Is Gerrymandering Unconstitutional?

[maxparrish]

That's been understood for years (and is one of the reasons Gerrymandering has become so aggressive in the first place). It happens when a large bloc of conservative voters all wind up in the same general area as a matter of demographic or socioeconomic concentration. If, for example, 90% of all Republican voters all live in two highly affluent congressional districts, the other 30% are spread too thinly to win races in other districts. In this (extreme) scenario, the Republicans would win both of those districts by huge landslides, but would loose in almost every other district by somewhat smaller margins, despite the OVERALL vote being relatively narrow.

Sans Gerrymandering, it would be VERY difficult to arrange a situation where the distribution of candidates actually went in the opposite direction of the popular vote. If, say, you have a very high and heavily democratic urban population -- the majority of the state's population -- all clustered into one district surrounded by a dozen sparsely populated Republican districts; but THAT would screw up the state's representation anyway, giving a huge number of people only a single representative and a small number of people MANY representatives. So long as the random districts reflect actual population densities (which they DO, in this study) that ceases to be a problem.

Finally, there is a bit of begging the question by the authors (and you). Who says that compactness and pinpoint accuracy in population divisions are only of interest? Compactness is appealing because it avoids some of the absurd gerrymandering; on the other hand it also makes sense to take into account community interests. How can one reflect "the will of the people" if the 51 percent are part of a winner take all? What happens to the will of the 49%?

One should consider that large groups of people in a particular community are more likely to vote together on a particular regional issue than not. For example, voters are more likely to support an anti-fracking candidate if they all live in a town whose water supply has previously been poisoned by fracking. It would be uniquely undemocratic to deliberately split that town into four different districts, dividing up its voters and forcing them to try and out vote pro-fracking voters from the other side of the state. At least insofar as LOCAL politics will be preserved by letting people vote together in geographic knots, the "winner takes all" problem is somewhat mitigated.

And such communities of interest often span party divisions. In California they include agriculture, environmental concerns, ethnic community, hi-tech industry, etc. Should not the lives and fortunes of those folks be included if practical?

Possibly, but it sort of depends on who defines what interests are relevant and what communities are included in them, while at the same time adjusting for representation by population. That potentially just boils down into gerrymandering again.

It would seem to me to be far more efficient to distribute congressional districts among organic, pre-existing political entities -- towns, urban districts, counties, etc -- with an acceptable margin for population difference. This way, you at least avoid the trap of slicing up demographic groups for some political advantage: if you want to get access to those 30,000 white republican votes in that wealthy suburb, you're also going to be stuck with 55,000 democratic votes from the neighboring college towns because a congressional district has to have at least 100,000 votes.

Sorry, this "random" districting still does not pass the smell test. The 2012 Presidential vote nearly even, yet it seems (in eyeballing it) the Democratic voters are highly concentrated in 30 of the 100 counties. Whereas Romney did not win any county with more than 50,000 GOP votes, not so for Obama. Obama drew most of his votes from a limited number of counties:

Wake 286K
Guilford 145,000
Buncombe 70K
Cumberland 75K
Durahm 109K (76 percent of the Vote)
Forsyth 91K

Seems to me that it is the Democrats who are concentrated and a random district generator should actually favor Republicans.

http://www.politico.com/2012-election/results/president/north-carolina/

 

[Davka]

Sorry, this "random" districting still does not pass the smell test.

Take it up with the Pew Center. Their methods and results are publicly available. "I don't like the results they got because I don't like the results" is not a valid argument.

 

[T.G.G. Moogly]

In personal terms, this means that - even though I live in the most liberal area in North Carolina - my vote was rendered null and void by the careful creation of safe Republican seats via dilution of the Democrat vote. Asheville, a college town and a hippie mecca, is represented in both the State House and in D.C. by Republican legislators.

This strikes me as an abrogation of my Constitutional right to proportionate representation. In fact, it strikes me as so blatantly in opposition to the spirit and letter of the Constitution, that I am considering asking the ACLU to take up a suit against the State of North Carolina. I would like top take this idea to a Constitutional scholar at a local University, and ask for help in crafting my initial argument. But even before taking this step, I thought I'd ask folks here what they think. If you were to approach a Constitutional scholar for help in suing the State, how would you word your verbal argument?

It certainly appears that you have standing.

Not sure how I'd word it because I'm not lawyer. Gerrymandering takes place to give one political bloc an advantage over another political bloc. This is not unlike laws that were on the books in the South - all legal - that prevented blacks from participating in government as laid out in the constitution. You would need to go back to before the districts were redrawn and demonstrate that they were redrawn for the sole purpose of giving one group of voters an advantage over another group of voters. Seems pretty simple to do and the data is all on your side as far as outcomes. The outcomes prove your case.

 

[maxparrish]

Sorry, this "random" districting still does not pass the smell test.

Take it up with the Pew Center. Their methods and results are publicly available. "I don't like the results they got because I don't like the results" is not a valid argument.

So you try to defend your blustering over GOP 'gerrymandering' with bogus popular vote "Democratic majority" claims for 2010 and 2014, and then a preliminary study, but when caught you refuse to defend the points by handwaving to PEW? The "I can't defend the results they got even though I like them, so go somewhere else" is not a valid argument either old chap. But it is a valid argument to note counter-factuals that either you are still unable to address, or choose to ignore.

a) You claim that as Democrats got 51% of the vote, their should be a proportional outcome. However, the study average of 7.6 seats for Democrats in that election is not a proportional outcome. Using your own benchmark makes the study statistically biased in some unknown manner.

b) Your complaint, apparently, was based on the presumption that the criteria of compactness and pinpoint accuracy is both neutral and valid. I did not see why community of interest is not also equally valid.

c) I pointed out that the basis of your complaint about NC and 'the majority voted for democrats' was wrong, both in popular vote totals for 2010 and 2014. I also pointed out the lack of proportional outcomes in 2010, in favor of Democrats.

PS Hopefully you can also point out where, at PEW, this study and background material is. Given that a preliminary search at there site did not turn it up, I'd hate to think you are not even aware of its source.

 

[Davka]

So you try to defend your blustering over GOP 'gerrymandering'

Do you ever actually bother to read posts before jerking your knee?

Didn't think so.

 

[fromderinside]

So you try to defend your blustering over GOP 'gerrymandering'

Do you ever actually bother to read posts before jerking your knee?

Didn't think so.

You post here and you are surprised when those responding to your posts don't get beyond the first hot point?

Naw

yeah? Naw.....

 

[Derec]

It nowhere mandates either proportional representation in general or single-member districts. Nowhere. At least as far as I can tell.

I stand corrected.

 

[Underseer]

So you try to defend your blustering over GOP 'gerrymandering'

Do you ever actually bother to read posts before jerking your knee?

Didn't think so.

It's only gerrymandering when Democrats do it (because Democrats are socialist dictators who hate America and hate our freedom). His use of the scare quotes is perfectly correct because any actions by the Republican party are necessarily for the purpose of preserving our freedom and defending us from the socialists, and thus can never be wrong and must never be questioned.

 

[fromderinside]

I stand corrected.

Why? Isn't the constitution like the bible. One can find succor for their view somewhere in it. Its has nothing to do with words it has to do with how one reads those words or how one reads the words he believes are in there.

For instance the supreme court in Westbury vs Sanders (1964) from:  Section 2.1 composition and election of Members see clause three

[C]onstrued in its historical context, the command... that Representatives be chosen ‘by the People of the several States' means that as nearly as is practicable one man's vote in a congressional election is to be worth as much as another's.... The history of the Constitution... reveals that those who framed the Constitution meant that... it was population which was to be the basis of the House of Representatives.... It would defeat the principle solemnly embodied in the Great Compromise-equal representation in the House for equal numbers of people-for us to hold that, within the States, legislatures may draw the lines of congressional districts in such a way as to give some voters a greater voice in choosing a Congressman than others.

My reading of that particular ruling is that vote is proportional and one man's vote must be equal in value to another man's vote.

 

[lpetrich]

I'd like some numbers on the 2010 and the 2014 US House races. Numbers comparable to what I could find for the 2012 House races at the Boston Globe, boston.com . Numbers like counts of how many votes for each of the candidates in every Congressional district. I once did an analysis of the BG 2012 numbers in gory detail.

Overall by Party -- --- --- --- [('GOP', 234), ('Dem', 201)]
Overall Proportional -- --- --- [('Dem', 215), ('GOP', 211), ('Lib', 5), ('Ind', 2), ('NPA', 1), ('Grn', 1)]
Overall Proportional MPX -- --- [('Dem', 213), ('GOP', 210), ('(other)', 12)]
Overall Proportional MP0 -- --- [('Dem', 220), ('GOP', 215)]
State-By-State Proportional --- [('GOP', 219), ('Dem', 213), ('Lib', 1), ('NPA', 1), ('NPD', 1)]
State-By-State Proportional MPX [('GOP', 219), ('Dem', 213), ('(other)', 3)]
State-By-State Proportional MP0 [('GOP', 220), ('Dem', 215)]

Overall proportional: Dem-GOP = +5
State-by-state proportional: Dem-GOP = -5
Overall by party (actual) Dem-GOP = -33

Here are the results for the 12 most populous states:

California
Votes -- [('Dem', 6221080), ('GOP', 3976901), ('NPD', 243756)]
Votes X  [('Dem', 6221080), ('GOP', 3976901), ('(other)', 243756)]
Party -- [('Dem', 38), ('GOP', 15)]
Prop --- [('Dem', 32), ('GOP', 20), ('NPD', 1)]
Prop MPX [('Dem', 32), ('GOP', 20), ('(other)', 1)]
Prop MP0 [('Dem', 32), ('GOP', 21)]

Texas
Votes -- [('GOP', 4230024), ('Dem', 2940085), ('Lib', 245304), ('Grn', 32768), ('Ind', 5345)]
Votes X  [('GOP', 4230024), ('Dem', 2940085), ('(other)', 283417)]
Party -- [('GOP', 24), ('Dem', 12)]
Prop --- [('GOP', 21), ('Dem', 14), ('Lib', 1)]
Prop MPX [('GOP', 21), ('Dem', 14), ('(other)', 1)]
Prop MP0 [('GOP', 21), ('Dem', 15)]

New York
Votes -- [('Dem', 3652498), ('GOP', 2029993), ('Grn', 37693), ('Con', 20066), ('SWP', 4874), ('Lib', 2621), ('Oth', 2164), ('CST', 346)]
Votes X  [('Dem', 3652498), ('GOP', 2029993), ('(other)', 67764)]
Party -- [('Dem', 21), ('GOP', 6)]
Prop --- [('Dem', 17), ('GOP', 10)]
Prop MPX [('Dem', 17), ('GOP', 10)]
Prop MP0 [('Dem', 17), ('GOP', 10)]

Florida
Votes -- [('GOP', 3825723), ('Dem', 3391037), ('NPA', 282615), ('Lib', 11165)]
Votes X  [('GOP', 3825723), ('Dem', 3391037), ('(other)', 293780)]
Party -- [('GOP', 17), ('Dem', 10)]
Prop --- [('GOP', 14), ('Dem', 12), ('NPA', 1)]
Prop MPX [('GOP', 14), ('Dem', 12), ('(other)', 1)]
Prop MP0 [('GOP', 14), ('Dem', 13)]

Illinois
Votes -- [('Dem', 2640819), ('GOP', 2161599), ('Ind', 71524), ('Grn', 31302)]
Votes X  [('Dem', 2640819), ('GOP', 2161599), ('(other)', 102826)]
Party -- [('Dem', 12), ('GOP', 6)]
Prop --- [('Dem', 10), ('GOP', 8)]
Prop MPX [('Dem', 10), ('GOP', 8)]
Prop MP0 [('Dem', 10), ('GOP', 8)]

Pennsylvania
Votes -- [('Dem', 2722560), ('GOP', 2651901), ('Ind', 45171), ('Lib', 6135)]
Votes X  [('Dem', 2722560), ('GOP', 2651901), ('(other)', 51306)]
Party -- [('GOP', 13), ('Dem', 5)]
Prop --- [('Dem', 9), ('GOP', 9)]
Prop MPX [('Dem', 9), ('GOP', 9)]
Prop MP0 [('Dem', 9), ('GOP', 9)]

Ohio
Votes -- [('GOP', 2315250), ('Dem', 2065814), ('Lib', 77465), ('Grn', 24890)]
Votes X  [('GOP', 2315250), ('Dem', 2065814), ('(other)', 102355)]
Party -- [('GOP', 12), ('Dem', 4)]
Prop --- [('GOP', 8), ('Dem', 8)]
Prop MPX [('GOP', 8), ('Dem', 8)]
Prop MP0 [('GOP', 8), ('Dem', 8)]

Georgia
Votes -- [('GOP', 1456400), ('Dem', 1439127)]
Votes X  [('GOP', 1456400), ('Dem', 1439127)]
Party -- [('GOP', 9), ('Dem', 5)]
Prop --- [('GOP', 7), ('Dem', 7)]
Prop MPX [('GOP', 7), ('Dem', 7)]
Prop MP0 [('GOP', 7), ('Dem', 7)]

Michigan
Votes -- [('Dem', 2324884), ('GOP', 2083613), ('Lib', 102437), ('Grn', 25336), ('UST', 16308), ('NPA', 16010)]
Votes X  [('Dem', 2324884), ('GOP', 2083613), ('(other)', 160091)]
Party -- [('GOP', 9), ('Dem', 5)]
Prop --- [('Dem', 7), ('GOP', 7)]
Prop MPX [('Dem', 7), ('GOP', 7)]
Prop MP0 [('Dem', 7), ('GOP', 7)]

North Carolina
Votes -- [('Dem', 2219165), ('GOP', 2143118), ('Lib', 24044)]
Votes X  [('Dem', 2219165), ('GOP', 2143118), ('(other)', 24044)]
Party -- [('GOP', 9), ('Dem', 4)]
Prop --- [('Dem', 7), ('GOP', 6)]
Prop MPX [('Dem', 7), ('GOP', 6)]
Prop MP0 [('Dem', 7), ('GOP', 6)]

New Jersey
Votes -- [('Dem', 1641671), ('GOP', 1342631), ('Ind', 51808)]
Votes X  [('Dem', 1641671), ('GOP', 1342631), ('(other)', 51808)]
Party -- [('Dem', 6), ('GOP', 6)]
Prop --- [('Dem', 7), ('GOP', 5)]
Prop MPX [('Dem', 7), ('GOP', 5)]
Prop MP0 [('Dem', 7), ('GOP', 5)]

Virginia
Votes -- [('GOP', 1833934), ('Dem', 1736164), ('Ind', 25522), ('IGr', 23189), ('Grn', 2000)]
Votes X  [('GOP', 1833934), ('Dem', 1736164), ('(other)', 50711)]
Party -- [('GOP', 8), ('Dem', 3)]
Prop --- [('GOP', 6), ('Dem', 5)]
Prop MPX [('GOP', 6), ('Dem', 5)]
Prop MP0 [('GOP', 6), ('Dem', 5)]

MPX = lumping all the third parties together
MP0 = ignoring all the third parties
Proportionality algorithm: Hare quota + d'Hondt topping off

So it's evident that several states have sizable disproportions, some favoring the Democrats, and some the Republicans. But overall, the disproportions favor the Republicans.

Here's an archive that contains my results, data, and analysis program: View attachment USHouse2012.zip