Thursday, June 28, 2007

Plan B: The best electoral reform for British Columbia may be the simplest

{ Here's a question: what would the results of the 1996 and 2001 BC provincial elections have been if between a quarter and a third of the seats up for election had been in the form of multi-member districts of say, 4 members each? (3 in the Lower Mainland 3 in the rest of the province, for a total of 24 seats?) Well, in 1996 chances are the Reform Party would have picked up a couple of seats , the Liberals would have wasted fewer votes in the South, the Greens might have picked up a couple. Certainly, a more proportional result, and almost certainly a minority government, although it is difficult to say whether Campbell or Clark would have become the premier. In 2001, a 57% vote for the Liberals in terms of first preferences still would have allowed at least 4 or 5 additional opposition MLAs under such a system. Another question: would the adoption of such multi-member districts be so radical a change that it would require another Citizens' Assembly or even a referendum? My learned friends Wilf Day and Raymond Lorenz--see their excellent comments below--remind me that the BC Citizens' Assembly did do its homework in considering my own preferred option--regional districts with open lists--before settling on BC-STV. But in the event that the BC-STV proposal again fails to reach the double threshold required in 2009, I propose a much simpler fall back position that anyone can understand. Just have 66 single-member constituencies, with 6 4-member regional districts. No preferential ballots, no fancy voting formulae. Just a lot of re-drawn constituency boundaries, for which a tolerably impartial mechanism already exists. --MC}

On October 10, 2007 Ontarians will follow British Columbia's lead and hold a referendum on a recommendation by its Citizens' Assembly for electoral reform. As in BC, the government has set a double threshold for approval of the proposal of at least 60% of the provincial vote, plus a majority of voters in at least 60% of electoral districts. I am betting , however, that unlike British Columbia in 2005, the Ontario proposal will succeed, for two reasons. First, the Mixed-Member Proportional ("MMP") system advocated for Ontario is much simpler and easier to understand that BC-STV is ("One Ballot, Two Votes" as its brochure says.) Second, the Ontario Assembly has strongly recommended that "a comprehensive, well-funded public education program, beginning in May and continuing through to the referendum, is vital." It would be difficult for the Ontario government to be as lax in educating the public as the BC government was in 2005.

But what if in 2009 a majority of British Columbians vote for BC-STV , but once again fail to meet the double threshold? Should this merely be the end of the matter? Electoral reform is a movement across this country, as evidenced by the spate of commissions, assemblies and even a Supreme Court case challenging the constitutionality of our present system. I suggest that a 50%+ vote for BC-STV should be taken as a mandate for more modest electoral reforms. These would take two forms.

Plan B: a la New Brunswick?

The best electoral reform for BC in this event would be similar to the one recommended for New Brunswick: that is, keep about 2/3 (i.e. 60-66 seats ) as single-member constituencies--so that the line of acountability with a single member would remain in place. Then distribute the other 24-30 seats between 4-6 multi-member districts--e.g. 3 in Lower mainland and 3 in the rest of the province. The BC Citizens' Assembly didn't go for this because they a) wanted high proportionality, which would not be possible under such a system; and b)were anti-party--i.e. didn't want a reform that could strengthen party organizations that would control the party lists, as in most forms of MMP, with their province-wide closed party lists.

I disagree with both of these positions. First, I don't necessarily want high proportionality, just a bit more proportionality, such that governments will be forced to be more representative and the bar for achieving a majority government will be raised (thereby avoiding the electoral disasters of 1996, when a party with fewer votes got a majority and of 2001, when a party with 57% of the vote received 77 out of 79 seats). Second, I don't necessarily want to weaken political parties or party discipline --I just wish to avoid making them stronger. The worry about giving too much power to central party organizations or party hacks can be addressed simply by making the multi-member ridings either "open-list" ( i.e. a form of MMP where voters do the ranking) or first-4-or-5-candidates past-the-post (which would not be MMP but which would still reduce the number of wasted votes). In this way, the effects on party strength and party organization would be comparatively neutral.

An End to Corporate and Union Financing of Parties

The federal government and Manitoba have ended corporate and union donations, increasing reliance upon both public funding and individual donations. This can only be salutary in a province where ideological polarization is regularly exacerbated by warring political elites backed by union and corporate treasuries.

Taken together, these two reforms will force all parties to work harder to appeal to a wider cross-section of British Columbians. Consensus-building and inclusiveness, rather than exploiting divisions and capitalizing on "split votes", would become relatively more important in electoral strategy. Something closer to real majorities, and not just artificial ones, would determine the ultimate parameters of political power. At least, to a greater degree than has typically been the case during much of BC's turbulent political history.


Anonymous said...

In Ontario, we have had the benefit of BC's citizens' assembly experience to guide us through this process.

Polls published by environomics are one objective way to measure how the word is getting out:

Democracy, without an informed decision, is no democracy at all. It doesn't make sense to save onpublic funds in such an important venture. (Overspending is not necessary either, when targetting for a specific level of voter awareness)

Raymond Lorenz

Mark Crawford said...

Thank you Raymond--you are right, it doesn't make much sense. But neglecting to fund education of the public is another way that politicians have (in additions to the double threshold) of reducing the likelihood of change. It will take some public pressure to ensure that the program is properly funded. But would it invovle funding the "No" side to an equal extent?

Anonymous said...


I remember seeing a BC public opinion poll where voters expressed their approval of the 60/60 referendum threshold (by a narrow majority). The poll was conducted just after the referendum. Unfortunately, I am unable to relocate that polling data.

In Ontario, I understand that the plans are both to inform the public about the mechanics (pros & cons) of the two systems, as well as about the citizens assembly process & its recommendations.

Although I have no objective proof, my feeling is that the more the public learns about the citizens assembly process (and the STV system), the higher the level of support for both the 60/60 referendum thresholds AND adoption of the new system.

I think most people would find it reasonable NOT to fund the yes or no sides beyond the necessary public awareness campaigns, say up to the 90 or 95% public awareness level by 4 weeks before the referendum.

Raymond Lorenz

Mark Crawford said...


Well, Ontario was guided by BC's experience, and now it is BC's turn to be guided by Ontario's! I have also heard that support for both the process and the BC-STV proposal generally goes up. So I hope that it passes. But if it fails a second time, we can safely pronounce it dead.

I was just speculating that a 50%+ vote again would signal the need for a "Plan B"--a less radical reform, such as the one outlined in my blog. If all wer are doing is adding 6 4-or-5 member ridings, and beinging in campaign finance reform, we should be able to do it without yet another referendum, particularly if these policy options have been discussed in the election campaign (of 2009 or 2013).

Anonymous said...

To make sure the referendum doesn't fail, the BC public should be asking all the local MLA's how they will be voting in the 2008 referendum and keep track of their answers. In the hypothetical case that the referendum fails again by a small margin, these answers can be used to make electoral reform an even bigger political issue in the next provincial election. (Voters can then choose between pro-reform and anti-reform candidates)

I don't think you will get much support for your plan B, especially from the CA alumni, because they presented the province with an all-or-nothing proposal. Plan B may be going against the original intentions of their BC-STV proposal.

A better plan B might be to immediately convene a citizens' jury to recommend to the government how to proceed next.

Raymond Lorenz

Wilf Day said...

An MMP model like New Brunswick's, but with open lists? That, of course, is exactly the MMP model the BC CA decided on, but did not finish the details before they decided by 80/20 that STV would be better for BC.

BC has far smaller ridings than Ontario's new 90 will be under MMP; about 1/3 the size. So a community that would have one local MLA under Ontario's MMP can have three local MLAs in BC, which is what made the BC CA say STV gives better local representation, which would not have been so in an Ontario-STV model. They were not just anti-party; they chose the same goals as the OCA, but they had way more MLAs per capita to play with.

But they never designed the details of open-list. This is not at all simple. Only one open-list MMP system exists, in Bavaria, and it works so poorly that no one has copied it. A citizen initiative for a really creative and complex open-list MMP model (that makes STV look simple) in the city-state of Hamburg has been derailed by political games and litigation, but shows the possibilities.

Who will design your open-list MMP model?

The BC Liberal caucus? That's what happened to Quebec's MMP model, which was a poor closed-list one-vote MMP model, panned by every political scientist as the world's worst.

Another CA? What makes you think it would be different than the last one?

A Commission like New Brunswick's? But then the governing party's caucus would re-write it.

The problems with the last BC referendum were two: no one could say what the districts or district magnitude would be; and the public education campaign was inadequate. Both should be remedied by the time of the next referendum, when I expect it was be accepted by a similar margin as I think Ontario will accept the choice of Ontario's Citizens.

Was the New Brunswick model more modest than BC-STV? It would have been slightly more proportional, with DMs of 14, equivalent in proportionality to STV 8-seaters (since two-thirds of a quotient of first preference will generally get you a seat under STV.) Simpler? An open-list MMP ballot would be quite comparable to an STV ballot in complexity.

I think you are on the wrong track. The CA process in BC was identical to Ontario's. Both produced models in line with the political culture of their respective provinces. Both choices should be implemented.

Mark Crawford said...

Wilf: When you say "only one open -list MMP system exists-in Bavaria-and works so poorly that no one has copied it", I suppose that is why Ontario, once it had chosen MMP over STV, went for closed lists. But maybe that is because they were thinking in terms of province-wide lists, and the Ontario Assembly found that voters didn't like ranking lists of that magnitude--as you say, that would be comparable to STV in complexity.

Now I explicitly adopted the "New Brunswick" label, so it is entriely fair and appropriate that you criticize my proposal on that basis. But I was really just referring to the idea of having maybe 6 regional multi-member districts, each with a magnitude of only 4-6 seats. My feeling is that voters would not mind ranking candidates in districts that small. But if I am wrong, we could just have the first 4-6 past the post elected. That would still deliver a modest proportionality improvement and increase voter choice without harming the quality of local representation much (and perhaps improving it).

I am aware that BC has more MLAs per capita than Ontario (which simply uses federal boundaries, I believe). But it seems to me that if BC-STV is defeated twice, it will have to be treated as dead.

I am thinking that in the event of that failure, having 6 multi-member districts would be an alternative proposal that could be undertaken by an electoral commission without having another citizens' assembly or another referendum. Treat it as a fall-back if you like.

Raymond: If you say that a citizens' jury or mini-assembly should convene in that event, consider my Plan B as the recommendation that I would make to such a group. Perhaps this discussion is premature, but I would like to see our leaders commit to doing something about electoral reform if we once again have a vote that falls between 50% and 60%. I would like to see that commitment made before May 2009.

Anonymous said...

For more details about the upcoming referendum visit the "Q and A" part of this site:

Raymond Lorenz

Wilf Day said...

"66 single-member constituencies, with 6 4-member regional districts."

Do you propose a compensatory system (MMP), or a parallel system 27% proportional, somewhere between Japan's (37% proportional) and Korea's (19% proportional)? A parallel system gives extra seats to a party already over-represented, which seems perverse.

If MMP, then BC would have six districts, each with about 15 MLAs: 11 local and 4 regional. Very like the model some of us thought the Ontario CA was heading for. But if the districts are unlinked -- no province-wide calculation -- you get regional sweeps or disproportional results, regional "overhangs," which hurt proportionality, just as occurs in Scotland. So, having chosen a low ratio of list seats, they went for province-wide lists.

But your model would work nicely, except for the unsolved question: how would an open-list model work? A lot of people might waste their personal preferences by piling them up on a "big name," totally wasted if the "big name" wins a local seat as he likely would. The winners of the open list contest would then be chosen by a handful of untypical voters. That's exctly what's wrong with Bavaria's model.

So we need some sort of transferable vote. Prof. Matthew Shugart has been working on this for some time. Please let him know when you have a good model. But it may look a lot like STV.

Mark Crawford said...

Wilf: Than you for that excellent comment. I think that it helps to show that a lot of thought and comparative research went into BC-STV (and Ontario MMP, for that matter). My own views may seem a bit ambiguous or even confused, but that is because I am wearing two hats: first, my own considered preference and preferred choice; and second, the more pragmatic compromise or "second -best" reform that could arguably be implemented even if BC_STV is laid to rest in the next referendum.

I think the difference between myself and the majority of the Citizens' Assembly is that I do not place quite as high a value on proportionality. In part, this is because the putative "equality" achieved at the ballot box doesn't always translate well into the parliamentary context with strong party discipline; small parties can sometimes wield disproportionate power. There may also a philosophical basis for preferring regionally concentrated preferences to diffuse ones, particularly if they tend to be about priorities and trade-offs concerning the whole range of issues confronting a community. I also tend to see raising the bar for the achievement of a majority government to be unambiguously good, whereas creating the conditions for permanent multi-party coalitions may not be.

I remember writing an essay on the implications of electoral reform for national unity for Alan Cairns about 20 years ago, in which I suggested doing the calculations for both the Broadbent formula (parallel provincial lists, if I recall?) and Irvine (Compensatory MMP, I think?) and simply splitting the difference between them. Broadbent would have abolished the Senate and attempted to institutionalize the principle of regional equality in the House of Commons, where it comes into conflict with the principle of representation by population, and, as you say, gives more seats to parties in their strong areas than in their weak ones. On the other hand, by using the provincial seats to compensate the weaker parties, Irvine would have punished the stronger party by depriving it of professional political careers. The old adage about the way to get a safe government job is to be a losing Liberal candidate in Western Canada would be institutionalized in Parliament--where members of weaker parties who win "compensatory" provincial seats might also lack legitimacy as regional spokesmen. There may be a conflict between the desire for proportionality and the desire for regional representation, party-building, and career-building. Instead I proposed neither Irvine's "topping up" formula (which gives the lion's share of 'career building' PR seats to the less popular parties) nor Broadbents "reflection of the popular vote" (which does not sufficiently correct regional imbalances int he party caucuses), but average of the reslts of these two formulae, reflecting our equal valuation of a)the fair distribution (and hence the enhanced legitimacy and effectiveness) of professional career oppportunities and b) correcting regional imbalances in the House of Commons.

But I digress a little in talking about the federal system. My "Plan B" really refers to what I think we can do in BC, even without approval by 60% in a referendum. It ain't STV, it ain't even MMP, it's just saying that you put one X by local candidate and up to 4 X's for the regional candidates. There may still be some regional sweeps (the example of Liberals of 2001, I estimate that they could have swept up to three of those regional districts) but there would still be an improved proportionality and an opposition caucus at least three times as large as the one we got. Surely everyone could see that as an improvement over our existing system, even if they love majority government?

Raymond Lorenz Ont., Canada said...


I have 2 comments to make. One is about a legitimate process, the other is about your electoral model.

In Ontario & BC, a lot of care has been taken to come up with a process that caries some legitimacy. (Politicians who legislate an electoral system can be perceived as acting in their self-interest, rather than acting for the good of the province)

A few elements of a legitimate process.
1. Learning (Advice of a variety of experts, seeing how things work in other countries)
2. Public consultation
3. Non-partisan guidance
4. Multi-party guidance (all-party committees, etc.)
5. A semi-judicial process (citizens jury, assembly, etc.)
6. Public consensus (referendum)

My 2nd comment is about your electoral model.

When elections cover large & populated jurisdictions, a party system is the norm. There is a practical reality to communicating & clarifying provincial policy to large numbers of voters, and among large numbers of legislators.

In this context, proportionality among parties is an important value for many voters.

Your electoral model "plan B" seems to have a mix of single-member FPTP ridings and "block vote" regional districts.

Under our current partisan system, I do not see any more proportionality offered by the "block voting" you have described. I expect that most voters would still vote along party lines, as they currently do.

Raymond Lorenz (Ont.)

Mark Crawford said...

Thanks Raymond.

Re: Your first point: I agree, this is a good summary of the elements of a legitimate reform process and happily precedents have been set for "system" change (e.g. from SMP to PR-List or PR-STV). But we are also stuck with a high threshold, a motivated opposition, and a reluctant political class. So what happens when we repeatedly fail to ratify system change? Is there no fallback or moderate reform that can be implemented?

Re: your second point, my proposal is not designed to be a "block vote" model, although the tendency to vote along party lines has to be acknowledged. But for example would the BC Liberals have really swept ALL SIX of the multi-member districts in 2001, achieving as much or more disproportionality than was actually the case? I have to think that more than 2 opposition MLAs would have been elected--New Democrats and/or Greens in East Van, 1 or 2 in the Kootenays and on the Island. If new age voters really hunger for greater choice as much as we like to say, I think that there would be a trend toward using that choice? I like to imagine a number of marginal voters in Nelson-Creston using 4 of their 5 votes to ensure that the NDP lost the election, but saving 1 vote for Corky.

Raymond Lorenz said...

A "fall back" position should come from the citizens themselves. (ie. a petition to accept the 2nd referendum results as binding, under extraordinary circumstances...two consecutive referendums signalling a strong mandate from the voter)

Don't let your politicians off easy...insist that they represent the will of their constituents. 77 of 79 your MLA's should already be questioned on this.

Be meticulous & persistent in the scrutiny of your governments' referendum procedures.

Raymond Lorenz

Mark Crawford said...

Well put,Raymond!

Raymond Lorenz said...

...shows how difficult any agreement on electoral reform can be...

Raymond Lorenz

Mark Crawford said...

I know that you and Wilf are more optimistic about BC-STV passing than I am, but I have to think that we missed the boat somehow in 2005. Had the electoral boundaries been drawn as they are now (so that people had known what they were voting for), and had there been a public education campaign like there presently is in Ontario, the BC Referendum would have met the 60% threshold in 2005.

Remember, "perfection is the enemy of the good", and the Citizens' Assembly went for a pretty radical proposal, even though they knew about the 60% threshold and knew that they would be relying upon a reluctant political class. Even "STV lite"--just 3-member districts in the south and 2-member districts in the north--might have had a better chance of passing.

Anonymous said...


I would not call a recommendation approved almost unanimously by the BC citizens' assembly and then by 57% of the voters as a "pretty radical proposal".

25% of the Ontario citizens' assembly also preferred STV to MMP and designed an almost identical system to BC's proposal, with DM's ranging from 3 to 7, and an average DM of 5.

I believe the BC-CA has probably gone about as far as it could, to find public consensus.

Raymond Lorenz

The Doctor said...

"just a bit more proportionality"
can be achieved with a single member preferential system.
If you wish more than that, it is case of going with an STV, the smaller the number of members per electorate the less proportional it will be.
Adjust to taste as desired!