GERRYMANDERING OCTOBER 14, 2013
“Beware of gerrymandering denialists!”
That’s the title of an article by David Weigel, who’s pushing back against myself and a few others who doubt whether gerrymandering is responsible for the shutdown and all the other problems that bedevil the republic. Since it was largely aimed at me, it shouldn’t be too surprising that I found the article thoroughly obnoxious: It juxtaposed my position with the obvious, failed to engage with my argument, and claimed I was fighting a strawperson, which the first page of Google search hits reveals to be untrue. Weigel himself wrote a whole post about gerrymandering and the shutdown!
On the other hand, perhaps his mischaracterizations were somewhat understandable. Most political writers, myself included, frame their pieces with respect to their perception of the conventional wisdom. For that reason, my pieces about gerrymandering relegate many of its substantial, but widely understood effects to caveats and parentheticals. And as a result, I guess Weigel reads my piece and thinks I’m some crazy “denialist” who doesn’t think that gerrymandering resulted in additional GOP seats in North Carolina, which it obviously did.
On the other hand, perhaps Weigel would agree with more of my arguments downplaying some of gerrymandering’s perceived effects if those arguments were placed in fuller context.
So in the interest of reaching an understanding about the effects of gerrymandering, or at least more completely representing my position, here’s the big picture on gerrymandering.
Yes, gerrymandering increased the number of Republican-held seats in the House of Representatives.
In Pennsylvania, Ohio, Florida, Virginia, Michigan, and North Carolina, GOP-led redistricting packed Democrats into a small number of heavily Democratic, urban, non-white districts, leaving the majority of districts to lean Republican. In those states, Republicans won a clear majority of districts, despite far closer tallies in the statewide popular vote. How much did partisan gerrymandering help the GOP? The estimates vary, in part based on the presumed alternative. But there’s no question that partisan redistricting efforts yielded additional GOP seats.
Yes, gerrymandering makes it more difficult for Democrats to retake the House in 2014.
The next time there is a Democratic wave, partisan gerrymandering will make it more difficult for the GOP to lose the House. Waves usually knock-out incumbents on enemy territory, but most survive on even slightly friendly turf—very few lose on safe turf. So in a wave election, a few points worth of partisan-lean can substantially improve an incumbent’s chance of survival.
And today, there just aren’t very many opportunities for Democrats. There are only 2 Republicans holding districts with a Cook PVI of D+2 or more. In 2006, there were 18 Republicans on Democratic turf and Democrats won ten of them. In other words, a 2006-esque wave might not get Democrats over the top in 2014.
Not all of this is due to partisan gerrymandering—Democrats have consolidated more of the D-leaning districts than they had in 2004, in part because they expelled many decent GOP incumbents in 2006 and 2008. But with such a tough road to 217, even modest effects from partisan gerrymandering would make a difference.
No, gerrymandering didn’t represent the GOP’s margin of victory in the House in 2012.
It’s impossible to re-run the 2012 congressional elections with different districts, so it’s also impossible to claim Republicans would have maintained control of the House with absolute certainty. But Democrats fell 17 seats short of retaking the House, and partisan gerrymandering probably didn’t represent the margin of victory.
In fact, I’m not sure I’ve seen a single study argue that gerrymandering cost Democrats 17 seats. The closest I’ve seen comes from Sam Wang, who argued that Democrats would have won 215 seats, 3 short of the magic 218. Other studies, which match more with my own more subjective count, find smaller effects: For example, 9 seats according to Nicholas Goedert, 7 seats according to Eric McGhee and John Sides.
Why was the effect of gerrymandering so modest? A process blind to partisanship would still give the GOP a modest advantage in the House compared to the national congressional vote. Democrats have a spatial inefficiency problem: The blue areas are bluer than the red areas are red, so Democrats are inevitably bound to “waste” votes in heavily Democratic districts. Here’s one way to look at it: after you’ve made your “blue” districts and your “red” districts, the votes remaining for the, say, 40-to-60th percentile of districts will lean Republican, even if you started with a Democratic advantage.
There’s another reason to believe Republicans would have held the House: Incumbency. Not only do many bipartisan redistricting efforts protect incumbents, but the advantages of incumbency would have allowed most incumbent Republicans to survive, even if a partisan- and incumbent-blind redistricting process put Republicans on more competitive turf.
Incumbency regularly allows representatives to significanty outperform their party’s presidential candidate—they build greater name-recognition, get an opportunity to co-opt traditionally hostile constituencies, raise plenty of money, and deter viable challengers. So if abstract calculations put the Democrats short of retaking the House, then they probably won’t even be close after considering incumbency. Indeed, McGhee’s study, which found that gerrymandering provided Democrats with 7 additional seats, determined that the entire apparent effect of gerrymandering vanished after accounting for incumbency.
If you need any proof, consider that most Republicans from gerrymandered districts won by a wide margin. This doesn’t prove that they would have won without gerrymandering—for instance, fairer districts might have led to stronger challengers—but the wide margins suggest that a few points worth of redistricting weren't decisive. To overcome incumbency, Democrats needed a wave election. 2012 was a modest Democratic win, but it wasn’t a wave.
And if that’s true, then the case that the shutdown happened because gerrymandering put the GOP in charge is quite weak.
Gerrymandering reduces the number of ultraconservative districts.
Although gerrymandering increases the number of safely Republican districts, it doesn’t produce ultra-conservative districts. After all, the goal of partisan redistricting is to maximize the number of seats held by a party. That discourages the GOP from constructing ultraconservative districts, which waste Republican votes making red districts redder, rather than making purple districts redder. So without gerrymandering, many Republican districts would get redder.
I think this is an area where reporters get confused, because, for many, “safe districts” means “gerrymandering.” For instance, Dave Weigel talked to Ohio Representative Jim Jordan, who basically conceded he’s more partisan because his district is safe. But why does Weigel assume that gerrymandering made Jordan’s district more conservative? It’s quite possible that it would be more conservative without gerrymandering.
Let’s look at Ohio. In order to maximize the number of Republican seats, Ohio Republicans concentrated Democrats in a handful of Democratic districts centered in Columbus, the Mahoning Valley, and Cleveland. Then, Republicans split up the remaining Democratic bastions (Dayton, Cincinnati, Toledo), allowing them to be overwhelmed by the Republican countryside.
Now look at Jim Jordan’s district (OH-4). For the most part, his district represents the safely Republican countryside, but it sneaks back up to northern Ohio to steal a few Democrats and make his district more competitive than it would be if it only represented the conservative countryside. Without gerrymandering, his district gets more conservative--even if some suburban Republican winds up moving into a more competitive district. And indeed, Jordan’s district became less conservative after redistricting.
So even if reporters are convinced that safe districts influence the behavior of Republicans, which is debatable, those seats may not be safe because of gerrymandering. And for the most part, they’re safe because the country is deeply polarized, and most Americans live in uncompetitive counties or precincts. In 2008, for instance, only 40 percent of Ohio precincts were won with less than 60 percent of the vote! It was probably worse in 2012.
So, yes, Jordan's seat is probably safe no matter what.
Is it tough to distinguish between districts that are safe due to gerrymandering, districts that would otherwise be safe? Yes. Absolutely. But perhaps that’s a good reason to defer to the preponderance of political science research.
There’s even a case that gerrymandering adds relative moderates to the Republican Party.
Where are the Republicans gaining the most seats due to gerrymandering? Mainly in the competitive or northern states, like Pennsylvania, Michigan, Virginia, Ohio, Florida, and North Carolina. Without those partisan gerrymanders, there would be fewer northern Republicans, who tend to be the relatively moderate. And those additional Republicans tend to be the most moderate of all.
Simply put: there’s no way to make Pennsylvania’s congressional delegation overwhelmingly Republican without plenty of lean-Republican districts. Indeed, 4 of Pennsylvania’s 13 Republicans come from R+1 or R+2 districts. 3 of Virginia’s 8 Republican districts come are R+4 or closer. 4 of Michigan’s 9 Republicans come from districts that are R+4 or closer. Most of these incumbents are relatively safe as a result, but they're also relatively moderate. 17 of the 30 most moderate Republicans from last cycle came from gerrymandered states.
To be sure, there would be more Democrats without gerrymandering. But the point is that those Democrats would tend to come at the expense of relatively moderate Republicans, which would leave the rump House GOP caucus even more conservative.
Now that doesn’t mean the aggregate effect of gerrymandering is toward additional moderation. Without gerrymandering, some relatively safe Republican districts get more competitive, and perhaps they wind up replacing the moderate Republicans who lost reelection in the previously mentioned scenario. It’s also possible that a more vulnerable House majority would be less likely to engage in tea party sponsored adventures, like the shutdown. But if gerrymandering does have an effect on polarization, it’s probably very modest and it’s hard to prove.
But there’s also a case that none of this matters at all.
Because those “moderates” don’t vote so differently from their colleagues.
So gerrymandering probably isn’t responsible for the shutdown.
Even without gerrymandering, the GOP would have held the House and the GOP caucus would remain extremely conservative. Maybe the House leadership would have been more cautious if control was more obviously at stake, but who knows. I'm open to that possibility. Either way, that’s a separate argument from the two main reasons cited for why gerrymandering caused the shutdown: That it leads to extremely conservative districts and gave the GOP control of the House. And no, neither are strawman.
I'm not sure whether this settles any disagreements. But here's something to keep in mind: I didn't really disagree with anything that Weigel wrote about gerrymandering in his last piece. Gerrymandering does increase the number of Republican seats and it does make some Republican seats safer, as it did in North Carolina. But in other places, it increases the number of moderate Republicans and makes safe districts modestly more competitive. And the overwhelmingly majority of ultraconserative Republicans would continue to hail from ultraconservative districts, either way. The aggregate effect on polarization is modest, at best.