Duke Mathematicians Investigate 2012 Election Results In North Carolina

[octoberlib]

Back in 2012, more North Carolinians voted for Democrats than Republicans in North Carolina’s Congressional elections. But Republicans ended up winning nine out of the state’s 13 seats that year. Those numbers piqued the interest of researchers at Duke, who decided to seek a mathematical explanation for the discrepancy. They recently published a study with their results.


"We share this hall with the physics department, physics starts somewhere right down there," Mattingly says. He opens the door to reveal an office that looks like it belongs to a math professor. There are books everywhere and a big chalkboard on one wall covered with half-erased equations. This is where Mattingly first got the idea to include one of his students, senior Christy Vaughn, in the mathematical conundrum of the 2012 U.S. Congressional Elections in North Carolina.

"One day I kinda had this idea that we should look at gerrymandering. And so I called her and I said I got an idea," says Mattingly. (Gerrymandering is the setting of electoral districts in an attempt to obtain political gain.)

"Right away I was very interested in this project because it’s just such a stark result that so few seats were awarded to Democrats when the popular vote was so different," chimes in Vaughn.

http://wunc.org/post/duke-mathematicians-investigate-2012-election-results-north-carolina

[drm604]

1. Gerrymandering is a HUGE issue.

Everything else flows from who is elected. If the winners aren't truly representative of the electorate then we don't have representative government.

It's difficult to solve, because even if you find a fair method of drawing districts, the party that is benefiting from the current boundaries is likely the one in charge so they won't want to adopt it.

[rickford66]

2. I have a solution.

A Constitutional Amendment stating that the minority party or parties in each state determine the legislative districts. This would be a self correcting system.

[bvar22]

3. I believe that California has instituted a program....

...that evenly divides the districts, and keeps them as close to a simple geometrical shape as possible.
As long as the software and hardware is Publicly Owned and certified by two independent agencies every two years, this could be a solution.

I don't live in California, so if I'm mistaken, please correct this.

[octoberlib]

7. This sounds like a good solution but we'd have to take back the General Assembly

first.

[Jim Lane]

8. The California solution is an independent commission, not software/hardware

Information about California's commission process is here and here.

There is no easy technological fix because there are multiple legitimate considerations and no objective method of weighting them.

For example, look at the map of South America as if it were a legislative districting. That border between Chile and Argentina is an obvious gerrymander, right? Well, no. It reflects the reality that today, and even more so when the line was drawn, the Andes Mountains presented a significant barrier to travel. It makes sense that someone on the Chilean side should be in the same district as someone 300 miles to the north rather than someone 50 miles to the east.

As that example shows, calling for a "simple geometrical shape" doesn't always make sense. In particular, there are a couple of factors that have a good claim to being objectively reasonable and yet tend to help Republicans. One is that district lines should, where feasible, respect existing political boundaries. "The First District consists of Skinnertown, Eladville, Earl G City, and part of Kosberg." If the municipal boundaries, dating back many years, are somewhat contorted, that district may be less than perfectly compact. Nevertheless, it makes it easier for people to know which district they're in, and it groups together people who presumably have some common interests. The trouble is that Democrats are often concentrated in larger cities. Therefore, this criterion will often produce a few urban districts that are packed with Democratic voters, leaving Republican majorities in a larger number of rural and suburban districts.

Another important factor is representation of ethnic minorities (in the U.S., blacks and Hispanics). History demonstrates that it's hard -- not impossible, but hard -- for people from these communities to be elected in districts in which they're the minority. The solution is the creation of "majority minority" districts. Linking together pockets of minority voters requires some districts that look preposterous, like NC-12 for blacks and IL-4 for Hispanics. Such districts also tend to concentrate large numbers of Democratic voters. (Those two examples have Cook Partisan Voting Index scores of D+26 and D+29, respectively.) That again enables Republicans to have the majority in more districts.

If you tried to redistrict based solely on a mathematical algorithm, you could split up some cities and abolish majority-minority districts.

[Botany]

4. In Ohio we have more Ds then Rs but the Republicans run this state and in 2012 both

President Obama and Sen. Brown won by big margins but only 4 out of 15
Congressional Districts went to the Ds.

[octoberlib]

6. It's not representative state government, that's for sure.

[valerief]

5. If a vote can be stolen through gerrymandering, it will be stolen. nt

[KamaAina]

9. So Duke is not a complete waste of oxygen after all?