Featured Post

Action Plan?

Over the weekend, you may have heard that our President-Elect decided to get into a Twitter-Fit (hereinafter “twit”) regarding John Lewis, the Democratic Representative to the House for Georgia’s 5th District.

Other folks have covered the story, with their own assessments of the intelligence of deciding to deride an actual Freedom Rider over MLK weekend, so I’ll leave that aside for the purposes of this post.

But on Twitter and Facebook (and elsewhere in social media), I saw a number of “resting on laurels/get rid of the bum” sentiments regarding Congressman Lewis.

While I think there may be more ideology than just “toss the bums out” involved, here, I thought I would do the folks expressing this desire a solid and tell them how to achieve their goal.

This table shows you the election results for the last two election cycles in Georgia:


Well, already we see the problem.  Lewis ran unopposed in 2014 and whupped his opponent in 2016 by an astounding 68.8 percentage points, in a year where Trump carried Georgia by 5.1 percentage points.

Hm, that’s tough.  Lewis sitting in a D+32 District… oh, wait, let me back up a second, for those who don’t know what PVI is.

The Cook PVI is a measure by which election regions are marked more or less partisan, and indicates the percentage point lean that district has towards a particular party.  A difference of 2 points in the D direction (D+2) means that all other things being equal, you would expect if you plucked 100 random voters out of the voter pool in an election, you’d get 51 Democrats and 49 Republicans.  The Cook PVI relies upon actual voters in recent elections, which means it’s *not* a measure of partisan leaning for a general populace or even a measure of partisan leaning for the registered voters.  Famously, not everybody shows up for each election.

Still, it’s a good benchmark for telling you how big of an 8-ball you are behind, if you’re running against someone of the other party… or how big of a leg-up you already have.

Georgia redistricted itself in 2013, following the 2010 Census.  You may recall that in 2013 the U.S. Supreme Court overturned a good chunk of the Voting Rights Act, granting Alabama (and by extension, Georgia) much more latitude to redraw its districts.

Georgia has a recent history of redrawing districts that (I’m certain “entirely coincidentally”) exclude sitting members of their Democratic House membership from running in their current district, by moving the residence of said Congressperson out of the current district lines.  It worked in 2014, as the GOP unhorsed John Barrow by moving him into the 12th District and giving Rick Allen a chance to knock him out of the delegation.

For entertainment value, notice that Athens (now at the northernmost border of the 10th district) used to be in the 12th district.

But, my GOP friends in Georgia… it’s going to be hard to do that to Lewis.  He’s right in the middle of Atlanta, the heart of the most populated area of the state.  Neighboring districts to the 5th include the 13th, the 4th, the 11th, and the 6th.  You probably can’t carve Lewis’s house out of the 5th district *and* move a majority minority district far enough away that Lewis can’t just sell his house and move a couple of miles and get re-elected.  You can’t outright merge the 13th and the 5th and let Lewis run against Scott, the population numbers don’t work out.

So those two options are out.

The alternative is to make it harder for Lewis to get elected by flooding his district with Republicans.  That’s actually feasible by redrawing district boundaries, just move some of the 5th into the 11th or the 6th!

Hm, but then we have a different problem.  Lewis has a pretty commanding advantage in the 5th.

You’ve got to move *enough* Republicans into the 5th to give someone a fighting chance.  But if you do *that*, you’re also moving all those Democrats from the 5th… into the 6th or the 11th.  About 100,000 Democrats would need to be exported from the 5th, replaced by 100,000 Republicans, just to get the PVI down into D+3 or D+5 territory and that’s no guarantee you’d get rid of Lewis.  Since urban areas are compact, you can’t just select voting districts by turnout, it won’t work that way.

And worse, if you put 100,000 active voting Democrats into the 6th or the 11th, you’re also taking 100,000 GOP voters from those places… shoot… even if you split them up, and draw equally from the 6th and the 11th, you’re putting yourself at risk of losing either/both of those seats in off-cycle years and you’re dropping the R advantage in both areas down closer to R+3 to R+5 territory… which means someone like Loudermilk is probably far less viable as a candidate, with his membership on the House Freedom Caucus and his former involvement in the Family Resource Council… well, here he is in his own words.  Price is 383rd on the bipartisan index, Loudermilk is 429th.

Wait… Lewis is 272nd?

I guess it’s time to fish or cut bait.  You can quite possibly get rid of Lewis by gerrymandering his district around.  But to do it, you might put two of your own most partisan members right into the meat grinder.

How bad do you want to get rid of him, again?

Staff Writer
Home Page 

Patrick is a mid-40 year old geek with an undergraduate degree in mathematics and a master's degree in Information Systems. Nothing he says here has anything to do with the official position of his employer or any other institution. ...more →

Please do be so kind as to share this post.

12 thoughts on “Action Plan?

  1. How *EFFECTIVE* is Lewis?

    How good is he at being a representative? Like, does he sponsor really good legislation, does he help leverage other people in his party to make the wobbly less wobbly? When he’s grilling people at a hearing, does he do a good job of making them look bad?

    Because, I’m gaming this out in my head and I’m thinking that if *I* knew that I would be in a situation with razor thin margins everywhere, I would do what I could to make my opponents put valuable, limited, resources into places where I’d never be able to beat anyway because those are resources that they wouldn’t be using against places where I might not beat or, god forbid, places where I myself are vulnerable.

    And if I could get them to put those valuable, limited resources into places where I wouldn’t be able to beat but also doesn’t really give them whatever the political equivalent of “victory points” would be, then that’s probably the best way to distract. Give your opponent an obvious and noisy victory that he would have had anyway to distract from thousand cuts he’s dying from in other battlegrounds.


  2. Who redrew the diatricts in Georgia in 2010? Was it the Republicans or some kind of nonpartisan commission? If it was the GOP, then they’re probably perfectly happy with the results because they’ve managed to cram as many minority voters into his district as possible, giving them a clear shot of victory in other districts. They can’t shut out Democratic voters completely.

    That’s how redistricting worked here in NC although it’s been tied up in the courts because they found clear evidence of racial bias. But, in the meantime, gerrymandering allowed them to change the composition of the NC delegation from 7-6 Democrat to 9-4 Republican. Worked for them.


    • I remember in the early 90s a friend got an internship with a law professor to help draw districts in the Florida panhandle to address civil rights issues. She would describe herself as a typical New York liberal Jew, but when people asked her what she was doing for the summer, she would say, ‘carving out Republican districts.’ Same thing. Its like the optical illusion where some people see the old lady and others the goblet.

      Illinois has three black majority districts and as partisan districting process as any in the country and here are the averages for those districts compared with Georgia’s four black majority districts:

      55.28% Black
      31.6% White
      D +19

      53.2% Black
      30.6% White
      D +31

      The racial breakdown seems indistinguishable to me. The effect in Illinois though is worse for Democrats, who drew the districts.


      • Which is why it’s probably not a good idea to let partisan politicians redraw districts. I know California has recently instituted some kind of nonpartisan redistricting scheme. It will be interesting to see how that experiment works out.


        • Given the way gerrymandering works, seeing a packed district in isolation makes it really hard to tell which side packed it. Depends on the context of the state and demographics of the region.

          That said, I like the “efficiency gap” method that was used to strike down Wisconsin’s state leg districts and really hope SCOTUS accepts it.

          You don’t need non-partisan boards (which, despite the label, can be really partisan or really bad at drawing districts) if you can express the resulting map in terms of deviation away from proportional results.

          The efficiency gap basically tallies up ‘wasted’ votes — that is, the number above the amount needed to win a district and does a statewide analysis. If you find one party is wasting a LOT more votes than the other, you can reasonably conclude there’s a partisan advantage baked into the pie — and even what areas the problem lies in.

          It’s straightforward,purely mathematical, and can be applied easily — and neatly forecloses any attempts to gain a seat edge your voting percentage doesn’t give you (not ungameable, but your returns are very small) and doesn’t require taking the district drawing away from legislatures to get fair representation.


          • I read the court decision and would bet a small amount of money that the efficiency arguement won’t be accepted. See Sam Wang here.

            But it wouldn’t help Democrats that much. In Wisconsin, proponents of the efficiency argument conceded that Republicans had a natural four point advantage because Democrats are concentrated in Milwaukee and Madson. In a perfectly efficient map, Democrats would still need to get more than 54% of total votes to get more than 50% of the seats.

            The proponents of the efficiency argument are not looking for perfect maps; they just want to consider presumptively invalid any map with inefficiencies greater than 1.5 standard deviations from the mean. Looking at past maps, they explain which states that would cover for Congressional districts:

            In each of the decades we analyzed, only a handful of plans had average gaps of this magnitude. Illinois and Texas did so in the 1970s; California (the first plan), New York, and Texas (both plans) in the 1980s; California, New York, and Texas (both plans) in the 1990s; and California, Florida, Illinois, and Texas (the first plan) in the 2000s. (It is too soon, of course, to compute average gaps for the 2010s.)

            There could be zero net effective from this approach on Congressional seats.


            • I think you’re overlooking something: The 2010 gerrymandering was insane by historical standards, because they had the computing power and programs to finally do it.

              Throw away ANY gerrymandering done before 2000 at all. It’s irrelevant to modern methods.

              2000 wasn’t anything to write home about either — computing power was slow and the tech revolution that made computing so ubiquitous had just ended. The software, computing speed, and techniques that were used in 2010 simply didn’t measure up.

              Simply reducing gerrymandering to a mere 1.5 deviations out, to reduce it to “merely” the effectiveness of the 70s, 80s, and 90s would be a vast victory.


              • Seriously. I live in a state where the Democrats just very, very narrowly captured the governor’s mansion and a majority in the state supreme court, while the GOP has a veto-proof supermajority in both houses of the state legislature that was elected simultaneously by the same voters. Something is very fishy about that sort of result.


              • The increase in the efficiency gap after 2010 is asserted in the link, but the creators of the test still concluded that based on the 2012 election results, probably only Florida, Ohio, Pennsylvania, and Virginia would have presumptively unlawful maps.(*) If Democrats had won just one more Congressional seat in each of these four states, there would be adequate efficiency.

                (*) More elections might give better data.


  3. In Virginia, the federal district court (which the court of appeals then affirmed) made the state legislature unpack some of the African Americans in Congressman Bobby Scott’s 3rd congressional district. Which had, like Congressman Lewis, a D+30plus PVI (The new district is D+27 per the current Cook report)

    So it’s possible that the courts may remove some of Lewis’s overwhelming advantage (but unlikely, as it hasn’t happened yet, and the Supreme Court isn’t going to set a national standard or even get involved for a good number of years from now, thanks to Trump’s SCOTUS picks)


    • It would probably violate the Voting Rights Act (as long interpreted by the SCOTUS) to dilute Lewis’ district by including enough non-African Americans such that the seat became competitive. There are court cases where packing majority-minority districts goes too far, but that’s not supposed to be to the detriment of minorities, its supposed to place a limit on gerrymandering among non-minorities under the pretext of aiding minorities.

      It looks to me like African-Americans have a right to four Congressional seats in Georgia unless demographics or SCOTUS decisions change, and so long as Lewis is popular among African-Americans he gets a seat or he goes to the courts.

      The paradox here is that the rule gives voice to minorities, but makes it more difficult for Democrats to gain majorities in Congress that would benefit minorities.


  4. Georgia has a recent history of redrawing districts that (I’m certain “entirely coincidentally”) exclude sitting members of their Democratic House membership from running in their current district, by moving the residence of said Congressperson out of the current district lines.

    Which was completely unpredictable. John Roberts is an honest, unbiased referee. Just like Tim Donaghy.


Comments are closed.