There is nothing to prevent a state from having at-large elections to elect Representatives. My question would be, what would be the effective difference between current elections by district and one where the top 7 vote-getters (using my state of Colorado as an example) go to Washington D.C.
Certainly someone popular in a large city like Denver or the Springs would get a lot of votes but a lot of candidates from those cities would split the vote while a few candidates from an area like SE Colorado running on water issues would split the vote fewer ways. Could minor 3rd parties sneak in and get the 6th and 7th spot more easily? Assuming the major parties would run 7-10 candidates in an effort to sweep the state, are they in danger of splitting their own vote or only getting the 2-3 most popular candidate elected if lets say Western Colorado really likes the candidate from Pueblo.
The devil is in the details, and the first detail is how many candidates a voter can cast a vote for. If there are 7 representatives to be elected, and each voter can cast 7 votes, then typically all 7 representatives will come from the same party, and each major party will nominate exactly 7 candidates. On the other hand, if each voter can cast just one vote, then each party will do some strategic nominating of exactly the number of candidates that it expects to elect, and urge its supporters to do some strategic voting to achieve the best number of candidates elected by the party.
With at-large elections, the fairest systems will be some kind of proportional representation, but they will be rejected as “too complicated” or “giving too much power to political parties”. (Even though, of course, neither is true.)
I can see how at-large elections with multiple votes (e.g., 7 votes for 7 candidates) would suppress minorities. However, using a PR system would make it much easier for minority candidates to be elected.
Wouldn’t it make it harder? Lets say there’s some visible minority that makes up a small percentage of the population, but it’s geographically concentrated, and assuming for fun that everybody’s a racist…they’ll only vote for a candidate that is of their own race. I know that in the real world, white people vote for black candidates and vice versa, but in this case, they don’t. With a district system, you can set up a district where the people of the visible minority is a majority, and guarantee a minority representative, but with an at large system, even with PR, the minority population is still going to fall below the threshhold for candidate election.
With single-member districts (using the 7 member state that I’ve been using as an example), to get one minority member elected, the best way is to set up one district with 1/7 of the state’s population, and at least one half of those members of the minority. So that’s 1/14 of the state’s population electing a minority member in that district. And you will do that at the same time as the party that controls districting sets the boundaries to favour their own party, of course.
With PR, you only need 1/8 of the whole state’s population to belong to the minority, and vote for a minority candidate, and automatically the minority will get a candidate elected. And you do that without ghettoising some of the voters in a minority district. (You also don’t set up a system where the minority candidate has to belong to a particular party: you’ll be getting Black and Hispanic candidates elected belonging to both major parties, as log as there enough minority voters across the state.)
So that makes it harder to get a minority candidate elected, because with the district system, the minority only has to make up 1/14th of the population of the state, but with your at large, PR system, the minority has to make up 1/8th the population of the state.
Yes, but it’s 1/14 in just 1/7 of the state. If the minority is not concentrated into a limited area, it might be impossible to get that happening. For example, I doubt if you could put together a Hispanic-majority district in Ohio, but across the whole state there should be enough Hispanic votes to elect one candidate.
There are 18 members of Congress from Ohio. 12.1% of that is 2.2 members – they’d be likely to elect 2 Black members, and might elect 3. (The mayors of the three largest cities in Ohio are all African American.)
I love the idea of rendering gerrymandering a thing of that past. But I’m not sure quite how the system should work. Ideally, any time you have a group that makes up n% of the population that votes precisely as a block, that group should be able to choose candidates and elect them and make up n% of the legislature. Which presumably means that if I just want to vote for democrats, I can pick some option which says “straight democratic ticket”, and if 55% of the population does that, and 45% picks “straight republican ticket”, then we end up with 55% democrats and 45% republicans in the legislature. That’s intuitively extremely fair.
But what if we have 40% who want “straight democratic ticket”, 40% who want “straight republican ticket”, and 20% who are like “well, I like Joe Smith(D) but not Bob Black(D) and really like Ron Wright ® and also Jim Fitch (I)” or what have you. I think you probably end up with a ballot that is WAY more complicated than what we poor Americans are used to… The straight tickets are tricky… you certainly don’t want some system where the democrats as a party get 10 thousand votes and can divvy them up as the party decides between candidates, and the republicans get 9000 to divvy up, and each is trying to game theoretically psych out the others as to where they’re going to put their votes.
Meh. Congressional races are run on district demographics. Why would we want 9 ‘senators’ in Colorado? This system would hurt the little guy candidate, as 1.) You’d have a hell of a time with primaries and 2.) the party could just pick the top 7 big name folks and go with it…where all 7 could be from the Denver area (mine).
It flies in the face of proportional representation. I don’t want some blowhard Republican from Colorado Springs not representing me because he’s focused on what got him elected in the first place (and it wasn’t Denver metro issues). Plus when candidates come home for town halls or whatever, they aren’t responsible for a district - they’d just visit the places where they are the most popular.
Me, I still favor Monte Carlo democracy, at least for many-member bodies (it’d probably be too swingy to use for something like the Presidency). If you want 500 legislators, then you choose 500 citizens at random, and each of them can pick whomever they want (including, if they want, themselves). There’s no paradoces of voting, no gerrymandering, no suppression of turnout, and fair and equal representation for all.
Why not apportion representatives alphabetically by last name rather than by geography? If protecting minorities is the goal, why is geography better than voting based on occupation, your hobbies, or even explicitly based on race or skin color?
I understand the historical reasons, I just wonder whether geographical representation is still justified now that it isn’t absolutely necessary.
The problem with a “top seven” approach is that it assumes all political parties are equal.
For example, suppose the election results are:
Smith 842,534 votes
Jones 751,847 votes
Allan 52,821 votes
Johnson 51,917 votes
Martin 51,055 votes
Scott 50,116 votes
Thompson 49,228 votes
Do you feel that the 95% of the voters who voted for Smith or Jones are going to feel it’s a legitimate result when they get two out of seven seats in Congress?
When we talk about 7 votes for 7 statewide candidates, I think we should allow a voter to divvy up his votes as he/she sees fit. He could cast all seven for one candidate, or could cast 1 vote for each of seven candidates, or 4 for one guy and 3 for another, etc.
This is called cumulative voting, and is a way to allow minorities a good chance to concentrate their votes to be sure they don’t get shut out of representation. Many corporations vote for their boards of directors this way and it seems to be a fair method.