianapolis. The only area where the Democrats have a lot of voters outside those areas where they can be properly put to use is South Bend and maybe the Muncie/Anderson area (with smaller districts like the legislature, you can get a few more safe seats, like one in Bloomington obviously and probably one in Terre Haute, but the rest are easy to cancel out.)

For all the talk about how bad Wisconsin is, compare Wisconsin 2004 to Indiana 2008, both razor thin Dem victories.

[Indy Texas]

It seems like most states favor the Republicans, due to the Rural-Urban divide. Maryland comes to mind as an extremely democratic state that nonetheless favors the GOP in redistricting.

Texas has a huge Urban population but a poor layout for the GOP. Reason being that the urban Hispanic districts, despite being effective % packs, pack in fewer actual votes and leave them with heavily overpacked suburban districts and counties.

Districts are drawn by population, not voter turnout. Whether voters in one district turnout more than voters in another one is up to the voters.

Anyway, as far as statewide races go, the Texas GOP should be sending rural Hispanics thank you cards in the mail for opting to stay at home.

[jimrtex]

It seems like most states favor the Republicans, due to the Rural-Urban divide. Maryland comes to mind as an extremely democratic state that nonetheless favors the GOP in redistricting.

Texas has a huge Urban population but a poor layout for the GOP. Reason being that the urban Hispanic districts, despite being effective % packs, pack in fewer actual votes and leave them with heavily overpacked suburban districts and counties.

Districts are drawn by population, not voter turnout. Whether voters in one district turnout more than voters in another one is up to the voters.

Whether persons in one district are eligible to vote depends on their age and citizenship status.  Whether eligible voters vote depends in part on their age and ethnicity and income.

[Benj]

It seems like most states favor the Republicans, due to the Rural-Urban divide. Maryland comes to mind as an extremely democratic state that nonetheless favors the GOP in redistricting.

Texas has a huge Urban population but a poor layout for the GOP. Reason being that the urban Hispanic districts, despite being effective % packs, pack in fewer actual votes and leave them with heavily overpacked suburban districts and counties.

Districts are drawn by population, not voter turnout. Whether voters in one district turnout more than voters in another one is up to the voters.

Whether persons in one district are eligible to vote depends on their age and citizenship status.  Whether eligible voters vote depends in part on their age and ethnicity and income.

Agree with this. While overall population is one way to distribute electoral districts, I don't think it has any kind of moral superiority to distribution of districts by voting-eligible population. (Distributing by voting-age population is problematic, though, because it's not at all clear why age should determine distribution when age is not the only factor in eligibility.)

Distributing by actual voters voting has serious problems with how to determine turnout.

[jimrtex]

Whether persons in one district are eligible to vote depends on their age and citizenship status.  Whether eligible voters vote depends in part on their age and ethnicity and income.

Agree with this. While overall population is one way to distribute electoral districts, I don't think it has any kind of moral superiority to distribution of districts by voting-eligible population. (Distributing by voting-age population is problematic, though, because it's not at all clear why age should determine distribution when age is not the only factor in eligibility.)

Distributing by actual voters voting has serious problems with how to determine turnout.

A State could require mandatory enrollment for all residents over the age of 18.  Non-citizens would be required to register, but would be given a distinctive ID.  Citizen felons who are not incarcerated would not be registered to vote, but would be given a standard ID.

Since registration would be tied to actual residence, there would be no need to depend on the census, and districts could be updated on an as-needed basis, as is done in Australia.

[Franknburger]

A State could require mandatory enrollment for all residents over the age of 18.  Non-citizens would be required to register, but would be given a distinctive ID.  Citizen felons who are not incarcerated would not be registered to vote, but would be given a standard ID.

Since registration would be tied to actual residence, there would be no need to depend on the census, and districts could be updated on an as-needed basis, as is done in Australia.

Or in Germany, for that matter. In Germany, there is mandatory requirement to announce any change of address (while doing that, they just fix a sticker with the new address on yout ID). Everything else (notification of tax authorities, vote register update, taking you off the records in your previous location of residence, etc.) is done automatically, so its actually quite easy and citizen friendly. One drive to the county office, filling the form, 2-3 minutes inside while they record your data,  print out and fix the sticker to your ID, and that's it. No fees, since registration update is mandatory.

[Brittain33]

It's better to ask which state DOESN'T have a natural bias toward Republicans because of the Democrats' concentration in urban areas.

You need states with heavily Republican areas and moderate areas. So Kansas, Nebraska, Texas, Utah... Then the New England states with sufficient districts (MA, CT) where the Republican vote is too dispersed to allow a district winnable with the current Southern oriented GOP.

[traininthedistance]

It's better to ask which state DOESN'T have a natural bias toward Republicans because of the Democrats' concentration in urban areas.

You need states with heavily Republican areas and moderate areas. So Kansas, Nebraska, Texas, Utah... Then the New England states with sufficient districts (MA, CT) where the Republican vote is too dispersed to allow a district winnable with the current Southern oriented GOP.

I don't know that Texas's population distribution benefits Dems at all per se. The usual "pack urban liberals" into a minimum of districts works just as well here, and Texas is so large that you'd expect a few such districts even given the state's overall lean. However, most feasible maps do possibly have a slight "artificial" pro-Dem bias for two reasons:

1) the heavily Dem areas have horrible turnout, relatively depressing the Dems' statewide results,

and

2) the VRA-mandated fajita strips (a maximally compact map would probably have two Democratic Valley districts rather than three)

[politicallefty]

I don't know that Texas's population distribution benefits Dems at all per se. The usual "pack urban liberals" into a minimum of districts works just as well here, and Texas is so large that you'd expect a few such districts even given the state's overall lean. However, most feasible maps do possibly have a slight "artificial" pro-Dem bias for two reasons:

1) the heavily Dem areas have horrible turnout, relatively depressing the Dems' statewide results,

and

2) the VRA-mandated fajita strips (a maximally compact map would probably have two Democratic Valley districts rather than three)

I think Texas is actually a state where the combination of the VRA and partisan gerrymandering works very heavily against Democrats. If you disregard race and political party, Texas could have a map with the parties near parity in terms of Congressional representation. I drew what I would consider to be a reasonable map that created 16/36 Obama districts, not to mention some McCain districts that could probably go Democratic with the right House candidate under the right circumstances. The Obama districts are four in the DFW Metro, four in the Houston area, two around Austin, two for San Antonio, three in South Texas, and the El Paso district. I think the weakest district was in Fort Worth at 55% Obama. Texas seems to have the opposite effect in that the rural areas are so extremely Republican that Democrats really should be overrepresented in a fair map.

[they don't love you like i love you]

One county where the distribution clearly affects the Democrats, even though it obviously doesn't in the state, is Hamilton County, Ohio. For some reason the black population is just randomly scattered throughout, including alongside the northern edge, and it's pretty hard to draw a consistently R-voting State House district in it without absurd gerrymandering, except for the western part of of the county (which also serves as an R pack.)