This week, the UK Coalition submitted its bill for reform of the House of Lords. The sponsor is Deputy PM Nick Clegg; it has has the public backing of PM David Cameron, though he has been warned that 100 or so MPs of his party will defy the whip and vote against, including possibly some quite senior members.1 Labour is also divided on the issue.
Despite earlier news items that suggested the reformed body would be called the Senate, it will actually remain the House of Lords. However, members will not be called Lords, though it is left to parliament to decide upon a title.
Highlights from the bill:
At the first election, anticipated with the general election of May, 2015, there would be 120 elected members. By the third electoral period that number would have risen to 360. There would continue to be 30 appointed members at each electoral cycle (thus 90 at steady state), as well as a declining number of Lords Spiritual. There will also be some ministerial members (appointed by the monarch upon nomination by the PM).
An elected member of the House of Lords serves a 15-year term.
The bill makes clear that the Parliament Acts of 1911 and 1949 will continue to apply. (These allow the Commons to override objection by the Lords after one year; there has been concern that a popularly elected second chamber would be more assertive.)
Elections concurrent for the two chambers, except in the case of a Commons election that happens within two years of the previous one. (The Coalition has already legislated fixed Commons terms, but there are still provisions under which an extraordinary election could be held.)
A list system of proportional representation in Great Britain, but Single Transferable Vote in Northern Ireland. (Earlier drafts had called for all members to be elected by STV.)
The districts will coincide with those used to elect Members of the European Parliament. Their number of elected members will range from 3 (Northern Ireland) to 16 (South East) at any given election (see Schedule 2). This will mean an average district magnitude of 10, and only three districts are set to be below this average. There is a provision for redistribution of magnitudes across districts.
The electoral formula in Great Britain is D’Hondt. Lists are flexible: a candidate can be assured election to an available seat for his or her party only upon obtaining preference votes equal to at least 5% of the list’s total vote; seats not filled via preference votes are assigned in “the order in which they appear on the party list” (see Schedule 3).
Former members of the House of Lords will be unable to stand as candidates for the House of Commons for four years.
- One example from the linked item:
Conservative aide Conor Burns said: “If I lose my job for something that was a mainstream view within the Conservative party in the last parliament which serving cabinet ministers held, so be it.”
[↩]
15 year terms? They had better hope it does not go to a referendum. I assume they went with closed list PR in Scotland/Wales/England to explcitily not have a direct link between candidates and voters?
I do enjoy the protests of one Lord Reed in the article linked that the new chamber would ultimately become the more powerful house. Because it would be (somewhat?) more representative? I also admire the fudge of having a House of Lords with no actual Lords in it.
Seed planted by DC — 29 June 2012 @ 05:39
Indeed. On the substantive question of ends, I see something of a… disconnect between “PR is a clearly less democratic voting system than FPTP” and “a PR-elected upper house may/ will come to rival, even overshadow, the FPTP-elected Commons.”
On the procedural question of means-es, I am bemused how many conservative critics (whether Tory or Labour) of the proposed reform are arguing along the lines of “Clegg’s changes would utterly disrupt Britain’s tried and tested, centuries-old constitutional settlement of the majority party in the Commons both having untrammelled legislative power and being unambiguously politically accountable at the next election. – Therefore these changes should not be enacted, at least not by ordinary Act. REFERENDUM NOW!!!!”
Seed planted by Tom Round — 29 June 2012 @ 06:14
Apart from the electoral system, it sounds good.
The opponents of the bill are often simply ridiculous: how can any change to the membership of the upper house undermine the primacy of the lower house when the latter can OVERRIDE the former under almost any circumstance?
Seed planted by JD — 29 June 2012 @ 15:10
DC, the proposed list is not closed, but flexible.
There actually was a piece in the Guardian today (29th) in which a Tory argued that the problem with the reform of the Lords was precisely that there would not be any more “radical” PMs a la Thatcher.
Seed planted by MSS — 29 June 2012 @ 23:18
One news item I saw days before the bill was introduced said that there would be a fused vote (i.e. the votes for members of the Commons would be summed to determine party votes for the “Lords”).
Such a provision would be difficult, though not necessarily impossible, to administer with the preference vote that is in the introduced version, so I assume there will be separate votes for the two chambers. If there was anything specifically on this point in the bill, I missed it.
Seed planted by MSS — 01 July 2012 @ 19:56
As long as the HoC can still override the HoL with a simple majority, JD is right that the constitutional balance would not change with an elected HoL. However, that’s sorta the point. Right now, the HoL enjoys only the ability to delay and go public with an argument against the government’s plans. But their authority, such as it is, derives from the fact that they’re not grandstanding, vote-grubbing politicians. Their opposition is seen, therefore, as principled, governments ignore them at their peril. Make them elected, and they’ve no more real power, but also now they’ll lose that “above politics” moral authority. The government would be even LESS inclined to pay any attention.
So if you want bicameralism, you need a form of bicameralism that actually grants the 2nd chamber the constitutional authority to veto (at least some types of) government bills. Then the argument that a 2nd chamber with real power should be electorally accountable makes sense.
But just what is the positive case for bicameralism in the UK? Is it regionalism and possible creeping ethnic federalism? That Scotland and Wales, etc need to be represented as such? That’s the usual justification for strong bicameralism, but I frankly don’t see the virtue of it for the UK.
The other reason for bicameralism in most places, of course, is vestigial aristocracy. “Upper” chambers came first, before the unwashed masses got their own chambers, and only became weaker than lower chambers over time. Denmark and Sweden eventually confronted the anachronism and got rid of their upper chambers. That the British HoL has survived this long, without any but moral authority, is odd, but I think is attributable precisely to its role as the “conscience” of government, and not a real, election-minded competitor.
Seed planted by Mike — 02 July 2012 @ 15:10
While Mike makes good points, reform of the House of Lords gives the British a chance to institute an electoral/ constitutional system that I wish more countries instituted, and which to my knowledge only Australia (sort of) has.
That is election of one chamber using single member districts (either SMP or SMM), and the second using proportional representation (probably open party list, but STV may be more appropriate for second chambers with smaller memberships or where the national electorate would have to be broken down into sub-electorates).
There is alot of debate on this blog and elsewhere about how to square the circle of instituting an electoral system for a single chamber that combines main benefit of single member districts, which is local representation one person clearly accountable as the representative of that area, and the benefits of proportional representation; eg a legislature more accurately reflective of national party support, a broader spectrum of opinion in the legislature, reduced likelihood of a reversal-majority. Its very difficult to do both of these in a single chamber without resorting to gimmicks.
I think instead of trying to do this with a single chamber, a more appropriate solution would be one chamber elected by single majority districts, and a second chamber elected by proportional representation. The two chambers would have different electoral bases, one would represent localities and one would represent currents of opinions. Assuming a parliamentary system where confidence and supply resides in the chamber of localities, which I think would be the norm, a party gaining a majority there with a low plurality popular vote would be prevented by the second chamber from ramming the more unpopular elements of its program through. The reverse arrangement might be more appropriate for a federal system, where representatives from the states can check a government based on a chamber nationally elected using proportional representation.
In the case of the specific situation in the UK, I would have preferred to leave the composition of the House of Lords as is, but to remove the automatic right of peers to vote (most peers would be able to attend but not vote). After each election to the House of Commons, voting writs would go out to selected peers after consultations with the party leaders, so that each party wound up with a number of voting peers proportional to their popular vote strengths in the last House of Commons election. The presence of the non-voting peers, other than being a pool for the voting peers, would allow the continuation of the various odd uses of the Lords in the British system, such as for honors and to import experts into parliamentary deliberations.
Its beyond me why past Conservative governments, since this reform would most likely appeal to a Tory government, is beyond me.
Seed planted by Ed — 02 July 2012 @ 17:11
One could make the case for one FPTP chamber and one PR chamber, but I’d expect them to be the other way around.
Seed planted by Vasi — 03 July 2012 @ 03:30
@Tom @2
I think the opponents of the bill actually have an argument, although they probably do not like where that argument goes. The moral authority of an elected House of Lords (what happened to the somewhat more inclusive ‘House of Peers’?) can only threaten that of the House of Commons if the lower house is actually a less than democratic body that merely serves to enact what the executive tells it to enact.
Now some might argue that is actually how the modern Westminster system drifts ever further into elective dictatorship. These critics actually have a case. It is just not a case they really want to make.
Seed planted by Alan — 03 July 2012 @ 10:09
Australia’s combo makes sense because Australia is federal. As Mike notes, there is no similar logic in the UK, because even as the UK creeps towards a quasi-federal construct (assuming Scotland stays in), no proposals for the Lords involve representing the “federal” units as such.
As far as combining the virtues of localism and accountability with a legislature reflective of partisan differences, it is true that these are tradeoffs to some degree. But not nearly as much as typically claimed. First of all, FPTP in parliamentary systems has been shown by much research not to offer much meaningful local representation or individual accountability. Secondly, one can get a decent amount of the benefits of both pure types via either MMP or small-magnitude PR.
I agree with Vasi’s point–as I understand and expand it–that it might make sense to have one chamber that is elected by a form of PR and which determines the government. Then one could have another, presumably much weaker, chamber elected by FPTP. The members of the latter could be responsible mainly for casework, ribbon-cutting and various other “local service” functions that some folks seem to find desirable. This is obviously not where the UK debate is heading, however.
Seed planted by MSS — 03 July 2012 @ 14:50
While federalism and bicameralism are historically linked (ie, unitary countries that tend to abolish their national upper houses, and sub-national units are, as far as I can tell, even more likely to be unicameral than national ones), they’re not inevitably a logical corollary of each other. Indeed, the American Founders copied British bicameralism, which was then centuries old, at the same time as they invented federalism. And other polities like France and Finland began with three or four chambers (or “Estates”) long before Madison was born.
If the goal is to maximise the power of veto for member States – weighted with no (or less) regard to their population differences – this could be done in a unicameral, per capita-apportioned federal assembly by putting in the constitution that a bill can pass only if a majority of individual legislators and (say) 60% of “State delegations”, both vote in favour. A State delegation is deemed to vote in favour if at least (say) three (or all, whichever is less) of the individual legislators from that State vote in favour. You then build in both a quorum requirement and a “majority of States” hurdle. Some trades union and student union federations have rules along these lines for their (unicameral) federal congresses, conferences or conventions.
I would prefer State autonomy be protected by allowing a minority of States a veto – but requiring a super-majority within each such State to register its veto – than the Swiss/Australian rule requiring simple majorities in a simple majority of States. Some US proponents of a national initiative & referendum have proposed copying it, but I think it is a bad idea. In Australian terms, I am happy to let (say) Tasmania veto a referendum proposal supported by all five of the other States, as long as (say) 70% of all Tasmanians are opposed to it, not just 50.1% of them, because then there are stronger grounds to believe the proposal really does discriminate against smaller States, and isn’t just a difference of political opinion.
Seed planted by Tom Round — 03 July 2012 @ 23:41
I’ve noticed a tendency in the last few comments to devalue the effectiveness of individual representatives from local districts and/ or from states or provinces.
Executives have gotten skilled at managing legislatures, in different ways in parliamentary and in presidential systems. And since World War I, governments in developed countries have gotten hyper-centralized by historical standards, again usually working around or co-opting the old (sometimes medieval) provincial and local levels instead of outright abolishing them.
So of course you have a situation where a co-opted legislature composed of individual representatives from shires and boroughs doesn’t offer more effective local representation than a co-opted legislature composed of nominees from regional or even national party lists. We are far from the days where legislatures acted as sorts of ambassadors from semi-individual boroughs or states. But I think the degree of centralization shown in the twentieth century won’t last forever and its important not to abandon the local component of representation.
Seed planted by Ed — 04 July 2012 @ 00:15
I think the majoritarian-lower-house + PR-upper-house actually makes sense, and may even be the logical direction of the Westminster system. The advantage of a majoritarian lower house, as seen from a common UK/Commonwealth perspective, is a very high level of government stability. The PR-basis of the upper house moderates this, (and prevents ‘elective dictatorship’), by making it more difficult for a gov’t to pass its legislative programme. Parliament will thus become more independent, and more legislation through compromise.
However, the latter can’t really be done if the lower house can override decisions made by the upper house by simple majority vote. The upper house therefore must have a real veto. (I would suggest a veto at least as powerful as in India, where disagreements can lead directly to joint sitting.)
In this system, the lower house keeps its primacy, but that primacy should stem mainly from its role in choosing the gov’t and holding it to account. (The Australian system of course deviates from this by giving the upper house a strong ability to block supply. I would suggest a less powerful upper house with regards to supply).
The reason I largely support the bill is because it is an important first step towards a stronger and more legitimate constitutional arrangement, and because its passage in itself is dependent on keeping the inter-house relationship at the status quo. Right now, even though the bill leaves the Commons with the power to override the Lords, it’s being met with opposition because it supposedly makes the Lords more powerful. Imagine what sort of opposition it would encounter if it actually did!
Seed planted by JD — 04 July 2012 @ 12:26
@MSS: Flexible list, yes. D’oh, you’d already told us that of course. Speed reading is not my forte.
As far as the role of the HoL goes, it seems to have acted almost uniquely as a break on Labour governments. This Total Politics article on the use of the Parliament Acts (1911 & 1949) only cites one instance of its use by a Conservative government, in 1991, to force through an international agreement on war crimes extradition:
http://tinyurl.com/d67xlno
The subjects on which the Act has been used are instructive as well-mostly constitutional, religuious or social issues. Would a different HoL be more assertive on other issues, like economic and social policy for example? One would have to assume so.
Even this reform, hedged as it is, would almost ecertainly change the dynamic between the two Houses, but the question I suppose is whether that is a bad thing…
Seed planted by DC — 04 July 2012 @ 12:51
“As far as the role of the HoL goes, it seems to have acted almost uniquely as a break on Labour governments.”
This is somewhat out of date, and one area where the two changes to the composition of the House did have an effect.
The nineteenth century House of Lords was made up exclusively of hereditary peers, plus a few Law Lords and bishops. Note that this was not the composition of the medieval House of Lords, where clerics predominated and the peerage was not necessarily hereditary (the secularization of the monasteries played a big role in turning the Lords into the preserve of the landed aristocracy). In 1885, the Conservatives gained a large majority among the hereditary peers when many of the Liberal aristocrats deserted the party over home rule for Ireland, and over the next twenty-five years the Conservatives used their majority to frustrate Liberal governments.
Hence both the composition and the powers of the House of Lords became political issues, and in the twentieth century there were two reduction of the body’s powers, in 1911 and 1949, and two changes to the composition, in 1959 and 1998. Note the reduction of powers came first, though the 1911 legislation, which the House of Lords voted for, envisaged changes to the composition as well.
The Conservative Macmillan government brought in (or revived if you take the long view) non-hereditary peerages, and all except a handful of peerages, one of the exceptions being for Macmillan himself, created since them have been non-hereditary. The Blair government removed most -not all!- of the hereditary peers and therefore life peers predominate.
So between 1885 and 1959, or 1998 at the latest, the House of Lords was controlled by a predominantly Tory hereditary aristocracy, supported Tory governments, and frustrated Liberal and Labor governments. This hasn’t really been the case since 1962 -the Thatcher government had problems with the House of Lords- and definitely hasn’t been the case since 1998 but that won’t show up in the data since the Conservatives haven’t had a majority in the Commons since that year either.
My point is that studies trying to show the partisan effects of the House of Lords that ignore the changes to its composition and powers, and more so rely mainly on 1885 – 1962 data, are misleading.
Seed planted by Ed — 04 July 2012 @ 15:50
A couple of nineteenth century British colonies considered trying to create hereditary chambers. The idea was labelled the bunyip aristocracy in Australia and there were bad jokes about the title of Prince of New South Wales.
What the colonies did come up with, life appointments by the executive, was as close as they dared go and arguably had the disadvantages of a hereditary chamber with none of the advantages. The members of legislative councils tended to be creatures of whoever controlled the government in the longterm. Life anointment did not make for an apolitical chamber.
The appointed chambers led to the phenomenon of ‘suicide squads’ where a government would appoint large numbers of MLCs sworn to abolish or reform the legislative council. Weirdly the Canadian senate is the only surviving unreconstructed chamber of appointees.
Seed planted by Alan — 05 July 2012 @ 03:33
I came to this blog to see discussion of the 5% personal-vote-threshold. None yet.
Sweden uses 8%, for districts with an average magnitude of 12. If all the voters cast personal votes split between 12 candidates, half would likely get more than 8%, and it’s unlikely a party would win more than 6 seats, so it’s effectively open-list. But of course some voters will vote for the list, so the list will have some effect.
For the Lords the average DM is ten at the first and each subsequent election. If even half of the voters cast personal votes between ten candidates, half will likely get over 5%. So this model is twice as “open” as Sweden’s. The list rank is less likely to have much effect.
Fans of open lists, including STV fans who can tolerate open lists but hate closed lists with a passion, should be pleased.
Seed planted by Wilf Day — 05 July 2012 @ 17:02
Hey, Wilf.
OK, here’s my comment, fwiw. (I’m an open list fan, as you probably remember.) I think thresholds like this are pointless, but people will insist on trying to manipulate systems with thresholds rather than just coming out and saying what they mean.
Personally, I would have no objection to a party being able to name a few candidates in a list system. Greece, for example, exempts current party leaders (and former prime ministers, I think), and also sets aside 12 out of 300 seats for closed national party lists. That’s seems like a good definition of “a few”.
Actually, you might just go with the 12 out of 300 seats, since it seems totally reasonable for a party leader to be a national representative, but that would prompt small party to run their party leader in a large district rather than the national list, and then they would still have to compete with the rest of their list in that district. That’s not too serious, in a sense: usually the leader of a small party is also the most well-known candidate, and is unlikely to have built up internal opposition the way that some former prime ministers might have.
Anyway, my point is that being explicit about who the party prefers and who the voters are (allowed to) prefer is more explicit and probably involves less corner-cases and difficult explanations.
Seed planted by rici — 05 July 2012 @ 21:15
As I understand the provision in the UK bill, it would not matter what percentage of a party’s voters simply voted for the list. In some flexible (or semi-open) lists, there needs to be a certain threshold of voters casting preference votes before any such votes can alter the pre-fixed list order.
Seed planted by MSS — 05 July 2012 @ 21:21
MSS, the bill says ‘A candidate is a “qualifying candidate” if the number of votes given for the
candidate is at least 5% of the number of votes given for the party’. So it’s more like it doesn’t matter how many votes the party gets in all (which is certainly better; it means small parties also get to have open lists). And actually, as Wilf says, it seems pretty unlikely that any otherwise electable candidate would fail to make the threshold.
But that’s precisely the sort of corner case I was talking about. Say the Closed List Party (“technocrats and proud of it”) runs a campaign based on “vote for the party; you must trust us”, then it becomes a race to see if more than five percent of their supporters will break the contract. And what’s the point of that?
Seed planted by rici — 05 July 2012 @ 21:56
Right. It’s an intra-party criterion: percentage of the party vote. But some flexible-list systems, such as the Czech Republic and Slovakia, do not allow a candidate to qualify for election unless another threshold is reached, defined as some percentage of the party’s voters casting any candidate-preference vote.
I try to keep the terminology tidy: it is not an “open” list unless preference votes are the only criterion determining which of a party’s candidates win, and in what order.
How easy is it for a candidate to reach 5% of the party’s vote? It just so happens that this very question has been addressed here at F&V!
Seed planted by MSS — 05 July 2012 @ 22:25
Hmm… so if Party A scores 1 seat out of 15 with just over 6% of the vote, its highest vote-getter could win a seat with 0.3% of the total turnout. Whereas if Party B scores 8 seats out of 15 with 51%, a candidate with 2.49% of the total poll could miss out on a seat, if s/he were very low down the list.
I think the Daily Mail sneers are already writing themselves. (Anyone remember “Under AV, BNP voters could get their votes counted as many as SIX TIMES!!!”).
I would myself suggest a threshold more like “half a Droop quota”, although this would favour the higher-ranked but lower-polling candidates on lists that just squeak in with their last seat.
Seed planted by Tom Round — 06 July 2012 @ 03:17
Lords reform: five days to the first (last?) vote
Scion grafted by The House Divided: Politics, Procedure and Parliament — 06 July 2012 @ 09:09
I’m probably approaching this too simplistically, but for open lists I prefer a double threshold, a threshold for the party and a second threshold for the candidates.
The thresholds don’t (probably shouldn’t) be explicitly written into the electoral law, but can be more of a function of the number of seats to be filled divided by the number of votes.
For example there are 25 seats to fill. A list is allocated a seat for each 4% of the vote it pulls, plus some mechanism such as greatest remainder to distribute the seats unallocated due to the fact that party percentages won’t divide into neat 4% increments. But if at least 4% of the voters vote for an individual candidate, that candidate is declared elected regardless of where they are ranked on the list. The party bosses’ rankings only come into play to fill the seats that aren’t filled by more or less direct elections.
Actually I would also allow voters to blackball candidates on party lists, effectively pushing those candidates down towards the bottom of the list order, but that is another topic.
You can’t do something like this as easily with MMP, which is why I’m not crazy about that system, though my objection is not strong since its worked pretty well in practice.
Seed planted by Ed — 06 July 2012 @ 13:54
I know I am harping on terminology too much, but if there is any threshold of the vote required to assure a seat within a list, it is not an open list. It is a flexible list, or as others prefer, semi-open list.
As you were…
Seed planted by MSS — 06 July 2012 @ 17:31
Tom: So how does a threshold help? With the threshold, Party A’s candidate Bob T. Banker gets declared elected with no votes at all, because Leyla Secretary just missed out on the threshold. And you think the Daily Mail will applaud rather than headlining “Banker Bob buys his way into the Lords”?
In other words, the threshold is saying, in effect, that the party hierarchy (or whatever) has the same democratic value as five percent of the party’s voters. You might believe that; the party might believe that; but the voters are also entitled to an opinion. Otherwise, why did you ask them for one?
By the way, MSS: you’ve convinced me that it is hard enough to reach a five percent threshold that it probably would make a difference. I checked my datasets from Peru and Greece, which both allow multiple candidate preference votes, and the threshold does indeed seem to kick in. As per above diatribe, I disapprove
However, I still think there’s value to the Greek system, which is really a mixed open list/closed list rather than a flexible list, since no list is flexible. In that case, the rules were entirely clear up front, and in any event the closed list is small enough to be reasonable. Although Bob the Banker still worries me…
Seed planted by rici — 07 July 2012 @ 00:13
How about this idea?
5% threshold for individual candidates to be chosen on personal votes
5% threshold to allocate all remaining seats
Seed planted by Derek — 08 July 2012 @ 04:12
I like Fair Vote Canada’s statement “A democratic voting system must encourage citizens to exercise positive choice by voting for the candidate or party they prefer.” A flexible (semi-open) system does just that. If voters like the list ranking, why should they have to cast a personal vote?
As to the weight to be given to personal votes, that is, how “open” is the semi-open system (if that’s poor terminology, what is the right way to say some semi-open systems are more semi-open than others?), that is a design issue in the MMP model recommended by the Law Commission of Canada and also implicity recommended in the MMP study by the Quebec Director General of Elections (“The aim of the flexible list system is to achieve a balance between voter choice, which is usually associated with open lists, and better representation of women and minorities, usually attributed to closed lists.”) If we must use a Swedish-style threshold, what should it be? It’s an oddly blunt instrument for the careful Swedes. My own “flexible algorithm” inclination is more toward the Belgian quotient model or the neat model proposed by the Netherlands Citizens’ Assembly (whatever percent of the party’s voters vote for the list, that percent of the party’s candidates is elected in list rank order, while the rest of the elected candidates follow the personal ranking.)
Seed planted by Wilf Day — 08 July 2012 @ 06:05
Sure, we can speak of degrees of openness, with (fully) open at one end of a continuum and closed (no preference votes at all) at the other end. The question Wilf asks–what weight should be given to preference votes vs. list ranking–is, in the literature, an open one, I I dare say so.
Again, I refer interested readers to a a thread where this specific matter was discussed (see link in my previous comment).
Seed planted by MSS — 08 July 2012 @ 15:36
Lind of an interesting take on SMD unicameralism.
Seed planted by Alan — 09 July 2012 @ 05:29
Alan @30, I assume “Lind” was a mistype but it was also serendipitous given the topic: see http://tinyurl.com/7mdo8ym
Seed planted by Tom Round — 09 July 2012 @ 08:51
The antidote to the party power inherent in the semi open list voting system is to allow voters an alternative short vote – ‘Independent’ – for a raft of ‘official’ independent candidates sponsored by an independent peers appointments commission in addition to the 90 independents planned.
Seed planted by Stephen Johnson — 09 July 2012 @ 09:40
yes, a parallel system with two ballot papers, 2/3rds of the seats allocated for the parties to fight over among themselves, 1/3 restricted to independents [must swear a declaration they have not been a member of a party for say the past 5 years]
On the “semi-open” list issue, it will surely behave more like a closed list than an open list, with the donkey vote accumulating more “personal” votes for those candidates at the top than for those at the bottom. Those at the bottom could even win more than the magic 5%, with no effect on list order!
The donkey vote effect will be ameliorated the fewer candidates are offered. Will the parties offer a list of n candidates for n available seats with a straight face? What kind of person worthy of election would agree to be nth candidate on the list?
Seed planted by RodCrosby — 09 July 2012 @ 13:43
I agree with Rod’s “donkey vote” claim, inasmuch as the experience of other flexible (semi-open) lists is that most preference votes are cast for candidates whose ranks were high enough on the pre-fixed list to be elected anyway. I should add, however, that this does not make the system no different from a closed list, in that parties may anticipate the popularity of their candidates when setting rankings (or may involve their base in the nomination/ranking process). Donkey voting and perfect party understanding of which candidates are most popular would observationally equivalent!
[paragraph deleted because I realized later that Rod was referring to something Stephen proposed, not to the actual bill.]
Welcome to the virtual orchard, Rod!
Seed planted by MSS — 09 July 2012 @ 13:55
To Rod Crosby,
I wasn’t suggesting parallel voting systems.
Rather that ‘Independent’ should be a possible short voting choice on the ballot paper alongside each of the separate parties, with a comparable list of approved independent candidates prepared by a statutory (independent) peers appointments commission.
Seed planted by Stephen Johnson — 09 July 2012 @ 17:39
Sorry, I misunderstood. But they would be allowed to compete anyhow under the current proposals. I can see how a list of independents might elect more individuals than independents standing individually, but would such a list actually be ‘independent’, or start to function like a party, given the party-centric mathematics of all list systems [a vote for any counts as a vote for all]?
I think my idea of 1/3 reserved seats for independents, elected separately, or in Matthew’s words ‘concurrently’, might work better.
Seed planted by RodCrosby — 09 July 2012 @ 17:56
To (try to) be clear, I use “concurrent” to refer to elections to different institutions happening on the same day (whether or not the ballots involve distinct choices, though they usually do).
I find the ideas on how to elect “independents” interesting, but I agree that once they run on a list they are essentially a party.
I realize I neglected earlier to welcome Stephen to the virtual orchard, so welcome, Stephen!
Seed planted by MSS — 09 July 2012 @ 20:14
To RodCrosby and MSS
Independents are allowed to compete under the present proposals, but going it alone is difficult for independent candidates (campaign resources, achieving individual public prominence)
Candidates approved by the Independent Commission would have a collective advantage over ‘wildcard’ independents. The voter would benefit from being able to vote for a ‘quality assured’ raft of independents with a short voting choice.
Such a group of individuals would be different from a party in that they would have no common policies or manifesto, no party mass membership, party hierarchy, party discipline, or party patronage, and there would be no equivalent group for them to work with, or be influenced by, in the Lower House.
Once elected, they would have no obligation to anyone except to obey the rules of the House, and in particular no obligation to the Peers Statutory Appointments Commission.
Ensuring the impartial independence of the Peers Statutory Appointments Commission would be important but, I think, not impossible.
Politically such an arrangement might be attractive to those who see the present proposals giving too much power to the Parties at the expense of Independents. It might also help to reassure some who fear an advantage for the Lib Dems or other third parties as the likely recipient of protest votes.
In due course, it might even be the basis for a wholly elected chamber.
Seed planted by Stephen Johnson — 10 July 2012 @ 07:01
Wilf, thanks for mentioning the Netherlands Citizens’ Assembly! I completely missed that back in 2006, and their flexible list concept is interesting. I have some thoughts on that, but I’ll share them next time the open/flex/closed-lists issue sprouts up, and let this thread focus on the HoL.
Seed planted by Vasi — 10 July 2012 @ 12:55
I also had missed that Dutch proposal, and it is a promising idea.
As Wilf described it at #28: “whatever percent of the party’s voters vote for the list, that percent of the party’s candidates is elected in list rank order, while the rest of the elected candidates follow the personal ranking.”
It would be somewhat similar to the “Intra-party D’Hondt” system that I have circulated. I’d have to game both of them out to see in what ways they would differ, and which might be preferred on various grounds.
The Dutch proposal is a vast improvement over what they were discussing when a delegation from the Citizens Assembly visited several of us electoral-systems specialists at UCSD. At that time, they were proposing SNTV with a compensatory list top-up.
Seed planted by MSS — 10 July 2012 @ 14:22
Colour me sceptical on the Dutch proposal.
One of the great advantages of simple PR-list systems (both open and closed) is that the meaning of a vote, and thus the elector’s voting strategy, are clear and simple. I just vote for the party (and candidate, if it’s open-list) which I think is best. [fn1]
Now consider the Dutch proposal. Let’s say I live in a district whose magnitude is such that it is most likely that my preferred party will elect two representatives, or anyway between one and three. Suppose furthermore that I really like A, the candidate at the head of the party list, but I’m aware that she’s not hugely popular; candidate C (at position 3) is much more popular, and even candidate B is likely to do better. Now what’s my best strategy?
One option is to vote for the party list. If enough people do that, then at least the first candidate on the list will be elected as a party choice, so I’ll be happy. But I could also vote for my preferred candidate. If enough people do that, then she’ll get elected on personal preferences. Great. Now, which one is best? If it’s a difficult decision — and in various estimations of probabilities, it’s going to be difficult — then it is entirely popular that some candidate A supporters will go one way, some will go the other way, and we’ll end up with candidates Mr. B and Mr. C because there were neither enough people voting the list nor enough people voting for Ms. A, although if we’d all managed to find the same strategy we would have ended up with Ms. A joining Mr. B.
fn1. I know it’s not quite that simple. I might have to balance these two preferences, and in some cases I might need to discard the idea of voting for some party and/or candidate whose election prospects are null. There are always some strategic considerations, particularly in systems with a high representation threshold. But in general, and for most voters, the best PR-list vote is the simple and sincere option.
Seed planted by rici — 10 July 2012 @ 22:29
The NY Times reports that Lords reform is dead. I’m sure Clegg will sputter about for awhile, but then roll over and do whatever the Tories want, yet again.
Seed planted by Vasi — 11 July 2012 @ 18:37
@41 The easy answer would be to give your first preference to your most favoured candidate, your second reference to your next candidate, and so on. Oh wait…
Seed planted by Alan — 11 July 2012 @ 23:46
Alan,
I guess out are getting at my question: why all this complicated mucking around when you can just have STV?
Seed planted by Chris Curtis — 13 July 2012 @ 02:28
It would seem that parties generally do not like STV. Of course, in those places where it has been implemented with a party-ticket voting option (“above the line”) it might be more acceptable to parties. But that seems to me a worse option for voter choice than all the alternatives we have been discussing.
Seed planted by MSS — 13 July 2012 @ 15:36
The Tories don’t like preference voting, as it allows voters to ‘gang-up’ against them. I suspect that is why New South Wales ‘partywise’ STV was ditched in favour of the List for the House of Lords.
Partywise STV would have been even more unattractive to them than standard STV, with voters voting 1,2,3 across parties instead of (usually) just voting 1,2,3 among their preferred party’s candidates.
Seed planted by RodCrosby — 13 July 2012 @ 16:24
The major parties generally do not like STV in Australia either and, seriously, do not really exhibit a very deep understanding of how it works. The ACT modified d’Hondt system was constructed as a ‘better’ alternative to STV by political operators on both sides against the advice of electoral officials and political scientists. The ACT adopted STV as a result of a popular movement in the territory backed the Proportional Representation Society of Australia.
The task of electoral reform is only partly to adumbrate an electoral system/ There is also the hard business of constructing a popular movement.
The advantage of a clear alternative system, as STV was in the ACT or MMP was in New Zealand, over a Frankenstein monster is that you can persuade people of oys virtues. You can’t advocate for an electoral system that is a creature on the table.
The AV referendum in Britain shows the creature on the table effect can prevent even very basic reforms. There was one notable interview where John Howard was being interviewed by the British media and the military dictatorship in Fiji was cited as an example of a country that had abandoned AV. The fact that Fiji has also abandoned elections themselves was not mentioned.
Even in this hallowed blog itself we have seen repeated statements about the impact of AV that simply do not correspond very much to the empirical reality and ignore the impact of the system at the level of states and territories.
Seed planted by Alan — 14 July 2012 @ 01:51
I quite like the former ACT system. It has some features in common with a system I’ve designed for the UK House of Commons, which involves first squaring the votes, using D’Hondt to round them, then a kind of STV to select candidates in small multi-member seats.
The outcome of recent UK elections under this system.
http://www.titanictown.plus.com/prsquared.jpg
Seed planted by RodCrosby — 14 July 2012 @ 14:16
To quote an extremely cogent critic of modified d’Hondt:
MDH could almost be the type case for an electoral system with a creature on the table problem.
It seems to me that the supreme criterion for an electoral system is that it be intelligible to the electorate. I do not mean every elector has to understand the math underlying the system. I do mean they should give results which are predictable and transparent and can be rationally related to the votes each candidate receives.
About the only good thing to be said for MDH is that at the second MDH election there was a huge majority for adopting Hare-Clark and entrenching it beyond the reach of the legislature.
Seed planted by Alan — 14 July 2012 @ 20:05
‘I do mean they should give results which are predictable and transparent and can be rationally related to the votes each candidate receives.’
Well that rules out FPTP, of course.
I’ve looked into MDH, and I think the problems were that Australia had already used pure STV, so a ‘bastardised’ version was bound to attract criticism, and the particular political circumstances in the ATC at the time (fragmentation) probably made it an unwise choice. But in principal there is nothing wrong with it. In fact something near-identical has been suggested for Malta, to cure the wrong-winner problem, which MDH certainly achieves.
Seed planted by RodCrosby — 14 July 2012 @ 22:25
The principal problem was and is that MDH is unintelligible. As The Canberra Times editorialised at the time:
What drove the adoption of MDH was that FPTP would almost certainly have produced an assembly where the ALP held 100% of the seats. MDH was designed to induce fragmentation, not avoid it.
However it was also designed to entrench the major parties because preferences could only transfer to parties that passed the threshold. A huge number of minor parties were simply locked out at Stage 2 of the scrutiny. The AEC noted that parties:
I suppose you could argue that MDH would work without a threshold, but it took months to count the 1989 election with over 47% of all candidates excluded. Without the threshold even a body as efficient as the AEC may well have still been counting when the term of the legislative assembly expired.
There is a point at which the transaction costs of a tinkered system exceed any possible benefit.
MDH has a unique record among electoral systems. It was put to a popular vote at the second election and rejected by over 63% of the electorate.
Seed planted by Alan — 15 July 2012 @ 09:43
Note that Canberrans have had no trouble understanding how Hare-Clark STV works. I lived there for a year and while I came across a few people (mainly Labor stalwarts) who preferred single-seaters or Senate-style “closed-list STV”, no one tried to claim it was incomprehensible.
Alan, if you are alluding to our Range Voting friends and their… interesting views about how preferential voting works in Australia: I am working on a detailed rebuttal but it will be quite lengthy and probably a separate post.
Seed planted by Tom Round — 15 July 2012 @ 21:40
How about a twist to STV? Let voters mark their preferences for either candidates or parties.
Seed planted by Derek — 10 August 2012 @ 09:46