But one small district and two large ones is in fact more compact than three largeish ones. The compactness of a geometric shape is the ratio of its area to its perimeter. The total area of a state is fixed, so the most compact map will be the one with the shortest boundary lengths. Which rewards drawing the tightest boundary possible around a group of people.
Draw the lines of Trianglia so that the median district’s partisan lean is the same as the mean district’s partisan lean. You may end up with three red districts or three blue districts once in a while as partisan lean shifts, but you will never end up with the majority of voters voting one way and the majority of the seats going the other way.
In the Virginia House of delegates election in 2017 the median seat was dead even (decided by a drawing of lots). The mean seat was D+10 (1,075,206 R votes to 1,306,384 D votes statewide).
Similar things have happened in MI, WI, PA and NC (and possibly other states).
This should be the focus of anti-Gerrymandering efforts (it isn’t). The median district should be representative of the state as a whole and efforts should be made to draw lines that satisfy this criterion.
This sounds like a good simple way to detect most blatant gerrymandering, but it doesn’t wholly solve the problem: You could still have a state that’s 55% Red elect up to 100% Red Congressmen. On top of which, it might seem like a strange technical constraint to those drawing up the districts.
What all this discussion points to is that: Drawing districts is fraught with problems; Proportional representation would be much better.
Single-seat districts are defended as serving an Ombudsman purpose. In fact, that ombudsmanship role in Congress is often exercised corruptly; proportional seats might serve the role better — someone who campaigned for votes statewide from a specific group might serve their interests better. (Or a completely different mechanism might be developed to serve an ombudsman function.)
For example… we can look at the Cook Partisan Voting Index of every congressional district and compute the skew of state’s congressional district map to see which states are the most Gerrymandered by this metric. Here we will define skew as the difference in partisan lean between the mean district and the median district.
Here are the worst twelve:
Missouri R+7.750000
TennesseeR +7.333333
New York R+6.259259
Wisconsin R+ 6.250000
Louisiana R+6.000000
Mississippi R+5.500000
Michigan R+5.428571
North Carolina R+4.692308
Texas R+4.500000
Oregon R+4.400000
Ohio R+4.312500
New Jersey R+4.083333
It’s not until number 13, New Mexico, where we see a D skew.
Overall there are 28 states skewed R, 10 states skewed D, and 12 states with less than 3 CD’s that can’t be skewed.
Maryland, the poster child for the, “But they do it too,” crowd, is skewed D+1. Similar to Pennsylvania’s R+1 with their court ordered new(ish) district maps.
I say fix this first and then see if the other problems you mention arise.
Also… here’s the skew of the four states where the median district is of a different party than the mean district.
Wisconsin R+6.250000
Michigan R+5.428571
Minnesota R+3.750000
Virginia R+3.727273
Four for four Republican. It’s not really a both sides issue.
nm
Nitpick: It wasn’t decided by drawing of lots. It was decided by the Republicans saying that they had drawn lots, and then declaring themselves the winners.
I’m glad to see the discussion has become more balanced while I was gone.
Indeed.
For those who staunchly oppose gerrymandering: would you really support outlawing it within your blue state, even if all the red states were continuing merrily apace? That is the very definition of unilateral disarmament, which by turn is the very definition of Bad Idea Jeans.
It’s already outlawed in my very blue state, which has just gotten bluer. The efficacy of the new method used( citizen commission ) is arguable, but it hasn’t done the Republicans any great favors.
In Canada, election rules are determined by a federal organization, not by subnational units. In the US, the problem has to be solved 50 times. I don’t see why states should have any impact on federal election rules.
Maybe, but that’s not really the point of my post. The point is that the district was as even as even can be.
Here’s a post of mine from a thread about the Wisconsin Assembly 2018 elections.
The mean district was D+8.5. The median district was R+12.5. Sickening.
And I wish we had a federal education system, as in France. But we are getting way beyond the bounds of anything remotely possible under our constitution. Anti-gerrymandering efforts are going to result in unilateral disarmament in blue states, plus maybe some slight amelioration of the most extreme gerrymanders in some red states if the SCOTUS so rules in a best case scenario. (Let’s also note that Stephen Harper’s Conservatives had a majority government for a number of years, despite getting a distinct minority of votes, so I’m not sure if Canada is a great model to emulate.)
You don’t think it would make any difference if your very blue state scrapped the commission and just “gerrymandered the shit” out of the state as I propose? Republican lawmakers would be no worse off? Really?
A Canadian political party winning over 50% of the vote virtually never happens. There’s a few provinces where you see that (Alberta, for instance) but otherwise, no. I’m not sure that has ever happened since the creation of the NDP.
The rules aren’t particularly unfair, however. Smaller provinces tend to have more representation than they “should”. Rural areas tends to have higher representation, which favors (favours) the Conservatives slightly.
There’s nothing blatant. Urban ridings are basically a few neighborhoods stuck together, organized by geographyand generally rectangular. Even the most obvious examples (such as Chinatown) aren’t really bad examples. Toronto’s Trinity-Spadina riding includes a big chunk of Chinatown… and also the homes of many university students living off-campus, but nearby. While the area has a fairly large Chinese population, and of course offending the community means you won’t be elected, you don’t have to be Chinese to run or win there, there are many more Chinese voters living outside of the riding, and nobody has changed Trinity-Spadina’s shape every election in an effort to either increase or decrease the Chinese population of the riding.
But…
The alternative (eg proportional representation) would be replicating the system in Israel and many European nations where every government is a coalition of dozens of parties, and nobody actually gets the government they want.
We also avoid the issue of having a political leader whose party does not control the House of Commons. (Trudeau is the leader of the Liberal Party, the majority part in the House of Commons. We will not have a situation where Trudeau is the leader but the Conservatives have the majority. On occasion we do have an issue where the political leader failed to win their seat, so it’s not perfect.)
Statements like this make it difficult to eliminate gerrymandering. As you (should) know, American districts are sometimes gerrymandered to protect the voice of minorities. If your main argument is that gerrymandering “thwarts the will of the people”, then you have an argument for “good” gerrymandering, not an argument for getting rid of it.
Consistently on these boards ‘Mercians bemoan their political environment not having multiple parties while at the same time getting themselves in knots over about the necessity of winning a majority of the popular vote.
Have you really thunk out how it all works?
Don’t look at me. I don’t advocate for third parties. I’m a Democrat.
ETA: Just so people understand: Harper’s Conservatives parlayed a distinct minority of the vote into a comfortable majority of seats. Not a plurality, an outright majority.
But that’s a function of a first past the post parliamentary system, nothing to do with gerrymandering. I wouldn’t be surprised if it was similar for almost every majority Canadian government in the last 80 years.
When did I ever claim it was a result of gerrymandering? 
This thread is about gerrymandering and you said “let’s not forget Harper got a majority government with a minority of the votes, so maybe Canada is not the model to follow”. That’s a non sequitur because Canada redistricting system had nothing to do with Harper’s majority. So yes, maybe you shouldn’t switch to a FPTP parliament but that doesn’t mean our redistricting is bad.