How I Teach Gerrymandering (opens in new tab)

You don't have to use FPTP to distort: There's been plenty of debate in Spain, which uses the D'Hondt method. There have always been more than two healthy parties, but ultimately only two get major representation, just like in an FPTP system. And unless we are applying the system to the entire country (which is not how it works), there's still a boost to regional parties.

A decade ago we saw two parties with pretty much the same number of votes, one of Catalonian nationalists, and a typical european leftist party. But the nationalists all got their votes in the same region, gaining very healthy representation overall, while the other party got less than a 5th of the seats, because all their voters do not live in the same place.

What we have to face is that ultimately, the way people are represented is an aesthetic, values-based choice. This doesn't mean that there isn't such thing as partisan gerrymandering (because there is), but that even if we let a computer decide representation, there's still more than one algorithm that is unaware of party affiliation, and each will favor different ideologies in different situations.

The moment I finished my comment I actually thought to myself - "hmmm, FPTP is the problem here" (which I was saying even well before the last election).

The gerrymandering is interesting though, and there's another video link in this thread that explains how you can use it to try to fix this issue (as you said).

> UK election outcomes have nothing to do with gerrymandering and everything to do with

> being a multi-party democracy using a FPTP electoral system, which tends to boost

> the representation of leading parties at the expense of smaller parties.

More precisely, FPTP tends to boost the representation of leading national _and local_ parties (the latter being best exemplified by the SNP) at the expense of smaller national parties. Small local parties are usually toast anyway even in proportional systems due to thresholds.

During the 2015 elections the system worked exactly as designed, by favoring parties that are strong in one or more nations of the UK at the expense of parties that are a bit meh in all of them.

The entire "Politics in the Animal Kingdom" series is great, I think... http://www.cgpgrey.com/politics-in-the-animal-kingdom/

Good video. Thanks!

My experience is that people go on about it all across the spectrum, including lots of people who not only don't think George W. Bush did steal the 2000 election, but who would have thought it was a good and justified thing if he had.

Gerrymandering is a real issue and a real thing that really effects election results, and, simultaneously, an effective distraction from the more serious problem of representation caused by FPTP in single-member districts. So, not only is it a popular target of the disaffected, its something the disaffected are redirected to vent their anger at by both sides of the partisan duopoly in different times and places, because while it is a real thing, none of the methods used to attack it fundamentally change the structures reinforcing the partisan duopoly, and in any given condition certain of the methods for fighting it will reinforce the position of one side or the other of that duopoly against the other within the basic duopoly-reinforcing structure.

Your response only makes sense if people identify on a state level.

Do voters in Ithaca, NY feel like they got attention when candidates visit New York City? How about Spokane, WA vs Seattle, WA?

When the US could tightly tie state views to voters, the electoral college MIGHT have made some sense, but now the issues are regional at best (coal belt, bible belt, etc), and not based on states. Now we're allowing a minority of the country to disproportionally influence elections (and legislation via the Senate).

To use your counterexample - Why stop in Boston indeed if you can hit Columbus? What difference does that make? I'm less worried about candidates _visiting_ places (because what good does that do?) and more about them being accountable to voters.

You are right about the consequences, but there are consequences to the system now.

> To use your examples of Massachusetts and Ohio, why waste time on a campaign stop in Boston when you can hit Columbus? Columbus has a much larger population and one can also hit Detroit and Indianapolis in the same day to maximize their face time with the public.

As opposed to the current system, where...why waste time on a campaign stop in Boston when you can hit Columbus? Ohio is a swing state and MA has voted Democratic for decades.

The point of a democracy is to represent actual human beings, not shapes on a map.

This logic confounds me. Is there some inherent reason to disproportionately focus our national choices on the areas of the country with the fewest people (rural areas)?

I've lived in both the most rural and urban parts of the US, and it's clear that the current system massively discounts the will of people in urban areas in this stated endeavor of providing some kind of rural / urban "balance".

Depending on what you consider urban, the US is roughly split 80% urban, 20% rural.

To me, it seems unlikely that a ground effort in rural America would be completely neglected in a proportional system. If a party did neglect the ~20% of rural voters, it would seem to put itself at a disadvantage to other parties, especially given the lower relative cost of rural media markets.

We see the same problem with the Senate, where populous states such as Texas and California stand at a remarkable disadvantage to Montana and Wyoming in terms of representation. The voice of a constituent in Wyoming is 66 times that of one in California.

How can we call this system representative?

That might have been the rationale when it was originally introduced, but for the most part it has the opposite effect now, apart from a few very select "swing" states. Convince ~50k voters in Florida to switch allegiance and you win the state, and probably win the electoral college. Convince ~150k voters in Arkansas - a large proportion of that state's electorate - to switch allegiance and (i)it wouldn't affect who wins the state and (ii)it probably wouldn't affect who becomes President even if it turned blue.

Just because it means that local concerns in a couple of large states aren't going to be much of a factor in the presidency (Texas is more likely to freeze over than vote Democrat in the next election) doesn't mean that very many people in smaller states benefit from the electoral system.

There's the Senate to ensure that smaller states get their say in national politics.

maybe we should just set up states. half blue, half red. each get to do what they want, and everyone blue or red moves to blue or red state. We come back in 50 years, and see what's up. The blue states are spending gobs on tuition and everyone is still stupid, and the red states, I don't know someone help. I'm colorblind

It's not a great picture. For one thing, it's unnecessarily provocative (Title: "How to Steal an Election") and not-so-subtly partisan: The use of blue for the 'natural' winner and red for the 'thief' is clearly meant to echo (and comment on) contemporary U.S. politics.

Yes, gerrymandering happens, and yes, it currently disproportionately benefits the GOP: Some lucky wave election timing put them in charge of more states during the post-2000 and post-2010 Census redistricting. But the benefit they get over their popular vote share is on par with the benefit the Democrats saw in the '90's[1].

And the benefit isn't that great. It's good for maybe 8 seats[2], far short of the 30 seats Democrats would need to swing to regain the majority.

The problem for Democrats in the House is that their districts are too natural: It makes sense to put urban voters with similar needs and issues into Congressional districts with each other, rather than diluting their influence with very different suburban areas. The Voting Rights Act even made concentrating minority votes to create majority-minority districts mandatory in some areas. But the result is winning Chicago wards 95-5 while losing out in the 'burbs 55-45. To have a chance at consistently winning the House, Democrats need to draw districts that look an awful lot like the 5-0 image in that picture. Practically and politically, that's a non-starter. That's why ITT and elsewhere you see calls for multicandidate districts, proportional statewide representation, etc. It's not to "fix" gerrymandering, it's to fix the "problem" of concentrated Democratic votes.

[1] http://www.centerforpolitics.org/crystalball/articles/gerrym...

[2] http://www.nytimes.com/2014/09/07/upshot/why-democrats-cant-...

This leads to a situation whereby elected representatives are assigned local constituencies, rather than having any particular link. Or atleast that's the argument which is always returned when people suggest a full PR system.

Personally I think larger multi-member constituencies under STV provide the best compromise.