Political Randomness

[Daehawk]

He doesn't have to concede. It wont change anything. Just makes him look like more like a dick than he already does.

[Kraken]

And yes, the #MeToo thing is veering into that territory.

How so? I'm willing to entertain the idea, but I'm following along for the most part and have not seen what you describe. Please elaborate.

Plus, the idea that this politician doesn't want to meet kids because some time in the future she might be ruined is ridiculous on the face of it. First, there is no expectation that she be alone with anyone. Presumably there would be additional adults chaperoning and organizing, including her own people. Second, and this is key, don't try to get into any of their pants and you won't have your career ruined years later.

I was extrapolating from pedophilia accusations to the whole pussy-grabber zeitgeist, and that's a false equivalency. There is a bandwagon effect with that that reminds me of the old daycare-center abuse scare, or the '70s-vintage all-men-are-potential-rapists paradigm, but it's not relevant to the case at hand.

[Scoop20906]

He doesn't have to concede. It wont change anything. Just makes him look like more like a dick than he already does.

Shadow Senator?

[pr0ner]

So election night in Virginia left the VA House of Delegates with a 51-49 GOP advantage (after a 67-33 advantage at the start of the night), with a few recounts pending. After one of those recounts, the democratic candidate is now the winner with a margin of literally one vote.

So as of now, it looks like it'll be 50-50 and the parties will have to figure out a power-sharing arrangement. There are a few more recounts, but while the margin in thin it looks like the GOP candidate is likely to hold on in each of them. There is a chance for democrats to win one more seat - basically in one of the races (where the GOP candidate leads by something like 83 votes), several hundred voters were routed to the wrong precinct by an election official. So democrats are likely to petition in court for a redo of that election - I assume that's unlikely, but who knows.

FYI, that race that was originally decided by one vote is now a tie, and will be decided by a drawing of lots.

[pr0ner]

The estimate in the Virginia elections is that about 200k more Democrat votes were cast than Republican votes. And yet the state is so badly gerrymandered that with all that voter support Democrats still only managed to eke out a tie. It's truly sickening.

To be fair, a) the state isn't *that* badly gerrymandered (the last attempt to have the district map ruled unconstitutional was shot down in the courts), and b) the Democratic party had a HUGE advantage in votes cast in unopposed races (ie races where there was only one major party candidate) by the tune of almost 344,000 votes. Republicans as a whole statewide collected more votes than Democrats in the contested races.

[Paingod]

Would it be possible to make a law that says a voting district cannot have more than, say, 8 sides to its shape? Let them redistrict, but don't let them create a fiddly tiny line that slips here and there to silence opposition.

[Holman]

He doesn't have to concede. It wont change anything. Just makes him look like more like a dick than he already does.

Shadow Senator?

If Moore were a normal politician, I'd assume he was positioning himself for a firebreathing outsider run for Guvner, especially since the Alabama governorship is still shakey from Bentley's resignation.

But he's even more of a religious fanatic than an opportunist, and I'm about 75% sure that he believes God will somehow reverse this election. It's a grand test of faith for him and his followers, but the Lord will see them safely to victory in the end.

[Max Peck]

The estimate in the Virginia elections is that about 200k more Democrat votes were cast than Republican votes. And yet the state is so badly gerrymandered that with all that voter support Democrats still only managed to eke out a tie. It's truly sickening.

To be fair, a) the state isn't *that* badly gerrymandered (the last attempt to have the district map ruled unconstitutional was shot down in the courts), and b) the Democratic party had a HUGE advantage in votes cast in unopposed races (ie races where there was only one major party candidate) by the tune of almost 344,000 votes. Republicans as a whole statewide collected more votes than Democrats in the contested races.

Having a few districts where opposition support is greatly concentrated and a larger number where the governing party has a substantial advantage in support is pretty much textbook gerrymandering. It's just that gerrymandering alone won't work when there's enough of a shift in support across the board to overcome the baked-in advantage.

[ImLawBoy]

Would it be possible to make a law that says a voting district cannot have more than, say, 8 sides to its shape? Let them redistrict, but don't let them create a fiddly tiny line that slips here and there to silence opposition.

Seems arbitrary, and causes problems when a district might logically follow a physical boundary such as a river or a street that winds a bit.

[El Guapo]

Intuitively it seems like the fairest way would be to agree on some neutral factors (contiguity, compactness, etc.), and then hire someone to put that into an algorithm.

[pr0ner]

The estimate in the Virginia elections is that about 200k more Democrat votes were cast than Republican votes. And yet the state is so badly gerrymandered that with all that voter support Democrats still only managed to eke out a tie. It's truly sickening.

To be fair, a) the state isn't *that* badly gerrymandered (the last attempt to have the district map ruled unconstitutional was shot down in the courts), and b) the Democratic party had a HUGE advantage in votes cast in unopposed races (ie races where there was only one major party candidate) by the tune of almost 344,000 votes. Republicans as a whole statewide collected more votes than Democrats in the contested races.

Having a few districts where opposition support is greatly concentrated and a larger number where the governing party has a substantial advantage in support is pretty much textbook gerrymandering. It's just that gerrymandering alone won't work when there's enough of a shift in support across the board to overcome the baked-in advantage.

So your default assumption with the results is just to say gerrymandered districts without taking into account things like voter turnout, or how Virginia's political divisions exist geographically?

For instance, my delegate, running unopposed, collected more votes by himself than many contested races did. He outgained that well publicized race between Roem and Marshall by 9,000 votes. If voters are going to turn out in heavily Democratic areas but not in other areas, how can gerrymandering be the default answer for how Virginia is in a virtual tie with so many more Democratic votes? There are more factors at play here.

[Defiant]

Intuitively it seems like the fairest way would be to agree on some neutral factors (contiguity, compactness, etc.), and then hire someone to put that into an algorithm.

There have been various algorithms proposed, although I don't know if there's ever going to be a perfect algorithm (I believe the problem is NP-hard). I would be reasonably satisfied if there were some way of measuring gerrymandered-ness, allow humans to come up with the maps, but reject them if any of the districts go over a certain value. If no one (on either side of the aisle) can come up with a map that fits, slowly raise the value until they do.

[El Guapo]

The estimate in the Virginia elections is that about 200k more Democrat votes were cast than Republican votes. And yet the state is so badly gerrymandered that with all that voter support Democrats still only managed to eke out a tie. It's truly sickening.

To be fair, a) the state isn't *that* badly gerrymandered (the last attempt to have the district map ruled unconstitutional was shot down in the courts), and b) the Democratic party had a HUGE advantage in votes cast in unopposed races (ie races where there was only one major party candidate) by the tune of almost 344,000 votes. Republicans as a whole statewide collected more votes than Democrats in the contested races.

Having a few districts where opposition support is greatly concentrated and a larger number where the governing party has a substantial advantage in support is pretty much textbook gerrymandering. It's just that gerrymandering alone won't work when there's enough of a shift in support across the board to overcome the baked-in advantage.

So your default assumption with the results is just to say gerrymandered districts without taking into account things like voter turnout, or how Virginia's political divisions exist geographically?

For instance, my delegate, running unopposed, collected more votes by himself than many contested races did. He outgained that well publicized race between Roem and Marshall by 9,000 votes. If voters are going to turn out in heavily Democratic areas but not in other areas, how can gerrymandering be the default answer for how Virginia is in a virtual tie with so many more Democratic votes? There are more factors at play here.

The thing is that "a lot of the excess votes came in unopposed races" is not inconsistent with gerrymandering (which is what I think Max was getting at). I'd have to look at the districts / history to know for sure, but there's a pretty good chance that so said democrats ran opposed *because* their districts were so gerrymandered that no GOP candidate would stand a chance (because the map packed a lot of democrats into a few districts).

[pr0ner]

I'm not so sure about that. Northern Virginia is heavily blue. The districts up here are very compact. How are you going to convince a Republican running here is worth the effort?

[Remus West]

I'm not so sure about that. Northern Virginia is heavily blue. The districts up here are very compact. How are you going to convince a Republican running here is worth the effort?

That is sort of the point. Compact all the minority power in one place while making sure you have enough other locations firmly in your grasp to control the elected body. Thats what gerrymandering is all about. If you can not break an area into something you can win you either break it apart so that it can no longer vote as a block and gets swallowed by areas you do control or you minimize the impact of losing that area by making sure the rest of the territory is under your control.

Northern Virginia has been ceded to the Democrats in order to be certain the rest of the state will reliably keep the legislature red.

[Max Peck]

The estimate in the Virginia elections is that about 200k more Democrat votes were cast than Republican votes. And yet the state is so badly gerrymandered that with all that voter support Democrats still only managed to eke out a tie. It's truly sickening.

To be fair, a) the state isn't *that* badly gerrymandered (the last attempt to have the district map ruled unconstitutional was shot down in the courts), and b) the Democratic party had a HUGE advantage in votes cast in unopposed races (ie races where there was only one major party candidate) by the tune of almost 344,000 votes. Republicans as a whole statewide collected more votes than Democrats in the contested races.

Having a few districts where opposition support is greatly concentrated and a larger number where the governing party has a substantial advantage in support is pretty much textbook gerrymandering. It's just that gerrymandering alone won't work when there's enough of a shift in support across the board to overcome the baked-in advantage.

So your default assumption with the results is just to say gerrymandered districts without taking into account things like voter turnout, or how Virginia's political divisions exist geographically?

For instance, my delegate, running unopposed, collected more votes by himself than many contested races did. He outgained that well publicized race between Roem and Marshall by 9,000 votes. If voters are going to turn out in heavily Democratic areas but not in other areas, how can gerrymandering be the default answer for how Virginia is in a virtual tie with so many more Democratic votes? There are more factors at play here.

I'm just saying that none of the factoids you presented preclude the existence of gerrymandered districts.

[pr0ner]

It's not accurate to say the vote played out the way it did simply because of gerrymandering, either.

[Defiant]

I'm only skimming through posts, but since people seem to be touching on this point...

I don't think gerrymandering should be defined by how much one party or another has a major advantage in a district, I think it should be defined by the shape and compactness of districts and how well they fit in with the population densities, geographic and topographical nature of the area. Otherwise the solution would be to create districts that had as close to a 50/50 split as possible and that would introduce more problems and would be harder to address.

[LordMortis]

https://www.dailykos.com/stories/2016/1 ... heir-stats

Doesn't look too bad to me, without knowing much about that big pale red strip in the middle. That pale red strip could be very telling if you could see how the cities/counties within vote differently.

Of course, when I look at our own districts, VA might not seem bad because what we have is absolutely shameful.

[GreenGoo]

How do the votes normally play out in previous elections?

[Max Peck]

It's not accurate to say the vote played out the way it did simply because of gerrymandering, either.

Did anyone say that it did?

[pr0ner]

It's not accurate to say the vote played out the way it did simply because of gerrymandering, either.

Did anyone say that it did?

Gilraen appeared to.

[pr0ner]

https://www.dailykos.com/stories/2016/1 ... heir-stats

Doesn't look too bad to me, without knowing much about that big pale red strip in the middle. That pale red strip could be very telling if you could see how the cities/counties within vote differently.

Of course, when I look at our own districts, VA might not seem bad because what we have is absolutely shameful.

That map is Virginia's Congressional districts, not their House of Delegates districts (which I posted in thread).

[GreenGoo]

I was extrapolating from pedophilia accusations to the whole pussy-grabber zeitgeist, and that's a false equivalency. There is a bandwagon effect with that that reminds me of the old daycare-center abuse scare, or the '70s-vintage all-men-are-potential-rapists paradigm, but it's not relevant to the case at hand.

We could get there, and there have been a couple of questionable accusations, but so far, despite the sheer bulk of the accusations, they seem to be factual. And remember these are accusations that cover decades of time, they are only coming out now, all at once, because the climate seems to be more supportive to the victims than it would normally be. Not to mention the media is more likely to print them today.

This has all resulted from the NYT's doing their exposé on Weinstein, I think. Once the dam springs a leak, no amount of Israeli former special ops guys can plug it, it seems.

[Enough]

Roy Moore continues to be an absolute ass. He's now going after Doug Jone's gay son. Which is a pretty odd look for a guy that has a son who has been arrested nine times.

[Smoove_B]

Roy Moore continues to be an absolute ass. He's now going after Doug Jone's gay son.

Which, to be fair, is exactly what Jesus would do. I think...my Bible verses are a bit rusty.

[Max Peck]

It's not accurate to say the vote played out the way it did simply because of gerrymandering, either.

Did anyone say that it did?

Gilraen appeared to.

I took her comment at face value as meaning that she believed gerrymandering played a significant role in the disconnect between the popular vote and the composition of the legislature. A cursory search for articles relating to Virginia and gerrymandering seems to indicate that this is not exactly an uncommon view.

For example: What Virginia tells us, and doesn’t tell us, about gerrymandering

Although the Virginia governorship was Tuesday’s marquee race, the Virginia House of Delegates produced the day’s most surprising result. Democrats picked up at least 15 seats and reduced a 66 to 34 Republican advantage to, at most, 51 to 49. A gerrymandered chamber thought to be safely Republican suddenly became a toss-up — and may yet flip to Democratic control after all the recounts are completed.

This unexpected outcome raises the question: Can gerrymandering really be such a problem if a party’s legislative edge can virtually disappear overnight? This question is especially important at present, as the Supreme Court mulls over Gill vs. Whitford, a potentially historic case about redistricting in Wisconsin.

The question also has a clear answer: Of course gerrymandering is deeply troublesome even if it can be overcome, at least temporarily, by a wave election.

Consider the following facts about the Virginia House of Delegates: In three previous elections (2011, 2013 and 2015), Republicans won 66 or 67 out of 100 seats. Republicans maintained this supermajority even though Democrats narrowly won every statewide race over this period. To secure (roughly) half of the House seats on Tuesday, Democrats had to earn well over 50% of the statewide House vote. This was Democrats’ best showing in more than 30 years. Had Republicans done as well, they would have won far more than 50 seats: close to 70, in fact.

[The Meal]

I don't think gerrymandering should be defined by how much one party or another has a major advantage in a district, I think it should be defined by the shape and compactness of districts and how well they fit in with the population densities, geographic and topographical nature of the area. Otherwise the solution would be to create districts that had as close to a 50/50 split as possible and that would introduce more problems and would be harder to address.

538's Gerrymandering Project.

I'm looking for (but unable to find) the specific article (not podcast) where they go into details (walk us through the algebra), but it tends to boil down to:

The metric at the heart of the Wisconsin case is called the efficiency gap. To calculate it, you take the difference between each party’s “wasted” votes — votes for losing candidates and votes for winning candidates beyond what the candidate needed to win — and divide that by the total number of votes cast. It’s mathematical, yes, but quite simple, and aims to measure the extent of partisan gerrymandering.

I, too, want more than just this efficiency gap (including geographic effects). But this turns out to be a decent mathy start.

[El Guapo]

I don't think gerrymandering should be defined by how much one party or another has a major advantage in a district, I think it should be defined by the shape and compactness of districts and how well they fit in with the population densities, geographic and topographical nature of the area. Otherwise the solution would be to create districts that had as close to a 50/50 split as possible and that would introduce more problems and would be harder to address.

538's Gerrymandering Project.

I'm looking for (but unable to find) the specific article (not podcast) where they go into details (walk us through the algebra), but it tends to boil down to:

The metric at the heart of the Wisconsin case is called the efficiency gap. To calculate it, you take the difference between each party’s “wasted” votes — votes for losing candidates and votes for winning candidates beyond what the candidate needed to win — and divide that by the total number of votes cast. It’s mathematical, yes, but quite simple, and aims to measure the extent of partisan gerrymandering.

I, too, want more than just this efficiency gap (including geographic effects). But this turns out to be a decent mathy start.

Bear in mind (as they get into in the Gerrymandering Project), an analysis doesn't stop with an efficiency gap calculation. The legal test that the plaintiffs are proposing in the Wisconsin case also incorporates evidence of intent in districting (of which there is *plenty* in Wisconsin* and other possible explanations for the gerrymander (including geography).

Also you can do things like produce computer-generated maps and test them against the actual map. (Again from the Gerrymandering Project) a professor created a district generating algorithm and ran a number of simulations generating Wisconsin electoral maps. Most (using objective criteria) generated maps with efficiency scores around 0 (i.e. even between the parties), although a quarter wound up generating maps with a pro-Republican efficiency gap - which suggested that (even done objectively) geography in Wisconsin slightly favored Republicans in terms of where voters lived (but that one could readily generate even maps accounting for geography).

(Also none of the GOP-friendly maps generated came close to the efficiency gap results of the GOP legislature map).

[The Meal]

Yeah, I got pressed for time just about at the moment I started that post.

[Max Peck]

I don't think gerrymandering should be defined by how much one party or another has a major advantage in a district, I think it should be defined by the shape and compactness of districts and how well they fit in with the population densities, geographic and topographical nature of the area. Otherwise the solution would be to create districts that had as close to a 50/50 split as possible and that would introduce more problems and would be harder to address.

538's Gerrymandering Project.

I'm looking for (but unable to find) the specific article (not podcast) where they go into details (walk us through the algebra), but it tends to boil down to:

The metric at the heart of the Wisconsin case is called the efficiency gap. To calculate it, you take the difference between each party’s “wasted” votes — votes for losing candidates and votes for winning candidates beyond what the candidate needed to win — and divide that by the total number of votes cast. It’s mathematical, yes, but quite simple, and aims to measure the extent of partisan gerrymandering.

I, too, want more than just this efficiency gap (including geographic effects). But this turns out to be a decent mathy start.

538 references this document for details on the actual formulae.

Efficiency Gap Equation
As an equation, the efficiency gap looks like this:

• Efficiency Gap = (Total Democratic Wasted Votes –Total Republican Wasted Votes) ÷ Total Votes

Simplified Efficiency Gap Calculation
If either party’s seat margin and vote margin for a given election are known, then the efficiency gap can also be calculated using the following formula:

• Efficiency Gap = (Seat Margin – 50%) – 2 (Vote Margin – 50%)

[The Meal]

Yeah, I needed to walk through a few examples before I was on board with the math. Unfortunately, while I thought I could dig in and defend that process, I no longer can.

[Defiant]

Yeah, I'm not terribly impressed with the measurement (I mean, I'm fine with using it as additional proof that districts are gerrymandered if only to show the supreme court that this is a big problem).

Using it as a measurement, seats that are won with a 75%-25% are the least gerrymandered, while those that are 51-49 or 99-1 are the most gerrymandered. While I think it might be useful in detecting a pattern across all districts together (by summing it up), I don't think it does it well for individual districts. It would also be a poor way to design the districts from scratch (since we don't know how individual people vote. And also how do you deal with people who split their votes on election day?).

Also, if you're going by wasted votes, how do you differentiate between results caused by gerrymandered districts and those caused by strong candidates (or wave elections)?

I would much rather not use any knowledge about the people when coming up with maps, except where people are located and what communities/neighborhoods/town/city they are in (since we presumably want to group people together by that, when possible).

[El Guapo]

Yeah, I'm not terribly impressed with the measurement (I mean, I'm fine with using it as additional proof that districts are gerrymandered if only to show the supreme court that this is a big problem).

Using it as a measurement, seats that are won with a 75%-25% are the least gerrymandered, while those that are 51-49 or 99-1 are the most gerrymandered. While I think it might be useful in detecting a pattern across all districts together (by summing it up), I don't think it does it well for individual districts. It would also be a poor way to design the districts from scratch (since we don't know how individual people vote. And also how do you deal with people who split their votes on election day?).

Also, if you're going by wasted votes, how do you differentiate between results caused by gerrymandered districts and those caused by strong candidates (or wave elections)?

I would much rather not use any knowledge about the people when coming up with maps, except where people are located and what communities/neighborhoods/town/city they are in (since we presumably want to group people together by that, when possible).

It's not meant to measure individual districts - it's meant as a measure of evaluating the overall map. Which makes sense - the whole point of partisan gerrymandering is to maximize the number of legislative seats for your party across the state.

Similarly you measure wasted votes because that's the means of producing a map tilted in your favor - you can't eliminate opposing voters through redistricting (though you can do that through voter ID, potentially), so you have to set up the map so that the fewest number of opposing votes count.

[Defiant]

But suppose you have a map with five districts that all go to one party by a 51-49 margin. Under the above system, it would rate the efficency gap as (49*5)/500 = .49, the highest you can get. But, independent of any other information, I'd be skeptical that any of districts were gerrymandered.

[El Guapo]

But suppose you have a map with five districts that all go to one party by a 51-49 margin. Under the above system, it would rate the efficency gap as (49*5)/500 = .49, the highest you can get. But, independent of any other information, I'd be skeptical that any of districts were gerrymandered.

I suspect that we're misapplying the formula here, since that can't be the right answer, for the reason you say.

[Defiant]

But suppose you have a map with five districts that all go to one party by a 51-49 margin. Under the above system, it would rate the efficency gap as (49*5)/500 = .49, the highest you can get. But, independent of any other information, I'd be skeptical that any of districts were gerrymandered.

I suspect that we're misapplying the formula here, since that can't be the right answer, for the reason you say.

It's correct - I went through the explanation in the paper, and also used their "simplified" method (I don't think it is simpler, myself) which gave me (100%-50%) - 2(51%-50%) = 48%.

The thing is, it does make sense, because if you have 51-49% across 5 districts, you would expect one party to win about 3 of them, and the other to win about 2 of them (but, of course, potential wave elections would likely mean that the tilt would all go one way). This would be less of a factor over time, though. With enough election over time, such that things like wave election years wouldn't be as much of a factor, or you would tend to have enough wave elections in both directions to cancel each other out.

Also, a minor flaw is that it misses out voters (on either side) that don't vote because their district is noncompetitive. I don't think it's a big deal, but it could mess with the numbers somewhat.

[Max Peck]

That paper just explains how to calculate the efficiency gap, but doesn't explain how it's intended to be used.

This one (by one of the guys who dreamed up the efficiency gap formula) explains how and why it is intended to be used.

If insanity is doing the same thing over and over and expecting a different result, then litigants who challenge gerrymandering must be mad. Last month, a federal court threw out the Texas Democratic Party’s claim that the state’s new congressional and state house districts are unlawful. This was the twelfth time in a row that this sort of claim has failed in the current cycle. Plaintiffs’ record of futility now spans at least three dozen cases over four decades.

It doesn’t have to be this way. Litigants keep losing these lawsuits because they keep proposing standards the courts have already rejected (such as partisan intent). They’re failing to capitalize on encouraging comments by the Supreme Court, which show that it’s open to a test based on partisan symmetry—the idea that district plans should treat the parties equally. In a forthcoming law review article, Eric McGhee and I lay out just such a test. If plaintiffs were to use it in litigation, they’d have a fighting chance at winning. And if they were to win, then the whole landscape of redistricting in America would be transformed.

[Defiant]

That paper just explains how to calculate the efficiency gap, but doesn't explain how it's intended to be used.

No, I meant I read the scientific paper they wrote referred to in the above. Well, a couple of sections of it I read in depth, while I skimmed the rest, just like in school.

And while I get what they're doing, I still think geography makes for a better standard. Still, if they can do both in an attempt to show egregious gerrymandering (both show that it's geographically absurd and leads to lopsided results where many people's votes don't count) then so much the better.

[Max Peck]

That paper just explains how to calculate the efficiency gap, but doesn't explain how it's intended to be used.

No, I meant I read the scientific paper they wrote referred to in the above. Well, a couple of sections of it I read in depth, while I skimmed the rest, just like in school.

And while I get what they're doing, I still think geography makes for a better standard. Still, if they can do both in an attempt to show egregious gerrymandering (both show that it's geographically absurd and leads to lopsided results where many people's votes don't count) then so much the better.

My take is that the efficiency gap isn't intended to be used by itself, but rather as a quantitative metric alongside other evidence of gerrymandering.