Thursday, August 16, 2007

Dawg vs. man: third bite

The MMP wars continue. David Graham is back, with reiterated arguments, and not a little indignation. He was unhappy with my restatements of his position in my previous post, although I think I did a pretty good job summarizing them, and I linked to his article in my very first sentence. Nevertheless, this time I shall simply quote his responses in italics, under the original headings, although it will make for a very long article.

1) MMP will create unstable coalition governments.

I never referenced PEI's population, I referenced its size. The diversity needed is regional diversity. Our MMP proposal has no provisions for this, allowing parties to create lists of indentured MPPs from any area, with little to prevent them from concentration on vote rich urban areas.

There are a number of things that divide voters in Ontario and in Canada as a whole. Canadians, certainly are divided by region, each with its specific issues and concerns. But we are also divided by ethnicity and by values, the latter asserted in various political positions. FPTP has done an abysmal job of representing the political views of Canadians, excluding many of them while distorting the support given to others. And it has done a poor job of representing our diverse peoples (e.g., Aboriginals). Further divisions--rich/poor, urban/rural and so on--do find their political expression under FPTP, but that is not threatened by MMP.

Some issues, in other words, are geographical; others are cultural; still others transcend such boundaries and are national, or, in the case before us, Ontario-wide. As noted earlier, FPTP emphasizes a narrow, local view, sometimes to the point of exaggeration. MMP offers the opportunity of a politics that is not solely regional, while retaining the regional component with 90 ridings.

The fear that parties will concentrate their list MPPs in urban areas is mere speculation. Let me offer some counter-speculation. Any party foolish enough to ignore Northern Ontario by choosing all of its list MPPs from Toronto will suffer devastating results in the next election. Not only the party, but also local candidates, will be punished for such a move. Rival party strategists will have a field day pointing to the narrow, divisive Toronto-centrism of the party in question, and will ensure that their own lists have proper geographic balance.

In order to achieve the diversity that MMP proponents inexplicably assert will magically happen under our proposal, there either needs to be a law mandating diversity on the party lists, or the parties need to create their lists in a manner that is not completely and openly democratic. For a party to disqualify anyone from any list entry based on their race, sex, religious background, or number of toes, is completely undemocratic and will be necessary to create diverse lists.

We shall have to agree to disagree on this one. No truly democratic system ignores questions of representation. A legislature without Aboriginals, or without women, or one that is lily-white, however elected, cannot be said to be democratic. It's that diversity of perspective again, which FPTP has failed so miserably at achieving. MMP is not a panacea here--the complex question of Aboriginal representation, for example, is not addressed by it--but it offers at least the possibility of better representation over-all. Don't consider this "disqualification" of beleaguered white males, Graham's subtext here: consider it qualification of many other groups and interests presently shut out. Ultimately, the voters will decide in any case.

MMP is inappropriate because it replaces the last vestige of independent representation with another layer of party oversight, breaking away from the core tradition of the Westminster model of representatives representing their ridings to the government. That's key. It should be noted that every country that uses MMP uses FPTP to select at least half its representatives, so while more than four European countries may use MMP, every country that does uses FPTP and all the problems that entails for riding elections. To say that FPTP is bad and retain it in our proposal is hypocritical.

This is somewhat self-contradictory. How would the system proposed for Ontario "replace the last vestige of independent representation" when it retains the riding system for two-thirds of the seats in the legislature, ridings in which anyone can still run? And to argue that a mixed system retains elements of FPTP is a truism--it does so by definition. Why this is "hypocritical" is a mystery, given that the objections to "pure FPTP" are what we are supposed to be discussing. MMP offers a cure for the worst ills of pure FPTP, without killing the patient.

[Dawg: the Ontario threshold is relatively low--5% rather than 3% is more common. But that can be fixed if it proves to be a problem--it's hardly an argument against MMP as a whole. Ditto for the lack of overhang provisions (addition of extra seats after an election if required to preserve proportionality), which could indeed result in rare majority governments with minority support. Far from being rare, though, such false majorities have been the rule in Ontario under FPTP.]

This is, of course, more a matter of opinion than argument, but I do not share MMP proponents' hatred of majority governments, even when they are held by a party other than one that I would support. My ideal, as long as we are condemned to a party system, is to have a mixture of majority and minority governments, where when all parties misbehave with majorities we can return to minority, but when minority gets too dysfunctional we can return to majority. Both types of government have their strengths and their weaknesses. The system of purgable majority is what we have as demonstrated by the minority governments in both Ottawa and Quebec City with no return to majority in sight in either case. The issue of "false majorities" is a bit of a misnomer. Yes, majority governments happen with, in some cases, less than the plurality of the popular vote, but I believe this owes more to a democratic failing in a lack of accurate, regular redistribution of ridings and of our lack of a preferential ballot than to any inherent weakness in the riding system.

Of course, this disingenuously misrepresents the supporters of MMP. What we have been saying all along is that stable government based upon coalitions has been the experience in European jurisdictions: that's majority rule by another name.

In Canada, false majorities are not "some cases," but the rule. They have nothing whatsoever to do with poor riding boundaries or the lack of a preferential ballot, but with the plurality system that characterizes FPTP. Indeed, even when on rare occasions a genuine majority government is formed, the results can be exaggerated: for example, Frank McKenna's victory over Richard Hatfield in New Brunswick, in which the Liberals won every single seat with 60% of the popular vote. How would redistribution have affected that one?

There are likely to be two classes of caucuses. The first is what we have now, with caucuses made up of riding representatives. The second is likely to be dominated by list MPPs. The Liberals and Conservatives will likely be the former, with the NDP and Green more of the latter. Within the parties where MPPs are dominated by representatives of ridings, current levels of internal discussion are likely to continue. In parties where the current party line is more important than the party principles, as in the NDP, the party-appointed members are likely to be the dominant voice within caucus and will not be as inclined to barter.

This is simply wild speculation, and caricatures the NDP--Graham's non-partisanship once more in evidence. What would the bartering consist of? Usually it has two axes, the first being contending regional interests, and the second being regional vs. national/provincial interests. In the case of a caucus where list MPPs predominate, why would the "party line," which seems to come out of nowhere in Graham's scenario, lead the caucus to electoral disaster by ignoring the regional interests that could be addressed to their advantage in the riding races? I submit that the discussions within such caucuses would not differ much from those in caucuses where riding MPPs predominate.

In Germany, everyone's favourite example, a 1996 study of MMP MPs found the following rather interesting bit of information:83.2% of German constituency MPs felt that they should represent all citizens in a constituency. By contrast, only 55.6% of list MPs in Germany felt the same way. This discrepancy goes to show that even German MMP MPs do not share this crazy notion that there is no second tier under MMP.

I'm not certain what it shows, frankly. As one of Graham's own commenters points out, how will a list MPP know how a constituent voted? So, in practical terms, when that MPP is doing constituency work, how would their private preferences (if that is what Graham's cited study indicates) find any real expression?

We have had numerous minority governments in the country's history, and extraordinarily few formal coalitions. Without a requirement to form a coalition before taking office, which is not something that I would support anyway, there is no obligation for a party to form a coalition with another if it feels it can govern on an issue-by-issue basis.

This is apples-and-oranges reasoning. Under pure FPTP, coalitions have of course been rare--who needs them, if you can get a majority in power with 40% popular support or even less? Under MMP, coalitions will become necessary, and the political culture will have to change from narrow partisanship to a more cooperative, win-win approach to governance.

Based on the last election, the Liberals should have had 62 seats with 2% popular overhang which is 3 seats short of the 65 seats needed for majority in a 129-seat house, still within the 3%/4-seat allocation margin for a fringe party balance of power.

I'd be interested in Graham's method for demonstrating this. As noted, the Fair Vote Canada analysis was that, in a 103-seat legislature, the Liberals (winning 46% of the vote) would have achieved around 48 seats under a proportional system. Extrapolating to a 129-seat legislature, a proportional system would have yielded 60 seats--ten short of a majority

2) MMP will allow small fringe parties to call the shots.

No other jurisdiction has a comparable MMP system to that proposed in Ontario for the reasons stated elsewhere here, such as an asymmetric list/representative set, low margin of entry, completely closed list, and the underlying political traditions and culture. What would happen here can best be described as "undefined behaviour" and my contention is that fringe party control will happen sooner or later, though not continuously.

That's simply speculation. In any case, as I noted earlier, MMP is not the Israeli "pure PR" system with its multitudinous parties and one electoral district: coalitions are more likely among parties with similar interests. Why would the NDP and the Liberals, for example, allow the Christian Heritage Party to call the shots with one or two seats? Better that the two parties join forces, swallowing hard, of course.

3) MMP will elect members who represent no one, and whom no one's ever heard of.

A party with pure motives and a completely democratic list, which may or may not come to exist, will challenge the list of another less perfect party in the election, without doubt. Where I have doubt is that any party will be sufficiently uncorrupt in its party list creation to be able to make such accusations without losing more than they gain. Even a party that has its membership in a party-wide vote select its list members is no less of a corrupt procedure than the already tainted party nomination procedures already in place in ridings across the province. Parties are made up of its members, and party memberships make up a tiny fraction of the province. Barely enough Ontario voters hold party memberships to achieve the MMP minimum threshold for being granted list seats were they all to vote themselves into their own party.

The notion of a balance of corruption is unsupported in any country with MMP save Russia. I suspect that, once voters who currently waste their votes feel that they have a real stake in an electoral outcome, party membership will increase. Much will depend, of course, on how the parties deal with the list question--we can agree on that. But Graham's equivalence--that the party nomination process under FPTP is "tainted," and so will the list process under MMP be--is hardly an argument for FPTP.

[Dawg: Graham goes on to display his fundamental misunderstanding of the MMP proposal by arguing that the 39 list MPPs will lighten the load in only 39 of the proposed 90 ridings, leaving 51 MPPs to do more work than their riding colleagues. But no one has claimed that a list MPP would be available only to the electors in a single riding. They could, and indeed are far more likely to, work in regions, not single constituencies.]

This is not my misunderstanding: it is based on Dawg's assertion in his original "lies" piece. What I wrote is, "if this were to be the case [...] with 39 list MPPs and 90 overgrown ridings, a maximum of 39 ridings will get additional representation from list MPPs seeking an alternate way into Queens Park." The case I reference is in his quote from Louis Massicotte, where he quotes: "Typically, a list member starts out by running unsuccessfully in a constituency. To run, he or she has to become familiar with the local issues. The person tries again in the next election. If his or her party comes to power, its number of list seats will decline noticeably and the only way to get elected will likely be by running in a constituency. For this reason, such a person will remain active in the constituency during his or her term of office and give such activities almost as much effort as a "directly" elected member." Massicotte, not I, is suggesting that a list MPP is only going to be interested in representing a riding in which that person has a chance of getting elected should the need arise, which in our imbalanced MMP proposal, means that fewer than half of constituencies would get a second representative, by Dawg's own argument.

Massicotte is, of course referring to MMP in Germany in which 50% of the members are list MPs. There is, of course, a riding-to-list fit there. Nothing in the Ontario proposal will prevent a list MPP from working regionally rather than in a specific riding: if he or she eventually runs, one assumes that it would be in a riding in that region.

Using European examples is not necessarily an applicable comparison to Ontario, whether it is Russia or it is Germany. If he would like to compare Ontario to Germany, we could just as easily compare it to Russia where the lists are openly corrupt. The system with which we will be creating our party lists is closer to that of Russia than of Germany anyway, as at least Germany has laws mandating that lists be created in a democratic way, something not part of the Ontario proposal and objected to by MMP's proponents.

The latter is simply not the case: here's one MMP supporter who would have had no objection to specific laws to ensure a democratic process, although the devil would assuredly be in the details. The view of the Citizens' Assembly was that the political process itself, including a legal requirement for transparency in list-formation, would accomplish that aim.

4) MMP is less efficient than Single Member Plurality (First-Past-The-Post)

[Dawg:Graham cites a frustrated Helen Clark, Prime Minister of New Zealand, who said in an interview, after trying to get her way on monetary policy, that it's hard to make tough decisions under MMP. This isn't an argument, but an anecdote. One would hope, in any case, that under a democratic system, tough decisions would be made with majority legislative approval.]

Tough decisions are always hard to get as party interests seldom match up with provincial or national interest. The party system, not the voting system, is my number one enemy in our democracy.

That's another debate. The party system itself is not under review. Maybe it should be, but, whether it's FPTP or MMP, the party system is what we have.

[Dawg: He continues his response to this point by claiming that reducing the number of riding MPPs reduces the effectiveness of riding representation, blithely ignoring the fact that there will now be list MPPs in the regions to shoulder the load. (While I can agree that Ontario, and Canada, have diverse regions that require region-specific attention, the same cannot be said of ridings. As an urban Ottawa voter, for example, I would be hard-put to name differences of interests between, say, the electors of Ottawa-Centre and of Ottawa-South.)]

I agree with the Ottawa example, which is why I suggested that MMP is less inappropriate for jurisdictions smaller than PEI, which Dawg interpreted as population, but I meant as physical size. When regions are as wide and diverse as Ontario, it is necessary to have regional representation. When our largest of over a hundred ridings is 50,000 square km, or around 10 GTAs, larger than the entire country of New Zealand, our issues are bound to be very diverse. That we will have list MPPs who will represent these sparsely populated vast regions of the province rather than the vote rich urban centres, like Ottawa, is not at all clear.

Nor is it unclear. See my response under [1]: there would be serious political risks entailed by ignoring the diverse regions of Ontario. But the geographical size of a riding is a bit of a red herring, in any case. People vote. Rocks and trees do not.

5) MMP does not require parties to explain how their lists are put together

[Dawg: Here, once again, Graham claims that list-formation will be a non-issue. See my response under (3).]

This is addressed under (3).


6) MMP will make contact with your representative more difficult

Voters votes will have put their parties into the legislature. Parties will have put the MPPs there. Try not to overlook this critical fact. The result is that list MPPs represent their parties to the voters and to their government, and not the voters to the government and their parties. In Germany, studies show that barely half of list MPs believe representing voters is high on their priority lists.

I'd like to see these studies, which run competely against Massicotte's findings, cited earlier. As he notes, list MPs today can be riding MPs tomorrow, so that list MPs do pay attention to their constituencies.

Tom Wappel is a disgrace without doubt for his handling of his constituency request from the supporter of another party. MMP takes Tom Wappel's behaviour and makes it the norm. MMP list MPPs especially will be expected to treat voters the way Tom Wappel did. The comment that MPPs tend to support the party line is a matter that needs addressing as part of democratic, rather than electoral, reform. One of my key objections to our system that is retained and even strengthened in MMP is the existence of whipped votes, which I find completely contrary to democracy. As I have stated in the past, if you need to tell your MPs that they have confidence in you, you do not deserve their confidence.

Again, as noted above, list MPPs will usually have no way of knowing how a person voted--the Wappel case is, in that respect, a bit of an aberration. The question of whipped votes has nothing to do with either FPTP or MMP: either system could exist without it.

7) MMP is confusing.

Dawg again takes what I wrote and only looks at a small part of it. I expressed a preference for Condorcet, but as I stated in my submission to the assembly, I believe that, for the sake of simplicity, Approval Voting would be the optimal. In between is Instant Run-Off, variants of which nearly every party uses for its own internal elections (whether instant run-off or just run-off).

That said Condorcet is not as complex as it is made out to be. I have fought three elections under variants of Condorcet, most recently last month and the voting is the simplest of any form of preferential, while the counting system is rather complex and more or less requires computer assistance. The results, however, are the most democratic of any voting system I am aware of.

I guess Graham concedes my point about the Scottish MMP balloting confusion. With respect to the Condorcet system, which, as he notes, requires computer assistance to perform the count, I don't believe that voters choose among candidates by imagining an endless number of person-against-person races. Voters like me are simple folk, who vote for the person or the party they like, and hope that both do well. Any system that encourages an array of second, third and fourth-best choices, and results in the election of (for example) everyone's second choice, is not one I would personally favour. As for the various run-off schemes, they produce a simple majority instead of a plurality of support for the winner, but they still result in large numbers of wasted votes for voters' candidates of choice.

8) MMP will produce two tiers of political representative

The point is that some MPPs directly represent ridings, while others do not. This is, by definition, two tiers. My question that he ignores completely is whether the two tiers is a good thing. It could be argued that it is, but at that point we should be considering a return to a bicameral government in which the proportional seats become a senate and so at least recognise the two tiers for what they are. I would argue that it is not a good thing, as all representatives should be equal.

9) MMP is undemocratic

Whether the party selection process for the lists are under scrutiny is clearly a matter of conjecture for both sides in this debate. Dawg believes they will be, I believe they will not be. That he would assert that either way is a fantastic assertion and that therefore the other is obviously right is more than a little conceited. I do not believe that the 39 names on each of the 4 major party lists and the several other lists that show up as additional parties join the fray will be under any serious scrutiny, nor do I believe that the process itself will be under serious scrutiny. Each party will have its own way of doing things, and short of selling off lines on the lists, I don't see the media caring enough to keep it on the front burner and make it an election issue. Parties may challenge each other, but in a sufficiently corruptible process, each party will adopt the same corruptible procedures.

See my note about the "balance of corruption" under (3).

Preferential balloting would be a meaningful form of electoral reform as it keeps what is good about FPTP and eliminates what is bad about it. It is no more wrong to say that voters will not make the "Natural Governing Party" their first or second choice than it is to say that voters will punish parties that have questionable entries on their MMP lists. Preferential balloting, except some forms like Borda, severely reduces strategic voting. It gives the advantages of the two-sided vote MMP gives us in spades, without the disadvantages of party lists. That all said, having everyone's second choice govern is still better than having everyone's fifth choice hold the balance of power.

Preferential balloting, as already noted, favours the Liberals because NDPers are unlikely to choose Conservatives as second-best, and Conservatives are unlikely to do the same for the NDP. It also forces second or third-best choices, when what we really want, I hope, is a system that favours first choices. I have already dealt with the "fringe party holding the balance of power" argument, and will not repeat myself.

Under MMP the FPTP's strategic voting model remains unchanged. The notion that it will not be suggests that MMP proponents believe the riding representatives are completely irrelevant which belies an agenda of pure Proportional Representation, an extraordinarily dangerous destination. Who wins in the riding will be just as important as who wins in the list seats to the overall outcome of the election. MMP also creates, on top of the FPTP strategic voting, a second layer of strategic voting as described.

Graham is right to point out that, so long as we have a riding system, strategic voting of the old kind will continue--I concede that point.[Not so fast--see UPDATE, below]* But there will be less concern about it, given that (for example) the NDP can now win list seats under MMP. Hence, the Hargrove solution--"NDPers, vote Liberal to keep out the Conservatives," will not threaten the political viability of the NDP, as it does under FPTP. Admittedly, though, a one-ballot MMP system (in which a single vote is for both party and candidate) would do more to reduce this kind of strategic voting--although that system has its own flaws, including the virtual elimination of independents.

10) MMP is divisive

A regional party will be able to leverage far more power with far fewer seats than is now the case under MMP. If the MMP proponents' prediction of near-constant minority governments comes to pass, which I don't believe is in doubt by anyone, a regional party with just a handful of seats will have far more power than a larger regional party would have today. That said, I don't see Ontario as particularly vulnerable to regional parties as compared to Canada as a whole and therefore am not terribly concerned about geographic-regional parties, at least in the short term, although I could definitely see a GTA-vs-non-GTA rift forming with myself being firmly in the latter, especially with the likely further GTAification of Ontario politics with the advent of MMP. My counterexample, as Dawg puts it, if he had cared to read it, was addressing the overall culture of cooperation within parliament, or lack thereof, and had nothing to do with regional parties.

Here, at least in the Ontario case, we appear to have some agreement about regionally-based parties. On the national scene, I don't agree that the same need for regional parties exists under MMP. I addressed this point already in my "Ten Lies" piece under point (10).

Graham's counterexample--the opinion of a retired police officer in New Zealand--is, as I said, not an argument but an anecdote. And I don't agree with it. MMP creates pressure for cooperation in the legislature, and against narrow partisanship, which would no longer be in the parties' best interests.

[Dawg: MMP, Graham asserts, will push us further away from such basic reforms as a legislatively elected Premier. It is not at all clear why this should be so. If the electors vote in MMP, they will see for themselves that the existing system is not part of the natural order of things, like gravity, but a structure subject to change. The debate around other reforms will be likely to grow, not dissipate.]

It might, but the debate is not exactly out of sight and out of mind at the moment. What passing MMP does is start us from further back when meaningful democratic reforms are next discussed. Having a legislatively elected Premier would be nice, but giving parties more power to create their own representation and weakening independent representation will only serve to weaken any efforts to make it ever happen.

I'm afraid that this doesn't follow. The parties have no vested interest in awarding the premiership to the winning party--the case under the current system, where the premiership is decided by that party and by one riding in the entire province.

I would welcome progressive change, but I do not consider MMP to be that progressive change and as I have said before, changing to MMP is not better than not changing at all. Changing to a preferential ballot and retaining a full or increased slate of representative ridings, and/or removing direct party control over the legislature would be progressive change.

I think we should try to demystify the notion of "party." Yes, they do become institutions, but they are really people who have agreed on common approaches to the issues of the day. That grouping will happen whether Graham gets his legislature of independents or not. One thing that MMP does offer that his preferences do not, by the way, is the ability for some MPPs to take the province as a whole into account in their political calculus. That can only be a good thing, and it militates against divisive, parochial politics.

Another concern of mine with the untimely passage of MMP is the probability that MMP will meet the "good enough" test of electoral reform and with previously ostensibly under-represented parties now holding a near-permanent balance of power, the issue will simply never be allowed to be re-opened unless it can be further modified to their benefit, which real democratic and electoral reforms will not be, as they would not be to the benefit of any party.

On the contrary, it will open up the debate. A Premier and Cabinet appointed by the legislature, for example, is in the interests of all parties. But, more than that, the electors themselves will be encouraged by their success in winning MMP to be active in this respect, and grassroots proposals for reform will not necessarily be ignored by the parties--they could, indeed, become political issues in election campaigns, now that the electors have sensed that they can actually change the natural order of things.

I maintain that MMP has no merit as an electoral system for Ontario, and nor do the arguments of the MMP lobby who have a great deal to gain at the expense of the province as a whole. One key point that proponents of MMP deliberately fail to grasp is that the representatives we send to Queen's Park actually have to govern the province through good times and through bad, and not just sit pretty.

I fail to see the logic or the point here. Under any system whatsoever, the problems of governance will remain. What does this have to do with the relative merits of MMP and FPTP?

I will leave out Graham's last paragraph, which consists largely of character assassination, and end here. There is only so much trouser-leg I can rip at one go.


*UPDATE (August 18): My friend Wilf Day of Fair Vote Canada reminds me not to concede this point after all. Under a proportional system, unless there is overhang, such voting is not "strategic"--you're just voting for the candidate that you prefer. If you go on voting for NDP candidates, rather than voting for your second-choice Liberal in a riding where the NDP has no hope, the electoral outcome (riding seats plus list seats) will still be the same, decided by the second (party) vote. Say there's a 40% party vote for Conservatives: they'll still get 40% of the seats, whatever NDP voters do for the Liberals in the ridings. If more riding Liberals are elected with NDP help, the Conservatives will simply pick up more list seats. So, unless there's a specific Conservative riding candidate that he or she simply can't abide, where's the strategy for an NDP voter in voting for a Liberal riding candidate?

No comments: