I filled out my sample ballot last night. There was not much to fill out, actually. While there are several propositions and I am voting on all of them, I will not vote in legislative general elections under California’s new top-two exclusionary electoral system. Unlike some districts around the state, I don’t have any races in which the only two candidates are of the same party, but I also can’t vote for a third party candidate. I often vote for Libertarian legislative candidates, not because I like that party, but because I don’t like the big two, the districts I live in are safe, and the only third parties that used to be on my general election ballot were Libertarian and the far-far-right American Independent. (I am willing to vote lesser of multiple evils, but not of two!)
I am not voting in the school board race because none of the candidates has anything appealing to offer. I am not voting in the hospital board because why should anyone? I am not voting in the Ramona Planning Group election for two reasons: (1) the electoral system is stupid (top-8 plurality, with 10 candidates) and (2) this board twice has unanimously voted against a solar-energy generating field in Ramona (though the County Planning Commission recently overruled the board) and none of the non-incumbents has a different position. I will vote for one of the incumbents on the Palomar Community College district board, but only because I know her (the wife of my former train conductor–how’s that for a real social network?).
Is there anything else on this ballot? Oh, yes, I will vote for President–or rather for a slate of 55 presidential electors–but not for Obama. Again, because the electoral system is stupid, but at least here I have a choice among irrelevant options. I am voting for Jill Stein. I did vote for Obama in 2008, and I think he’s been all right under difficult conditions. I look forward to another four years, given how bad is the alternative that Ohio swing voters could inflict on us. But I have a long list of grievances against the Obama government, I live in one of the safest of safe states, and I am a dissenter against the partisan duopoloy.
I wish I could vote Green for congress and have it matter. The only Green on my ballot is Jill Stein and her running mate, and they have my vote.
I voted absentee in New York. It turns out I am still here due to bizarre circumstances but I honestly expected to be elsewhere now. Anyway I noted that the only races where there was an option other than Democratic and Republican were for President and federal Senator. Also, none of the races, including for the 29 presidential electors, are expected to be anywhere near close, so protest votes across the board would have been sensible if they had been possible.
Isn’t the new California system a disguised version of the French runoff system? But many French voters vote in the first round and skip the second round.
Seed planted by Ed — 06 November 2012 @ 15:26
Its interesting that the new California electoral system has created situations where a Democratic candidate has adopted the strategy of appealing to Republican voters in one district (the Berman-Sherman race in north-central LA), and a Republican candidate is trying to appeal to Democratic voters in nearby district (where the Democrats screwed up and didn’t get a candidate into the runoff in a Democratic-leaning district), where the runoff featured candidates from the same party.
Seed planted by Ed — 06 November 2012 @ 17:22
Two remarkable things about the re-election of the USA’s only socialist Senator, Vermont’s Bernie Sanders: 1, no American news media report that he is a self-described socialist; 2, he got 71% compared to Obama’s 67% in Vermont.
Seed planted by Wilf Day — 07 November 2012 @ 02:45
Winner-take-all rules in the electoral college: Even Maine, where its electoral votes are allocated in part by congressional district, gave all four votes to Obama (while 55% of voters in Maine voted to legalize same-sex marriage). And even in Nebraska, the other state where voters can split their support, all five of Nebraska’s electoral votes went to Romney. What a boring system. Thank heavens Washington and Colorado voters have made their states the first in the nation to legalize recreational pot use, with a measure setting up a system of state-licensed marijuana growers, processors and retail stores, where adults over 21 can buy up to an ounce. Adults could grow up to six marijuana plants in their homes. Some one claims this will put Colorado “to the left of the Netherlands” when it comes to marijuana policy.
Seed planted by Wilf Day — 07 November 2012 @ 03:13
What is going on in the north east of America-they now have two (admittedly Democratic-leaning) Independent Senators, and an Independent came a creditable second in the most recent Maine governors race. Three-party (or two-party plus non-party) politics doesn’t seem to have left these states smoking ruins.
Seed planted by DC — 07 November 2012 @ 06:49
Keep in mind that New England is greatly overrepresented in the federal Senate (in partisan terms, this cancels out the overrepresentation of the prarie and mountain states). Small rural states clustered near the Canadian border are really not representative of the rest of the country -there is even a secession movement in Vermont- so this is no trend.
The way U.S. politics works, effectively Sanders has been roped into the Democratic caucus and it will almost certainly work the same with King.
There was an old phrase, “as Maine goes, so goes the nation”, but that is from the time when states voted in the general election at different times (!). Maine voted in September, so its results were analyzed as exhaustively as the results of the New Hampshire primary are analyzed today. Actually both Maine and Vermont were rock-solid Republican states and not indicative of anything until they started moving to the Democratic column in the 1960s.
Voters in Maine have also developed the unusual habit of voting for the loser in close presidential elections (1960, 1968, 1976, 2004, though not this year), during its few decades as a swing state.
Seed planted by Ed — 07 November 2012 @ 14:56
The “its the electoral vote that matters” mantra is so ingrained now that its actually hard to get information about the popular vote. Right now, though not all the precincts have been counted, Obama has won a majority of the popular vote and Wikipedia has him up 1.9% over Romney.
This is actually on the fourth election since World War II that the Democratic candidate has won a majority of the popular vote (the others are 1964, 1976, and 2008). Republicans have done better, getting majorities on seven occasions.
Historically, when presidents in the U.S. get re-elected, their margin of victory increases. Obama’s margin shrank from his 7.2% margin over McCain. I could find three cases where the President won re-election with a reduced margin. One was FDR’s re-election for a third term in 1940, when he was defending a 20% + margin, plus the idea of the third term itself of course was controversial. The second case was Truman in 1948, who was defending FDR’s margin from 1944. The third case was Woodrow Wilson in 1916, of course the non-Wilson vote in 1912 was badly split and Wilson did increase his popular vote percentage. All these other cases had special circumstances so Obama’s re-election is something of an anomaly.
Gary Johnson is at 0.93% and I am rooting for him to get over 1%. In the U.S., anything over 1% is considered to be a good result for minor party candidates!
It looks like the re-election rate for incumbents, always high in the U.S. compared to other countries, was even higher than normal.
Seed planted by Ed — 07 November 2012 @ 15:07
Get rid of the top2. I’d prefer to have a 2-round system, first round with all the candidates and final round with the top n facing off.
Another example would be to have the first round with only the third party and independent candidates, allowing the top n qualifying for the final round with the Democrat and Republican already qualified.
Seed planted by Derek — 07 November 2012 @ 21:53
What’s the difference, Derek?
Seed planted by JD — 07 November 2012 @ 22:00
@Ed @7
Why rely only on majorities of the popular vote? The system relies on pluralities and if the candidates were campaigning for a majority of the popular vote they would campaign differently.
Why exclude 2000? Irrespective of the disgraceful shenanigans in Florida and the Supreme Court, I am not aware of anyone who contends that Bush received more votes than Gore in the national count. The more interesting statistic is perhaps that in the last 6 presidential elections the Democrats have achieved a plurality of the popular vote in 1992, 1996, 2000, 2009 and 2012.
Unless the Republicans abandon their focus on a limited section of the electorate (I accept of course that would mean an internal civil war with elements like the Tea Party) it would seem that each election will grow more difficult for them into the future.
I suspect one outcome of the various referendums is that opposition to same sex marriage will cease to be a viable campaign tactic. It’s not immediately obvious what could replace it.
Strangely enough the ALP has been moving along a similar trajectory since the middle 80s with a marginal seat strategy replacing the focus on battleground states that the electoral college forces on both US parties. The most dramatic example was the 2010 election where the ALP’s approach to migration and refugee policy was largely determined by focus groups in outer suburban marginal electorates.
The unhappy result was a sharp fall in the Labor vote in long established Labor bastions like Italians and Greeks and the fall of ‘inner-ring’ seats where those votes were determinative. A similar process saw the ALP actually lose the indigenous vote in the Northern Territory this year.
Slicing and dicing the electorate only works for so long.
Seed planted by Alan — 08 November 2012 @ 00:06
The problem with the variations of electoral college reform that have been presented in this thread is that they all come with huge transaction costs. Unless you imposed a uniform method of proportional representation on the states you would only be opening an extraordinarily fertile method for manipulation of district boundaries, voting methods, and seat calculations. It would be Ohio, Florida and (we have already seen this in California) attempts to have blue states vote proportionally while red states remained winner-takes-all.
Reforming presidential elections would be a massive undertaking and it is impossible to see any gain to democracy in plans that retain some form of electoral college without:
1. a uniform method of proportional representation in the electoral college
2. drastic reform to the contingent election procedure.
Arguably the most urgent reform is transferring contingent elections from 1 state 1 vote in the House to a joint session where each representative and senator has one vote.
Seed planted by Alan — 08 November 2012 @ 00:16
The issue that no-one mentions is that the House of Representatives itself is severely distorted. Think Progress is claiming a reversed majority. If that claim is sustained it only means that reforming the contingent election procedure takes on greater urgency.
Seed planted by Alan — 08 November 2012 @ 00:21
Steven Taylor at OTB made the great point the other day that the stagnation in the House is another instance of Framer-failure. It was widely assumed at the time that the House would have a rapid turnover due to the intemperate mood swings of the rabble (its why the Sneate and the EC were constucted the way they are). Instead 90%+ of incumbents can rely on being in Congress as long as they like, provided they can survive primary challenges (that rarely succeed anyway).
Seed planted by DC — 08 November 2012 @ 06:27
The difference between my proposal and top2 is that the average runoff won’t have only 2 candidates.
BTW, in order for the Libertarians or Greens to make waves in the US, and to start moving away from being a “third party”, there’s a strategy that could be used.
-Focus on states where it’s heavily pro-Dem or pro-GOP: for Presidential elections, states like Texas, New York, Alaska or Vermont are always ignored because these are either safe GOP or safe Dem; this is an ideal strategy for third parties, since they can appeal to Dem voters in Texas or GOP voters in Vermont, letting them know that their vote isn’t going to elect a Dem or GOP candidate in their safe state and they’re better off voting for a third party.
I believe that the Greens in San Francisco and DC used a strategy to become a second party in those heavily Democratic cities. This should be followed for all races.
Seed planted by Derek — 08 November 2012 @ 11:46
Meanwhile, in other news, I believe Xi Jinping is doing extremely well in early returns from the Chongqing province level municipality…
Seed planted by Alan — 09 November 2012 @ 02:39
Alan;
I would think the most urgent reform, in light of the reversal in the House, is fixing it and making sure it can’t happen again. That might not be enough; the House Republicans will spend the next two years preventing as much of the public business from being done as they can, rather like the Optimates during Caesar’s year as consul. You know how *that* turned out. I’m not expecting anything that dramatic, of course, but it’s still pretty bad. We shouldn’t let ourselves get used to the idea that the House doesn’t and shouldn’t matter — even though I’m rooting for Obama to act in exactly that way.
First, make the states use impartial commissions to do all their redistricting.
Second, make all the districts multi-member and proportional. I like STV. I like CPO-STV even better. (If you really wanted to, you could add lists to CPO-STV, since Condorcet methods allow ties.)
Third, require all districts to elect the same number of seats, and elect any remainder at-large.
And while you’re at it, expand the House.
I know all this is much more political reform than Americans are used to, but if the Democrats are smart they’ll go for it anyway. The reform will produce an immense amount of howling, but mostly by people who wouldn’t vote Democratic anyway. Once in place it’ll quickly be accepted, and the opposition will look bad in retrospect.
Seed planted by AaronArmitage — 09 November 2012 @ 16:54
@1 @8 @9, the Louisiana/Washington State/California top two system is not even remotely a “disguised version” of two round runoff. It is embarrassing to have to explain this again on a blog devoted to electoral systems.
The purpose of top two is to weaken political parties. The purpose of that is to make big money more effective outside the party organizations. Neither has anything to do with any purpose I know of for two round runoff.
Seed planted by Bob Richard — 09 November 2012 @ 18:14
@Aaron
I do not think an equal district magnitude rule would work. In a system that is going to contain districts like Wyoming where M=1 the rule would essentially mandate M=1for the whole country. It would be better to specify 1 vote 1 value and allow larger states to have districts with larger magnitude. Indeed you would probably specify a single electorate in each state below a certain number of seats.
Apart from the NPV, I’d imagine the main game in the next 4 years will be opposing Republican efforts to have blue states vote by some version of the proportional plan while red states continue to vote winner-takes-all.
In the long run I suspect efforts to suppress the vote rebound badly on the guilty party, but that long view may be a way over the horizon to party operatives for a little while yet.
Seed planted by Alan — 10 November 2012 @ 04:56
@Aaron also, but on a separate issue
I would not specify a particular form of STV. Doing so converts a debate on principles into an incredibly detailed exercise (that would invariably be called fuzzy math) about the benefits of the Hager-Bischendorf quota versus the Droop quota etc etc etc.
I think the language in the Irish and Indian constitutions has it about right with one tiny change. Where Ireland says:
I would say:
Seed planted by Alan — 10 November 2012 @ 05:13
Back towards the topic after re-reading MSS’s opening line… What’s a sample ballot for? I ask because anything California has now, Australia will adopt in another 20 years.
Seed planted by Tom Round — 10 November 2012 @ 05:26
Alan;
I wrote unclearly. The idea is that district magnitudes within a state must be uniform. So, for example, a state with 21 seats might have three districts of seven, but if it has 22 seats it might have four districts of five, with two seats elected at-large. That way, even if the redistricting commission is compromised it can’t give areas where one party has a little more than half the people an odd number but stick the other party with even numbers.
I am imagining this being done as a statute, so the actual rules would normally be spelled out, not just general principles. But maybe you’re right. Current federal law merely requires single-member districts. At about the same level of specificity, the new law would just require proportionality in some form and leave the rest up to the states. We badly need to add anti-gerrymandering laws, but strictly speaking it isn’t necessary to put the actual procedures into federal law.
Seed planted by Aaron Armitage — 10 November 2012 @ 14:03
With STV there is no real need for districts to return the same number of members so long as there is a uniform ratio between electors and representatives. There are perhaps, however, political advantages, to using districts of identical magnitude. Smaller states would each be one electorate. Where exactly to draw the line is problematic. I’m a bit radical and I’d go for a highish number like 12 as the cutoff.
It would probably be even more important to require impartial commissions to administer the elections as well as the redistricting. You could either have surplus representatives elected at large or you could allow one smaller electorate to take up the surplus representatives.
STV has a great advantage over MMP in that STV can be applied also to electing presidents, senators and representatives of states with only one representative.
Seed planted by Alan — 10 November 2012 @ 15:04
I’ve posted this before, but for the federal House of Representatives, nationwide STV would probably work as follows, assuming the current constitutional provisions on apportionment remained unchanged:
1. The half dozen states with one at large representative would continue to elect one at large representative by plurality vote.
2. States with two representatives would not use STV. They could either continue using two single member plurality districts, though I think now a change to two at large members elected statewide by plurality vote, with each voter having two votes, would be backed better by historical precendent.
3. States with between three and nine representatives would elect the entire state delegation, on a statewide basis, using STV.
4. Its with the thirteen states with at least ten representatives that would present problems of drawing smaller district boundaries. They account for almost half the members of the House.
While tullymandering is not as effective as gerrymandering, I think we can assume it would be attempted except in California, Washington, and maybe Florida. So I agree with Aaron that you probably want to constrain the mapmakers with a rule that the districts elect an equal number of representatives as much as possible. Also, the fragmented and convoluted nature of American local government boundaries do not lend themselves to clean electoral boundary drawing.
But I would avoid using at large Congressmen -imagine electing three at large representatives from the state of California- and instead adopted the rule that the district with the smallest number of representatives (“the remainder district”) would have to include the state capitol. This would be a single member plurality district or a double member plurality district in some cases. In California, the logical way to do this would be a three member district comprising Sacramento, Solano, and Yolo counties, though it would be possible to gerrymander a district stretching from Sacramento to western Contra Costa county certain to elect three Democrats.
And yes, you could use AV to elect the members from the remaining single member and double member districts. I’m assuming nationwide STV but no other changes to the system.
Seed planted by Ed — 10 November 2012 @ 17:16
That’s actually a very good point about 2-member states, and about including the locating the remainder district around the state capitol.
Historically Australia allowed lower quotas in remote districts and my thought was to locate the remainder district there, but with the exception of Tasmania, Australian states tend to have very clean transect where the closer you are to the coast, the more people you have.
STV works best where M is an odd number. It may be desirable to require that M is an odd number in the states and only the remainder district can have an even number of representatives.
Can I ask why you would retain plurality voting in the 1- and 2-member states?
Seed planted by Alan — 10 November 2012 @ 20:41
“Can I ask why you would retain plurality voting in the 1- and 2-member states?”
As I stated at the end of my comment, I’m assuming a switch to STV but otherwise no changes in the system. Using AV in one or two member districts would make more sense, but its hard to exaggerate the conservatism of the United States. Even when fairly sweeping changes happen, its actually the minimum change possible and all sorts of anomalies have crept into U.S. law that way.
I agree that odd-numbered STV districts work better. As it happens, of the states with ten Congressmen or more as of the last apportionment, New Jersey (12) and Ohio (16) would most cleanly divide into even-membered districts, though other states could divide into even-membered districts unless it was specified that odd-membered districts be used if possible.
There are eleven states in the 3-9 Congressmen range that currently happen to have even numbers of representatives, six of which are 4 congressmen states so it would be impossible to subdivide them into smaller multimember districts.
Seed planted by Ed — 11 November 2012 @ 02:49
How about open party list with the smallest proportional districts being 2 seaters?
States with 3 members must elected a three member district, 4 members must elected 4 member districts, 5 members must be elected 5 member district, 6 members must be elected 6 member districts, 7 members must be elected 7 member district, 8 members must be elected 8 member district, 9 members must be elected 9 member district, and then after that states can subdivided their districts if they have 10 members or more, but the minimum size of a district is no less than 5 members, then this rule other wises preserves proportionality and at least prevents extremes of gerrymandering.
It may be better to leave the districts as permanent, and allow the House to increase and decrease in size as time goes on and abolish the need to redraw districts, and districts as communities of interest,
Seed planted by Suaprazzodi — 11 November 2012 @ 03:26
@Suazzaprodi
You cannot use open list for electing presidents, senators or representatives for m=1 districts. There is no advantage to mixing up different systems in the same election. Moreover there is no benefit of open list, particularly at smaller magnitudes, that cannot be had more easily and more elegantly from STV.
Until quite recently I thought open list the better system, but the closed list advocates on this bog have persuaded me otherwise. The serious democratic weakness in open list is that an elector must vote for a party’s list without knowing the content of that list. Personal votes for candidates on open lists turn on FPTP within the list and that is just not a very precise indicator of electoral support.
You could perhaps argue for full preferential voting within the list but then you are asking the elector to cast a preferential vote among list candidates and then a non-preferential vote between lists.
Google Canberra, modified d’Hondt for the unhappy result.
Seed planted by Alan — 11 November 2012 @ 05:15
I got my math wrong on some of the points I made earlier.
First, with 12 congressmen, you can have four three-member districts from New Jersey. There is no need to have even-member districts.
California divides into 13 four-member districts, plus a one-member district that would comprise Sacramento and a few neighboring towns. If you required odd-member districts instead, it would divide into 17 three-member districts and a two-member district comprising Sacramento County.
California and North Carolina are the only states with more than ten congressmen where the number of representatives in the delegation are also a prime number. Prime numbers should get rarer as delegation size increases. In each case the smallest remainder would be single-member districts drawn around Sacramento and Raleigh. This would leave four member districts as the only solution for California, if you used the principle that the remainder had to be as small as possible. The other large states except Ohio could be divided into same size odd-member districts.
Seed planted by Ed — 11 November 2012 @ 06:34
On Alan’s other point, the open list system I’d like to see are one where the candidates are ranked, but if a voter votes for the list they are asked to vote for an individual candidate on the list (votes for the list only are counted as individual votes for the top ranked candidate).
If a candidate received a number of individual votes equal to some threshold, they are elected regardless of their position on the list. Essentially the rankings come into play for the candidates on the list you don’t get enough individual votes to reach that threshold, but may still get in to fill the quota of seats achieved by the list as a whole.
One merit of this is that this would mean at least a few members of the assembly would be elected as individual candidates, even if most got elected by means of a position on a party list.
I’d like to see some means where a voter could block his vote electing a candidate on the list, with the preference flowing to the next spot on the list, but that probably would make things overly complex and it would also mean the list system mimicking STV in a backhanded way. You might as well just have STV.
I greatly prefer STV myself, but there may be instances where electing large numbers of representatives from one large electoral district is desirable, though I can’t think of any offhand.
And also for US House of Representatives, as long as representatives are apportioned among states, STV would be the only way to get any sort of proportional representation, since most of the electoral districts would be too small for party list to be a good idea. That may be a rule for federal systems with a low average representative to state ratio (which in the US is 8.7).
Seed planted by Ed — 11 November 2012 @ 06:47
NSW is the largest STV electorate in the world. The legislative council has 42 members with half elected by the whole state every 4 years.
There was a rather manipulative ticket voting system, which was reformed after spectacular preference deals between microparties made the thing a farce. Tickets have been retained but they are limited to a party’s own candidates and it is for the elector to vote an order of preferences among the tickets. The elector retains the right to vote for individual candidates. I once made it all the way through 254 candidates but I accept i may be too much of an election junkie to survive any rational sampling.
Not very relevant to US House elections but it could work as a model for senates in the several states.
Seed planted by Alan — 11 November 2012 @ 07:28
Alan: there are lots of models state legislatures could use. The lack of variety in this area is appalling.
Seed planted by JD — 11 November 2012 @ 13:28
I am not sure what this has to do with the blank spots I left on my ballot, but then again, talking about my ballot is not of any general interest.
Yes, you can use OLPR and M=1. It is done for legislative elections in some districts in Finland (Aland Islands), Peru (Madre de Dios), and perhaps some other cases. It was formerly the system for electing the president of Uruguay, and once (from memory, 1985) for the president of Honduras.
I actually think there could be many reasons for preferring OLPR over STV, depending on one’s preferences (so to speak) over what an electoral system is meant to do. But this should be a topic of a planting all its own. I think we have been over it before, but one of these days I may just take up the question again.
Seed planted by MSS — 11 November 2012 @ 17:56
Alright, erm, what’s so stupid about using multiple-vote plurality to choose 8 out of 10 candidates?
Seed planted by JD — 11 November 2012 @ 18:24
OLPR would certainly work with M+1, but I cannot really see either major party nominating more than one candidate for one president. Doesn’t OLPR simply revert to FPTP under those conditions?
Seed planted by Alan — 12 November 2012 @ 03:11
@22 Alan writes, “With STV there is no real need for districts to return the same number of members so long as there is a uniform ratio between electors and representatives.” Voting power is indeed measured by the uniformity of the ratio of representatives to voters. But it is also be measured by uniformity of the threshold of representation. Consider two small parties, one whose support is concentrated in urban areas with relatively large district magnitudes and the other whose support is concentrated in rural areas with relatively small magnitudes. The former might well get representation while the latter gets none.
This is a compromise that is probably well worth making in order to make PR workable in countries or provinces with large sparsely populated areas. But that doesn’t make it completely right.
Seed planted by Bob Richard — 12 November 2012 @ 21:05
> “make PR workable in countries or provinces with large sparsely populated areas”
… or with a constitutional requirement (explicit for Australia, implied for the US) that districts not cross State borders and that districts be based on voter/ population numbers.
BTW, some proposed variations of STV and AV I have seen put forward for the US (eg, San Francisco) would cap the number of preferences allowed. If “number up to 7 candidates in order of preference” were the rule in a 7-seater, it would hardly change the degree of proportionality (most Tasmanian and ACT voters only number the 5 or 7 candidates of their preferred party, and stop there). But translated to “number one candidate for this one seat”, this form of STV would “reduce” to FPTP.
Seed planted by Tom Round — 12 November 2012 @ 22:03
On the topic of remainder districts being based in state capitals by law, that isn’t likely to happen in the US because most state capitals tend to be far more Democratic than the state as a whole (California may be an exception given how populous and Democratic SF and LA are). If that resulted in a two-seater in a two-party system, it’s essentially disadvantaging left-leaning voters in the capital.
For instance, Texas had 32 seats prior to the 2011 reapportionment (now it’s 36, which divides evenly into 3, 4, and 6, so it wouldn’t be as much of a problem). If M=3 were the norm, with 10 3-seat districts and a single 2-seat district, most of the state would have the opportunity to elect 2 R, 1 D or 2 D, 1 R (and perhaps in some rural areas 3 R), Austin-area voters would be stuck electing 1 D, 1 R despite the fact that it’s one of the most Democratic areas of the state.
I would think that with modern districting software, a requirement that the two-seat district be “the contiguous geographic area in which the average number of votes for the two largest parties in the previous four general election is closest to equal” would be fairer than an arbitrary requirement that it be based in the state capital.
Seed planted by Chris — 26 January 2013 @ 06:06
I actually like Alan’s idea of making the remainder districts the district with the lowest density population in the state, though it may be harder to write that into practice.
The situation in Texas, where Austin is a deep blue island in a red sea, is something of a political outlier in terms of US states. In fact the American tradition has been deliberately to put the state capital in an out of the way rural area so the legislators are not exposed to the corrupting influence of the big city. There are even a few states, Pennsylvania probably being the most prominent example, where the state capitol area is more Republican than the state as a whole.
Seed planted by Ed — 26 January 2013 @ 08:18
Perhaps an overstatement. All of the state capitals near TX that I know of (Little Rock, OK City, Santa Fe, Baton Rouge) are more liberal than the rest of their states, though OKC is still conservative, just more moderate than the rest of OK, and
Seed planted by Chris — 26 January 2013 @ 10:46
Sorry, got cut off…
Little Rock is very liberal for Arky, and at one point had a Green state legislator. I’d always assumed the high number of govt employees made most capitals lean left, though now that I think of it, perhaps it’s only the states where the capital is one of tbe largest cities where they lean left.
As for the density idea, I think it works for m=1, but for m=2 in a two party system, it’s always going to advantage the smaller party in the district unless it’s drawn so 2/3 of the vote goes to one party.
I actually really like the doubling rule of the Chilean binomial system for two seat districts where there are more than 2 effective parties/coalitions- I think that model could actually work really well for a place like Britain with 2.5 to 4 parties (4 in Wales and
Seed planted by Chris — 26 January 2013 @ 10:57
Chile’s binomial system is actually D’Hondt with two-seat constituencies (and open lists)…
Seed planted by Manuel Alvarez-Rivera — 26 January 2013 @ 22:33
Yes, I knew that. I don’t think the flaw in Chile’s democratic process is necessarily the Binomial system, though, but more the districts that are intentionally both malapportioned and permanently gerrymandered to favor the right, and that the Constitution basically requires the right’s consent to rectify this.
It doesn’t help that there are permanent coalitions of more than two parties and that each coalition cannot nominate more than two members per seat. In a seat where one list doubles the second place, this gives voters no choice in their legislator, and often means voters for center left parties don’t even have a member of their own party on the ballot; if the list does double, Socialist voters often see a Christian Den take the second seat though there are enough Socialist voters to elect two socialists.
Chile’s best remedy could just be for the Concertacion and the right to both agree not to run in their coalitions for one election and to let voters decide which individual parties they want. Apart from the malapportionment, this could create a very exciting race and wouldn’t be a tremendously unfair way to vote in a multiparty system.
In Britain, this too could be a fairer way to choose reps in a there party system, with a rising 4th national party in UKIP (though we’ll see if Cameron’s referendum plan gets them to vote Con), and strong regional parties in the SNP, Plaid, and Greens in places like
Seed planted by Chris — 27 January 2013 @ 19:53
Brighton. It may not be as fair as MMP, the additional member system, or larger list OR, but in a country like Britain, which desires overrepresentation for large parties to form strong governments, the top two Binomial system might be a possible compromise to determine representation between there major parties.
Seed planted by Chris — 27 January 2013 @ 19:56
[not sure what this discussion has to do with the original thread, but whatever...]
“Binomial system” is a term we should retire. The word, “binomial” just means “of two names” but tells us nothing about the actual electoral system. As Manuel already noted, the Chilean system is D’Hondt OLPR (though not proportional due to being applied in districts of) M=2.
Calling it “binomial” may make sense in Chile, where D’Hondt OLPR was in use already before 1973, and thus the one difference in the mechanics of the system from 1989 onward has been that lists can contain only two candidates, and all districts elect exactly two.
It is NOT a “top two” system. That would be SNTV (or MNTV). It is a list system, so the winners do not have to be the two with the most individual votes.
It tends to over-represent the second most popular list. It also tends to hollow out the center, because in most districts there is no incentive for candidates to compete for the district median. Rather, the incentive is to compete for your list’s median voter (in the vast majority of districts where each of the main lists can expect one seat).
While it served its purpose in Chile in the 1990s (as someone who was an adviser put it to me in the 1980s, that purpose being “to screw the left”), it is hard for me to imagine any reason why it should be considered a model to be emulated.
Seed planted by MSS — 27 January 2013 @ 21:14
We will have to resist the temptation to start speaking about monomials and polynomials.
Seed planted by Alan — 27 January 2013 @ 22:31
I think it went tangential when I was talking about STV in 2-member districts not being very fair.
I’m pretty sure “binomial” came to English from the Spanish. In most Spanish-speaking countries, “circunscripción uninominal” means SMD and “circunccripcion plurinominal” means multi-member district (and is often used for the list portion of an MMP or supplementary member system). Hence “binomial” means “two-member district.” I don’t think it’s related to the fact that lists only contain two candidates.
I was using top-two in reference to parties/electoral coalitions, though the d’Hondt system means that one list can win both seats if it has twice as many votes as the second-place list. However, I’m pretty sure that either the Constitution or Electoral Code (can’t remember which) actually just says that if the list with the most votes wins more than double the number of votes than the list with the second-most votes, that list wins both seats; otherwise the two lists with the most votes each win one seat and the candidate on the list with the most preference votes is elected. While this has the same result as d’Hondt, technically d’Hondt is not specified.
“Doubling” is me translating what I’ve seen referred to in Chilean academic and news articles as “doblaje,” meaning a list winning both seats.
The system isn’t as fair to coalition outside the top two coalitions as a system with larger lists, or STV in consituencies where M is greater than 4. However, it’s far fairer than FPTP, and lists outside Concertación and the right generally win at least a few seats at every election. It does generally overrepresent the second-most popular list; Chile’s gerrymandering and malapportionment compound this to a level that’s not inherently a part of 2-member d’Hondt OLPR.
It may not be ideal for a two-party system, because it generally overrepresents the second party, but in a country like the UK, with 3-4 major parties and where the top two are not the same in every region, could do well to implement the “binomial system.” The UK is not likely to implement any form of large list PR for election to the House of Commons. I don’t think their voters want a system which would make coalition governments more likely (as most list proportional system, including MMP or additional member, would), and they aren’t likely to vote for a system that gives groups like the BNP, or even George Galloway’s Respect, significant numbers of MPs. STV might have a shot, but I think you’d need at least 3-seaters to make that effective, and Britons seem to like their small constituencies. The “binomial” system would be more proportional than FPTP, likely giving the Lib Dems more seats, but still with a strong chance of the strongest party/electoral coalition winning a majority of seats (as Concertacion did at every election from 1989-2005), allowing a government to be formed without a coalition. I think that, while it’s not on the table now, people who want to see a more proportional House of Commons would do well to take a look at the binomial system.
Chile’s electoral system is flawed because the districts are gerrymandered and malapportioned, and the binomial system makes it extraordinarily difficult for the Congress to change that. But the two-seat d’Hondt OLPR system itself is not flawed, in my opinion.
Seed planted by Chris — 27 January 2013 @ 23:27
Though overall, if there are two remainder seats, the fairest way to distribute them may be to create two SMDs. That may favor the largest party, but at least it’s competitive, whereas 2-seat STV makes it almost a given that the second party will be substantially overrepresented.
If a polity is going to use the principle of two single-member seats for remainder seats, though, the lowest population density rule may not be wise, as it would tend to unfairly favor conservative parties.
Seed planted by Chris — 27 January 2013 @ 23:32
I think the Brits, due to the fact that they like their small constituencies, would use multi-member districts between 2 and 4 seats each, using STV.
Seed planted by Derek — 28 January 2013 @ 04:26
It is not yet a constitutional convention, but the UK has a clear record of referendums on constitutional and electoral matters. This is a country that regarded the alternative vote as a radical, dangerous and fearsome innovation.
What do you suppose the chances are of a referendum adopting an electoral system whose opponents can, without too much overstretch, describe as an invention of Augusto Pinochet?
Seed planted by Alan — 28 January 2013 @ 05:13
Probably not strong. It would take either a third party demanding its implementation without a referendum or simply being branded as 2-member proportional while hoping Pinocho doesn’t get brought up.
Though mentioning General Pinochet could very well make the Thatcherites more likely to support such a plan; they are, after all, nearly identical economic systems as well as similarly bellicose and with very strong executives supported only by a minority of the population (The Baroness Thatcher never won more than 43.9% of the vote in an election).
Seed planted by Chris — 28 January 2013 @ 06:12
“Binominal” would make more sense than “binomial”, for the reasons Chris notes ( generalization from uninominal). As for “binomial”, it is well established as an English word in both mathematics (as Alan alludes to) and biology, and I assume both fields got it straight from Latin rather than by way of Spanish.
Again, my point is that it is not specific enough. Either binomial or binominal makes (some) sense for describing the presence of two candidates of a given party/coalition or two legislators per district. But the essential feature of the Chilean system is as much the use of D’Hondt as it is M=2.
Seed planted by MSS — 28 January 2013 @ 15:56
‘Binominal’, combining a mix of Greek and Latin roots, would be macaronic, and therefore inherently evil. ‘Duonominal’, please.
Seed planted by Alan — 28 January 2013 @ 16:04
Last (I hope) comment on Chile’s two-seat D’Hondt: It is true that it has permitted a majority of seats for the leading coalition in most elections. However, usually the leader had a majority of votes, too (which would not be the case in the U.K.). In 2001 it manufactured a majority, but it was a close call: seats split 62-57 when votes split 47.9%-44.3%.
In 2009, there was a plurality reversal. The system inherently overrepresents the second largest party/coalition, and hence manufactured majorities likely would not be common, and plurality reversals an ever-present threat. I suspect (but won’t claim to prove) that such outcomes would be even more prevalent with M=2 STV or SNTV.
(I can’t tell how much of the Chilean pattern stems from malapportionment, but the patterns described, even if attenuated, would be inherent to the system even with fair apportionment.)
Seed planted by MSS — 28 January 2013 @ 16:13
John Carey has an article on the virtues of the Chilean/Pinochet system: http://www.scielo.cl/scielo.php?pid=S0718-090X2006000100016&script=sci_arttext#a
It uses “sistema binominal,” so it would appear that the English “binomial” is simply a mistranslation of the Spanish for “two-member system.”
I would absolutely agree that it isn’t a system to be copied on the whole (I’m frankly surprised the GOP hasn’t proposed it for congressional elections), but I would argue that the biggest flaw in the system is extreme malapportionment, and that the system would likely perform admirably were that not the case. In particular, the “reasonable mix of representativeness, governability, and individual legislator accountability” is something that most systems do not achieve, either in being far too proportional, and therefore inefficient (Israel’s current Haredi-versus-Lapid coalition negotiations come to mind), or in the FPTP flaw most prevalent in the UK, Canada, and the USA of not representing the votes of the majority of the population in the legislative majority. Were Chile fairly apportioned, I think the system would end up reflecting a Concertación majority until the right actually won more votes.
Seed planted by Chris — 30 January 2013 @ 04:52
As for not representing the center well, as alleged earlier, I think that’s unfair in that the center in Chile is largely the PDC (Christian Democrats), which have chosen to cooperate with the Socialist/Radical left over the right. They don’t have enough support as an individual party to take many seats, but then again, no party in Chile has enough support outside a few districts to do well on its own; if, as it has from time to time been suggested, the PDC and Renovación Nacional (the more moderate of the two main right-wing parties) formed an alliance, I think said alliance would take one seat, if not two, in most districts.
I think the M=2 system’s impetus to form pre-election coalitions, rather than contentions negotiations after elections, is one of the system’s strongest virtues.
Seed planted by Chris — 30 January 2013 @ 05:00
Not to belabor the Chilean point, but I was unaware that a vote to eliminate the binominal system was held last week. It received 23 of the 38 votes in the Senate; 25 were needed to enact the change.
http://www.infolatam.com/2013/01/23/chile-fracasa-de-nuevo-el-proyecto-de-reforma-del-sistema-electoral/
Seed planted by Chris — 30 January 2013 @ 06:09