Gerrymandering and a cure - shortest splitline algorithm

http://rangevoting.org/GerryExamples.html

by mmastrac • 14 years ago

83 82 14 years ago

If only the immense amount of political problems we have could be solved through simple, rational technical solutions! What a fantastic world we'd live in right now!

A bit more on-topic is that I feel some issues like gerrymandering wouldn't be such a big problem if more root-level problems were addressed, like implementing instant run-off voting to help weaken the two-party system in the first place.

ianb3• 14 years ago

I don't see how instant run-off voting would change much of anything, unless perhaps you got rid of the endorsement process at the same time so that people of the same party could more easily run against each other.

weeksie• 14 years ago

It removes the fear of throwing away a vote on a third party. Look at the Australian government for example, there are a ton of third parties represented.

JoshTriplett3• 14 years ago

I agree with you that voting systems represent a root-level problem (and in particular that first-past-the-post causes the two-party system, per Duverger's Law), but IRV is pretty much the only voting system worse than first-past-the-post. It allows people to vote for a third party, but completely ignores their preferences among other parties until eliminating that third party; thus, it does a good job of convincing people to vote third-party, but actively works against the preferences of people who do so.

taejo• 14 years ago

> completely ignores their preferences among other parties until eliminating that third party

I'm afraid I don't follow. Eventually, the third party either wins or is eliminated, and then the preference between the two main parties is effected. Am I missing something?

ccleve2• 14 years ago

Nice sentiment, but the implementation is inept. The courts have determined that existing boundaries, including county and village boundaries, matter. Physical features matter. Existing "communities of interest" matter. Population equality matters. And, sadly, race matters a lot.

The only sort-of mathematical awareness that the courts have concerns contiguity and compactness. Contiguity is easy. Compactness has a million definitions, and the courts haven't settled on a meaningful one. Come up with a compelling, mathematically-sound definition of compactness and you might have something.

Also, the author ought get at least a basic familiarity with the literature on political map-drawing. If he had it, he would know that maps are drawn on census blocks and that houses are never "split".

cduan• 14 years ago

I wonder if this algorithm would still present biases given the fact that large cities tend to vote differently from less densely populated areas. My hypothesis is that this algorithm would favor city voters over rural voters, for the following reasons:

1. Successfully gerrymandering in favor of one party generally involves stuffing the other party's voters into one district, while splitting your party's voters into multiple districts.

2. The proposed algorithm favors straight-line divisions through regions such that about half the population of the region is on each side of the division.

3. Thus, chances are, the straight lines will run through the middles of cities, cutting cities into multiple districts.

I'd be interested to see someone test this out--it's likely that my hypothesis is wrong, but it also seems pretty likely to me that there will be at least some urban/rural bias, one way or another.

3JPLW• 14 years ago

My favorite example is Illinois' 4th Congressional district[1]. But: there can be positive outcomes to such gerrymandering. Chicago is very strongly majority democratic. Splitting the city with straight lines would not change this. All of the congressional districts that include the city are already held by democratic incumbents.

But this funny shaped district connects the two large latino neighborhoods: predominantly Puerto Rican in the North and Mexican in the South. And as such, Luis Gutiérrez has represented the district since 1993, when the district was defined as such. He was the first Latino to be elected to the House from the midwest.

More information on the general 4th district wikipedia page[3].

1. https://upload.wikimedia.org/wikipedia/commons/a/a0/IL04_109...

2. https://en.wikipedia.org/wiki/Luis_Guti%C3%A9rrez

3. https://en.wikipedia.org/wiki/Illinois%27s_4th_congressional...

ccleve2• 14 years ago

Just curious: who are you?

I'm also a Chicago software developer political type. And I was involved in a lawsuit challenging the constitutionality of the 4th CD in the 90s.

frozenport2• 14 years ago

Its not immediately clear disproportional representation is a positive outcome.

mjn3• 14 years ago

It's closer to proportional rather than disproportional representation in this case. Latinos make up about 30% of Chicago's population, but they're spread out so they wouldn't form a majority of any single compact district, so wouldn't be able, as a community (if indeed it's a coherent community), to send a representative to Washington to represent their views. However, if you connect demographically similar areas, then the community gets representation closer to their actual share of the population.

Of course, we could just use proportional representation outright and achieve a similar result, or elect candidates from multi-member districts and a slate of candidates (e.g. put all of Chicago in one district, and elect the top 4 candidates). But that's a more radical reform that's less likely to happen.

jacques_chester• 14 years ago

In Australia electoral boundaries set by the Electoral Commissions. Officers of those commissions are by law required to be politically neutral. For example, because I have been a member of a political party, I am ineligible to work for them.

There's a modest amount of argy-bargy every time the boundaries are redrawn, but in general seats are contiguous and have shapes that follow the population distribution.

ianb3• 14 years ago

In the case of a lot of states with one major city, the algorithm basically just splits that main city, for instance New Mexico: http://rangevoting.org/Splitline2009/nm.png – when the city is big enough to have a couple districts of its own it's not as bad, like: http://rangevoting.org/Splitline2009/mn.png

I'm kind of unimpressed. There are mechanisms to remove partisanship and political gerrymandering without making the whole thing into an algorithm.

kenthorvath• 14 years ago

What's wrong with algorithms?

brownbat• 14 years ago

Splitlines are very likely to cut through urban areas. Since the urban/rural distinction is meaningful in American politics, this has interesting consequences.

I'm not sure "arbitrary" always gets you "unbiased," or "unbiased" always gets you to "fair."

Fairness requires tallying up the numbers of disenfranchised in the new method, comparing it to the old. Splitline will exclude some urban voters from the political process for the foreseeable future, and some rural voters. It's not obvious which group loses more (it's highly contingent on the population distribution). But it's worth tallying up, just to see.

If disenfranchised is really our concern, why not just try to minimize it more directly?

Let's say someone is defined as "disenfranchised" if they identify with a party with no chance of victory, or if they contribute overflow votes that will never be necessary for their party to win the district. If you're in a minority party that gets less than 40% of the vote, you're disenfranchised. If your party has over 60% of the vote, then everyone beyond that 60% is disenfranchised. (We can quibble over the definition, or generalize it for more parties, but let's just assume we have some operational definition approaching "disenfranchised," without any definition, we couldn't evaluate the success/failure of splitline anyway.)

Now just draw all the constitutional district maps, and pick the one that disenfranchise the fewest number of voters possible.

Ok, that may be computationally infeasible. So instead, just have each party draw it's own recommended map, and then pick the one of those offered that disenfranchises fewer voters (using data from the last election, maybe).

[1] A man-made definition is exactly what splitline recommends against, but even imperfect but vaguely reasonable definitions of "disenfranchisement" in this method will tend to be more "fair" (in terms of reducing disenfranchisement) than purely arbitrary dividing lines.

ianb3• 14 years ago

> Splitlines are very likely to cut through urban areas. Since the urban/rural distinction is meaningful in American politics, this has interesting consequences.

Thinking about it more, in a place like New Mexico (and there are many similar states) this means rural voters will be systematically disenfranchised, as they the state ends up being divided into regions around the central city, and in each region the urban area is larger than the rural area. If combined the rural area could be at least a seat of its own, even if more overall seats went to urban/metro districts.

Well... thinking about it, it could also go the other way, depending on what the overall rural/urban divide is – if there is a larger but dispersed non-metro population then it could be the urban population that is lost in the rural population. This might actually be the case in someplace like New Mexico or Idaho.

New Mexico's actual congressional districts look very reasonable: http://en.wikipedia.org/wiki/New_Mexicos_congressional_distr...

eli6• 14 years ago

This would have the practical effect of reducing minority representatives. Some people think this is OK, but it's worth at least mentioning.

w1ntermute4• 14 years ago

That could be easily solved by adding some proportionally elected representatives.

eli6• 14 years ago

Wouldn't that then result in less geographic diversity or representatives who don't represent just your district? Or am I misunderstanding?

mikeash• 14 years ago

Gerrymandering is only a symptom, and cannot be cured without tackling the underlying disease. That disease is a political culture in which politicians will do whatever it takes to win, regardless of popular sentiment. No, I have no idea how to fix that.

gizmo6866• 14 years ago

Set up the system so that the politician's best chance at winning comes from doing as good a job as possible. This means providing as few oppurtunities as possible to game the system.

nathanstitt• 14 years ago

These type of articles annoy me. Not the article necessarily, but the thinking that goes into it. The solution presented is technically sound and would absolutely solve the problem of crazy gerrymandered districts.

The possibility of something like this actually being implemented: zero.

Does anyone seriously think that congress would allow their re-election prospects to be controlled by a software program? It sounds like they are attempting a ballot initiative route, but good luck explaining this to a skeptical public while facing negative campaigning that makes it sound like 'computerized death panels' all over again.

I see the same thinking all the time - a new developer shows up all bright eyed and tries to talk everyone into switching to his favorite 'Framework/Language X' without even considering considering if it's at all feasible.

I'm not saying that things can't be improved. But if you're proposing a change you have to be smart enough to consider it's chances and propose ways to overcome the inevitable pushback.

graeme• 14 years ago

Not to mention that it ignores the motives for Gerrymandering. Politicians surely know that Gerrymandered districts are 'bad', if a logical electoral system is your priority.

Politicians Gerrmander because they see an advantage in it, and they don't have compunctions against it.

I suspect good, non-Gerrymandered districts could easily be generated using traditional means as well. But the officials in charge have signalled they aren't interested.

danso3• 14 years ago

There is nothing easy about drawing districts...what fundamental rule is there that says it is better for my household to be grouped with the politically-polar-opposite area south of me rather than the more politically similar group east of me? Is it bad if each district contains demographically homogenous groups that vote as a bloc? How else to introduce diversity than by organizing districts differently than traditional neighborhood boundaries?

Obviously, there are some who will attempt to game the system? But one candidate's gerrymander is another's fair apportionment...to dismiss it all as a big conspiracy controlled by indefeatable evil interests is to practice the worst kind of intellectual ambivalence.

alexqgb6• 14 years ago

That's an understatement. Here in California, politicians reacted to the idea with an incandescent, frothing-at-the-mouth rage peppered with (political) death-threates aimed at anyone who dared to make an issue of the issue. But thanks to the direct-initative process, they didn't have to power to actually block it. The direct initiative system has been the source of a lot of grief, and frankly, it has a lot to answer for. But in this case, it proved to be worth its weight in gold. Indeed, it may be the thing that leads to the end of gerrymandering throughout America.

Until very recently, California was polarized even more sharply than Washington DC, and seats were filled with ideological firebrands backed by primary voters who would punish "compromise". This chronic disunity meant we unable to use the power of our size effectively. But now that we're steadily replacing those folks with people who can safely operate together, our reps can safely ignore their parties to vote as a uniform block that only cares about itself. Coming from a state as big as California, that represents a major a threat. In another election cycle or two, California will have enough unity to become the swing vote on anything. And yes, you can expect it to use this power ruthlessly. Nothing will move unless there's something in it for California, and given the size and diversity of the state, it'll be a rare issue that passes without California extracting some major concession from the other States. It'll be all take and no give, all the time, on everything.

Power this substantial and unchecked will be deeply resented and increasingly abused. For other big states, the only way to stop the pain will be to start operating the same way themselves. Florida is well on its way to ending gerrymandering. Governor Cuomo is starting to make an issue of it in New York. If they can start operating as post-partisan blocks, they can work together to neutralize California's power. Once Texas and Illinois get into the game (in that order, I suspect), the political dynamic will be Big States vs. Small States Further Disadvantaged By Intractable Partisan Rifts. Applying pressure of this magnitude to what will be perennial losers should cause the whole sordid institution to collapse entirely, subordinating politicians to people rather than parties in all 50 States. In retrospect, the madness we're living through now will look a lot like the crime-ridden hell scape of New York City in the 70's, as seen from the 90's on.

aneth44• 14 years ago

While I understand the sentiment that it's wasted effort to propose impractical solutions, it is important to know what our ideal is before deciding what is possible - otherwise we are condemned to incremental change based on pessimistic assumptions of what is possible that are baked into our ideation process.

kscaldef• 14 years ago

I'm not even sure that the solution is technically sound, as it appears to ignore the possibility that multiple people live at the same location.

gizmo6866• 14 years ago

They would likely have the support of the party which is being dissadvataged by Gerrymandering. Combine this with the fact that (I think) this type of solution has a natural advantage based on its merit, and passing it does not seem as unlikely as you think.

jfb• 14 years ago

Plus, if you want an unrealistic but legal way to end Federal Congressional gerrymandering, a solution exists: proportional representation.

danso3• 14 years ago

I'm sorry but this is the kind of "solution" that gives analytical thinking a bad name. Ignoring the political will to implement this, this solution's biggest problem is its willful ignorance of physical boundaries. A four lane highway can be as stark a dividing line between groups as miles of physical space. This article simply dismisses this reality by saying "well politicians never consistently considered those boundaries either, so there"

Ironically, this ostensibly logical argument ends up going further into anti-intellectual territory than the politicians it derides. The thinkers behind this can't think of any other way that math might be used to accommodate or weight the effect of these different boundaries? Really? Or would the resulting boundaries be so aesthetically unappealing that its just better to go for the elegance of straight lines? That is among the stupidest rationales I've read in the whole gerrymandering debate.

The whole post just smacks of incredible naïveté. No wonder engineers and scientists have a bad reputation in representative politics.

gizmo6866• 14 years ago

Putting aside the problems of districted voting, is this naive attempt really that bad? Why is it a problem if I am in a different voting district than my next door neighboor, we both still go to vote, then have our votes evaluated by the system resulting in a winner. The goal is to make said system as fair as possible (while maintaing the concept of districts). I do not see why physical boundries are relevent at all. As cduan pointed out, this algorithm might have a bias favoring urban or rural voters, which is a problem.

I have an even simpler algorithm that I think solves all problems of bias. Say you want a state to have N districts. Create N district labels (probably the numbers 1 through N). Now assign each voter in the state to a randomly selected district. For each district, take all but the largest contious regions and re-asign each voter within them. Repeat until all districts are continuous. Or we could move to a popular vote.

mjn3• 14 years ago

As far as the relevance of physical geography, I do see some benefits for the original idea of representatives actually representing an area in Washington. That's more likely to happen if boundaries are drawn along somewhat coherent lines that delineate cities or neighborhoods. On the other hand, those concerns are clearly not the main thing that goes into drawing today's gerrymandered boundaries anyway, so maybe losing that notion for geometrically drawn boundaries isn't a large price to pay.

AnthonyMouse• 14 years ago

Congratulations, the two of you have aptly demonstrated why these things are never implemented.

There is more than one way to algorithmically draw district lines. The different ways will benefit different people. The people who benefit from a particular method will then conspicuously be found arguing that the way that benefits them is the One True Way, and hire a bunch of lawyers and the RIAA's copyright mathematicians to come up with plausible-sounding arguments for why they're right.

And since neither of them is actually "right" because there is no One True Way, it immediately degenerates into a cynical political battle to choose the best automated gerrymandering algorithm for one's own political party.

btilly5• 14 years ago

A random note for those who might be inclined to think that gerrymandering does not have a significant political impact. In the 2012 elections, Republicans carried Congress by 233 to 200 seats. Yet across all Congressional elections they lost the popular vote by about 1.35 million. (http://skeptics.stackexchange.com/questions/14315/did-the-de... is where I ran across this.)

The cause of the discrepancy is largely gerrymandering.

Unfortunately lots of states have laws about trying to draw boundaries that put ethnic groups together, etc in non-trivial ways. There are algorithms that can produce maps satisfying those restrictions, but they are more complicated. And, for obvious reasons, they often have the goal is producing such maps with a definite electoral bias...

danso3• 14 years ago

So is the 1.35 million number just the absolute difference between all votes cast nationally? Doesn't there have to be so e adjustment for districts in which many D voters elected an easy winner but in which excess votes don't really do much since it is an easy win?

And of course, there's other explanations besides gerrymandering...it could be that the RNC and its financial allies spent more savvily in tight races and pulled through. The problem of money is of course its own discussion

chimeracoder2• 14 years ago

> And of course, there's other explanations besides gerrymandering...it could be that the RNC and its financial allies spent more savvily in tight races and pulled through.

Yes, this issue is incredibly complicated to analyze - for example, in the House, contentious bills like the ACA are oftentimes passed with ~220 votes (usually ~217 defines a majority). That doesn't mean that only 220 people would have been been willing to support the bill; just that the winning party didn't need to waste any resources whipping another ten votes in excess of what was needed to secure a victory.

Just looking at the final vote tally is misleading, because if the threshold for passage were higher, it's very likely that the actual vote tallies would miraculously adjust accordingly.

btilly5• 14 years ago

The 1.35 million is just the absolute difference between all votes cast nationally. The result of all of the adjustments you could try to make, taken to the limit, is the 233/200 margin of Republican control of Congress.

Incidentally I credit the win in the popular vote for Democrats to a very good get out the vote operation for Obama. I am sure that future incarnations will specifically target every marginal Congress seat as well.

waterlesscloud• 14 years ago

You had districts in California that were Democrat vs Democrat, meaning Republicans couldn't pick up any popular votes in those districts. That alone makes for a mighty skew in the popular vote totals.

thethimble• 14 years ago

Interesting question: Isn't having laws that prohibit putting ethnic groups together in an algorithmically generated way discrimination in itself? For example, if the algorithm generates a predominantly black district (that's algorithmically generated and thus unbiased), breaking it up because it is predominantly black is an act of discrimination.

btilly5• 14 years ago

Isn't having laws that prohibit putting ethnic groups together in an algorithmically generated way discrimination in itself?

Sorry I was unclear, it was the other way around. If you can, you have to keep ethnic groups together. So, for instance, if you can put all of Chinatown together, you have to.

The political effect is kind of complicated. On the one hand it gives members of those ethnic groups a chance to win office. On the other hand it creates very safe districts. Given the way the math works, creating very safe districts for your side leads to improving the odds that the other side will win more seats than their share of the popular vote.

Activists are very focused on the first effect and tend to ignore the second. Furthermore politicians encourage this because they like safe seats. But the math is inescapable - an effective electoral gerrymander is created by creating a modest number of ultra-safe seats for the other side, and many more safe but less extreme seats for your side.

So all black seats that guarantee black representation in Congress leads to seats that vote 80%+ for the Democrat which in turn leads to Republicans electing more than their share of the popular vote.

kbutler• 14 years ago

> The cause of the discrepancy is largely gerrymandering.

Proof needed.

The discrepancy simply means that democratic candidates won by wider margins than republican candidates, which can easily happen without gerrymandering.

Dove• 14 years ago

I'm not apologizing for gerrymandering in the least, but . . .

http://catb.org/jargon/html/koans.html#id3141241

. . . just because you didn't put bias into a system doesn't mean it isn't biased.

alwaysdoit2• 14 years ago

Districts should be deprecated altogether. Instead we should elect representatives via proportional representation.

When we we had approximately 30,000 citizens per representative there was a chance at your representative actually knowing you, but at this point we are more likely to have our views represented by a third party.

vidarh3• 14 years ago

Norway is below your 30,000 citizen per representative number, and frankly even then, while there certainly is a chance, not many people do know their representative even there.

In Norway, some regional connection is arranged by grouping seats into multi-seatconstituencies by region. Then more proportional results are obtained by taking any votes "leftover" for a party after assigning whole seats and assigning seats from a non-geographic pool with some caveats.

It satisfies those who want a regional link of sorts, and doesn't skew the proportionality too much.

alwaysdoit2• 14 years ago

For the US, I think that electing representatives by proportional representation within states would work similarly and would work pretty well with our existing constitutional framework.

mjn3• 14 years ago

Some interesting mathematical background, on whether formal convexity measures can work as a gerrymandering metric: http://mathdl.maa.org/images/upload_library/22/Polya/Hodge20...

aneth44• 14 years ago

Let me challenge some assumptions about gerrymandering. The most common misconception of gerrymandering is the assumption that close races are required for the voters to have a choice. While in current practice this may be true, the reason is due to other factors in our voting habits and electoral structure, not inherent in gerrymandering. The solution is not to get rid of gerrymandering, but to make other more fruitful changes to our electoral habits and systems. Gerrymandering is neither good nor bad.

Gerrymandering works by making certain districts overwhelming in a single party's favor, leaving few competitive districts. While this may sound like it undermines voter choice, that is not the case.

A district dominated by one party does in fact have a choice in their primaries. In fact voters there have a better choice than in competitive districts. Voters in a single party district get to focus on whether their representative precisely reflects their views rather than whether their party can win in the general election. Take for example a 90% Republican district. We know that this district will be representative of a swath of Republican voters - that's a good thing. Those voters can now focus on the more subtle leanings of their candidates during the primaries. Voters and primary candidates can openly debate about social conservatism and gun control, rather than what the small percentage of "swing" voters care about. Essentially, party factions become the choice, exactly what this country needs to evolve two parties that do not do a good job of representing the country. This is how I foresee democracy emerging in China, and is common from time to time, though not necessarily implemented to the voters advantage, in frequently defacto single party states, such as Mexico, Japan, and Canada.

Pushing the impact of our representative system into more subtle debates in a primary election is only a bad thing because our primary system has some fixable issues. Fewer voters participate in primary elections, and there are not enough primary challengers. Making primaries more competitive and getting voters more involved in the primaries of the gerrymandered districts would help, and this leverages the positive aspects of gerrymandering rather than wasting effort trying to ban the practice.

In some ways the issue should be self-correcting. The more guaranteed a district is for a given party, the more candidates should emerge in the primary, the more competitive the primary should be, and the more voters should care. This is because winning the primary is nearly as good as winning the general election.

Perhaps a better solution to competitive primaries is something similar to what is being tried in California - non-partisan primaries. These shift the more subtle intraparty distinctions in single party districts to the more widely attended general election where they should be. In the above-mentioned Republican district, the district would likely have two Republican candidates in the general election. Such a system does both a good job of representing the constituents and of providing voter choice. The impact is similar to what one gets in a proportional representation or many-party system.

The reverse of gerrymandering - engineering all highly competitive districts - also has major disadvantages, the largest being that every district has a choice between two distinct parties who could have radically different views. The winner of the election represents a slim majority, is chosen by a very small percentage of "swing voters," and small changes in "swing voter" sentiment can lead to major changes in representation. While that may have the advantage of throwing the bums out quickly, it can also lead candidates to appeal to their bases and push for voter turnout rather than having substantive distinctions. Sounds familiar.

Conclusion? Stop being outraged about gerrymandering and practice some political Aikido. Push for active primaries and support legislation for non-partisan primaries. These actions don't require Supreme Court intervention or Constitutional amendments, are practical, and may have a better result than banning gerrymandering.

alexqgb6• 14 years ago

This is wrong on so many levels.

First and foremost, the idea that "elected representatives" get to pick who votes for them represents the anthesis of democratically elected representation. It's like a parachute that opens on impact.

Your second error is thinking that gerrymandering simply seeks to place voters from each party in their own districts. What it actually aims to do is advantage one party over the other by marginalizing the other entirely. This can be the case even when the marginalized party actually represents a clear majority of the voters, as happened in the last election.

The third problem is that once a party controls a district exclusively, it inevitably becomes more extreme. Since there's no danger that a firebrand zealot who goes too far will attract challenger from the other party, the contests in "safe" districts select for ideological purity rather that the ability to negotiate with others - which is what's actually the most important thing in people whose whole job is to identify workable compromises that both sides can live with.

An even more insidious effect of the "safe" seat is that it aligns the interests of a party and its donors against those of individual representatives and the represented. Under the current corruption-facilitating system, a politician that pisses off donors or party leadership by putting constituent's interests first can be safely challenged from within his own party without the party risking the loss of the seat. Enforcing "discipline" becomes a lot riskier when retaliating against a "wayward" member could hand the seat to the other party.

You're correct to note that open primaries in California have been a positive development. But you mustn't forget that they were introduced as part and parcel of a broader electoral reform effort that also ended partisan gerrymandering, first for statewide offices, then for national representatives. Of the two reforms, the end to gerrymandering was the more important, by a wide margin.

You may disagree, but that's just a reflection of your own limited understanding. The savviest politicians in the state (correctly) recognized and end to gerrymandering as the greatest thereat to their illegitimate power, and fought against that far harder than they opposed open primaries.

Of course when it comes to open primaries and an end to gerrymandering, electoral reform is not either / or. It's BOTH. And while we're at it, we need to bounce any Senators who block filibuster reform. Getting all three done together would give the people vastly more power over Congress, and by extension, the entire Federal government.

aneth44• 14 years ago

How do we decide on districts? There will always be a possibility of bias and there is no "fair" way to divide districts. There is more random - that's about the best we can do. No districting system ensures a majority vote wins a majority representation, nor is it intended to.

I agree with your point on gerrymandering's biggest issue being the overall bias toward a party, however a proper primary system I believe moderates the parties, or at least provides one moderate and one extreme choice from one side of the electorate rather than two extreme choices from both sides. This ultimately represents people well.

Frankly I don't want a choice between a right wing extremist and a left wing extremist. I want a choice between two reasonable politicians, and I believe fixing the primary system would have a far better impact than worrying about gerrymandering.

drcube• 14 years ago

Why do we have to use geography at all? Can't we apportion representatives alphabetically, by last name? Or numerically, by SSN? I don't really see what geographical representation buys us, except gerrymandering and pork.

freehunter• 14 years ago

It gives us neighborhood-specific representation (or at least should). This neighborhood may have different needs and wants than the neighborhood next to it, and both should be represented by a politician who understands those needs.

5teev• 14 years ago

This article reminded me of a Flash game I played with several years ago:

http://www.redistrictinggame.org/

If you work through the levels, you'll see that voting districts sometimes must have highly irregular shapes to ensure maximum representation of the various groups of constituents. You'll definitely see cases where "objective" boundaries leave some groups completely unrepresented.

Gerrymandering definitely happens, and it's a big problem, but it's not a simple one.

Gerrymandering and a cure - shortest splitline algorithm

Related Stories