Thanks for that. Very interesting. I think that with the fact that it doesn’t really impact partisan advantage, on the surface it seems reasonable to give the latinos a voice.
But back to Oregon, the boundaries do serve to gain partisan advantage, In all districts but #2, where they cede it to R’s, all the other districts pull in major population centers with rural areas.
The representation shows the results. 20% more D’s than R’s - and been lagely in D control since 1993 when they redrew the district boundaries. It fits the definition of what gerrymandering is.
Yes, because districts are required to have roughly the same number of constituents.
So to take your suggestion, if you wanted to make a single district including the entire west coast, then you would also have to include an urban center like Eugene. There simply aren’t enough people to form a coastal district otherwise.
Look at Oregon’s 2nd district, which includes more than half the state by area. That’s how big a district has to be if it has no urban areas. There isn’t enough room in Oregon for another one.
The math doesn’t work. I just did some calculations. If you take all of the costal counties, including the entirety of Lane and Douglas which means including all of eugene breaking your commitment to keeping urban and rural groups separate for voting, you only get ~670,000 people. For example Tillamook county has about 24,000 people, Clastrop 39,000 and is the most populous one not including significant interior cities. The state of Oregon has 5 districts and a population of 4.2 million, which means that it needs to average about 840,000 popped per district.
If you add in Josephine and Benton counties you get fairly close to the 840,000 people per average district. But the problem is that almost 40% of those people are from Eugene and the surrounding towns.
I’ve looked more at the numbers, because numbers are inherently interesting to me, and it is not possible to construct two rural districts in Oregon. The population is just too highly concentrated to a few major population centers. Outside of the few cities and major towns most counties have populations that are sub 50k. The entire population of people living west of the coastal range in Oregon is right around 200,000. And that is a generous estimate.
Oregon is not gerrymandered. Oregon is suffering from the same thing seen in places like Kansas and Nebraska. Both are about 60-40 GOP, but have 3-0 and 3-1 splits respectively. The very nature of small numbers of districts means that disproportionate representation is inevitable. Drawing districts that closely match overall state trends is difficult. And though they both disproportionately favor the GOP, I can’t honestly tell you they are gerrymandered. They genuinely are drawn in reasonable ways in compact districts. They don’t play games like splitting their urban centers across multiple rural districts a la Texas. So even though they both have one or two close districts, the math makes uneven representation likely.
Which is why, along with ranked choice, I am a big fan of multi seat districts. Have larger geographic areas with three representatives. It makes gerrymandering impossible, because a 60-40 split can’t be massaged into 3 safe seats. It would create one 2-1 district.
It doesn’t fit. This is actually just the split would empirically expect to get with fair districts in a two-party first-past-the-post system with fair districts.
The latest election had Democrats get 57.45% of the votes and Republicans 38.02%. If you plug those numbers into the cubic approximation formula you get (57.45/(57.45+38.02))^3/(38.02/(57.45+38.02))^3 =~ 3.45. So the expectation is that the majority party would be getting 3.45x as many seats as the minority party. For a state with five seats, that means an expected split of 3.88 seats for the majority party, 1.12 seats for the minority. Since you can’t have fractional seats, 4:1 is very fair.
It’s pretty gross to try to equate Oregon with something like Ohio or North Carolina.
Yes, you might need to come east a bit, but stay away from Salem, Hillsboro & Portland. Take a look at the 4th - it’s half the coast already. Combine Roseburg, Medford, Grants Pass + Ashland and it’s doable without Eugene (you might need Corvalis). Then you can give Eugene back to the 2nd in return.
Really, the state should be doing more, but they don’t want to because they’re happy with all the D representation. We can’t be satisfied if the rural areas are so under-represented, and it’s just going to get worse as Portland grows.
Wait, so your solution to the “gerrymandering” of Oregon is to make a district that snakes from the NW corner to the SW corner, studiously avoiding an urban center in between, and then cutting sharply east along the California border? And then taking a city that lies 60 miles from the ocean and packing it into the eastern half of the state?
Tman, your problem isn’t that Oregon is gerrymandered. Your problem is that it’s not gerrymandered enough.
Unfortunately it empirically is not. Eugene and the surrounding cities are about 290k of the 840k in my hypothetical coastal district.
The other problem is that Medford and Grants Pass are a significant portion of the 2nd district. So take those areas and you need to find another 150k or so people to add to the second district, which you posit to be adding Eugene. But this moves the problem, additionally Eugene and nearby are larger than the Medford Grants Pass area by a significant margin. And my hypothetical district already included Corvallis to get to 840.
No matter how you slice it, you are adding a massive city to one of the rural districts. In our current map that is Eugene into a southern Oregon district. You are adding Eugene to the huge eastern district. How is that better? And how is that also not worse? Either Eugene runs roughshod over the eastern half of the state, as Eugene and surrounding would make up about 40% of the district population, or basically condemn them to irrelevance by pairing them with eastern ranchers? Is not the current district, which is at least geographically close, better?
No matter what you can not make two majority rural districts in Oregon, the population distribution doesn’t work. The larger the state, the more populous it is, the more possible such notions are. But you basically can’t make that work until you get at least 8-10 districts with first past the post. Which is the root of the trouble. FPTP doesn’t provide even results, it skews very heavily and at increasing speed, the further from 50/50 you get. Unless you are working with very large number sets (50+) or deliberate result manipulation, then that’s what you get. A 60/40 split giving 75/25 representation is mathematically consistent with how these distributions work.
This is a a pretty cool tool to look at gerrymandering. Oregon is lightly gerrymandered compared to other partisan states, WI has 1 competitive district currently, Oregon has at least 2.
That’s a cool tool. And it does drive home one thing, that Oregon’s districts are fairly reasonable. The partisan mix has two safe D, one safe GOP, and two districts that are at 52 and 54%. And anything other than the deliberate partisan map resulted in something fairly close to what exists. That there is a 4-1 split is reflective of what happens with close districts. Small changes in partisan participation make large swings.
Now could you make a map that is more likely to result in a 3-2 split? Sure. Mostly involving moving more southwest Oregon into 2, trading Hood county into Portland.
But the map is pretty good considering.
It ain’t got boo on Pennsylvania, North Carolina, Texas, or even California (Which is hard to draw a pure proportional map that isn’t prone to massive overrepresentation in a wave election like 2018. The proportional map would be +3 dem over the current map even, due to 2018)
Which is kind of the inverse of the Oregon situation. The percentage of state population concentrated in urban areas is so high that there simply aren’t enough on non urban voters to have rural voters matter outside the second district.
Oh no, a horrific form of Gerrymandering I had never even heard of. Prison populations being counted in the prison district for census purposes,e.g. to swell (R.) districts, but of course not having local voting rights. This tweet thread makes me feel sick.
There’s injustice here, but that’s only half the story. The Census is used for more than just determining how to allocate districts - though that is a very direct result. It also provides information to the Federal, State and County governments on how to equitably distribute funding for everything from roads to sewage to Medicare. That’s why this whole citizenship question thing is so onerous.
So if you shifted the census results so that the prison population was counted against their previous home (or place of arrest or whatever), you’d be short-changing the communities that surround the prisons and which would otherwise struggle to provide services for them. Again, this has echoes with the undocumented/non-citizen community being counted or not.
And note that disenfranchising felons is a state-by-state decision. Felons in Maine or Vermont can still vote, so you’d hope that they would be counted in the Census for the district that contains the prison. Felons in Mississippi may never vote again, so should they be counted as residents of the district that they were arrested in, or where they currently live?
Finally, the elected official should be representing the people of his or her district. The prison is in their district, so they ought to be looking out for the well-being of those people. Sure, there’s not a lot of incentive for them to do that if the incarcerated people cannot vote, but there isn’t a lot more incentive for some representative several states away to be concerned about that person either.
