142

This image has been shared on social media

Gerrymandering in North Carolina

Transcription:

Gerrymandering in North Carolina

  • 1,747,742 votes for Democrats = 3 Congressional seats
  • 1,638,684 votes for Republicans = 10 Congressional seats

Example sources: [1], [2]

Are these numbers correct?

Community
  • 1
DJClayworth
  • 57,419
  • 26
  • 209
  • 195
  • 4
    Asking whether the numbers are correct is a legitimate question, but I'm tempted to ask another question about whether the seats in NC have been gerrymandered. It might be a duplicate of https://skeptics.stackexchange.com/questions/40256/does-gerrymandering-have-a-significant-effect-on-the-number-of-congressional-sea though. – Andrew Grimm Nov 14 '18 at 21:29
  • 5
    You call *this* "gerrymandered"? Son, you wouldn't know gerrymandering if if jumped up and kicked you in the behind. You want gerrymandered? Look at the Ohio congressional districts, in particular the Ohio 9th and 11th (my district). These are "designer districts", intended to capture many of the Democratic voters in two districts which between them span nearly the width of the state, and keep the surrounding districts "safe" for Republicans. – Bob Jarvis - Слава Україні Nov 15 '18 at 00:32
  • Do you know the total popular vote for Ohio in the House elections? – DJClayworth Nov 15 '18 at 01:16
  • 1
    Maryland is at least as bad as Ohio. Look at almost any [Maryland congressional district](https://en.wikipedia.org/wiki/Maryland%27s_congressional_districts). Most are not even contiguous. – President James K. Polk Nov 15 '18 at 05:20
  • Just look at the statewide contests. Democrat justice John Arrowood (who is openly gay btw) won by over 50% of the vote. Dems absolutely took the popular vote. – James Jones Nov 15 '18 at 17:08
  • @BobJarvis - some numbers for state house races in Ohio: https://www.cleveland.com/expo/news/erry-2018/11/0f32e762411182/ohio-democrats-outpolled-repub.html – Matt Burland Nov 15 '18 at 19:07
  • 2
    I predict the democrats have a majority in densely populated urban areas and the republicans have a majority in thinly populated rural areas. So a map may show tiny blue areas and large red areas. To counteract this, the voting districts would have to extend very far outside city limits, which may appear unfair also. – Chloe Nov 15 '18 at 22:38
  • Could you please add link to where the image was posted? I am interested in context, acceptance and other background – Croll Nov 17 '18 at 19:44
  • @Croll Links in the question. Also multiple Facebook shares. – DJClayworth Nov 17 '18 at 21:31
  • Impressive showcase of gerrymandering. I wonder how far the range could go if one would draw the voting district lines differently. – NoDataDumpNoContribution Nov 18 '18 at 16:41

4 Answers4

167

Yes, the numbers are correct (within an error margin – probably due to different sources and time of capture).

According to the 2018 House election results (I used this handy Washington Post page), adding up numbers for NC, will give you the total of 1,748,173 votes for Democrats and 1,643,790 for Republicans – very close to the claim.

Ten of the seats went to Republicans and three to Democrats (Districts 1, 4, and 12), with most Republican wins being quite narrow and Democrats wins overwhelming.

+-------+-----------+-------+-----------+-------+--------+
| dist. |     D     |  D %  |     R     |  R %  | Winner |
+=======+===========+=======+===========+=======+========+
|   1   |   188,074 | 69.8% |    81,486 | 30.2% |   D    |
|   2   |   148,959 | 47.1% |   167,382 | 52.9% |        |
|   4   |   242,002 | 75.0% |    80,546 | 25.0% |   D    |
|   5   |   118,558 | 42.8% |   158,444 | 57.2% |        |
|   6   |   122,323 | 43.4% |   159,651 | 56.6% |        |
|   7   |   119,606 | 43.4% |   155,705 | 56.6% |        |
|   8   |   112,971 | 44.6% |   140,347 | 55.4% |        |
|   9   |   136,478 | 49.7% |   138,338 | 50.3% |        |
|  10   |   112,386 | 40.7% |   164,060 | 59.3% |        |
|  11   |   115,824 | 39.5% |   177,230 | 60.5% |        |
|  12   |   202,228 | 73.0% |    74,639 | 27.0% |   D    |
|  13   |   128,764 | 46.9% |   145,962 | 53.1% |        |
+=======+===========+=======+===========+=======+========+
| Total | 1,748,173 | 51.5% | 1,643,790 | 48.5% |        |
+-------+-----------+-------+-----------+-------+--------+

Note: One caveat is that the Republican representative for District 3 ran uncontested. That is, it would be more appropriate to say that the result is 9 vs 3, as the total numbers don't include the voters in 3rd district.

Nat
  • 4,111
  • 2
  • 27
  • 36
sashkello
  • 3,762
  • 3
  • 28
  • 37
  • 33
    According to https://www.ncsbe.gov/ncsbe/, the unopposed Republican candidate in District 3 (Walter Jones) received 186,353 votes. So perhaps one ought to say that the total was 1748173 votes for Democrats and 1830143 for Republicans. Excluding the unopposed seat and calling the total 9 vs 3 seems a little bit like cherry picking. – Nate Eldredge Nov 12 '18 at 05:30
  • 68
    @NateEldredge I don't see it as cherry picking - "unopposed" means we can't really compare numbers properly, as we have no reference to what would a Dem candidate get there. In ideal world, in a randomly split 50/50 territory, we'd expect to get an equal number of representatives for each party. We just select a smaller territory, excl. district 3. Nothing wrong with that. You are welcome to introduce an edit with a possible alternative take on this, it doesn't change the answer in essence really, I don't mind... – sashkello Nov 12 '18 at 05:38
  • 1
    And do notice that the numbers by district tell a lot more, with the dems winning a few very large districts while the reps win a lot of smaller ones, thus skewing the numbers even further. – jwenting Nov 12 '18 at 06:32
  • 20
    @jwenting: They're not really "larger" - US Congressional districts (within a given state) all have roughly equal population, or did as of the last census. All we're seeing here is the number of people who voted - that turnout is some fraction of those eligible, which in turn is some fraction of the total population (non-citizens, minors, etc). Those fractions of course can vary between districts. – Nate Eldredge Nov 12 '18 at 06:48
  • 63
    It's more than just size. The Democrat-held districts all had massive majorities, with almost all the votes going Democrat. The Republican held districts had comfortable but much smaller majorities. That's exactly the sort of textbook distribution you try for in a Gerrymandering scheme. https://en.wikipedia.org/wiki/Gerrymandering – DJClayworth Nov 12 '18 at 14:23
  • 3
    @djclay Gerrymandering might be easier to prove if there was a picture and analysis of the districts map. –  Nov 12 '18 at 15:20
  • 87
    @fredsbend: There's nothing to prove. They openly admit to gerrymandering, and [even made it part of their public election strategy](https://en.wikipedia.org/wiki/REDMAP). It's not illegal, despite nearly everyone on both sides agreeing it should be, because the people who vote on the laws are the ones who directly benefit from it. – BlueRaja - Danny Pflughoeft Nov 12 '18 at 16:08
  • 13
    FiveThirtyEight has an [excellent tool to look at the impact district boundaries have](https://projects.fivethirtyeight.com/redistricting-maps/north-carolina/). You can see simulated boundaries that would favor each party, favor maximally compact districts, favor competitive elections, and more. Definitely helps put the current maps in context. – Toast Nov 12 '18 at 17:14
  • 21
    It's clearly *not* a terribly inequitable district breakdown, as the 538 maps show. It's just that most of the districts are rural, because most of the state is rural. The only way to swing it so that the Democrats could win is to split cities among as many districts as possible, so that the cities get as many representatives as possible while the neighboring rural areas lose all of their vote power. – hobbs Nov 12 '18 at 18:02
  • 10
    @hobbs, that comment cuts right to the heart of the matter. I wish I could award it a bounty. Also, FiveThirtyEight has a great article ["Hating Gerrymandering Is Easy. Fixing It Is Harder."](https://fivethirtyeight.com/features/hating-gerrymandering-is-easy-fixing-it-is-harder/) which backs up your point. – Wildcard Nov 12 '18 at 19:36
  • I think gerrymandering is a difficult charge to prove. Saying some line is more valid than another is surely impossible to evaluate unless a truly uninterested party were to look at it. I often get the impression that gerrymandering is a scapegoat for just not running a better campaign. –  Nov 12 '18 at 23:02
  • 8
    @fredsbend: Suppose that a manager hires only his friends and family. You may not be able to *prove* beyond a shadow of a doubt that he's hiring them out of nepotism (after all, maybe he just sincerely thinks that these are the best candidates?), but reasonable people can look at the situation and understand what's happening. – ruakh Nov 13 '18 at 01:41
  • 6
    @Wildcard A perfect fix to Gerrymandering may be hard, but having a neutral organization organize electoral boundaries instead of partisan ones would be a really good first step. – DJClayworth Nov 13 '18 at 01:47
  • 42
    @hobbs I think you came to almost the exact opposite view as me based on the FiveThirtyEight data. Their simulator shows that this is literally as inequitable of a split as it is possible to make--the current districts represent a more-or-less perfect Republican gerrymander. There is no way to split the districts in a way that gives the Republicans more seats, and literally every way they tested that wasn't an explicit Republican gerrymander gives them fewer seats. – Toast Nov 13 '18 at 02:50
  • @ruakh *"maybe he just sincerely thinks that these are the best candidates"*. That is *the* definition of nepotism. If gerrymandering is similar, proof is a matter of proving intent. However, many are prosecuted regardless of intent, having committed [crime x] in fact or effect. If gerrymandering can exist in fact (i.e. without intent), I've never heard its definition. –  Nov 13 '18 at 02:58
  • @Toast well... that's true, and as the article says, parties do tend to be pretty mercenary about districting. But it doesn't make terribly much *difference*; districts drawn on the basis of compactness would only have a probable change of one seat. – hobbs Nov 13 '18 at 03:15
  • 3
    @NateEldredge Excluding an unopposed district makes sense, as unopposed elections typically have much lower turnout than seriously-contested ones. Comparing vote totals from a seriously-contested election with an uncontested one would be misleading, as people who would have voted in a contested election would be more likely to stay home in an uncontested one, particularly if there aren't any other important contested elections in that district. – reirab Nov 13 '18 at 05:56
  • @jwenting That's just factually incorrect. Dem's won in the largest, 6th largest and 2nd smallest. – Kevin Nov 13 '18 at 13:52
  • 12
    @hobbs That's not what their data shows. Democrats won three seats this year, whereas under compact borders they should have expected 7 given the performance we saw last week. That's a four seat swing. Even under a "normal" year, Democrats would expect to win 4 or 5 seats under that map. – Toast Nov 13 '18 at 16:14
  • 13
    @fredsbend http://www.redistrictingmajorityproject.com/ - enough evidence for you? It's _literally_ part of the Republican party's campaign strategy. The politicians redraw districts to give a strategic disadvantage to their political opponents. They brag about doing it. You can find dozens of quotes where they assure their fellow legislators that their gerrymandering is over _political_ goals, not racial or sexual; they don't even try to deny it. – Nic Nov 13 '18 at 18:24
  • 1
    Isn't counting votes the wrong metric for gerrymandering? The results could be a turnout factor or voter sentiment issue more so than a gerrymandering one. For example, in district 3 the percentage of votes cast for the Republican dropped by 10% over 2016. Did the makeup of that district really change that much in 2 years? It would seem that the real measure of gerrymandering must be how many registered voters for each party there is - not for whom the constituents cast their votes. – CramerTV Nov 13 '18 at 18:39
  • @nick I did not say there's no proof of someone's intent to gerrymander. I said that a charge of gerrymandering necessarily includes that there must have been intent to gerrymander. I said exactly *"If gerrymandering can exist in fact (i.e. without intent), I've never heard its definition."* Without an accused party's confession, I find it very difficult to prove that gerrymandering did occur. –  Nov 13 '18 at 20:28
  • 13
    @fredsbend That's... literally what I just gave you. The Republican party explicitly admits to redrawing district lines for the sole purpose of giving themselves political advantage. Is that gerrymandering? It is. To say otherwise is a lie, because you clearly know better. Go on, then. Lie. Try to claim that, somehow, that's not gerrymandering. Pull another No True Scotsman. I'm done here. – Nic Nov 14 '18 at 07:27
  • @CramerTV Keep in mind that what was District 3 two years ago might not be the same District 3 as today. The whole point of gerrymandering is that they redraw the lines. (I have _not_ checked that, mind, as I couldn't find historical data. I'm a bad Googler. If District 3 has stayed the same, you're absolutely right.) – Nic Nov 14 '18 at 07:29
  • @sashkello In your ideal 50/50 world, we'd *expect* (in the sense of expected value) 6D and 6R out of 12 seats. However, in that same world, an outcome with any side winning 3 or less of these seats would still have probability of almost 15%, thus be far from any relevant level of significance - in fact, if the 50/50 situation held for 50 states (and there were 12 contested seats in each), we'd *expect* to find 7 states where the results look at least as suspicious as in the OP (though we'd expect some of these to be pro-D and some pro-R) – Hagen von Eitzen Nov 14 '18 at 10:07
  • @nic Calm down, bro. I'm talking about proving gerrymandering *generally*, not in this specific case. Again, I said I don't see how it could ever be proved without a confession. We all spent a lot of time looking at lines and numbers before you linked to a confession. –  Nov 14 '18 at 15:43
  • 2
    Oh, in case you need to hear it so you can move on, yes, this is gerrymandering. –  Nov 14 '18 at 15:46
  • 3
    @fredsbend Then why did you respond to my linked confession with "but how do we _know_?" – Nic Nov 14 '18 at 17:23
  • @NicHartley, I selected a bad choice for an example. Regardless of whether or not the district was redrawn, isn't the proper metric for gerrymandering the number of registered voters for each party as opposed to the actual vote tally? I get that 'votes cast' is a proxy for party preference but these numbers *could* vastly underestimate the scale of the issue. If many Republicans voted Democratic as a protest vote or a personal preference for the Dem, then it would much more clearly show the disparity - especially in the closer races. – CramerTV Nov 14 '18 at 18:14
  • @CramerTV You can look at registered voters if you're interested. As of November 10, 2018, [North Carolina reports](https://vt.ncsbe.gov/RegStat/Results/?date=11%2F10%2F2018) 2,688,975 registered Democrats and 2,112,229 Republicans, plus another 2,252,817 voters who are unaffiliated, and about 40,000 who are affiliated with some other small party, mostly Libertarian. It's broken down by county, not district, so you can compare mostly apples-to-apples if you want to see affiliation shifts over time (but not how that matches up to districts). – 1006a Nov 14 '18 at 18:14
  • @1006a, that would be an interesting statistical exercise but given that we know the answer in this case (apart from the numbers) my interest is not up for that level of analysis. – CramerTV Nov 14 '18 at 18:19
  • @CramerTV I mean, my only point was that districts are redrawn frequently -- that's how politicians optimize their gerrymandering. The thought process goes like, "Hmm, the margin on District 9 was a little low -- let's squeeze some more Democrats into 12, so we can put a few more Republicans in 9 and ensure our victory." That's part of why gerrymandering is so hard to quantify -- very little is constant for long enough to see a definitive trend. – Nic Nov 14 '18 at 19:40
  • 1
    @nic I think we were talking at cross purposes. I'm focusing on the meta question *how to prove gerrymandering?* You were focusing on the specific question *"how do we prove this instance was gerrymandering?* By linking to the confession, you demonstrate my point. I don't see how we can prove gerrymandering without a confession. Lines can look suspicious, surely, but without an overt statement of guilt, I think any line can be argued as legitimate and reasonable. –  Nov 14 '18 at 22:04
  • @Nic Check out what Dunk has commented under the community wiki answer below. –  Nov 14 '18 at 22:11
  • 1
    @fredsbend ...Oh, gotcha. Yeah, you're right, we were arguing at different things. Sorry about that. Though I'd still argue that you don't need to _intentionally_ gerrymander to still, effectively, gerrymander -- it's just as harmful either way, and IMO it doesn't matter if it was an accident, for an unrelated reason, or intentionally for political gain. IMO, if you draw district lines to do anything but group communities together, something is wrong. – Nic Nov 15 '18 at 02:29
  • @sashkello, yes "we can't really compare numbers properly". So we should through out what we can't compare. Or, at the very least, add disclaimer. – Paul Draper Nov 15 '18 at 09:00
  • @PaulDraper In my opinion, my last paragraph is a reasonable explanation. If you think more info is needed, then you are welcome to edit the answer. – sashkello Nov 15 '18 at 22:47
  • There is mathematical work from Duke Univ. aimed at quantifying the extent of the NC gerrymander by comparing the partisan results of the map with about 25k comparison maps drawn with Markov Chain Monte Carlo. The findings are that the map is an extreme outlier in the resulting distribution. https://arxiv.org/abs/1801.03783 – benblumsmith Nov 17 '18 at 02:26
  • Here is the Duke group's website: https://sites.duke.edu/quantifyinggerrymandering/ – benblumsmith Nov 17 '18 at 02:27
  • @NateEldredge - If one wants to sub in data from the 3rd district, IMO it makes more sense to use data from an election in which that seat was contested, such as 2016. It's imperfect b/c the totals aren't pure 2018 numbers anymore, but imo it's better than adding in voters who were not choosing between the parties. I did this and gave the percentages here: https://twitter.com/benblumsmith/status/1062168569025388544 – benblumsmith Nov 17 '18 at 02:39
56

This is a community wiki supplement to the other answer, which makes the columns easier to read and shows vote difference for each district. 3rd party or other votes are not included. 

District       D          R           Margin       Total Votes   Majority %
----------------------------------------------------------------------------
   1        188,074     81,486   (D) 106,588         269,560    (D) 69.8%
   2        148,959    167,382        18,423 (R)     316,341        52.9% (R)
   3              *          *             * (R)     186,353*       100%* (R)
   4        242,002     80,546   (D) 161,456         322,548    (D) 75%
   5        118,558    158,444        39,886 (R)     277,002        57.2% (R)
   6        122,323    159,651        37,328 (R)     281,974        56.6% (R)
   7        119,606    155,705        36,099 (R)     275,311        56.6% (R)
   8        112,971    140,347        27,376 (R)     253,318        55.4% (R)
   9        136,478    138,338         1,860 (R)     274,816        50.3% (R)
  10        112,386    164,060        51,674 (R)     276,446        59.3% (R)
  11        115,824    177,230        61,406 (R)     293,054        60.5% (R)
  12        202,228     74,639   (D) 127,589         276,867    (D) 73%
  13        128,764    145,962        17,198 (R)     274,726        53.1% (R)
------------------------------------------------
Total     1,748,173  1,643,790   (D) 104,383

* = uncontested, no votes are listed, same as Washington Post source.

Democrat candidates received 104,383 more votes than their Republican opponents. However, Republicans received 81,970 more votes overall (1,830,143 total), when including districts they were unopposed in. (Since the there was no challenger for district 3 it is impossible to calculate a meaningful Democrat-to-Republican margin for the total count. More or fewer people may have voted, some of the cast ballots may have gone to a different party, etc.)

Data from Washington Post.

Raleigh is in district 4.
Charlotte is in district 12.

North Carolina congressional districts

BurnsBA
  • 1,513
  • 1
  • 9
  • 22
  • As a supplement, including 3rd party/other votes and District 3 votes would be informative. Unclear why those are left out here. – chux - Reinstate Monica Nov 12 '18 at 16:11
  • @chux I found the data table difficult to read so reformatted it to make it easier. If you think it should include different data, please engage in the discussion in the original answer which already addresses why at least district 3 numbers are not included. – BurnsBA Nov 12 '18 at 16:38
  • 40
    Wow. That "margin" column paints more of a picture than the actual colored map. – PoloHoleSet Nov 12 '18 at 18:45
  • 1
    @polo Percentages might illuminate more and would be more appropriate for comparing, if someone wants to do the math. –  Nov 12 '18 at 23:04
  • 3
    Districts 4 and 12 suspiciously look like packing, while district one looks suspiciously like cracking. –  Nov 12 '18 at 23:07
  • 6
    @fredsbend-Or if you had a clue about the makeup of NC you would say that the districts are divided into very similar regions of concerns. District 12 is the city of Charlotte. Splitting the city of charlotte into any of its surrounding regions would mean that the people in the rural surrounding areas would get zero representation for their particular needs. Region 4 combined Raleigh/Cary/Chapel Hill and then throw in region 1 and you have the research triangle. 5 & 11 covers the mountains... – Dunk Nov 13 '18 at 18:54
  • 1
    In looking at the boundaries the reality of the situation is that these boundaries are extremely well designed to group people based on similar concerns. This is actually probably the fairest set of boundaries for providing the most people representatives that will look after their region's particular needs of any state in the country. – Dunk Nov 13 '18 at 18:56
  • 8
    I'd also add that if one were to split Charlotte which is surrounded by conservative regions then it is quite possible that nobody will end up getting elected to represent Charlotte (proper). And that is the problem that can't be solved without gerrymandering. How do you guarantee minority representation without it? – Dunk Nov 13 '18 at 19:04
  • @chux from district to district the number of parties that had a candidate on the ballot varied from 1 to 4. A total of 5 parties (Republican, Democratic, Libertarian, Constitution, and Green) had a candidate on the ballot in at least on district. 186,353 is the number of votes the Republican got in district 3. See my answer for further details. – DavePhD Nov 13 '18 at 23:48
  • @dunk You bring up valid points that tie in well to my comment exchange with Nic Hartley above. Any line can be argued valid and reasonable, so no charge of dubious line drawing for political gain (gerrymandering) can be proved without a confession. –  Nov 14 '18 at 22:10
  • 16
    @Dunk So it's reasonable that the 1.6 million republicans got 10 people to represent them but the 1.7 million democrats got 3? – Tim B Nov 14 '18 at 22:50
  • Could someone add and explanation of what the lines in the map are? (I presume the colours indicate the electoral districts.) – PJTraill Nov 14 '18 at 22:51
  • 1
    @PJTraill the lines are counties. The colors are districts. Image is from https://en.wikipedia.org/wiki/North_Carolina%27s_congressional_districts – BurnsBA Nov 14 '18 at 23:14
  • @PJT Counties are also permanent lines, or, it takes significant processes to change them. Districts are "regions of representation" which can change with little effort or public input. –  Nov 15 '18 at 02:04
  • @TimB That's american politics 101. Winner takes all while representation must come from population *and* region. Sometimes looks unfair, but often only if you neglect other things. –  Nov 15 '18 at 02:06
  • 1
    @Dunk " And that is the problem that can't be solved without gerrymandering. How do you guarantee minority representation without it?" You are actually arguing _against_ gerrymandering. Gerrymandering is about disenfranchisement not about districts that are shaped strangely (although strangely shaped districts can be an indicator of gerrymandering). I don't know enough to judge your claim that the NC districts are drawn fairly: the way to do that would be to ask the people who live in them (this is what redistricting reform is all about). – jberryman Nov 15 '18 at 23:16
  • Also the problems with redistricting are made worse by the fact that there are too few House seats (districts need to be more granular) https://www.nytimes.com/interactive/2018/11/09/opinion/expanded-house-representatives-size.html – jberryman Nov 15 '18 at 23:21
  • The margins chart may suggest Gerrymandering. The social media image doesn't. It only shows the reality of First Past The Post voting, that is sometimes called a "wasted vote" effect. Here in British Columbia, Canada, we are currently having a referendum on weather to keep our First Past The Post system or switch to proportional representation. https://elections.bc.ca/referendum/about-the-referendum/what-are-we-voting-on/ For instance our last election had disproportionate results but Gerrymandering isn't widely argued: https://en.wikipedia.org/wiki/British_Columbia_general_election,_2017 – rusl Nov 17 '18 at 01:18
  • 1
    @Dunk A research group at Duke Univ. aims to quantify the NC gerrymander by comparing the partisan outcomes of the map to 25k randomly drawn alternatives (drawn with Markov Chain Monte Carlo). They find that the NC map yields outcomes highly atypical in that distribution. https://arxiv.org/abs/1801.03783 – benblumsmith Nov 17 '18 at 02:45
  • Here is their website: https://sites.duke.edu/quantifyinggerrymandering/ – benblumsmith Nov 17 '18 at 02:45
  • 1
    @Ben Random lines would be to completely neglect real regions. I don't see what worth that could have. –  Nov 17 '18 at 04:52
  • @fredsbend - Read the paper; it's serious work. They use voting precincts as the basic unit of the map. They start the Markov chain with a real map, and the search algorithm prioritizes maps that avoid county splits, are geographically compact, equalize population, and satisfy a measure of adherence to the Voting Rights Act. The resulting maps are generally at least as good as the real map with respect to these standard principles of good redistricting. – benblumsmith Nov 17 '18 at 23:05
  • @Dunk No map I can find of Charlotte or Mecklenburg County has that distinctive notch in it, which looks way more like they were excluding the Republican parts of the Charlotte area than making a district to represent Charlotte. – prosfilaes Nov 18 '18 at 00:56
  • @benblumsmith-The main problem with that study (of which there are many) is deciding what is a 'fair' and 'impartial' definition of non-gerrymandered districts and how is that to be measured. The 'efficiency gap' is absolutely not a good measurement IMO and is certainly not something to be striven for. The efficiency gap is merely a way to add more weight to cities than they already have. Thus, any claims of 'atypical distribution' requires one to agree with the measurement heuristic being used before any credibility can be given to purported results. – Dunk Nov 26 '18 at 18:37
  • @prosfilaes-I have no idea what your point is but if you want to see examples of TRUE gerrymandering for political gain then you need look no further than the historical NC district boundaries drawn by the democratic party to give themselves an advantage in many districts. It is a mess of meandering interwoven snakelike boundaries. At least the new boundaries look reasonable. BTW, NC legislature was run by democrats for 100 years or so until 2010. So you can't blame republicans for any of those historical boundary monstrosities. – Dunk Nov 26 '18 at 18:51
  • @prosfilaes-Just guessing at your definition of 'distinctive notch' but it seems the 'notch' is disproportionately drawn on the above diagram and it seems to cover the south-westernmost cities in Mecklenburg county, in particular Mathews. If including people living in the same city as part of the same district is gerrymandering then you would have a point. – Dunk Nov 26 '18 at 19:05
9

According to the North Carolina State Board of Elections & Ethics Enforcement, the results of the 2018 election are as follows. (Parties are ordered by number of votes):

District 1
    Democratic Candidate: 190,445
    Republican Candidate: 82,209

District 2
    Republican Candidate: 170,050
    Democratic Candidate: 151,966
    Libertarian Candidate: 9,654

District 3
    Republican Candidate: 187,901

District 4
    Democratic Candidate: 247,067
    Republican Candidate: 82,052
    Libertarian Candidate: 12,284

District 5
    Republican Candidate: 159,915
    Democratic Candidate: 120,462

District 6
    Republican Candidate: 160,636
    Democratic Candidate: 123,601

District 7
    Republican Candidate: 156,797
    Democratic Candidate: 120,804
    Constitution Candidate: 4,665

District 8
    Republican Candidate: 141,371
    Democratic Candidate: 114,057

District 9
    Republican Candidate: 139,246
    Democratic Candidate: 138,341
    Libertarian Candidate: 5,130

District 10
    Republican Candidate: 164,969
    Democratic Candidate: 113,259

District 11
    Republican Candidate: 178,012
    Democratic Candidate: 116,508
    Libertarian Candidate: 6,146

District 12
    Democratic Candidate: 203,974
    Republican Candidate: 75,164

District 13
    Republican Candidate: 147,570
    Democratic Candidate: 130,402
    Libertarian Candidate: 5,513
    Green Candidate: 2,831

Total

    Republicans: 1,845,892
    Democrats: 1,770,886
    Libertarians: 38,727
    Constitution: 4,665
    Green: 2,831

(Note: results are not yet official)

DavePhD
  • 103,432
  • 24
  • 436
  • 464
6

This graphic from the question leaves off the results from district 3. District 3 cast 186,353 votes for the Republican candidate and none for a Democrat (the Republican was unopposed). That flips the total to 1,830,219 Republican votes to 1,748,018 Democratic votes (a margin of 82,201). That's 50.5% to 48.2%. Presumably the other 1.3% went to third party candidates.

Source: Wikipedia.
Original citation for district 3. As that is the official source, someone could get the rest of the districts from there as well. Javascript required to change districts and view results.

Remember that the original claim was that Republicans won ten of thirteen races with fewer votes. That's demonstrably untrue, as the graphic only includes the votes from twelve of the districts. If it were leaving off the uncontested races, it should only have been nine of twelve contested races.

If the claim is instead adjusted so that it only compares the seat proportion to the vote proportion, there are several other states where it's the Democrats who won a higher seat share than their vote share. E.g. three out of four in Iowa with only 50.38% of the vote; five of five in Connecticut with at most 64.4% of the vote; nine of nine in Massachusetts; or California, where Republicans won more than a third of the vote but no more than half as many seats (two still undecided).

It also may be worth noting that in North Carolina in 2016 and 2014, the Republicans won by about 300,000 rather than less than 100,000. In 2010, Republicans had over 236,000 votes more than the Democrats but only won six of thirteen seats.

Brythan
  • 10,162
  • 5
  • 46
  • 53
  • 31
    Answers should stand alone, so the explanation that the Republican in District 3 ran unopposed is crucial and absent here. Whether or not, and how, the votes in District 3 should be counted for this comparison is debatable, but let’s give readers all of the information required to understand what is happening and make their own judgments. – KRyan Nov 13 '18 at 04:50
  • 7
    This non-answer is misleading at best. It's literally impossible for democrat votes to be counted in a district that didn't have a democrat running. So you're just assigning 100% of the votes to republicans. I assume that there are many democrats that voted, but didn't give a vote in that race. Are you going to count non-votes for democrats ? Otherwise you're just falsifying statistics. You're counting 100% of republican votes in that district but discard 100% of democrat votes. – xyious Nov 15 '18 at 21:49
  • 2
    @xyious But that's what the graphic claims: that Democrats beat Republicans in thirteen districts (not the twelve competitive districts). And nationally, there are more races with only Democrats, including races in California with only Democrats. Even in races that have both Democrats and Republicans, many aren't actually competitive. People often don't bother to vote if they know it won't affect This makes the national popular vote misleading at best in evaluating who would have won a proportional election. – Brythan Nov 15 '18 at 23:51
  • 5
    The point of the graphic, since you missed it, is that 51% of the votes are Democratic,but they only got 23% of the seats. In other words, a state that is technically marginally Democratic is 70+% Republican because of where the borders are drawn. This is a direct consequence of how voting regions are laid out. Give one party a few seats in exchange for a lot of seats. This is what people who point out gerrymandering usually refer to. Not that Republicans won districts with fewer votes, simply that the state's total vote doesn't reflect what the seats represent. – phyrfox Nov 16 '18 at 08:50
  • I realize, though, that's not actually proof of gerrymandering, but you can imagine how, say, if football fans got upset because their team got 51% of the touchdowns but still lost the game somehow. Anytime you have a representative (vs direct vote) system, you'll always end up with accusations of that sort. No matter how you look at it, it's definitely a flaw in the system; no one side should get 75% of the representation with only 49% of the votes. It's simply something that doesn't make sense to normal, reasonable people. – phyrfox Nov 16 '18 at 09:04
  • I fail to see how your last two paragraphs are relevant to the question. – Julien Lopez Nov 16 '18 at 10:19
  • @Brythan I see what you're saying. I just don't think counting the votes is better. I think saying the Republicans won 10 (instead of 9) is less misleading than saying the Republicans got 189k votes while democrats got 0. – xyious Nov 16 '18 at 16:30
  • 2
    District 3 isn't a race, if there is only one contestant. Hard to say someone won against someone else, when there wasn't a someone else. – Edwin Buck Nov 17 '18 at 18:09