About that 2012 Congressional map for Iowa

You doggone liberals are just too nice... (0.00 / 0)

I understand the need to clarify, but if you absolutely must speak to a conservative, please remember appropriate prophylaxis.

And I've tried, I really have, honestly, to get my mind around this whole redistricting thing, but I don't see how this can possibly jibe with my notion of an Iowa Autonomous Region.
First, get the U.S. out of Iowa, then negotiate tarriffs and treaties. Speculating on redistricting only encourages them.

I know, I know, most of your readers are liberal and dem, but not all progressives are liberal and/or dem.

---

by: EltonDavis @ Tue Mar 10, 2009 at 10:51:21 AM CDT

Krusty . . . (0.00 / 0)

I think was APPROVINGLY called your map gerrymandering . . .

---

by: phatkarp @ Tue Mar 10, 2009 at 10:57:41 AM CDT

Frankly, I'm tickled by all the controversy (4.00 / 2)

Let me first say that my numbers are merely 2007 county population estimates. Unfortunately, 2008 numbers are not yet available, and when they are, those lines are subject to change. It's all speculation at this point anyway! Newsflash: the actual Census is nearly 13 months away.

Second, my intention with all these redistricting diaries has not been to produce Democratic gerrymanders. On the contrary, I'm trying to get a feel for what kinds of maps will come out of actual redistricting after the Census, and that means some states will be Democratic gerrymanders, others Republican gerrymanders, some bipartisan incumbent protection plans, some nonpartisan commission plans, etc.

Anyone who saw my maps for Ohio, Michigan, Nevada, or Texas can plainly see I did not draw the most favorable possible lines for Democrats. That's because, of the six states I've covered so far, only one (Massachusetts) is expected to be run by the Democrats when remap time comes around in 2011-2012. Goodness, I eliminated lefty favorite Dennis Kucinich in Ohio and protected GOP incumbents like Pat Tiberi, Thad McCotter, and Pete Sessions whom I would prefer to see defeated. Obviously, I'm attempting (as best a private citizen with nothing more than Excel, Paint, and a calculator to work with) to approximate the political environment in each state. I think that's a lot more informative and illuminating than merely drawing my fantasy district boundaries.

Point three: if Krusty's numbers are correct, my largest district (Steve King's District 4) has 18,866 more people than my smallest (the Latham-Boswell District 3). I remember the numbers being far closer to equal than that, but it's possible one of us messed up with a calculator. I certainly don't purport to be a math whiz.

But if population equality is such a fault with my map, Krusty's isn't much better. His 4th has 14,645 more people than his 3rd. That seems a fairly significant divergence too, if those numbers are accurate.

As for an apology, it's not necessary. This kerfuffle is quite entertaining to me.

---

by: Nathaniel90 @ Tue Mar 10, 2009 at 11:01:03 AM CDT

fair enough (0.00 / 0)

Thanks for this comment.

So maybe Krusty doesn't need to apologize to you, but he probably should apologize to 20th Century Fox for using the name "Herschel Krustofsky" and the image of Krusty the Clown from the Simpsons.

Invite other Iowa political junkies to join us at Bleeding Heartland.

---

by: desmoinesdem @ Tue Mar 10, 2009 at 11:21:22 AM CDT

[ Parent ]

one-man, one-vote (4.00 / 1)

trumps (almost) all.

If you want to examine this w/o taking political considerations into account, you could use algorithms to generate possible configurations and take it from there.

1. "grow" four initial clusters of counties. Typically, something like the 'greedy' algorithm is used where you start w/ 4 counties picked at random and let each county grab the largest neighbor until each cluster hits the max population permitted per district.

2. This gives you four crude districts to start with. By repetition & randomization, you zip thru all (or near all) possible starting configurations.

3. Use an optimization method to swap counties between districts. The two possibilities are (1) Monte Carlo simulated annealing, a method called 'local' because swaps are done one step at a time. After each step, the deviations are calculated, and the step/change is accepted if the new configuration lowers the abs min dev. Every once in a while, the algorithm accepts an increase just to shake things up and not get stuck, literally.

Second method is called a genetic algorithm, which is global instead of local -- all changes are made in one pass based on genetic rules that improve the 'fitness' (abs min dev) of the district boundaries.

You design to eliminate silly possiblities, like a district w/ a long octopus arm to keep things reasonably compact and to force contiguity.

Last time we discussed this, I thought it would be interesting to pursue this once the numbers come out in 2010.

As far as this particular configuration is concerned -- I think it's pretty good. The Iowa State Data Center does have population projections for 2010, and I believe (by eyeballing) that this model deteriorates if you use the projected numbers vs 2007 estimates. That's where fine-tuning by algorithm kicks in. However, by nudging a county here or there, you might land up generating a new but different configuration w/ better properties, per Iowa Code.

I imagine that the IRC uses redistricting software to get things going as well.

I don't think Krusty's map is more compact than Nathaniel's either, but like my Republican father used to say, "Reasonable minds can differ."

I recall a quantitative definition in the IA Code. If there's interest, let me know, and we'll do a Nathaniel vs Krusty run-off!

---

by: ragbrai08 @ Tue Mar 10, 2009 at 12:38:52 PM CDT

may the best geek win! (0.00 / 0)

I would be curious to see how the compactness of the Krusty and Nathaniel maps look from the perspective of the Iowa Code.

I'm sure things will change between now and the census--exurbs in Dallas and Warren counties won't grow as fast from 2008-2010 as they did between 2000 and 2007. The software presumably keeps improving too.

Invite other Iowa political junkies to join us at Bleeding Heartland.

---

by: desmoinesdem @ Tue Mar 10, 2009 at 12:46:09 PM CDT

[ Parent ]

Interesting conversation (4.00 / 3)

and I'm pretty thankful that our redistricting process isn't as politicized as it is in other states.

I like both maps I've seen (by you and Krusty) but I've come up with a third that is quite different than yours.

Quickpost this image to Myspace, Digg, Facebook, and others!

This one pretty much divides the districts into North, South, East, and West.

Here's the population estimates.

District 1 - Orange - 750,000 (24.98%)
District 2 - Green - 747,000 (24.88%)
District 3 - Blue - 751,000 (25.02%)
District 4 - Red - 754,000 (25.12%)

A couple of things

>  Des Moines and its suburban counties haven't ALL been in the same Congressional District for the past two redistrictings.  So I've maintained that.
>  I've also kept Linn and Johnson counties together to maintain the metro area there.
>  I came up with my own estimates for population.  obviously they won't be perfect, but they are educational guesses.
>  Finally, each district has two larger metro areas (with the exception of the exceptionally large Polk)

I like my model it's the very geographically neutral.  For example, it the model proposed by Nathaniel, it appears that King's district is getting ready to consume Latham/Boswell's third district.  My map has tried to avoid that.

I think this might be a solid model for next year.  Whatcha think?  By the way, this is my first post here.  Nice digs.

---

by: ghbraves @ Fri Mar 13, 2009 at 14:46:55 PM CDT

ha (4.00 / 1)

I have a diary going up Monday comparing the two maps. Make that three, you just upped the ante!

I came up with my own estimates for population.  obviously they won't be perfect, but they are educational guesses.

This might be a problem. Stop by Mon and find out.  

---

by: ragbrai08 @ Fri Mar 13, 2009 at 15:10:01 PM CDT

[ Parent ]

Hey thanks (0.00 / 0)

It's all speculative, so it should be fun.  I look forward to Monday.

---

by: ghbraves @ Fri Mar 13, 2009 at 15:15:21 PM CDT

[ Parent ]

looking forward to it! (4.00 / 1)

Judging from my traffic this week, there's a lot of interest in this subject.

Invite other Iowa political junkies to join us at Bleeding Heartland.

---

by: desmoinesdem @ Fri Mar 13, 2009 at 15:23:55 PM CDT

[ Parent ]

As for the population estimates (4.00 / 2)

I imagine that all of our respective pop. estimates won't be too different.  As for me, I simply took the most recent county population figures and educationally estimated a logically appropriate increase in population.

I don't think it'll be a huge difference, though.

---

by: ghbraves @ Sat Mar 14, 2009 at 10:56:42 AM CDT

[ Parent ]

Krusty and Nathaniel (4.00 / 2)

were also using the 2007 county population figures, though I don't know what they did in terms of estimating future growth.

A big question mark will be Dallas County, which grew at a ridiculous pace between 2000 and 2007. Presumably the collapse of the housing market will limit further growth from 2008-2010, but who knows?

Invite other Iowa political junkies to join us at Bleeding Heartland.

---

by: desmoinesdem @ Sat Mar 14, 2009 at 13:11:34 PM CDT

[ Parent ]

Dallas County = lots of growth (4.00 / 2)

58,000 in 2007.  I estimated 65,000 for 2010, but it'll be hard to say.  We'll see what happens.

I'll be blunt: I find congressional reapportionment very fascinating.  But, then again, I'm pretty geeky.

---

by: ghbraves @ Sat Mar 14, 2009 at 14:35:31 PM CDT

[ Parent ]

we welcome geeks (4.00 / 1)

at Bleeding Heartland!

If you like good reapportionment diaries in general, you should check out Swing State Project from time to time.

Invite other Iowa political junkies to join us at Bleeding Heartland.

---

by: desmoinesdem @ Sat Mar 14, 2009 at 15:25:27 PM CDT

[ Parent ]

welcome, and thanks (4.00 / 1)

for posting an intriguing variation.

Invite other Iowa political junkies to join us at Bleeding Heartland.

---

by: desmoinesdem @ Fri Mar 13, 2009 at 15:25:02 PM CDT

[ Parent ]

Very interesting third option... (4.00 / 1)

Awesome. This is a completely new approach to the puzzle that I hadn't even thought of.

The first thing I notice is the news IC/CR/QC district. That would become the most "urban" of the four districts it seems--replacing the "urban" Des Moines area district others envision. Would the Republicans go for that though? Seems like it would be a lock for Dems.

Of course, to balance that advantage, you removed the QC from the 1st D. and greatly expanded it to the west. That would make Braley's job much harder, so maybe it's a wash. Braley/Latham '12 seems like a very tough fight.

It looks like the new 3rd would be a lock for Bozwell, since a lot of that territory was his before the last redistricting. And since you're not throwing Latham into the mix, it looks like this district could stay Boz's as long as he wants.

And yeah, the 4th would stay King's. But you're right. It looks a lot less like like a hungry Pac-man now.  

---

by: American007 @ Mon Mar 16, 2009 at 11:00:54 AM CDT

[ Parent ]

great minds, (0.00 / 0)

and all that ...

---

by: ragbrai08 @ Mon Mar 16, 2009 at 14:43:29 PM CDT

[ Parent ]