Possibly more. I'd be interested in seeing the outcome of this algorithm, which could probably predict congressional makeup using a different algorithm.

it needs the permission of that 100 congressmen and their overlords.

So states could adopt these programs on a piecemeal basis.

The resulting problem, of course, is that states controlled by Republicans would be less likely to do so, so your larger point remains valid.

under mostly republicon congressional rule, but things are so gerrymandered it is pretty hopeless to expect to see much change.

http://bridgemi.com/2014/06/michigan-residents-below-average-in-real-income/

They've held a lock on redistricting much of the country, at least in the states with the larger electoral colleges.

I'd love for Pennsylvania to get this treatment, rather than protecting GOP assholes with lifelong, undefeatable appointments.

[hr][font color="blue"][center]"If you're bored then you're boring." -Harvey Danger[/center][/font][hr]

'compactness' isn't necessarily the best measure in my books. I would actually redraw to make as many districts competitive as possible. As it stands, many people are simply disincentivized from voting, because they know going in that their votes don't matter. 'Unslant' the districts to make as many as possible a tossup, and everybody suddenly has an interest in taking GOTV seriously, and being politically active.

there is surely a way to slap an unbiased grid on every state and let the chips fall where the may.. The USPS could be helpful since they know every address in the USA

Our constitution mandated a representative (house) for every 30K...until we changed in back in 1911 (IIRC)..

There is no reason why 435 should be a sacred number.. we are a HUGE country, and we need MORE representatives..

Though, they'd need to find more space in the Capital!

but it never gave that as a target.

https://en.wikipedia.org/wiki/United_States_congressional_apportionment#House_size

it was already over 120,000 by the end of the Civil War.

which means you can't just do it with this algorithm.

And the section about "communities of interest" is also important; you have to decide if what you're looking for is geographical proximity (what the program has achieved), or if things like town, city and county boundaries matter to you. And if they do, and you have, say, a city that should get 2 representatives, do you divide it as 'inner' and 'outer', or 'north' and 'south' (or some other geographical division), or 'inner' and split the suburbs up between a few surrounding seats?

There are, as you note, the "communities of interest"and the smaller political subdivisions. There's some merit in respecting both.

There's also some merit in paying attention to the previous district lines. If last time there was a roughly north-south line, and this time the computer says you can achieve slightly more compactness with an east-west line, that means that about half the people would find themselves moved to a different representative's district. There's some merit in just tweaking the old line even if the result is less compact.

Political lines aren't the only considerations. Areas served by newspapers and broadcast stations would be a legitimate consideration, because conforming a district to those lines will reduce the cost of advertising.

Any of these methods (computer algorithm or independent commission) will tend to give the Republicans an advantage, because Democrats are more likely to be concentrated in urban areas. This produces a "natural gerrymandering" in which a lot of Democratic votes are wasted because the Democrat wins the district with 80% of the vote, while two or three Republicans are scoring narrower wins in the suburbs.

and also rewards respecting city and county boundaries, etc.

Gerrymandering robs all of us. It harms all of us.

I'd rather have a truly representative congress (well, representative of who our two choices are in any given election, but better than what we have now).

I think California has some bipartisan commission that sets boundaries also.
-
Joking a little: quadrupling the amount of Congressmen/women might cripple the buying power of the elites a little. Bankrupt 'em (as much as possible !)

IIRC, it has a a rep or two from each major party, but most of the members are just residents of the state. It's not perfect - it slants towards people who can afford to spend time in Sacramento - but it has broken up some previously "safe" districts.

that started in Malibu and ran 10 miles down either the beach or the center divider of PCH to Santa Monica.

needs, their need for lights and phones and their propensity for feathering their own nests.

I can't see it bankrupting billionaires to buy them because the $#@s seem to sell out for a lot less than I ever assumed. What a few billion earn overnight will more than cover it. However, I can see it resulting in more tax obligations for those who don't deduct or offshore their way out of tax liability. And more borrowing from China.

Not as though the 435 we already have exert themselves on our behalf, either.

drawing up boundaries based strictly on the distributions of natural communities and having as equal populations as possible.

I think I would prefer that to an algorithm, algorithms are dumb and can be "tweaked".

I think we should adopt it, but (sadly) it'll never happen.

While I can appreciate the neutrality of a computer program, drawing districts isn't just about numbers. The Voting Rights Act has generated court rulings that state, quite bluntly, that it's not enough just to let people vote...you have to make sure they're vote actually counts for something. Dividing purely by the numbers potentially splits underrepresented minority groups into multiple districts, diluting their votes. This practice, known as cracking, is a form of gerrymandering (interestingly, the opposite action, where you create special districts to enhance those underrepresented voters, is also a form of gerrymandering that has been declared illegal by the U.S. Supreme Court).

District boundaries need to not only represent numbers, but represent shared cultural, economic, and social interests. The person being sent to "represent" that district needs to have something cohesive to represent.

This absolutely needs to happen. This idea, along with a Campaign Finance Amendment to the Constitution, and all of the sudden the political representation becomes a much more aligned with the electorate.

Considering the current public polling data, it would be interesting to see what this nations "big" political issues would become. Would we still be debating gun registrations, CO2 emissions, and tax policy or would we just want to outlaw public twerking?

Ahh, a man can dream,
Waitwut (the other one)

Garbage In, Garbage Out.

The garbage in is the census data.

that a computer/algorithm/program would be unbiased (unless tampered with) precludes this from happening in the current political climate. If legislators were truly interested in things like fair play, equal votes for all, and equal representation we wouldn't be having this debate. Don't hold your breath people. This is another one of those things we're going to have to fight for.

Pretty much ALL minority Black and Hispanic districts would be eliminated. This is the basic reason gerrymandering has withstood court challenges up to this point.

I looked as its results for Pennsylvania: http://bdistricting.com/2010/PA_Congress/
As well as the map, they've listed the racial demographics. Under the current scheme, there are 16 districts where the majority is white, 1 with a white plurality (705,688 total, 331,240 white, 250,543 black, and others), and 1 with a black majority. Under the redrawn scheme, it ends up with 16 white majority, 1 with a black plurality (705,888 total, 345,555 black, 226,179 white, and others) and 1 with a black majority (just - 705,819 total, 355,755 black, 260,314 white, and others).

So in this case, 1 district becomes slightly less black, though still a majority, while a 2nd becomes a black plurality from a white plurality (I don't know if it's valid to say these are the 'same' districts, however - probably not). But it could be different elsewhere (and 'Hispanic' isn't a category in the stats they give, so that can't be compared).

into the Witness Protection Program before the wingers put a hit on him.

This is a very doable reform, and much needed.

Simply draw equal population districts with the shortest borderlines possible. That's how you measure "optimally compact." You might end up with hexagons rather than rectangles. But you won't get any salamander-shaped districts.

If it's done by computer algorithm rather than political game playing, it will come out fair, which is the politicians' worst nightmare. And since politicians set the rules for their own games, guess what will happen.

The last time they gerrymandered NC, this is what my district ended up looking like:

The added a lot of new Republican voters to the west and cut off a lot of Democrats from Down East.

Borderline is about how much the edge zig-zags, even at a small scale. I think that's probably not the best measure.

Instead, compare the ratio of the largest circle you can get in to the smallest circle you can bound it it.

I hadn't thought about zig-zags. But with streets and neighborhoods, you probably have to have a little of that. Otherwise, you'd end up someone's dining room in a different precinct from the living room.

The problem is persuading politicians to do something rational, and especially persuading Republicans to play fair when it comes to voting.

Which is the need for a program to collect evidence on, and facilitate the arrest of the people drawing these districts. You can show how they could be until you're blue in the face, but until the criminals are out of the process, no changes are possible

Just needs to be acted on

"How hard can it be to create an algorithm to draw legislative districts after each census" - Creating an algorithm to do just this would be fairly easy. Keeping it accurate, through correct and up-to-date input would be hard. Point of example: Lets say you have two areas that need redistricting... well call them area A and area B. Area A, which is more populated than area b, is on the ball with getting all their statistical data in... though the Handler for that data "misinterprets" some of those numbers, either due to an accident or a strategic donation to the handler-of-data-retirement fund (bribe), are fudged in a particular direction. Area B has a bit less time for things like statistical data because they're busy drinking wine and living the good life, and so their data submission isn't quite up-to-snuff... but one good-lifer in Area B didn't want to give up his wine, and so had the idea to give someone with an important job over in Area A some extra cash in exchange for a "small" favor. The Handler fudges the stats, and inputs them into the algorithm and BAM!!! An incorrect redistricting has been accomplish... let the gerrymandering begin!


TLDR version; Human input is still needed, and there-in lies the weakness of the system. People can just scapegoat the system once something goes wrong, and any actual wrong-doers will get away.

None of this "trade secret" crap!

TPTB want the status quo and will get what they want. They always do. I wish they would allow it. It is bound to be better than this bullshit we have now.

As it stands now, I despise the fact that my solidly blue county is being lumped in with Rowan County, with people who support this kind of bullshit:
http://www.city-data.com/forum/north-carolina/1832863-nine-state-house-members-joined-republican.html

but we still have districts (Davis and Issa's come to miind) that are very safe and it is quid pro quo.

My only down side to it would be the protection of minorities. Otherwise. it MIGHT solve some of the issues we have.

Regardless with 20% turnout, I care little how fair the maps are.

Computer-generated non-gerrymandered districts for each state could be drawn, publicized and used to shame the politicians involved. I suspect it's more than my budget can afford, but someone or some organization might be able to do it, at least for one or a few key states.

I think various media outlets would love to see comparison maps, with gerrymandering vs. without, and that could put pressure on the politicians to try to avoid the embarrassment of being caught playing their corrupt games.

It's been done; the media knows about it. But that on its own doesn't help much; as replies in this thread have said, you also have to follow the Voting Rights Act, which insists of a bit of 'good' gerrymandering; and there's a value judgement about whether you should use a program like this which ignores existing community boundaries (eg it splits Pittsburgh into 3 different districts, each with significant rural areas) or whether you should keep cities together.