Quo Vadis GOP?

[DGuller]

It's not apparent to me what percentage of districts *should* be controlled by groups that constitute 39 and 12% of the population.

More than 18%, for a start.

We don't know that.  It depends on how the 39% group is clustered.  If every single household has 39% Hispanic people, then you can't have more than 0% of districts controlled by Hispanics.

Yes, I suppose one could choose to make up unrealistic scenarios.

I'm not choosing to make up unrealistic scenarios, I'm making a point by going to the extreme.  That's a very common technique when making mathematical arguments. 

What Yi was getting at is that you can't actually infer what the fair numbers would be for the second column.  The quiet implication is that the second column numbers should match the first column, which is a subtle way of lying by statistics.  As I said countless times before, it's possible to support a valid point with invalid arguments, explicit or implicit.

[The Larch]

There's no need to get into bizarre theoretical arguments when what happened is known. From the article I quoted:

Minority voters accounted for 95% of population growth in Texas over the last decade but there are no new majority-minority districts in the new plans. Texas gained two new seats in Congress because of its high population growth over the last decade.

Republicans who control the redistricting process drew the lines to shore up their advantage across Texas, blunting the surge in the state's non-white population. The new maps give Republicans a hold on 25 of Texas's 38 congressional seats and help them maintain their majorities in the state legislature.

(...)

The suit, filed in federal court in El Paso, claims Republicans "deliberately" reconfigured the 23rd congressional district in west Texas, where Latinos currently have a fair opportunity to elect the candidate of their choice, to make it more white and Republican.

Republicans also "surgically excised minority communities" from the Dallas-Fort Worth area and attached them to whiter, rural areas that vote Republican, where the power of their ballot would be diminished. Courts have repeatedly found Republicans undertook similar efforts in the same part of Texas during previous redistricting cycles.

Lawmakers also failed to draw districts that accounted for the growing Latino electorate in Harris county, home of Houston.

[DGuller]

There's no need to get into bizarre theoretical arguments when what happened is known.

What's so bizarre about pointing out a fallacy?  Yes, we all know what happened in Texas, that doesn't mean that we should abdicate our critical thinking.  If an argument doesn't follow, it doesn't follow, even if it supports something that's obviously true anyway.

[Valmy]

Regardless of such percentages, it does create a field where it makes it significantly less important to cater to non-white voters, even though they're in the majority.

Well the idea is to keep and expand Republican control. Any other things that result from that are just side-effects.

[garbon]

It's not apparent to me what percentage of districts *should* be controlled by groups that constitute 39 and 12% of the population.

More than 18%, for a start.

We don't know that.  It depends on how the 39% group is clustered.  If every single household has 39% Hispanic people, then you can't have more than 0% of districts controlled by Hispanics.

Yes, I suppose one could choose to make up unrealistic scenarios.

I'm not choosing to make up unrealistic scenarios, I'm making a point by going to the extreme.  That's a very common technique when making mathematical arguments. 

What Yi was getting at is that you can't actually infer what the fair numbers would be for the second column.  The quiet implication is that the second column numbers should match the first column, which is a subtle way of lying by statistics.  As I said countless times before, it's possible to support a valid point with invalid arguments, explicit or implicit.

Yes I got that you were making a commentary about what you can and cannot infer from the numbers Yi had reacted to. As Larch said seemed an unnecessary exercise given what he had posted.

I do understand thought why numerical literacy is a bugbear for you.

[DGuller]

Yes I got that you were making a commentary about what you can and cannot infer from the numbers Yi had reacted to. As Larch said seemed an unnecessary exercise given what he had posted.

I do understand thought why numerical literacy is a bugbear for you.

I don't think it's unnecessary.  A bad argument in support of a good point can undermine the point.  Yes, logically speaking, if you make 999 bad arguments in support of a point, and one ironclad good argument, the 999 bad arguments shouldn't matter as long as the one good argument puts away the matter for good. 

Realistically speaking, however, when you make bad arguments, you risk losing credibility, and you also give a convenient rebuttal opportunity.  Your opponents can focus on disproving your bad arguments, ignore the good arguments, and then subtly imply that all your arguments are bad become they showed some of them to be bad.

[Berkut]

Do we want to win arguments or elections?

If we want to win elections, it behooves us to avoid bad arguments ourselves that distract from the point.

[Josquius]

Its definitely true that just seeing a statistic that 12% of the state are black (incidentally I'm surprised at that, I have the mental image of Texas being very white/hispanic) doesn't necessarily equate to 12% of the seats should be black. In a perfectly integrated land this 12% would be scattered all around the state and probably be a minority everywhere.

....but I somehow doubt this is the reality anywhere in the US. Black communities do tend to cluster.

[The Minsky Moment]

I'm not choosing to make up unrealistic scenarios, I'm making a point by going to the extreme.

No you are attempting to make a point by postulating a factual scenario known to be false.  That's just sophistry.

We know for a fact that racial identifications are not evenly distributed across households.  In fact there the opposite tend to be true - housing patterns continue to show significant racial concentration and clustering.  Measures of black-white concentration have increased in Texas over the past decade; Hispanic concentration has remained steady at high levels.

An argument - even if posed in the form of a mathematical gedanken experiment - is a bad argument if premised on false underlying factual premises.

[garbon]

Do we want to win arguments or elections?

If we want to win elections, it behooves us to avoid bad arguments ourselves that distract from the point.

Does it? The Republicans make bad arguments all the time and that doesn't seem to matter.

[Berkut]

Do we want to win arguments or elections?

If we want to win elections, it behooves us to avoid bad arguments ourselves that distract from the point.

Does it? The Republicans make bad arguments all the time and that doesn't seem to matter.

That's because they have baked in stupid and crazy into their brand. We don't have that luxury, nor should we want it.

It isn't a fair playing field. Pointing out that Republicans are shitbags doesn't mean we can be shitbags. They've rigged the deck, and we have to unrig it, or at least stop them from rigging it even more.

[The Minsky Moment]

It's not apparent to me what percentage of districts *should* be controlled by groups that constitute 39 and 12% of the population.

Framing the issue as an entitlement of a certain percentage of racial seats by racial identification is less than ideal.

The broader point holds though - this is a blatant attempt to dilute the voting power of the affected demographic groups.  There is a reason why the Texas GOP districting map has districts that turn into pretzels put through an industrial tumble dryer the moment one nears an urban area and it isn't because they are devotees of modern art.

[DGuller]

No you are attempting to make a point by postulating a factual scenario known to be false.  That's just sophistry.

I am exploring the limiting conditions to show why the implied argument is false.  That's not sophistry, and I'm pretty sure that you know it and are frankly being dishonest about it.  That's how math works, and if you're going to present arguments that contain numbers, then unfortunately you have to understand how math works.  My point was that we don't know how the minorities are clustered, and that for extreme levels of that unknown variable the outcome may be far more unintuitive than first believed. 

Yes, obviously every household can't be 39% Hispanic, just due to integer constraints if nothing else, I thought it would be too obvious to mention.  What we don't know is how far from ideally dispersed minorities are for the purpose of this number.  How far do you have to go from that ideal dispersion before the 18% figure is actually fair?  I don't know, and it's impossible to know without very detailed information.

[The Minsky Moment]

My point was that we don't know how the minorities are clustered.

Really, you don't know that?  Perhaps you should inform yourself before accusing others. There is plenty of data out there.

Otherwise you risk making argument along the lines of - if we assume that every household in Texas consists entirely of 11 year old girls, then the redistricting will have no effect because there are no eligible voters in the state.

[DGuller]

My point was that we don't know how the minorities are clustered.

Really, you don't know that?  Perhaps you should inform yourself before accusing others. There is plenty of data out there.

Otherwise you risk making argument along the lines of - if we assume that every household in Texas consists entirely of 11 year old girls, then the redistricting will have no effect because there are no eligible voters in the state.

What we don't know is how far from ideally dispersed minorities are for the purpose of this number.

We don't know whether minorities are dispersed enough for an 18% number of be unfair.  I do know that they are dispersed, but I don't know whether that dispersion would imply 5% or 35% to be the fair number, and neither does Zoupa or you.