« Good Stuff | Main | Priorities »

Gerrymanders

I've seen it before, but a fresh look is always bracing. Via Mark Schmitt the bizarre map of New York City congressional districts. Note not only the odd, squiggly shape of the districts, but their habit of leaping over bodies of water. At times, this is justified. The 13th encompasses all of Staten Island, but that's not enough people for a congressional district, so it also includes part of Brooklyn. But those no good reason for Manhattan to contain portions of four or five different districts, none of which are wholly based in Manhattan.

January 6, 2005 | Permalink

TrackBack

TrackBack URL for this entry:
https://www.typepad.com/services/trackback/6a00d8345160fd69e200d8342f392e53ef

Listed below are links to weblogs that reference Gerrymanders:

Comments

What does Matthew mean by "good reason"?

Posted by: Al | Jan 6, 2005 3:01:47 PM

Redistricting has been something on my mind for quite some time now. Here in Ohio, we have a Presidential race that's 50-49 GOP but our representative distribution is 12 to 6 GOP. Some of us are thinking about this and looking towards how Iowa does redistricting.

Paul

Posted by: Paul | Jan 6, 2005 3:03:11 PM

" Here in Ohio, we have a Presidential race that's 50-49 GOP but our representative distribution is 12 to 6 GOP."

I'm all in favor of the Iowa method, and I'll be happy to sign your petition, but unfortunately, that will only solve part of the problem.

Because of the heavy clumping of Democratic votes in urban areas, there are many more fairly drawn districts that are 80/20 Dem than there are 80/20 GOP. And that means in an evenly split state, there are going to be more 60/40 GOP districts than 60/40 Dem districts, and thus more GOP districts overall.

In short, until Dem candidates run a bit better in rural and exurban areas, we're going to have trouble taking back the House.

Posted by: Petey | Jan 6, 2005 3:12:16 PM

Here in Ohio, we have a Presidential race that's 50-49 GOP but our representative distribution is 12 to 6 GOP. Some of us are thinking about this and looking towards how Iowa does redistricting.

You mean the Iowa that had a Presidential race that was 50-49 GOP but whose representative distribution is 4 to 1 GOP?

Posted by: Al | Jan 6, 2005 3:12:58 PM

"You mean the Iowa that had a Presidential race that was 50-49 GOP but whose representative distribution is 4 to 1 GOP?"

You should be well aware that fairness in opportunity is not the same as fairness in outcome.

Posted by: Petey | Jan 6, 2005 3:17:10 PM

Arnold wants to take redistricting power and give it to a panel of retired CA judges:

http://www.nytimes.com/2005/01/06/national/06arnold.html

I say it ought to go to computers with census data. Ignore voter registrations, race, etc completely.

Posted by: John i | Jan 6, 2005 3:18:33 PM

Redistricting reform is not going to be a panacea for the Democrats. It is at least as important at the state house and senate level. Here in Ohio, we have 99 of the former and 33 or the later.

There are other issues at least as important, including distribution of voting machines, machine design, etc.

Just because proposal X won't completely solve problem Y, doesn't mean proposal X is bad.

Paul

Posted by: Paul | Jan 6, 2005 3:29:43 PM

Texas redistricting guinea pig here. My new district (25th, represented by Lloyd Doggett) connects me in East Austin to Rio Grande City, 275 miles away.

Y'all betta recognize true achievement in crapulence when you see it.

Posted by: norbizness | Jan 6, 2005 4:25:21 PM

Here's norbizness's district.

Posted by: JD | Jan 6, 2005 5:00:20 PM

You should be well aware that fairness in opportunity is not the same as fairness in outcome.


True, but that's why the fact that Ohio's delegation is 12-6 does not prove that the lines have been gerrymandered there. All I'm saying is that the mere fact that the presidential results for state X do not match its congressional delegation does not, by itself, indicate that the district lines have been gerrymandered.

Posted by: Al | Jan 6, 2005 5:07:50 PM

Because of the heavy clumping of Democratic votes in urban areas, there are many more fairly drawn districts that are 80/20 Dem than there are 80/20 GOP.

Also, Petey, I'm curious if you have any actual, you know, evidence for this statement. I'm not saying it's not true, but I'd be interested to see where you get it from.

Posted by: Al | Jan 6, 2005 5:09:06 PM

"Redistricting reform is not going to be a panacea for the Democrats."

Got that right. Not the least, because, among many other reasons, there are states where it would be more of a panacea for Republicans.

Posted by: Brett Bellmore | Jan 6, 2005 5:12:30 PM

Al, unfortunately, it's true: Democrats ARE more geographically concentrated than Republicans. Any redistricting that was done completely ignoring information about partisan voting tendencies WOULD result in an advantage for Republicans, for that reason.

Essentially, you gerrymander yourselves, by where you chose to live. This isn't all that contraversial, though what to do about it is. Personally, I'm a fan of at large proportional representation. Makes gerrymandering completely moot.

Posted by: Brett Bellmore | Jan 6, 2005 5:24:30 PM

"Also, Petey, I'm curious if you have any actual, you know, evidence for this statement. I'm not saying it's not true, but I'd be interested to see where you get it from."

I don't have a link handy, but I remember reading some interesting stats after this November's elections on how there are many, many precincts that vote 95% Dem, and almost no precincts that vote 95% GOP.

Given the urban clumping phenomena, that makes sense to me.

I've read some items further back that talked about the phenomena at the Congressional district level, but the details are pretty fuzzy in my mind.

If you wanted to see this phenomena for yourself, I'd suggest checking out a map of Pennsylvania with the Presidential results broken down by Congressional district. The entire Kerry state margin comes from overwhelming victories in 3 or 4 Congressional districts in and around Philly.

Posted by: Petey | Jan 6, 2005 5:25:12 PM

"I don't have a link handy, but I remember reading some interesting stats after this November's elections on how there are many, many precincts that vote 95% Dem, and almost no precincts that vote 95% GOP."

Yeah, I recall that, and I can look up the link for you, but I should point out that he later corrected the bug in his program... It IS skewed, but that that radically.

Posted by: Brett Bellmore | Jan 6, 2005 5:30:28 PM

Darn itallics!

Posted by: Brett Bellmore | Jan 6, 2005 5:30:59 PM

Given the urban clumping phenomena

Obviously urban districts are more geographically compact than rural. But what you are really saying is that the typical urban district has a higher percentage of Democrats than the typical rural district has a percentage of Republicans. That is, for example, the typical district in Philly has X% Democrats, and the typical district in central PA has Y% Republicans, where X>Y. Is that true? Perhaps it is, due to, say, racial segregation, given that blacks vote so overwhelmingly for Democrats. Dunno, and would be interested to find out. It is obviously a powerful argument against "objective" districting.

Posted by: Al | Jan 6, 2005 6:06:16 PM

"But what you are really saying is that the typical urban district has a higher percentage of Democrats than the typical rural district has a percentage of Republicans."

I strongly believe that to be the case.

And it's not only a racial thing. My urban CD is majority white, and it was over 80% Kerry this year.

I'll bet that if you picked a random rural CD in central Missouri or Ohio, it would only be around 55% - 65% Bush.

Posted by: Petey | Jan 6, 2005 6:31:00 PM

So how about you just divide Texas and California by 32(Texas) equidistant vertical lines? Ignore population,geography,race,etc. The guy representing 76 people in the Panhandle would be essentially an at-large representative. In fact most of them would have such diverse constituencies as to be barely representative.

I kinda like proportional representation too, thinking it would make reps closer to constituences. But if the Founders didn't do it, I gotta look twice.

Posted by: bob mcmanus | Jan 6, 2005 7:51:21 PM

Petey writes:
>
> I'll bet that if you picked a random rural CD in central Missouri
> or Ohio, it would only be around 55% - 65% Bush.

Ooh, bad example, at least for central Missouri. We're pretty gerrymandered here, too. We actually have two small metropolitan areas here in central MO (corresponding to Jeff City and Columbia), and I believe those are parts of 3 or 4 CDs. My CD extends over 125 miles, all the way to the Mississippi River. If we had the most natural CD here (the two metros and the mostly closely related areas along US-63), I believe you would have had a CD that went for Bush 55-45 or so this year, led by the largest county (Boone) that Bush apparently won by the whopping total of 35 votes.

I actually don't believe in geographically defined congressional districts at all, since I've never been in one that made much "sense". Of course, I also don't see why Wyoming should have it's own CD, and I don't understand why CDs can't cross state lines. Frankly, I think the biggest single step you could take to reforming congressional district boundaries *and* solving the nutty electoral college issues we have is to set the number reps to the Census population divided by the population of the smallest state, rounding up. So Wyoming would still be worth 3 EVs, but a place like California would be over 100. With about 1000 districts, it would be much harder to justify the drawing of really scary looking ones (if that's what you don't want).

But in the meantime, I'll enjoy the zest green air on my home planet.

Posted by: Jonathan W. King | Jan 6, 2005 10:44:21 PM

Ignore population...

Unfortunately, this is unconstitutional. Baker v. Carr, I believe. All districts within the same state must be 100% equipopulous. Deviations of even 1% are not permitted.

Of course, the vertical line method isn't necessarily inconsistent with equipopulosity. The lines just wouldn't be equidistant. The district going through downtown Dallas and Fort Worth would be awfully weird though. Five city blocks wide but running the entire length of the state.

Posted by: JP | Jan 6, 2005 10:46:24 PM

Kevin Drum has a good point on redistricting.

Posted by: TomB | Jan 6, 2005 11:07:11 PM

"The district going through downtown Dallas and Fort Worth would be awfully weird though."

Good lord, nobody lives in downtown Dallas. Like, y'all think this is a real city or something. Dallas Center is a combination strip mall and office park. Hardly need cops, gets so quiet after dark. The rats and cockroaches have moved to the suburbs to get food.

Posted by: bob mcmanus | Jan 7, 2005 12:48:02 AM

JD,

That's the old, pre-Delay-gerrymander 25th. The new 25th doesn't go anywhere near Houston and no federal site I've seen shows the new maps. Texas state government has it up somewhere.

Posted by: Brittain33 | Jan 7, 2005 9:08:36 AM

Heh, point taken. Also, Cowboys suck.

Texas state government has it up somewhere.

Can't find the full map, but this article shows what a few of the districts look like, including the 25th. The word "fubar" comes to mind.

Posted by: JP | Jan 7, 2005 10:04:48 AM

The comments to this entry are closed.