Probably the most important thing to happen in this last election was the election of state legislators. This year is a Census year, and next year, the various state legislators will be reallocating Congressional districts (and state districts as well) based on the results of the census.
Classically, the party in control of the state legislature creates new districts that concentrate opposition voters in a few districts, while creating districts with small loyalist majorities everywhere else. The result is that opposition candidates win a few districts with 80% of the vote while loyalist candidates win everywhere else with 55% of the vote. Done correctly, and even a state that is 50-50 split between Dem and GOP voters could have like a 80-20 split in who is actually elected.
Is allocation of districts by state legislatures a good way to set Congressional districts? Is there a better way to do this? I for one find the process now icky, but I’m having trouble thinking of another way.
Help me out here, people.
Sua
There’s no good way to do this – or at least no way that someone wouldn’t find objectionable.
From time to time the courts are asked to rule on redistricting issues – e. g., an African-American neighborhood that suddenly sees itself split into four different districts. The results of letting the court do it are every bit as skewered as letting the legislature do it.
Nor would doing it strictly by census tract be an improvement, because some will argue that the tract lines are even more aribtrary and randomly-drawn than anything the legislators would come up with.
Given the choice between three bad alternatives, I’ll stick with the legislature. At least they make an effort to see that I will get grouped with my own little suburb (since I live on the northern edge of my town, in some proposals I’ve seen, I’d get lumped in with the people who live north of me.)
I saw a page that outlined the districts in either NC or SC. There is this one district that is MILES long, and sometimes only hundreds of yards thick. It takes twists and turns to gobble up a goodlooking street here and there, and avoid ones that are not quite as agreeable.
I will hunt around for it, but I thought it was a little crazy.
There is no explicit Constitutional requirement that Representatives be elected from districts (as opposed to being allocated among the states on the basis of population (save for the requirement that every state have at one Representative))
However, electing Representatives at large as a body has obvious problems as to how well they would represent the people within a state, and would likely be rejected by SCOTUS (as at-large city and county legislatures have been). Cumulative and preferential voting would almost certainly be rejected as violating the literal interpretation of “one man, one vote”, and proportional representation, whilst it might pass legal muster, would likely be rejected politically as insufficiently democratic (since the emphasis must be on the slate, not the individual candidate).
Districting, like democracy itself, is viewed as the worst method of selecting a government except for all of the alternatives.
But why not something simple and geometrical, like bands across the state? As population increases or decreases, you widen or narrow the bands.
I know the immediate objection to that is that districts should make geographical sense because, e.g., the people of Pittsburgh face different local issues than the people of Philadelphia.
But gerrymandering as it exists now pays only lip service to that goal. Even if we ignore some of the more egregious examples like the NC district noted by Freedom2, plenty of districts consist of a (predominately GOP) suburban region with a slice of a (mostly Dem) city, or vice versa, to dilute voting strengths. The issues of suburbia and urbia, despite geographic closeness, are often starkly different.
You give this issue to a bunch of neutral geographers, and they will give you a dozen different fair district divisions, depending on the criteria of “fair” used. As it stands, it’s probably the most naked power grab tolerated under our democracy.
Akatsukami, do you have a cite that suggests the interpretation needs to be strictly literal?
It seems to me that “one man, five votes” would be just as acceptable, provided it really was five votes per every man (and woman, of course). Unless they’ve already addressed this issue, I can’t see why the Courts would have a problem with allocating an equal, multiple number of votes to everyone, so long as one particular group of people wasn’t being singled out for benefit or disadvantage.
This is going a bit off my own topic, but I’ll use some (non-existent) OPer privilege. I don’t think “one man, one vote” is necessarily interpreted literally. Most local judge/school board elections I’ve voted in have a “vote for five” or something similar. I’m one man, allowed to cast more than one vote for a particular contest.
I think the Supreme Court case that established “one man, one vote” as a principle in apportionment was Baker v. Carr, 369 USC 186 (1962)
You can read it here:
It established the principle of “one man, one vote” for state legislatures. I don’t profess to know more about it other than that. I had a hard time understanding it.
States, from time to time, have elected all of their representatives “at large” if district lines couldn’t be agreed on.
At home, our redistricting as mandated by the Commonwealth Constitution is done officially by a 3-person committee, made up of the State Chief Justice plus two appointed members who must each represent a different party, which is given the budget to hire the actual technical staff (and borrows a lot of them from the tri-partisan State Elections Commission). The redistricting is then sent to the legislature for a straight up or down vote.
I agree that 1-man-N-votes, where N is constant for all men, may be a workable alternative for introducing proportionality to the ballots. AFAIK the big principle is that you vote directly for the person you want, that your vote(s) count as much as anyone else’s and that the election is won by the top vote-getting person(s) in the district. If someone decides to not use more than one of their N eligible votes, well, they are no more disenfranchised than if they stayed home. One possible objection may be from voters from traditionally-districted states, where they would habe been “stuck” with an either-or choice rather than being able to hedge; also, a political objection may come from M-O-R politicians, who’d worry that a population would throw their (N-(N-1))th vote en-masse to, say, a labor candidate, or a Green candidate. And from legislators like Congressman Gutierrez of Chicago, who totally control their little fiefdom and like it that way.
We have in our state legislature a combination of local-district and at-large membership. Each citizen gets to vote for two district Senators, one district Rep, and one each at-large Senator and at-large Representative (of a total of 11 in each house). For the at-large, each party presents a slate of between 1 and 7 candidates, the voter gets to vote for only one, and the members are seated in order of votes received. The parties rotate the names of the at-large candidates on the ballot in each precinct so as to give everyone a fair shot at the less-discriminating voter
Thanks BobT, but I don’t think that supports Akat’s contention either. Baker v Carr was about a Tennessee government that apportioned seats to its state legislature by county rather than by population (I’m seriously oversimplifying, but that’s the gist of it), with the result that the people in the less populous counties effectively had more votes, or stronger votes, than those in the more populous counties. Clearly in this case “one man one vote” is meant as opposed to “one man one vote, another man ten votes”, not as opposed to “one man N votes” with a constant N.
Assuming we’re going to stay with Congressional districts, here’s the sort of proposal that a mathematician might come up with. (I yam what I yam. ;)) My goal, in this plan, is to come up with the plan that, overall, has people in districts where they’re as close to everyone else in the district as is possible, for all the state’s Congressional districts.
In the Census data on which redistricting is based, every housing unit is geocoded down to the block level. So the legislators doing the redistricting know how many people are in each block. (Trust me that this applies to rural areas as well as cities.) Treat each person in the block, mathematically at least, as if they lived at the geographical center of the block.
Now, for any redistricting proposal, you give every person equal weight (you know their locations already) and calculate the mathematical center of mass of each district. Then you calculate how far each person is from the center of mass of their district. Sum up those distances for all the people in the state. The plan that minimizes that sum wins.
And if you followed that, take two aspirin and call your doctor in the morning if you’re still not better.
The problem with your plan is that it is based on fairness, while the plan it competes with is based on securing advantage. If legislators were concerned with fair representation then they would have passed similar laws long ago.
Now you’ve seen my mathematical approach to redistricting, I’ve got to say I’m a big fan of Lani Guinier’s proposal, alluded to in part by Akatsukami, that if a state has N reps in the House, each eligible voter could vote N times - and could allocate votes amongst candidates as they pleased, voting for N different candidates, voting for one candidate N times, or anything in between, so long as they cast no more than N votes for Congress. The N candidates with the highest vote totals, win.
I think her idea isn’t quite that simple: while smaller states would fit that description, larger states would be broken up into more manageable pieces so that Californians, for instance, wouldn’t have to decide how to cast fifty-some votes for Congress. But that’s the basic idea.
The reason I like it is that you no longer have the problem of being part of the ‘40’ in a district divided 60-40 between the parties, where your say is effectively nil. If you’re in a region that sends five reps to the House, you and your like-minded citizens can concentrate all your voting strength behind one or two candidates, and get some representation that you can support. And I suspect that it would, before long, lead to more ticket-splitting, and to more successful centrist candidates: people near the center would vote for candidates of both parties that they found acceptable, and not vote for the wingnuts.
(Even though I’m decidedly not a centrist, I think this is a good thing: I don’t want to see a system where my side shoves its ideas down the other side’s throat; I want my side to state its case in a way that causes the center itself to move, over time.)
Anyhow, that’s one of the radical ideas that got Guinier into so much hot water. I think she was a hell of a good ‘out-of-the-box’ thinker.
If Guinier’s proposal would mean that people in California could vote up for up to 52 different people for the House, then we could have a problem. I doubt few people know 52 people (let alone 5 people) qualified to serve in the House.
I remember when I was a student at Berkeley and there was a student Senate election. Proportional representation was used there. I was given a ballot and told to rank the candidates something like 1 to 16. I gave up after 3.
I agree that candidates for fifty-some seats would be far too many to keep track of. Hell, ten might be too many. Even a state like VA or NC would have to be broken up into 2-3 pieces for purposes of electing Congresscritters in this fashion.
