Fusion voting is an extension of FPTP that relies on parties. Parties are not essential to FPTP, but they inevitably creep in, and fusion takes advantage of this. In classical FPTP at a general election, each party nominates a candidate, and the winner is he with the most votes.
Fusion voting allows two parties to share the same candidate. So if the Green Party decides that they are splitting votes off of the Democrats (and enabling Republicans to win), they ordinarily have one alternative: Don’t run.
With Fusion voting, they can nominate the Democratic candidate as their own. Now voters can “send a message” and vote for Gore on the Green ticket. If Gore gets 60% of the votes and half of them are Green-based, he can see that environmental issues are very important to a third of the country.
Advantages: “Sending a message”, as above.
Disadvantages: From my perspective, fusion doesn’t really have any disadvantages, but lots of folk question whether it will do any good in MA. Democratic legislators are rarely in any danger of losing their jobs to Republicans, so even if you send them a message, they probably won’t listen because they know they’ll get re-elected anyway. Competition comes in the primaries, where fusion voting (being based on parties) doesn’t help. There’s two state-wide jobs that Republicans can win (Gov and LG) and certainly fusion voting will help these races… but pretty much no others.
IRV stands for “Instant Run-off Voting” and is a form of “preferential voting”. Preferential voting means that instead of voting for one candidate, you rank the candidates and your second-place, third-place or even fourth-place preference actually matters.
In IRV specifically, if no candidate gets 50%+1 of the votes, then the candidate with the fewest first-place votes is eliminated from all ballots. (So if your prefence was A > B > C, and B is eliminated, your ballot is now treated as if it said A > C.) Lather, rinse, repeat. Eventually, a candidate will be selected.
Advantages: Vote-splitting and the tactical “vote for the lesser of two evils” are eliminated. You can put the Green as your first choice, the Democrat in second, and if/when the Green is eliminated, your vote will count for the Dems.
Disadvantages: As with all preferential voting systems, the ballot can get very large. Also, it’s possible (but not very likely) for a candidate to be second place for everybody and first place nowhere, and thus lose straight off even if he’s the one everybody likes (but not loves). Tactical votng still exists, but the effect is lessened.
Condorcet: Another preferential voting system, the ballot would look exactly the same as in IRV. Rank the candidates, and cast your ballot. Behind the scenes, candidates are compared pair-wise. Almost always there will be a candidate, A, who wins (head-to-head mano-a-mano) against any other single candidate.
Almost always? Yes, almost always. It’s possible to have three (or more) candidates where A beats B, B beats C, C beats A, and then there are a number of different mechanisms for breaking this three-way tie (or “Condorcet cycle”). For example, you might choose A because A’s margin over B is greater than B’s over C or C’s over A. Or you might run an IRV considering only A, B and C. Or…
Advantages: Avoids IRV’s possibility of dropping the candidate everybody likes (but nobody loves). Has all of IRVs other advantages.
Disadvantages: It’s complicated to explain; people may not “get it”. Tie-breaking may seem arbitrary. Fortunately, Condorcet cycles pretty much don’t happen… but if they do, the wrangling over the 2000 election might seem like child’s play. Cannot be extended for multiple-seat races (see below). A different forms of tactical voting known as “burying” can be used here.
Single transferrable vote: Another preferential system which can be thought of as an extension of IRV: You can elect more than one candidate at a time. Cambridge city council elections are run using this method. Nine councillors, all of them elected at-large (no districts), city-wide.
Instead of 50%+1, the goal is to get (1/(N+1))+1 votes. With nine councillors, you have to get 10%+1 of the votes. When a candidate is successfully elected, they are dropped from the ballots just like an eliminated candidate, except… well, if the criteria is to get 1000 votes and you get 1500, it hardly seems fair that 500 votes are “wasted”. So under STV, they are not vasted.
Two common ways of handling this: All 1500 ballots that elected candidate A are treated as 1/3 of a ballot going forward. After all, you can’t say which thousand ballots of the 1500 elected the candidate… OR, you can do as Cambridge does and randomly pick 500 ballots out of the 1500, and throw them back in the bin while the thousand that elected the candidate are removed from the process.
You can see the entire process here.
Advantages: Well, if you’re like me, you might sometimes question the validity of geographical-based candidates (especially when the district is heavily gerrymandered). I have a lot more in common with a liberal computer programmer in Springfield than I do with my next-door neighbor here in Natick. STV allows multiple districts to “merge” and jointly elect the candidates.
Disadvantages: Very complicated. Very, very, very complicated. Has all the disadvantages of IRV, plus is a bit harder to “get”.
That’s about it for this quick summary. There are many other voting systems out there, but none that are in use in this state or being proposed for the state.
For more information on systems I’ve mentioned and systems I haven’t, I highly recommend Prof. Douglas Amy’s “Proportional Representation Library“.
A quick note on tactical voting: all systems are vulnerable to some form of it. The question becomes how obvious is it, and how willingly will people do it. Arguably, the method is pretty obvious under FPTP and it’s quite common. One question to consider is whether any other system will make it rarer…
Great introduction to voting systems, dcsohl!
<
p>
I would like to add an explanation of Proportional Representation. For starters, Proportional Representation is an idea, not a system. Proportional Representation (PR) is the idea that the legislature should represent the various views of the public, in proportion to the prevalence of those views in the public. Basically, if some people like pepperoni on their pizza and some like mushrooms, why not get half and half?
<
p>
If 60% of the public wants liberals in the legislature, 60% of the legislature should be liberal. If 30% want conservative, 30% should be conservative. If 50% want women legislators, 50% should be women. If 15% want black legislators . . . If 70% want legislators from urban areas . . . if 25% want environmentalists . . . if 10% are libertarian, etc. That’s the idea of PR: the public if fully represented. There’s a nice animated illustration of this here: http://msnbc.com/modules/mockracy/
<
p>
To get PR, you first need multi-member districts. That is, districts have to elect multiple people at once, which is fairly uncommon in the US. The only multi-winner elections in the United States are at the local level. Town committees in Massachusetts are elected all at once, as are the at-large members of city councils.
<
p>
Once you have multi-member districts, there are several voting systems that can attain proportional representation for that district. Used for the Cambridge City Council and School Committee elections, Single Transferable Vote (also called Choice Voting) is one such system. There are also party-list systems, which are popular in Europe. Another system, called Cumulative Voting, has a limited history in the US, most notably in Illinois.
<
p>
I think the best system and the system that’s most politically viable in the United States is STV. It’s currently in place in Cambridge. It used to be in use in New York City, but it succeeded in representing too many minorities (racial and political), so the powers that be did away with it. It’s also been under serious consideration in Davis, California. As dcsohl said, STV is like a multi-winner version of IRV — IRV is exactly STV when limited to elect one candidate — and IRV has gained a good amount of traction nationally.
<
p>
Because of the traction of IRV, the viability of STV, and the synergy between IRV and STV, I think the long-term strategy for the electoral reform movement should be to promote IRV for single-winner elections, STV for multi-winner elections, and convert existing single-winner districts into multi-member districts.
<
p>
Also, starting at the local level has had by far the most success. Since the 2000 presidential election, when IRV rose in prominence due to the ‘Nader’ factor, we’ve seen several cities around the US adopt it, most notably in San Francisco. It was used in March of this year in Burlington, VT for the first time to elect their mayor, and I predict VT will be the first full state to adopt it in the next 3-4 years.
You said: “I think the best system and the system that’s most politically viable in the United States is STV. It’s currently in place in Cambridge.”
<
p>
Is Cambridge known for good governance?
<
p>
I’d always thought the opposite.
In Cambridge, STV has ensured that minority populations are sufficiently represented on the city council. Fairly representing the public is a necessary, but not sufficient condition for “good” governance. As for whether Cambridge qualifies as good governance, I’m not qualified to say.
<
p>
Moreover, my argument was not that it’s the best system because it’s used in Cambridge. It’s the best system because of its ease of use, strong resistance to tactical voting, and effectiveness at proportionally representing the public.
<
p>
I mentioned its use and history in the United States, including Cambridge, as evidence of its politically viability in this country. I don’t think any of the party-based systems would fly here. As for cumulative voting, it’s highly susceptible to tactical voting, and it isn’t synergistic with a good single-winner system like STV is with IRV.
PR has been a miserable failure to respond to change, and the Cambridge City Council a remarkable evidence of the fraud implied in equity voting. It’s not even that susceptible to tactical voting, since it’s “tactics” rely on a slate, and the slate died after rent control. Ironically, it is most representative of a local, bi-party, adversarial condition like Cambridge experienced during rent control: it depends on adversity to mobilize participation, and thereby exacerbates the differences in order to stabilize its most effective conditions. When those differences are economic, as they once were in Cambridge, they find expression in issues and they divide owners from tenants, old from new residents, rich from poor. As those differences elide, and as the city finds new and increasingly disdainful homogeneity through gentrification, the minorities intended to benefit from PR find themselves relentlessly excluded by a dictatorial and inaccessible City Manager system.
<
p>
PR depends on an overpaid and bloated city bureaucracy which exists largely to feed itself while claiming to help a smaller and smaller population in need. It undermines comity by establishing a dependency on a benevolent, un-elected, and increasingly authoritarian, elitist, and enabling bureaucratic class inaccessible to voter accountability or even to commercial controls. They, in turn, retain control by bullying and bribing elected officials with high salaries and incredible perc’s for less and less work in overseeing the public’s investment in its local government.
<
p>
No where is this more visible than in the city’s public health system, where increasing numbers of people of color die of despair and fear of health costs while the city appropriates increasing millions to their care. Death by cancer of all causes among black males has increased tenfold since Cambridge City Hospital was privatized into the Cambridge Health Alliance, while that same mortality ratio declined by a third among white males. Yet holding the bureaucracy accountable to outreach to the most needy frustrates even the most progressive members of the City Council and School Committee, many of whom are outraged at this oblique racial and economic genocide.
<
p>
This pattern is also evident in the high school, where more money means less service; in the police department, where higher levels of skill and police presence results in higher rates of suicide and homicide than in Somerville (an otherwise parallel community demographically); and in housing, where the poor are squeezed mercilessly into placements outside their own city by their own housing authority.
<
p>
Look carefully at the results of PR before observing its purported benefits.
Maybe it’s just Cambridge? Gentrification isn’t caused by PR, it’s caused by economic conditions. What leads you to believe that an at-Large or district system would be so much better? Just because you have districts doesn’t mean those districts are well (competently) represented.
You’re complaint is against the City Manager form of government, not STV or PR. PR has nothing to do with the City Manager form of government. You can have PR with a City Manager government or you can have PR with a strong mayor government. They are two completely different things, so please don’t conflate them.
<
p>
I happen to agree with the City Manager form of government, but that’s a separate argument for another thread.
JoeB writes: “PR has been a miserable failure to respond to change, and the Cambridge City Council a remarkable evidence of the fraud implied in equity voting.”
<
p>
I believe it’s fair to say that the Cambridge City Council is a failure on MANY levels, but I don’t believe it’s the voting method that’s to blame. Is the Boston City Council (or any other municipal legislature) any better? In Boston, I’d point to the “strong mayor” form as the main reason their city council is so impotent. In Cambridge, I think that just about everything the city council does (or doesn’t do) has so little impact on most residents of the city that they don’t really care who’s in office. The one notable exception has been in housing, once in the form or rent control and now in the form of “affordable housing” programs.
<
p>
Now back in the golden days of political patronage (and there’s still a little gold lying around), it really did matter who sat on the city council. It could mean the difference between a job for you and your uncle Jimmy and no job at all. Now that’s a good reason to get out and vote – PR or no PR.
<
p>
Joe B writes: “It’s not even that susceptible to tactical voting, since it’s “tactics” rely on a slate, and the slate died after rent control.”
<
p>
I’ll agree that slate voting in Cambridge died with rent control, but that just means that now everyone runs a campaign like the old “independents” did and there’s less gaming of the system. In itself, that’s not such a bad thing. There are, arguably, some other negatives to this. Primarily, slates were a good way to assist voters who didn’t spend the time to learn about all the candidates and issues and who appreciated some advice (regardless what you may think of the advice). I believe the best quality of slates is that their sponsors recruited and helped candidates who might otherwise not run or who wouldn’t be all that great at campaigning. Often these are the people you would LIKE to have elected but who might otherwise not get elected. In an ideal world, political parties would do this, but I believe they rarely do.
<
p>
JoeB writes: “the minorities intended to benefit from PR find themselves relentlessly excluded by a dictatorial and inaccessible City Manager system.”
<
p>
Again, I point to Boston or many other cities and towns in Massachusetts. Are you saying that Cambridge’s city manager is any more dictatorial or inaccessible than the norm? I’d say the fact that he’s a bit too comfortable after 25 years on the job and the weakness of the individuals on the city council are the prime causes of his unchallenged authority.
<
p>
You blame PR voting system for far too much. At some point, you have to hold the elected (and unelected) officials responsible for their actions. You can’t always blame the system. That said, I do agree that in Cambridge the manager and the city council are very prone to dipping into city revenues to lavish each other with salaries and perks.
<
p>
JoeB writes: “No where is this more visible than in the city’s public health system, where increasing numbers of people of color die of despair and fear of health costs while the city appropriates increasing millions to their care.”
<
p>
I can’t comment on this specific charge (and it is getting a bit far afield from the topic of voting systems), but I have to point out that just about every other city hospital simply closed its doors. At least the City of Cambridge has found a way to maintain a hospital and clinic system, albeit one that is frightenly dependent on continued funding through the free care pool.
<
p>
When Cambridge adopted its Plan E charter in 1940, there were two major changes. Yes, the election method was changed from plurality and district councillors to an at-large system with proportional representation. However, the decisive factor in that election was the change to a city manager form of government. The fact that the city council couldn’t pass a budget that year and that the mayor was on the verge of being indicted were also factors.
The voting system is all the public has to express its interest or disinterest in government, and the failures of the Cambridge GOVERNMENT are really the problem. Since all there is is voting, that bears the weight of transformations ranging from gentrification to racism, class, and dictatorship. The system – whether through PR or its inept practitioners – fails to maintain a community of diversity, vitality, and comity. That failure may be more because of a City Manager, a burgeoning tech-driven economy, or a restrictive zoning system that limits height and building replacements, but all we have to correct it is a City Council and that’s what gets the heat. Unless there is a broad-based coalition and a referendum on the Manager, which would – as it was in the last election – be the focus of substantial city public investment to kill it, there is no other means to bring about change.
<
p>
That’s why all Bob’s arguments are fine, but moot. A city council will protect a manager who buys city councilors. Funny that way.
<
p>
Incidentally, affordable housing is no better than any other city program through a political system itself flawed at the root. As long as rent control kept the city diverse, it hid the increasing concentration of money and limited the expansion of housing to replacements for existing units. Now zoning is pretty much all that’s left, and pretty much a failure at generating diversity. As long as affordable units are as expensive as they are in Cambridge, the community’s diversity is at best moot, at worst a total sham. If the Council had any guts they’d use North Point to re-generate a middle class, but they won’t because the Manager sees where the money is.
<
p>
We might, therefore, be more interested in how the City Manager has so effectively twisted city government to meet his goal of a lifetime tenure at over $200,000 a year. Since the Council won’t – or can’t – change it, there is much more hope that the guy will die and no one with quite his skills will succeed him. It’s more than ironic that a city that prides itself on left-leaning politics has the most autocratic form of municipal government, but such ironies are the pleasure we who left take at the expense of those who stayed.
<
p>
One tactic might be to limit the salary of city councilors to a ratio in proportion to the average of all adjacent communities. That would limit the Manager’s capacity to bribe them. Another might be to limit terms on the Council. That would at least increase the pool of bribed city officials and redistribute some of the public’s money. Another might be to mandate slates – for Bob’s comments are quite right about slate building, and, except for the defeat of rent control, there once was a dynamic pattern of city council elections. Limiting candidacies to slate candidates, and mandating no more than three or four slate-generating “parties” might make the system more responsive.
<
p>
Given how complicated such systems get, however, a strong mayor and a strong council is a much more accountable system. Change does occur when the prize is open to public scrutiny and public accountability. Public financing of municipal elections (using, ideally, the Bonifaz model at NVRI – http://www.nvri.org/), along with ANY kind of open voting on key and critical officials, is surely better than what exists in cities as diverse as Cambridge, Somerville, Everett or even Boston. All this discussion of the variants of the forms of governing is wearying and irrelevant if people don’t ever get their teeth in the meat and potatoes of how to spend their money to achieve real community. And the key is not the mechanism, but, rather, the vote itself.
Joe, do you have a cite for the stat on cancer – the tenfold increase?
<
p>
I know an oncologist considering working at CHA and found your observation interesting.
Hospital privatization + voting system = black male genocide by cancer! This has Haliburton’s fingerprints all over it.
I’m not in favor of an election system that requires multi-member districts, either in philosophy or based on experience. As was discussed, in Cambridge we have nine city councillors who are elected city-wide. To put it snidely, I have never figured out what most of them do. To put it more specifically, the “roles and responsibilities” seem to get distributed quite randomly, and according to the preferences of each councillor. This is no way to run a railroad. Example: in 2004, when Cambridge re-jiggered its property assessment equation and instantly doubled thousands of people’s property taxes, a lot of people got in financial trouble. One Councilor, David Maher, became the “tax abatement guy”, and helped people get abatements to ease the shock. While Maher’s action was good, this was not a good solution for Cambridge. People who got abatements were people who knew Maher, or whose “go-to” councilor was Maher. If your “go-to” councilor happened to be someone else; or if you didn’t know any of the nine of them, which is quite easy to do; you probably didn’t get as much help, if you got any at all.
<
p>
This leads to the philosophical point: the importance of having one elected offecial who is solely responsible for a given set of citizens. A city with nine at-large councilors is about as smart as a company with nine CEO’s. No one knows where to go for help, and the elected officials trade constituent service for ribbon-cuttings to an extreme that just isn’t as easy in a single-member district.
<
p>
PS, this dirth of accountability and a chain of command is one reason why Cambridge is run so exclusively by the City Manager: nine at-large, ineffective Councilors simply create a power vacuum.
How do you fell about mixed member representation? That is where there is a combination of single-member district representatives with multi-member at-large representatives. That’s how many city councils are structured around Mass, including in my city of Somerville. The district representatives provide the “go-to” person quality you’re looking for, and the at-large can be elected with PR to ensure all constituencies are fairly represented.
<
p>
I don’t take a strong position on MMP (mixed-member proportional system) versus pure PR systems.
I guess the mixed-member districts aren’t bad. I’m not 100% sure how exactly the at-large representatives (aldermen, in Somerville’s case) “ensure the constituencies are fairly represented” … I always thought their purpose was to give people with a popular incumbent a chance at a spot 🙂
Your summaries are good — I also recommend the Wikipedia articles on the various systems. Let me repeat a few points that Grace Ross of the Green-Rainbow party made the other day (see my recent post):
<
p>
1) She thought fusion was not a good way to build a third party. Suppose the GRP endorses a Democrat for an open legislative seat under fusion and they win with 48% Dem votes and 6% GRP votes. This is supposed to be the ideal situation for fusion because the winner knows she was elected with GRP votes. But in the next cycle she is an incumbent and has far less to worry about. In the meantime she is beholden to the Democratic party for committee assignments, etc. — there is little reason for her to pay much attention to the GRP. Under IRV, she would still have been elected but there would have been a GRP candidate earning that 6% and spreading the party message directly by campaigning. (This begs the question of whether building a stronger third party on the left is good for the Democrats or for the left in general.)
<
p>
2) The simplicity of IRV is a great advantage, particularly with the more intuitive name of “instant runoff”. If you already have a system that requires a runoff if it’s needed to get a majority, then you can sell IRV as a simple cost-saving measure.
<
p>
3) The Cambridge system is really more like proportional representation, which you didn’t mention, than like IRV even though it looks like IRV. Its great advantage is that it allows the formation of coalitions unknown to the voters to represent fractions of the electorate. If in Cambridge, 1/3 of the voters ranked every left-handed candidate above every right-handed candidate, at least three of the nine councilors would be left-handed even if no voter knew about the others. Cambridge does have the technical problem that the result can depend on the order in which the votes are counted, but it assures broad representation just as PR does.
<
p>
PR is the most generally used system in Europe. We might imagine choosing the state House like Israel chooses the Knesset, with the whole state being one election district. If the Democratic party remained intact, there would be enormous power in the hands of whoever ranked the Democratic candidates.
(If the Dems got 60%, the top 96 reps on the list would be elected.) But probably the party would fracture — we might even be able to elect one or two Netroots Technogeek Party reps. You lose geographical representation, though you could have a hybrid system with a bunch of election districts with, say, 10-20 reps each. We’d probably have to deal with more Republicans than we have now, but it would reduce the disconnect between the compositions of the legislature and of the electorate.
Of the many huge obstacles facing minor parties, none is greater than their inability to recruit candidates and thus present voters with a steady opportunity to vote on their line.
<
p>
The inability of minor parties to recruit candidates for down ballot races is particularly notable.
<
p>
Voting is a habitual behavior. If I decide to vote Green in a certain election then I’ll be looking for a Green candidate in the next election. And if there are no Green candidates, I’ll probably revert to my habit of voting (liberal) Democrat. In the last two legislative election cycles, voters in only a handful of Massachusetts legislative districts have been able to vote Green. So the majority of the 76,000 voters who voted Green in the ’02 gubernatorial race had limited opportunities to vote Green in 2004.
<
p>
Thus, minor parties may get a little buzz going when they run a presidential or statewide candidate, but they are unable to sustain their voting base over time because their supporters have so few opportunities to vote on the party’s line!
<
p>
Additionally, because so few of their candidates ever actually get elected, they can never play a role in actual governance.
<
p>
Fusion solves the problem that minor parties have with recruiting candidates by allowing voters who may share their ideologies to vote for another party’s candidate whose ideology is probably close to their preferred party’s (or they wouldn’t have received that party’s endorsement!).
<
p>
Contrary to Stein’s assertion, fusion is a good way to build a minor party, precisely because it gives voters regular opportunities to vote for that party, even if their vote is also cast for the member of another party. The greatest period of third party growth and relevance in our history is that period of time when fusion was universally legal and widely employed (1880-1920).
<
p>
Furthermore, under fusion minor parties can actually participate in governance because candidates that ascribe to their views actually get elected.
<
p>
Finally, with regard to committee appointments, progressive Democrats are often shunted aside by leadership. And if Stein thinks that a cross-endorsed Democrat would have a difficult time getting meaningful committee assignments, what makes her think that a candidate that get’s elected solely on a minor party’s line would get better committee assignments?!
My sincere apologies to all……..
I grew up in Cambridge, and I have always hated the Cambridge voting system. It’s just too confusing – everyone understands the notion of ranking votes according to preference, but I would wager that very few residents understand how those votes are counted.
<
p>
I am a big fan of cumulative voting for multimember districts. It seems to me much easier to understand than the Cambridge system or other variants; it’s easy for communities of interest to create coalitions; and it’s already widely used in the private sector. I have never understood why this system is not more widely used in public elections.
First, on the Cambridge voting system. I don’t think its necessary that a significant portion of the residents understand the algorithm for counting the votes. They see the system at the endpoints: 1) it’s easy to use and 2) it produces fair results. I also know there’s ample documentation and tutorials to understand more of it if I want. I don’t have to be a cobbler to know that my show fits.
<
p>
Another analogy for you. I have absolutely no idea what’s going on inside my car. I know how to drive it, and I know it doesn’t break down very often. I also know that there are smart people that I trust that have looked into, and I know their track record of decision making. As a result, I feel qualified to buy a car and drive it. Most people don’t know what’s going on inside a television or cell phone either — that doesn’t mean they cannot make good decisions about which tv to buy.
<
p>
As for cumulative voting, the potential for tactical voting can be mind-boggling. Sure, the mechanics of cumulative voting are easy and the counting is trivial, but the strategies for voting in an optimal way can be very complex. That’s why so few advocate for its adoption. Voting systems that require a lot of strategy disadvantage those that don’t have the time to learn the appropriate strategy.
<
p>
Under STV voting on the other hand, the optimal strategy 99% of the time is the obvious one: rank your candidates in order of preference.
Or at least they shouldn’t be. So I don’t buy your analogies. The less transparent the system, the more opportunity there is for mischief.
<
p>
As for STV vs. cumulative: here’s a question that I don’t know the answer to. Which system makes it easier for, let’s say, a small Latino community in a largely white multimember district to elect a Latino candidate?
I don’t really disagree with you on that point. The more transparency the better. But transparency is largely an orthogonal issue. We need voter-verifiable paper trails and random partial recount regardless of the voting system. Without transparency, any election is subject to serious mischief and error; and with it, it is are not. Our plurality elections are currently subject to a lot of mischief and error right now due to lack of transparency.
<
p>
Consider San Francisco for instance. SF recently enacted IRV for their city and county elections, and they’ve held two IRV elections so far. For transparency, they post all the raw ballot data online for anyone to inspect. Several people have independently downloaded and checked this data with their own vote-tallying software.
<
p>
So we need STV and greater transparency.
There are these little men who….
Another preferential voting system, the ballot would look exactly the same as in IRV. Rank the candidates, and cast your ballot.
<
p>
This isn’t quite my understanding. Generally, I thought in addition to ‘>’ you could use ‘=’. So, I could go A>B=C>D=E. It’s a bit of a nuance, but important.
<
p>
Two common ways of handling this: All 1500 ballots that elected candidate A are treated as 1/3 of a ballot going forward. After all, you can’t say which thousand ballots of the 1500 elected the candidate… OR, you can do as Cambridge does and randomly pick 500 ballots out of the 1500, and throw them back in the bin while the thousand that elected the candidate are removed from the process.
<
p>
I hate that Cambridge selects randomly. Really — what’s the rush? I understand the elegance of random selection: statistics and whatnot. Still, if you were interested, we’d randomly select 1000 voters in Cambridge and just have them vote. But we don’t — we count every vote. Likewise, Cambridge should too, using the “1/3” example you use.
<
p>
Thank you for educating the masses on this one. Good stuff. I’ll likely comment more later.
This isn’t quite my understanding. Generally, I thought in addition to ‘>’ you could use ‘=’. So, I could go A>B=C>D=E. It’s a bit of a nuance, but important.
<
p>
My only practical experience with STV/IRV is Cambridge city elections, where you strictly rank the candidates in your preferred order. You can’t rank two candidates equal.
<
p>
In your example, if A gets eliminated, does that ballot then count for B, C or 1/2 for each?
Condorcet. In Condorcet, if you have A>B=C>D=E, you’re doing pairwise comparisons. A lot of them.
<
p>
So, you know
A>B
A>C
A>D
A>E
<
p>
B=C
B>D
B>E
<
p>
C>D
C>E
<
p>
D=E
<
p>
For each pairwise comparison, you have your data. In the case of B=C, you’re not “voting a preference” for that particular pairwise comparison. This isn’t unreasonable either. I’m indifferent when it comes to choosing the KKK or the Black Panther Party. I’d put a big ‘=’ between them, because I just don’t prefer one over the other.
stormv writes (in discussing a preferential ballot) “Generally, I thought in addition to ‘>’ you could use ‘=’. So, I could go A>B=C>D=E. It’s a bit of a nuance, but important.”
<
p>
There are different implementations of the single-transferable vote (STV). The Cambridge system ignores all overvotes, i.e. two or more candidates getting the same rank. Thus in your example, the ballot would be initially assigned to candidate A and it could never be transferred to B,C,D, or E under any circumstance. If candidate A was elected with a surplus, such a ballot would not be transferred. If candidate A was defeated at some point, the ballot would be exhausted.
<
p>
The “Cambridge rules” are a special option for the tabulation software Cambridge uses. The default option uses fractional transfer methods and is independent of the order in which the ballots are counted. It also splits the weight of a ballot in the case of overvotes. There are “switches” in the software that allow you to use it for a wide range of variants of STV.
<
p>
I would love for Cambridge to change its system to a fractional transfer method, one that is independent of ballot order. In case of a recount, the law effectively requires that the paper ballots be matched one-for-one (as best as humanly possible) with the electronic records that were processed by the tabulation software. This is a nightmare which comprises almost all of the time spent in the recount (two weeks in 2001!). If ballot order didn’t matter and if an electronic recount were permitted, the whole process could be reduced entirely to visual verification of voter intent.
<
p>
Unfortunately, it would take a Special Act of the state legislature to make this change. There’s also the complicating fact that the enabling legislation for PR elections in Massachusetts (M.G.L. Chapter 54A) was repealed in 1972.
Does that mean that Cambridge has been operating its elections illegally since 1972? Oh how fabulous that would be!
David writes: “Does that mean that Cambridge has been operating its elections illegally since 1972? Oh how fabulous that would be!”
<
p>
‘fraid not, David. The state legislature tossed out Chapter 54A in 1972 and asserted that Cambridge would have to change its voting system. This was challenged, and the Mass. Supreme Judicial Court ruled that the state couldn’t end Cambridge’s voting system. The ruling is known as the Belin Decision (named for Don Belin who led the challenge). The decision is:G. D’Andelot Belin vs. Secretary of the Commonwealth, 362 Mass. 530. [1972].
<
p>
I have this and several other references (including Chapter 54A) posted at http://rwinters.com/elections/ballot2001.
<
p>
The Belin Decision was an important decision for several reasons and serves today as a precedent for cases involving home rule authority in cities and towns. In effect, if many Massachusetts cities had PR voting, then the state legislature could have abolished it statewide. However, if only one or two cities had it (one in this case), they retain home rule authority. This interpretation came up (in the opposite way) with the 1994 statewide abolition of rent control. Three municipalities had rent control then, and that was enough to make it a statewide issue (the unanimous Ash Decision). That set the stage for Question 9.
Thanks for the post, very helpful. I have decided to support the Condorcet system because it has by far the most impressive inventor. For a fulsome biography of the man, well worth reading, from the 1911 Enclopedia Brittanica, click here.
We don’t elect governments as unstable as Italy (50 since 1949.) Or as divided as Greece. Or weak as Germany. Or as far-far-far-left as Cambridge (what system does Berkeley, CA use?)
<
p>
The First-Past-the-Post works fine.
<
p>
It’s clear cut and places control into the hands of a single responsible politcal entity (party or mayor or governor.) No messy and ineffectual coalition governments. No premature collapses. Screw up, and out you go next election cycle.
<
p>
The complicated methods described might work for a tiny sliver of engaged primary and general election political junkies, but for the average Joe and Jane, why fiddle with what works?
<
p>