If the supreme court abolishes gerrymandering nationally, how many seats do the dems pick up

Supposedly the abolition of gerrymandering in Pennsylvania means about 3 house seats will be picked up by the dems. I don’t know how many state level seats will also be picked up in that one state.

If the supreme court abolishes gerrymandering, and mandates that maps in all 50 states be fair before the 2018 election, how many extra seats will the democrats pick up?

Would an abolition on gerrymandering mean that districts will have to be written to not undercount votes in big cities as well as end gerrymandering? Isn’t that one of the big drawbacks the democrats have, their voters are all located in large cities? If they are using some kind of ‘vote wasting’ software test, would districts need to be rewritten so that large cities do not congregate all votes in one district?

In an electoral system which had no bias, the outcome in terms of seats would tend to match the outcome in terms of votes, as closely as the number of seats allows, with an equal chance of divergence favouring the Democrats and divergence favouring the Republicans. How many seats the Democrats could expect to gain at the 2018 3election under such a system would depend on their share of the vote in 2018, which is as yet unknown.

Obviously, though, there can be biases other than gerrymandering at work in the system, and therefore the elimination of gerrymandering will not necessarily remove all bias. An answer to the question raised in the OP requires us to identify, and then quantify, the biases other than gerrymandering that feature in the US electoral system.

In addition to gerrymandering and structural bias from city demographics, there’s also some bias due to the granularity of the House and the fact that districts are bounded by state boundaries. Thus, for instance, Wyoming residents have about twice the representation per capita that most Americans do, since the state can’t have half a Representative.

Right, we would have to unfreeze the number of representatives and increase the total number during the 2020 census. This would allow very large population centers to get cut into smaller pieces, and that would probably favor Democrats.

It’s admittedly a tough problem. If you increase the number of representatives every ten years you eventually end up with unworkable numbers in the House. If you maintain the freeze, you continually dilute the representation of densely packed population centers.

I do think the Representative freeze (or the end result as noted by Chronos) is pretty clearly against the intent of the Great Compromise. Citizens in sparsely populated states are supposed to have more per-capita representation in the Senate, not in the House.

I did a rough calculation just based on pure proportional assignment based on the 2016 vote. So if the Democrats got 50 percent of the vote in a state, they would get 50 percent of the seats. Based on that, they lose 18 and gain 38 for a net of 20.

Of course, there’s lots of rounding there. In Arkansas (4 seats) they got 12.77 of the vote, giving them .51 seats. Does that mean they should get a seat in Arkansas? There’s no way to create perfectly “fair” system.

The problem with city demographics is not about cities not having enough representation per capita. It’s about Democrats having overwhelming majorities concentrated in these areas, versus Republican voters who are smaller majorities in bigger areas.

So for example, suppose for simplicity that election districts were 1 million people apiece. You have one district comprised of a city which has 900K Ds vs 100K Rs. There are 3 other districts, each of which has 600K Rs to 400K Ds. If everyone votes based on the above identifications, the Ds will have 2.1M votes and win one seat, while the Rs have 1.9M votes and win 3 seats.

Breaking up the cities into smaller pieces won’t change that. The only way around that is to deliberately gerrymander the districts and combine some of the city people in with R-leaning areas.

It should be noted, though, that the city concentration is very helpful for GOTV efforts, and accounts (IMHO) for the D advantage in this area.

Don’t forget the I’s

Why would that be a gerrymander? Given an urban area surrounded by suburbs, you can either organise electoral districts on a doughnut model - one or more inner urban seats, surrounded by a belt of suburban seats - or on a slice-of-pie model - a number of seats, each of which includes some part of the urban area, plus the adjacent suburban areas. Why would one of these be a gerrrymander but the other not?

Is there any constitutional or other high-level rule in the US which requires elections to be conducted on the basis of one member, one district? If you had larger districts, each returngin a number of members, the task of securing representation proportional to vote share is hugely simplified.

There was a big discussion about this a while back (I believe in GD), and someone linked a study that nationally the difference would work out to eight seats (or sixteen votes when you consider relative gains). I remember that specific number because somebody claimed that 16 votes in the House wouldn’t make a difference and I posted the thirty-plus votes decided by 16 or less just in the year prior to that.

Because one follows “natural” boundaries and the other is being deliberately engineered to produce a certain political outcome.

Do you mean smaller districts?

Obligatory 538 link.

I don’t see how you could get multiple members per district by making the districts smaller.

There is nothing natural about the urban/suburban boundary; it’s entirely constructed, the outcome of things like zoning and planning codes. And even if we grant that it’s a “natural” boundary in terms of land use, it doesn’t follow that it’s a natural boundary in terms of electoral representation. In terms of social geography people live in the suburbs but work, recreate, etc in more urban areas. My life largely plays out in a slice of pie, not a doughnut.

The decision to use the doughnut model for electoral districts is just as much an attempt to secure a particular political outcome as the decision to use the slice-of-pie model.

No, larger. If the ratio of representatives to people is, say, 1:500,000, then a district returning, say, 5 members should have a population of about 2.5 million. You’ll have the same number of representatives returned from a smaller number of larger districts.

538 did a whole series on gerrymandering that was very interesting. The linked article concludes that there are lots of things causing problems and gerrymandering is probably a very small one of them. There was a different article that estimated the change in seats but I couldn’t find it quickly.

Cities and their suburbs are certainly distinct political entities, so separating them is perfectly “natural” in a context of politics.

And while many people do live in the suburbs but work in the metropolis, that doesn’t lead to pie slices. You can have one person who lives in a western suburb and works in a particular building downtown, while someone else lives in an eastern suburb, but works in a different building downtown, a few blocks west of the first person. In a pie-slice districting, one or both of them will still work in a different district from where they live.

Right. As far as I’m concerned, the constituents of a representative democracy are the demos. We don’t live in a metrocracy or a politocracy or whatever. As such, better districts result in representation that mirrors proportions of the people, not outlines of states. If 58% of people want a Republican representative, the ideal district maps result in somewhere between 57% and 59% of representatives being Republican. We should adjust every ten years to keep that as accurate as we can.

And this isn’t some whackadoodle idea: the wasted vote concept reflects it and has shown up in Supreme Court arguments.

Cities and suburbs don’t have to be separate political entities; it’s perfectly possible for local government to be arranged with large metropolitan areas that embrace both the city and the suburb, on the grounds that they are interdependent and governing them in isolation makes no sense.

If you think it’s optimal at the local government level to separate them, you are doing that precisely so that the city and the suburb will be separately represented and governed. Taking that division and applying it for the purposes of state or federal elections makes no sense, since you are no longer choosing people to run separate city and suburban governments; you are choosing representatives for the state or federal government. There is no reason why the local government boundaries should be carried over into state or federal elections since the rationale for having those boundaries in the first place is not relevant at state or federal level.

And, in democracies with less clunky electoral machinery than the United States, a high priority is generally not according to having national electoral boundaries follow local government boundaries; it’s generally considered more important to draw boundaries which will secure equality of representation (electoral districts having the same ratio of voters to representative) and equality of voting weight (no party secures an inbuilt advantage). If you want to call engineering those political outcomes “gerrymandering”, well, I’d argue that’s a very benevolent form of gerrymandering.

And such can be accomplished with a variety of proportional representation or ranked choice runoff systems … but most of those run into problems complying with Voting Rights Acts requirements (a result of not having minority majority districts).

In any case the only reason to pie slice a city and its surrounding regions is to dilute the populations in an effort to decrease voter efficiency, be it diluting the minority population in the urban core, or the Democratic one, or both.

Short of proportional representation systems structural features will commonly result in partisan voter inefficiency (see the 538 series now linked to twice) … we may need to accept that. But maps that blatantly increase voter inefficiency for partisan gain should be disqualified. Increasing voter efficiency, i.e. deceasing unfairness of the system and increasing the degree to which our representatives actually represent how we vote on a party basis, should be a major goal … just not the only one.

Please no. PR promotes loyalty to party over loyalty to electorate.

Arrangements like that might make sense, but then you don’t have suburbs any more, you just have neighborhoods.