Down with Proportional Representation

Over at The American Conservative, Jon Utley makes the case that proportional representation is a bad deal for democracies:

The rules for economic development and effective government are proven and well known; what's less understood is why many societies are unable to adopt them. The failure is often blamed on their cultures or on corruption, but a common affliction is their political structures: nearly all have proportional representation (PR).

He argues that PR makes reforms difficulty and notes that the new Iraq democracy is based on PR. Worth a read.

Whole thing here.

Comments

[Isaac Bartram]

My father was a big fan of Tasmania's Hare-Clarke system of proportional representation. He also attributed many of Europe's problems to the system of Party List PR described in this article.

Each party runs nationwide, and its candidates are determined by lists controlled by each party?s machinery?usually old-timers who are owed favors and remember grudges.

Hare-Clarke is used to elect Tasmania's lower house and is also used to elect each state's slate of Federal Senators. It has no resemblance to the system described in this article.

[joe]

The problem he identifies doesn't seem to be caused by PR, so much as by a lack of a democratic process in choosing the party leadership.

There's no reason you can't have a PR system in which the parties have primaries to rank people on the ticket.

[tomWright]

I think the problem lies more with a parlimentary system than with Proportional Representation, (PR).

If we had PR in our U.S. system, where elections occur at fixed intervals and can not be called at other times, except for rare situations, and can never be postponed, I think it would be very stable.

Also add, that seats earned by a party would be apportioned to candidates based on vote totals, not time in office, and I see a system more stable than the parlimentary systems, and also permitting change of office holders.

Parlimentary systems are inherrantly unstable, in my opinion, the fact that they also permit PR allows the bad attributes of parliments to wash over and affect people perceptions of PR.

As far as national candidates representing everyone and therefore no one, that is a very valid point. I think it could be addresed either by keeping our current state-by-state groupings of Representatives and Senators, or maybe even by using multi-member districts. In each case using PR within those areas to apportion seats.

All systems have difficulties, I have not seen one that gets very close to perfection yet, however perfection may be defined. But I think some form or PR, that gives real voice to minority viewpoints, is needed.

Tom

[joe]

PR could be used effectively in the US by keeping the House as SMD, and making the Senate PR.

Given the role the Senate is supposed to play, the "old man" problem could be a benefit.

[Isaac Bartram]

My point was that the method of PR described in the article is not the only method out there. The Party List method the author describes has all the faults he attributes too it. The Hare system has virtually none of them.

Tasmania has both PR and a Parliamentary system. It has extremely stable government. The flaw is, of course that it is small (pop 400K). But before people begin dimissing PR they should at least examine the Hare system.

We will have to wait for thoreau to return from his dissertation defense to get an really knowledgeable comment on this. He should be starting any minute now, if I've calculated correctly.

Joe, I think you have your houses reversed. The Australians use PR to elect Senators to make it so one party cannot dominate a state's delegation (Australia has 12 Senators per state, regardles of population, so litle Tassie - pop 400k has the same as NSW - pop 4mill).

The one complaint levelled at Hare is that it is so mathematically complicated. After all we have voters that cannot master the device used to punch a chad out of a ballot, and you're going to try to explain to them that they have to pick several candidates in order of preference in order to select the one they want to have as their representative.

[Isaac Bartram]

Also, it's pretty hard to "proportionally" elect one Senator, although a preferential ballot* (aka "instant runnoff") would be a good thing IMO. It works in Oz 'coz each State elects six at a time.

*The method used to elect the Australian House of Representatives. The Aussies have been responsible for many electoral advances. Not just "The Australian Ballot".

[Isaac Bartram]

tomWright

I am not an advocate of the Parliamentary System, but I think the UK, Canada, Australia, New Zealand and (now) South Africa demonstrate there is nothing inherently ustable about it.

The article is correct to the extent that it identifies Party List PR as the cause of many of Europes woes.

[joe]

Isaac,

In my grand plan, the Senate would be elected by party, nationally.

[joe]

BTW, if England, Canada, South Africa, Australia, and New Zealand all have PR systems, they author's reference to "Anglo Saxon countries" spurning the system doesn't make any sence.

Because as everyone knows, we are a multiethnic country.

[Douglas Fletcher]

I agree. Down with proportional spacing! Bring back Courier. If it was good enough for Dan Rather, it's good enough for me.

[Isaac Bartram]

joe, they don't.

I said those countries have Parliamentary systems NOT PR.

I know that UK and Canada both use "first past the post". Australia alone uses PR to the extent that I noted. And again it is not the same system he refers to.

One weakness with the Parliamentary System is the fact that one party can control the govt with a relatively low percentage of the vote. I recall your concern that Shrub was claiming a "mandate" with 51%. Imagine that you're a UK Tory or LibDem being told by Toady Bliar that with 37% of the vote "Labour has a mandate to carry out its programme."

[temujin334]

I recall your concern that Shrub was claiming a "mandate" with 51%. Imagine that you're a UK Tory or LibDem being told by Toady Bliar that with 37% of the vote "Labour has a mandate to carry out its programme."

Didn't Clinton declare a mandate from one (or both) of his elections? Doesn't everyone claim a mandate even if they don't use the actual word?

[Chris]

I think the 17th Amendment should be repealed. Let the state legislatures elect each state's Senators.

http://www.articlev.com/repeal17.htm

[joe]

Oops, my bad.

[Isaac Bartram]

This illustrates one common complaint against PR is this. Even with candidate choice methods.

http://www.michaeljennings.blogspot.com/2003_03_30_michaeljennings_archive.html

"It seems that racist populist Pauline Hanson may ultimately end up being elected to the New South Wales upper house."

My father told me that PR was used many, many years ago in New York City, but it was abandoned because a couple of Communists got elected. Heaven forbid that people might get the representatives they want.

[doctor thoreau :)]

My take on PR:

PR is absolutely essential if you want to see more than 2 factions represented. Some argue that it encourages extremism, but the same could be said of elections from heavily gerrymandered single member districts.

I support combining PR with bicameralism: Elect one chamber of the legislature by PR (in the case of Congress I'd say each state should serve as a single district, or large states like CA and TX could draw a handful of districts) to make sure more than 2 parties can gain representation according to their support. But elect the other chamber from single-member districts so that the big parties have an incentive to not stray too far from the center.

[Isaac Bartram]

Again, Utley is absolutely right about the PR he describes. It is responsible for all the ills he lists.

Party List Proportional Representation should not even be considered by any group seeking to improve a government.

And one problem is that if you bring up PR everyone assumes your talking about Party List. That is if you don't get a blank stare.

[thoreau]

Isaac-

What about open list? Where you vote for a party AND a candidate, so that the most popular candidates in each party get elected?

I know that it isn't as ideal as, say, the Hare system or various other methods that election algorithm afficionados have dreamt up. But it is much easier to implement than the Hare system. And if you keep the district size small (no more than 10 seats per district) and let the voters support multiple candidates on a list (to avoid a single popular candidate monopolizing all the votes so that unpopular members of the list also get seats) then you can have meaningful competition between candidates.

The Swiss do something like that. They also have single-winner elections for the other house of parliament, and the executive branch serves for fixed terms to avoid instabilities like Italy used to have.

[Larry A]

The statement in the article that lept off the page for me was the assertion that "The rules for economic development and effective government are proven and well known."

Yeah, right. Like religion, everyone knows what the proven rules are, they just can't agree on what they are.

[Isaac Bartram]

thoreau

I would have to take the time to get familiar with that (and I'm lazy so it'll take some time to get around to it), but that sounds good. If the Swiss use it, it must be OK.

My main point is that the party dominance, and especially the establishment dominance that Party List PR enshrines (and Utley describes) is especially harmful.

I suppose the fact that my father was a mathmetician might have had something to do with his liking of Hare-Clarke. One of the legitimate complaint against Hare is its complexity. Open list sounds like it shares the same flaw. There is a danger that people might feel that their government is in the hands of intellectual elites who are the only ones who can divine the portents of the mystical ballot.

Another objection to PR is that it makes reps to distant from there constituents. Smaller districts overcome some of this.

I was hoping to hear some input from you since you are one of the first people I've encountered in years who's interested in PR.

I also thought I might provoke some comments from Australians (I seem to recall some Aussie Hit and Runners)who can tell us how bad Hare-Clarke sucks or how great it is. I don't know how widespread it is now but I understand several more states have adopted Hare since I left 40 years ago.

*Hare was the Irish barrister who initially designed the system. I believe Clarke modified it by creating Multi-member districts rather than making it nationwide. I'm winging so I might be wrong.

Google "Hare-Clarke". There's a lot of interesting links.

[Isaac Bartram]

Larry A at June 4, 2005 04:25 PM

Excellent point.

We all know what electoral unfairness is too. But most peoples' definition is "my side didn't win".

[joe]

Dr. thoreau, me too. One house by PR, one house by SMD. Best of both worlds, checking each other.

[thoreau]

Isaac-

As an afficionado of election methods, here's my take on PR methods:

First, idealists can argue that every group should get a number of representatives that's exactly proportional to its share of the electorate. So let's have an election for 500 seats and anybody with 0.2% of the vote gets a seat.

The only problem is a ballot with a few thousand candidates. That's no good. Not to mention that individual candidates can't really get scrutiny.

So skeptics might say we should just stick with single member districts. The only problem is that then you are all but guaranteed a political duopoly, which leaves little room for competition from other factions or perspectives or ideas. So that's no good.

So districts with, say, 10 members (give or take)seem to be a reasonable compromise. And to further moderate the effects of PR, make sure that the other chamber of the legislature is elected by some sort of single-member district method.

Now, what method to use? I find that most PR methods can be put into 3 categories:

1) Very, very simple methods: Things like cumulative voting (google it for more info). These methods are, well, simple, and they put the focus on individuals rather than parties. The problem is that they don't do a very good job of guaranteeing some semblance of proportionality.

2) Hare and other complicated methods (google for, say, proportional approval voting): These methods guarantee proportionality and they put the focus on individual candidates. The problem is that they're very complicated to implement and complicated to understand. In practice, most people who use Hare rank their top handful of candidates individually and then either rank the rest along party lines or stop ranking.

Decent, but complicated. In an era where people raise (perhaps legitimate, perhaps not) questions about the reliability of even our simple voting method (one vote per voter, one winner per race), Hare just seems like a non-starter.

3) Party list methods: The simplest is one where you just vote for party and then the corporate donors...um, I mean, party leaders, decide who gets the seats. But that's not the only option. There are also methods where you vote for a party AND you vote for one or more candidate within the party. There are all sorts of variations on that theme.

The Swiss even have a version where to elect N seats you have N votes. You can vote for N separate candidates, or even combine multiple votes behind individual candidates (e.g. give 2 votes to one candidate, and 1 vote apiece to three other candidates, with 5 seats at stake). Each vote for a candidate also counts as a vote for the party. Seats are apportioned among the parties, and then the most popular candidates in each party are elected.

All sorts of variations are possible. The key point is that open list methods are proportional, they're simple, and yet they give voters the final say on who wins, not party leaders.

Needless to say, my preference is for open list methods. In an ideal world I would probably prefer some really complicated method, but in the real world I think some version of open list is good enough.

[Isaac Bartram]

Thanks for that input, Doctor. A banquet for thought. I will see if I can digest it.

Your note 3 is right on target. I believe that party list is just plain evil, resulting as it does in enshrining the position of party bosses. And yes it exacerbate the influence of donors as well.

joe, you'd like Tasmania, then :). The House of Assembly (lower, the one where bills originate) is elected by PR (5 seven-member districts) and is Parliamentary, so the terms are not fixed while the Legislative Council (the review body) is elected from single member districts with a preferential ballot for fixed term of six years.

Every few years the party in power tries to change the system because like pols everywhere they think they can game the system. I think it requires a referendum to change it and they never seem to be able to convince the people that a change is necessary.

[Isaac Bartram]

Not that a good electoral system guarantees anything in particular. About 10-12 years ago the Tasmanian Parliament increased the penalty for sodomy from 20 to 25 years in prison. I think it was later overturned in court, still one wonders what the people and their reps were thinking.

[Anton Sherwood]

thoreau: Hare and other complicated methods (google for, say, proportional approval voting)

I doubt you'd find many hits on that phrase; approval voting is neither complicated nor proportional, though it does weaken the two-party system.

[Anton Sherwood]

Well what d'ya know, I spoke too soon. I wouldn't have thought the concepts could be combined.

[mark anderson]

The problem is the 'assumptions' of what 'economic development' and 'political stability'. There are many governments far older than the US, that would be the definition of stability. Economic development typically means what WE like as economic development, meaning two demographic groups get stinking rich while the others live close to poverty.

So essentially the argument comes down to 'our government performs best, here's why others don't perform as ours does'. So all that's shown is that other countries are DIFFERENT, not better or worse. Government and economies are so vast and face such different situations there's really no way to say 'THIS' is what works for everyone. Keep in mind even democracy was never designed to be cost effective, dictatorships are better at that, but there are other considerations.