Topp's Coalition Memoirs, part2

105 posts / 0 new
Last post

Layton and the rest of the NDP crew seem to be on a roll like never before.

With Moe Sihota back in the saddle with the BC NDP on the Left Coast, the NS NDP Government on the East Coast, and the increase in NPD polling results in Quebec, I'm starting to finally get hopeful for the NDP in the next federal election. 



The rumblings of discontent within the Liberal caucus may have been overblown in the press, but everyone in politics knows that only the winner gets to smoke the real cigar. When Grit MPs privately acknowledge missing the political professionalism of Stephane Dion, it is time for a serious change of course. Iggy continues to sail steadily for the rocks.

At every major policy juncture, Ignatieff has forfeited the chance to distinguish his party from the government. He has backed the Harper position on Afghanistan, on fighting the recession, and most recently on the unpopular harmonizing of provincial and federal taxes. In the process, he has handed the political dividend to be had from opposing these divisive and unpopular issues to the NDP.

Even on those major policy issues like pension reform, where he did take a position, he ended up looking suspiciously like Jack Layton in less of a hurry - and with considerably less sincerity. The NDP have been early and consistent on their stands, making the Liberals look like timorous imitators in search of an issue.


NR, I agree the NDP is doing well.  I just hope your only criteria for whether they're doing well or not is NOT simply their appearance in columns and stories cited on National Newswatch.  It's the coverage in communities across the country when Layton has visited or our MPs are hard at work that will wind up being much more important.


Outrageous is right.


Bankers getting $8.3 billion in bonuses

NDP slams 18% increase despite bad economy


Now that we have all had to a chance to read Topp's accounting of what it is really like trying to work with the Liberals it should be clear to all that the NDP needs to avoid any dealings whatsoever with the Liberals. Just like you would with the plague.  


I would rather work with libs than cons.


NR, I'm not sure I would come to exactly the same conclusion as you; never say never.  But he sure does point to the inherent problems, alright.  My take: if the Liberals have as disastrous a campaign as I think they're going to in April, they'll jettison Michael Ignatieff, and be knocking on our door with a new leader looking to make a deal.

Wilf Day

Remember, Ignatieff did sign on to the coalition. And then he convinced the Liberal Caucus and the Liberal Executive that the party had to have an instant new leader to become Prime Minister after Dion blew it. He did it so well that Bob Rae had to make it unanimous. Outfoxing Bob Rae requires serious talent.

And then he changed his mind. Partly because of reaction from the West, which Ralph Goodale had presumably failed to anticipate.

So who is to say he can't change his mind again? He's a student of Mackenzie King, who never let his on the one hand know what his on the other hand was doing.

Malcolm Malcolm's picture

Why, Thorin?  Given a choice between a far right party that is honest about being a far right party and a far right party that lies about it, why would you prefer to work with the far right party that lies about it?


How's about we simply build an electorally viable NDP and form a bloody government ourselves?


I do like that option best of course, Malcolm!

But now, let me ask you another question: would entering into a government coalition assist with achieving that goal or serve to stall it, do you think?


Malcolm, there will be no NDP government until the Liberals self destruct.



The coalition rally in Vancouver at Canada Place was one definitely the most genuine political events I have experienced.

It had such a hopeful, non-partisan atmosphere.  It made me giddy with its novelty and potential.  It was good to be in a room of 100's of people who seemed earnest in trying something new, something important.  I believe people across Canada felt similiarly in other coalition events.

Whatever it was, I felt it was 'the real thing' and is something I would like to vote for and could help break out of the cynicism and entrenched interests that pervades Canadian body politic.

Steve_Shutt Steve_Shutt's picture

Coalition, phase two.

I loved Topps' review of the behind the scene negotiations and feel that the coalition model still has potential - for the NDP, Canada and even (ironically) for the Liberals.

I think Jack's fundamental conclusion that the NDP needs to find some way to reorder the game is the correct one.

Federally we are doing much better than we did under the two Alexa's, sometimes skirting 20% in the polls. Nothing to sneeze at but, ultimately, it limits what we can do to pursue policies that we support. We can be a conscience in Parliament, an organizing ally for NGOs in their lobbying efforts; we can even win small victories by leveraging our power in minority Parliaments (with differing levels of success depending on who we are trying to work with). What we cannot do is govern.

So, how do you change the game? One option (hope) is that we somehow supplant the Liberals through a combination of Liberal self-destruction, Liberal scandal, a broad collective political re-alignment of the electorate towards the left, and an unprecedented combination of NDP electoral success coupled with Liberal electoral disaster. Perhaps this is still a realistic goal (surely the Liberals have never been more ineptly led than they have in the Dion-Iggy era) but like almost every horror film you have ever seen, this undead corpse is very hard to finally kill off.

I am inclined to believe that even were we to cobble together the ingredients to bring about such a switch of position between the NDP and the Liberals we would run the very real risk of losing our identity. Much like the Labour Party in Britain under Blair, the need to occupy the broad chunk of the political spectrum ostensibly represented by the Liberals, we would, IMHO, lose our distinctly left positioning.

If you can't re-order the game by changing the players, then perhaps you can re-order the game by changing how it's played - double down on the coalition plan. Much of the outside criticism of the coalition plan centered on the role of the Bloc and the poor PR efforts of a divided Liberal caucus and its embattled leader, Dion, specifically the handy-cam from hell. Presumably both could be better handled if given a chance to do it again. The one criticism that had a ring of "truthiness" to it if not actual validity was that it was a plan that had never been put to the voters.

I agree that we elect a Parliament of MPs who legitimizes the government of the day through votes of confidence - the coalition was constitutionally valid but there was something to the point that on Election Day the prospect of NDP ministers in a Liberal PMs cabinet hadn't been put to the electorate. If we are going to take out the Tories through a coalition next time, then we need to do so expressly.

Now we can simply leave open the option to enter into a coalition should the policies and numbers justify it post-election (presumably with a combined Liberal-NDP majority without the need for Bloc support). I believe, however that the election campaign would become bogged down in coalition talk (forcing the Liberals to categorically deny any plan to enter into a coalition all the while fielding off, with Jack, "secret deal" questions from the media). Indeed, I imagine that whenever the writ is dropped the Tory attack-dogs will fire off Iggy's signature on the coalition deal to every fax machine in the country. If we (and the Liberals) are going to have to deal with this anyway, why not make a virtue of necessity (though I think there is inherent value in the notion in any event) and run on a progressive coalition agenda with an aim to not only deny Harper a majority but to replace him as the government. In order to truly do that the coalition would need to do more than have a joint program but would need to prepare an electoral coalition as well.

Sitting MPs, Liberal and NDP, would run unopposed by their coalition colleague and in seats where one coalition’s candidate ran a close second last election (the exact amount to be negotiated but say within 10%) that party's candidate would also run unopposed. Kitchener-Waterloo would have not NDP candidate; Surry-North would have no Liberal candidate and, hopefully, neither riding would be burdened with a Conservative MP next election.

So why would the NDP do it? For the same reason as we had last year. The prospect of bringing down Harper's government and getting a chance to actually govern - implement NDP policies at the national level -isn't something to be casually dismissed. Isn't gaining power the reason we run candidates in the first place? At a broader level, it becomes a game changer long-term. At some point the coalition will falter and we have to go back to fighting the Liberals but we will have established ourselves in ridings we had never held before and will have established a precedent that, presumably, NDP ministers are able to govern and govern effectively.

Why would the Liberals do it? They like to govern (at times they seem to think it is their god-given duty to do so) but the fact is they hold little prospect of doing so for the foreseeable future. A year ago at this time Iggy might have saw the chance to waltz into the PMO after a disastrous economic meltdown and create an election at a time and place of his own choosing, that plan stands shattered at this point on a couple of fronts (the economy is not as bad as it was presumed to be and Iggy's skills in orchestrating his political moments has been similarly exposed as less than originally advertised). Perhaps he can create the image of himself as a PM, but it is much easier to do so from 24 Sussex as opposed to Stornaway.

The grim fact that the Liberals need to face, recently remarked upon by Chantal Hebert, is that the Liberals have been unable to beat a unified Conservative Party since Trudeau bested Clark! The Chretien majorities were great times for the Liberals but those who have looked at the numbers must see that even with a weak NDP presence, it was not the Liberal strength that brought them their majorities but Conservative division.

The Liberals, maybe even more than the NDP (as we're doing close to as well as we have ever done) need to change the game too.

Bookish Agrarian

A very thoughtful post.

One question though, what would you reccomend, for the sake of discussion, in ridings were the supposed second place party is running a very lacklustre candidate and the third place party has managed to recruit a real barn-burner of a candidate?  Should the NDP who would be more likely to be in that situation just leave the playing field in honour of the previous runner-up even though the new candidate has a bum arm?  Would that serve progressive voters?  What about right-leaning Liberals who have little spitting room between them and a Conservative opponent?

Steve_Shutt Steve_Shutt's picture

Thanks for the compliment Book, something not always available here :-)

I agree that my plan is only one of many possible options for dividing up "winnable" ridings. I think it unrealistic, and unnecessary, to divide up every riding. The truth is Red Deer is not going Liberal or NDP no matter who we put forward.

One issue that would need to be considered too is how this scheme would impact the financial bottom line of the parties if they are giving up votes in uncontested ridings would they presumably be able to recoup them in ridings where they are unopposed by a coalition candidate?

As for your point main, I'm in a hockey pool, and have been for years, so my first fall back alternative would be to have a draft of the available "winnable" ridings. Given both current seat count and vote % last election (and even current polls) there would be more Liberal picks than NDP. Not sure which system would be more favourable to which party - though in the end if the goal is to defeat Conservatives I suspect that there will be a great deal of similarity in the seats assigned to each party's candidates either way.

As for who the Liberals put up for candidates in the ridings we would not be challenging them in (remember in my plan existing MPs would be unopposed) there may be the temptation to pick a more right-leaning candidate if there is not left-flank to defend but if the Liberals picked someone patently unpalatable to most NDP supporters - would they come out and vote for that person? I would think not.

The reverse is true too. Just because we would not have a Liberal opponent in a riding does not automatically deliver us every former Liberal vote - in fact I suspect our voters would be more reliable coalition voters than Liberal voters - so it would be wise for us to nominate candidates that don't frighten away the very voters we hope to pick-up. That said, I suspect the prospect of recruiting "star candidates" increases if they know that they would be coming into a riding where it would just be them and the Conservative, non?


The proposal might have a chance of working if both parties campaigned in favour of it, but I've seen no evidence the Liberals will ever do this.  And if they don't, we have a better chance of picking up seats from the Conservatives out west and in southwestern Ontario, with the Liberals *in* the race, not out of it.

Steve_Shutt Steve_Shutt's picture

I too don't see the Liberals being outwardly interested either but things can change, and change quickly, in politics.

The Liberal dominance through the 20th century was based on many things but one key factor was their near monopoly in Quebec. (Yes, Diefenbaker and later Mulroney found ways to temporarily pull in the soft, and sometimes not so soft, nationalist vote as a counterweight but it didn't ). When the Mulroney coalition splintered the Bloc realignment in Quebec fundamentally changed the game for the Liberals - moreso than for us or the Conservatives.

They were, under Chretien, able to capitalize on the Reform-PC split to create their "false" majorities but with the reunification of the Conservatives, I don't see how the Liberals can hope for more than a minority win in the forseeable future. They may be prepared to live with that prospect for now but the Liberals don't fare well in the wilderness and I'm thinking that the attraction of regaining power, even if they have to share it with us, may prove overwhelming.


A lot of us agree with you that we expect the Liberals have a newfound willingness to consider various governing arrangements, but that doesn't require or imply the kind of proposal for how the parties campaign. Not to mention that various permutations have been much discussed here, and that kind of proposal is making no headway against the sense there are 6 different ways it just cannot work.


Not least of which is the huge number of voters that would be VERY angry that their vote choice was taken away from them, and that they would NEVER vote for the Liberal/NDP candidate instead.

Leaving aside all the number crunching that says those numbers would blow out of the water any net gains in seats that simplistic addition presumes to exist... all of us have a hard enough time motivating voters, and really don't need to give them another reason to be pissed off.

Steve_Shutt Steve_Shutt's picture


I don't disagree with what you say per se, but I have to ask if there has ever been any polling to back up the "pissed off" reaction you anticipate? I'm not as convinced that the reaction would be there.

Certainly there would be a % of Liberal voters who would rather vote Conservative or stay at home as opposed to vote for the NDP, and a % of NDP voters who would never support the Liberals. I don't know how those numbers play out but I still believe that the net positives (in terms of seats and even in terms of overall votes, if the ridings are divided with some forthought) far out weigh the negagtives on this.

It would be an interesting question to insert into a poll after your generic "if an election were held today..." query - if the "Liberal (to Liberal supporters)/NDP (to NDP supporter) candidate had withdrawn in accordance with a Liberal/NDP coalition agreement which party's candidate would you vote for?"

Lou Arab Lou Arab's picture


Shortly after the last Alberta election, SEE Magazine published a poll of Alberta Liberal and NDP voters asking them what their second choice was.  Among Liberal voters, a slim majority picked the Conservatives.  Among NDP voters, a signifigant minority picked the Conservatives.

What this means (at least in the Alberta context) is that with one fewer opposition party, the Conservatives increase their majorities.  In Edmonton Calder, there was a Liberal candidate, but his campaign was very weak.  As a result, the Liberal vote dropped by about 10 percentage points. Most of that vote went to the Tories, who increased their share of the vote by 7%.  NDP MLA David Eggen saw his share of the vote increase by 3%, but it wasn't enough and he lost.  As a result, the opposition lost a seat to the Tories.

A similar result happened in Peace River, where the Liberals were unable to nominate a candidate.  The NDP more than made up for the Liberal votes (it helped that the NDP had a good candidate) but the Tory vote increased from 55% to 65%.

I don't know how these results would transfer today, given the rise of the Alliance.  But I suspect the Alliance could capture many votes that once were for the Liberals and NDP, making an arrangement between the two parties all the more dangerous. 

Obviously, these results might not translate to the federal scene.  But it's folly to assume that 1+1 will equal two when it comes to electoral coalitions.  It could easily lead to 1+1 equaling zero.

Steve_Shutt Steve_Shutt's picture

Hi Lou,

I live in Alberta Lou so you don't have to tell me the unique nature of politics here but while I'm not sure it is the best place to draw broad analogies from, you are absolutely correct that the arithmetic is not going to be obviously beneficial in all instances.  Even so, I still believe that if done with some forethought we can pull more wins out than losses.

Pulling an example out of Edmonton, let me ask you about Edmonton-Strathcona.  Does Linda Duncan's 42.6% to 41.6% lead grow or shrink if the 9.1% who voted Liberal is re-distributed?  A fair question to ask I think, and yes there is the risk that that vote goes disproportionately to the Conservative however my sense from the Provincial Edmonton-Strathcona experience with Raj Pannu (and now Rachel Notley) is that once it became obvious who the anti-Tory champion was, subsequent elections were much better at consolidating the Liberal vote under the NDP banner.

I would think a similar process would be at work in ridings like BC's Surrey North (which the Conservatives won over us 39.4% to 36.2% with the Liberal taking 15%), NS's South-Shore St.Margarets (with the Conservatives at 36.0%, us at 33.7% and the Liberals at 23.9%) or even Oshawa (with the Conservatives taking it at 41.4% to our 34.7% and the Liberal taking 16.0%).  Conversely maybe the Conservaties extend their lead in a place like Saskatoon-Rosetown-Bigger (where they one by a single percentage pt and the Liberals polling at under 5%) or Vancouver Island North (Conservatives got 45.8%, we got 41.4%, the Liberals 4.2% and the real "spoiler" for us might well be the 8.0% who voted Green - though the claim would be that they might have stayed home but for the Green candidate).

My preference would, not surprisingly, be for us to adopt a proportional rep voting system for a host of reasons but with limited prospects of that happening anytime soon I'm looking for ways to get us out of the electoral logjam we find ourselves in.  Jack has us doing well.  In terms of seats we are close to our historical high (with significant inroads into Atlantic Canada and the prospect of a continued, meaningful presence in Quebec, something we have never been able to experience before).  Incremental growth is good but it doesn't change the final equation that we are the 4th party in Parliament with little prospect of exercising power directly in my lifetime.  I'm not satisfied with that outcome and am looking for ways to think outside this box.

Malcolm Malcolm's picture

@ Ottawa Observer - Governing coalition deals are fine - AFTER and election.  That said, while the Liberals are more LIKELY to give us what we want in a coalition, it would be stupid of us to assume that the Liberals are the only potential coalition partners.  If, in coalition negotiations, the Conservatives offered more, we should not rule it out.  Even though we are more LIKELY to get a better deal, in the end, from the Liberals in that circumstance, ruling out other scenarios weakens our negotiating position.


@ JRoothman - The best possible outcome for Canadian politics is the utter destruction of the Liberal Party as a viable electoral force.


@ Steve Shutt - The problem with your proposal is that it is utterly unrealistic.  As Lou and others point out, Liberal voters, absent a Liberal candidate, are more likely to vote Conservative than NDP.  And while NDP voters, absent an NDP candidate, are more likely to vote Liberal, a significant minority (particularly in Saskatchewan, rural Manitoba and rural BC) are more likely to vote Conservative.  This does not account for the significant number of both Liberal and NDP voters who would simply stay home.


Using the last election as a basis, I have crunched the numbers.  There are only 64 constituencies where the combined Liberal - NDP - Green vote exceeds the votes of the winner - 55 Conservative seats and 10 Bloc.

* 45 Conservative seats where the Liberals were second

* 8 Conservative seats where the New Democrats were second

* 2 Conservative seats where the Greens were second

* 8 Bloc seats where the Liberals were second

* 2 Bloc seats where the NDP were second


I then calculated the net proportion of the other "coalition" parties vote the leading "coalition" would need to retain in order to take the seat.  That is NET retention.  If the leading "coalition" party lost by 500 votes, they need 501 NET votes to win.  If 100 votes shift to the Conservatives, the leading "coalition" party now needs 601 votes to win.


In the example of Simcoe North, for example, the Liberals trailed the Conservatives by 22%.  The combined Liberal, NDP and Green vote was 22.7%  The Liberals would need to gain a NET 96.92% of the NDP and Green vote in order to win.  If even 3.09% of NDP and Green voters stay home, and every single other NDP - Green voter votes Liberal, the Conservatives still win.  If 1.55% of NDP and Green voters vote Conservative, the Liberals cannot win.


Repeated Canadian Election studies suggest that the proportion of voters likely to stay home if their candidate is not on the ballot FAR exceeds 3.09%, and that the proportion of NDP or Green voters likely to choose the Conservatives over the Liberals is significantly more than 1.55%.  (As discussed, Liberals are actually MORE likely to choose the Conservatives over the NDP or Greens.)


I then calculated the electoral result using a 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100% NET retention.  It isn't until we reach 70% NET retention that the "coalition" forms a majority government.


This does NOT take into consideration how the electoral deal plays out in other constituencies.  Given (as oft noted) that Liberals would be MORE likely to vote Conservative than NDP, we can presume as practically a given that the NDP would lose at least nine seats (London - Fanshawe, Sault Ste. Marie, Thunder Bay - Superior North, Welland, Edmonton - Strathcona, Western Arctic, Burnaby - Douglas, New Westminster - Coquitlam*, Vancouver Kingsway).  [Data used general election results - in the recent byelection, the NDP broke 50% in NW-C.]


More can be found at:


Frankly, the data would work out similarly in any election campaign.  Thus I conclude that there are three types of people who argue in favour of these daft and unrealistic electoral deals:


* People who don't understand politics

* People who don't understand arithmetic

* People who support the Liberal Party and would like to see the NDP and Greens commit electoral suicide.


A much simpler solution would be to have an "Australian-style" ballot whereby all parties run candidates in all ridings but people have to number everyone on the ballot in order of preference. That way all the parties can keep competing with each other just as they do now - but people would be able to indice their second and third choices.

Its not perfect, but i think its far better than expecting major national parties to not run candidates in half the ridings in Canada.


How's about we simply build an electorally viable NDP and form a bloody government ourselves?
That's been the $64,000 question for the party's entire history. 

I think the NDP & libs should gang up on the Tories (and the Bloc where necessary if possible), change the electoral system, hold a new election and then never have to think about "strategic" voting again. Yes, I know, the voters are supposed to be "election-weary" and "the liberals will never agree" to PR. Those are obstacles to surmount, not  necessarily deal-breakers in an of themselves.

Malcolm Malcolm's picture

The only other national party that would EVER agree to electoral reform is the Greens.  Both the Conerverals and the Liberatives know that the current system is the ONLY way they can win - and the current system gives the Bloc a disproportional number of seats.


If you want to waste your life begging the Liberals to be nice, be my guest.


I can think of any number of more useful ways to spend your time.


Frankly, ten years of constant masturbation would accomplish more than wishing the Liberals were somthing they're simply not.


Until the last few years I would have agreed that the Conservatives were the second preference of most Liberal voters. That view seems prevalent in this discussion but I don't think it is valid anymore.

Certainly it was the case in the CCF years. Witness the Lib/Con provincial coalition in British Columbia and how it took seats from the CCF.

But in the last few years there has been a greater difference betweens the Cons and the Libs and a lesser (arguably perceived) difference between the Libs and the NDP. After the cooption of the Conservatives by Reform and certainly during the Harper government, the Cons have lurched to the hard right. No one but the Cons call the NDP socialist anymore. And the Liberals have appealed in the dying days of the last several elections, often successfully, to NDP voters to defect to them as the best progressive alternative to the Cons.

In other words the perceived differences between the Cons and the Libs have grown and those between the Libs and the NDP have shrunk.

Indeed, in the last election, I remember polls showing the NDP as the most common  second choice of all other parties. Other polls have shown that up to 30% of Candians feel the NDP best represents their values, athough only 2/3rds of them end up voting NDP. Perhaps someone on Babble has access to those polls.

I don't doubt that a significant number of Liberals would vote for the Cons over the NDP in a straight fight but I strongly suspect, subject to some regional variations, that the NDP would get the lion's share. A bigger portion of the NDP vote would be trransferred to the Liberals where they are the alternative to the Cons.

As for the Australian model mentioned by Stockhom, it would certainly cost the Cons quite a few seats. One drawback might be that in Lib/NDP contests the Con vote would largely drift to the Liberals.


What is your basis for that nicky?

The numbers gathered, right up through the 2008 election, is that the consistent second choice of Liberals is the Conservatives by a strong majority. This has been very stable, and it shows directly in the very large sample surveys that have been done for decades under the same conditions. Its also backed up by looking at the movement of raw vote numbers in the ridings where races are close... which is where you would have seen it if there were results different than the large general national trends.

Are you going on anything other than your perception of greater difference between the Liberals and Conservatives? Because it looks like that in your surroundings says nothing about what is generally true. And suppossing you could do a very large poll that measured people's perceptions of amount of distance between the Liberals and Conservatives, establashing your hypothesis that there is a significant difference established over recent years, that would still only hold out the possibility that significantly more Liberal voters would therefore shift from choosing the Conservatives as second choice to choosing the NDP.


. Ken S seems to relying on his own perceptions as well.

Maybe we can resolve this by actually looking at the polls  from the last election which i remember as showing consistently that the Conservatives were the last second choice for voters for all other parties. I don't think i am misremembering this. Can someone actually produce the figures?


You are right that someone else needs to provide a reference to the voter studies- too big to just post the numbers.

But we're not talking about polls. I'm not sure what you are remembersing, but polls that include second choices are epheremal at best, and far from rigorous. [And I don't think many of them even ask the questions you are talking about.] The studies I'm talking about have been done after every election for decades, and ask people how they actually voted, what were there second choices, and some other metrics of party identification [or not].

The studies have been discussed in this forum before. And numerous times Ottawa Observer and others have given careful analysis of actual movements in raw votes in close race ridings.

But none of that ever seems to dim in the slightest the passion that it must really be otherwise.


Like I said above, looking at polls is not definitive. And there are definitive studies that are aimed at establishing these questions.

That said:

nicky wrote:

....looking at the polls  from the last election which i remember as showing consistently that the Conservatives were the last second choice for voters for all other parties. I don't think i am misremembering this. Can someone actually produce the figures?

I doubt that you or anyone else can find more than maybe a single outlier poll showing the Conservatives NOT to be the most numerous second choice of Liberal identifiers. Which wouldn't stop it from being a sustaining myth.


Nicky's memory is correct, if you read his/her post as meaning second choice for all voters counted together (which is how I read it). [url= article[/url] is also ambiguously written, but clearly shows the Liberals (at one point in the last campaign anyway) as being the second choice of 28% of all voters, and the NDP and Greens tied at 25% - which mathematically would put the Cons behind all three.

It also notes that the NDP was leading the Greens at the start of the campaign (26% vs. 22%), but slipped.

Ken, I don't know what you mean by "definitive studies" conducted after an election. Surely the outcome of an election would influence voters' perceptions of how they might have voted, let alone who might have been their second choice?


Nothing is perfectly definitive of course. And of course what people say about how they voted is going to be coloured by the actual results. But thats part of what the question process of the Canadian Election Studies surveys are about. They have the focus and resources to both ask a higher quality and more 'inquisitive' panel of survey questions, and to ask it of a large sample of voters.

There is no other kind of surveying, let alone polls, that can touch that.


Unionist wrote:

Nicky's memory is correct, if you read his/her post as meaning second choice for all voters counted together (which is how I read it). [url= article[/url] is also ambiguously written, but clearly shows the Liberals (at one point in the last campaign anyway) as being the second choice of 28% of all voters, and the NDP and Greens tied at 25% - which mathematically would put the Cons behind all three.

After reading the article, it looks like testimony to the enduring power of this myth.

At a minimum: nowhere in the article does it say that Liberal voters would most frequently choose the NDP as their second choice. The article itself only says who gets the most second choices period. [And by the way, of course the Conservatives are going to come in behind on that. Since they are the first choice choice of 37% in the same poll, they are ruled out as second choice for that big chunk. While the Liberals are only ruled at for 24%, and the Greens only ruled out for 12%... which has a lot to do with their high second choice number.]

So this article comes by peoples eyeballs briefly. It does not say what people attribute to it. It might at best give people a tenuous shred of something to hang onto. [Emphais on might: that is if the poll numbers were delved into, one might at best find some very tenuously supported argument.]

By comparison, the Canadian Election Studies get discussed here numerous times. As do numerous analyses of changes in raw votes in close race riding campaigns.

And guess which item gets remembered and people store as reflecting reality?


KenS wrote:

At a minimum: nowhere in the article does it say that Liberal voters would most frequently choose the NDP as their second choice. The article itself only says who gets the most second choices period.

Isn't that what I just said (and you quoted me as saying):

Unionist wrote:
Nicky's memory is correct, if you read his/her post as meaning second choice for all voters counted together (which is how I read it).

Now, what are these studies you're referring to, and what exactly do they say? Got a link? A summary?


Scott Piatkowski Scott Piatkowski's picture

nicky wrote:
Certainly it was the case in the CCF years. Witness the Lib/Con provincial coalition in British Columbia and how it took seats from the CCF.

And how is that unlike the current situation in BC?


No link or summary to Candian Election Studies. When its not within the range of my short term memory- now where did I put hammer / pen / coffee cup- I'm never much good for that.

In this case, Ottawa Observor or someone else will read this and respond eventually.


From the home page of Canadian Election Study

"The 2008 Canadian Election Study consists of a survey with nearly 4500 eligible voters conducted during the second half of the election campaign. 3689 of these respondents completed a post-election survey as well. 1238 respondents who had participated in the 2004-2006 panel study were also interviewed after the election. All of the interviews were conducted by telephone. The final component of the study was a self-administered mail-back survey completed by..."


I can't recommend a summary. And there are a number of follow-up articles.


Data from the Canadian Election Study (CES) are used by political scientists and cited in academic articles.  If you have SPSS (or use an open-source replacement for it, like PSPP), you can download their data files from their website.  There's a long history to the validity of some of the questions, from what I understand.  So, while you can poke around the data, I believe it takes some expertise to understand properly.  Long afternoons in the university library ...

Pogo Pogo's picture

Looking a little deeper at the second choice data reveals that Conservatives are likely still number one.  While the article indicated polling numbers for all three parties, it did not indicate second choice levels for Conservatives and Bloc and the data appears to have been removed long ago from the pollsters site.   Two assumptions were therefore necessary. The total of the five parties support came to 98% so I took that number as the sum for the second choices.  The Bloc Support was 9% but I assume that it was far less likely that they would be as high in second choice support and allocated them 3% only.  Using these two assumptions I calculated the Conservative secondary support to be 19%.  This assumes that nearly all respondents made a second choice which may or may not be the case. 

Based on this assumption Conservatives received 30% of the available secondary support (100%-37%), Liberals 25%, NDP 22%, Green 22%. 


There's a second question to consider, which is how likely people are to exercise that second choice rather than stay home. Another research question to consider: are the stated second choices of a party's firm supporters different than the second choices of their soft supporters?

I don't know the answers to those questions, but find the questions very interesting. Does anyone know of any studies that have looked at them in Canada recently?


Bomb the bridge


And these days Premier Wall has particularly good reason to be worried by the experienced and wily NDP Leader across from him in the Saskatchewan Legislature. After all, Mr. Lingenfelter now has some beautiful material to work with. The Wall team, eastern-establishment darlings though they might still be, stand revealed this winter to be as fiscally incompetent as the last Saskatchewan Tory government was (I wrote about some of the details a few weeks ago).

Definitely time to try "bomb the bridge."

However, for this one-trick strategy to work, it needs to blow on a spark of truth. And it seems unlikely that citizens of Saskatchewan, who well remember the price they paid for fiscal recklessness and irresponsibility under conservative rule in the 1980s and 1990s, are going to ignore Premier Wall's inability to manage the finances of the province because - ludicrously - Mr. Wall wants voters to believe Mr. Lingelfelter isn't enthusiastic enough about the province.

It's the Wall government's enthusiasm for billion-dollar "mistakes" that are worrying the voters these days.

When the real issue is government incompetence, "bomb the bridge" doesn't seem to work.

Just ask former Ontario premier Ernie Eves, who p

Malcolm Malcolm's picture

Nicky, the ONLY available source which specifically discusses the likely behaviour of Liberal voters absent a Liberal candidate is the Canadian Election Study.


What the CES has CONSISTENTLY shown in election after election after election is that, absent a Liberal candidate, Liberal supporters are significantly more likely to vote CONSERVATIVE than New Democrat (if, in fact, they vote at all.)  HOW much more likely has varied from election to election.  The FACT of it has varied not a whit.


But even if we go along with your wishful thinking, that still accomplishes nothing.  In any riding won by the Conservatives, overcoming the Conservative margin means that the second party's NET gain must be in excess of the margin.  Thus, if the margin was 1,000 and the Liberal vote 2,000, the NDP would need to take 75% of the Liberal vote to have a net gain of 1,000 (1,500 to 500) - and that's assuming that every single Liberal would cast a second choice ballot.  If, say, 500 Liberals chose to stay home (or to spoil their ballots or to vote for a minor party candidate), then the NDP would need to carry at least 1,250 votes to 250 to have a net gain of 1,000 - an actual retention of 83.3% of Liberal voters who actually voted.


As I indicated earlier, there are three possible reasons you are advocating this "strategy:

* You don't understand politics.

* You don't understand arithmetic.

* You are Liberal who wants to see the NDP commit political suicide.


In Australia, you MUST rank every single candidate on your ballot. If you only picked your favourite and didn't prefernce your second, third and fourth choices etc... you ballot is considered spoiled.

The other thing to consider is that if we had preferential voting, chances are that parties would make "deals" in exchange for preference directions. For example, maybe the NDP and the Liberals make a deal whereby they will actively encourage supporters to preference each other ahead of the Tories and the election eve vote at cards will explicitly encourage people to fil out their ballots tha way? 

What if we start to see rallies and facebook campaigns from coast to coast where people chant "Tories last! Tories last!" and starts to be a mantra that people should fill in their ballots from the bottom up - give the Tory the lowest number and then worry about who you support afterwards.

My points s that if we had preferential voting, it would change our political culture.

Scott Piatkowski Scott Piatkowski's picture

Since the thread is almost at 100 posts, maybe it's time to bring it back on topic.

Topp's book, [url= We Almost Gave the Tories the Boot: The inside story behind the coalition[/url], is available beginning February 1.


And it already has its own thread and everything !


To return to the debate about second choice voter support:
I have argued in other posts that:

  1. Should the Greens fade away their support would go largely to the other opposition parties

  2. Should the opposition parties combine, either by running joint candidates or through a run-off voting system, the anti-Conservative vote would be be consolidated.


Others have argued that there it is just as likely that the Conservatives would benefit or that cooperation among the opposition parties would have little effect on the outcome.


I have found a notebook in which I jotted down a few poll findings during the last Federal election campaign. I believe they support my position.


  1. The final EKOS poll of the campaign: second choices: Con 8.3%; Lib 17, NDP 19.5, Green 4.6, Bloq 17.6 (presumably just in Quebec)

 2.  An EKOS poll for British Columbia . Sept 23. Second choices. Con 9, Lib 22, NDP 20; G 18

  3. A Harris Decima poll. Sept 30. Second choices: C 14; L 26; NDP 30; G 24, B 3. 

This was further broken down by parties:

Among Con voters the second choices were: L 38, N 29, G 25

NDP voters: C 20; L 44; G 30, B 5

Green voters: C 18, L 37, N 38, B 3

Bloq voters: C 5; L 21; N 43; G 28


I did not write down the Liberal second preference numbers but a reasonable extrapolation , given the other numbers, might seem to be: Cons mid-teens, NDP mid-20s; Greens about 20

  4.  An unattributed poll of second choices among Greens: L 24, C 14, N 26, B 6.

  5.  A Strategic Council poll Oct 7-9 on second preferences in Ontario;

 Overall: C 13, L 21; N 25; G 18

Con voters: L 27; N 24; G 13

Lib voters: C 23; N 39; G 21

NDP voters: C 15; L 40; G 31.

Green voters: C 19; L 27; N 35





Ottawa Observer I beleive did some calculations based on chewing through numbers. If so, could you link to that?


I notice people have not gone cranky in this thread.

Whats wrong, think you are better than people in the other threads!


KenS, that wasn't me, it was Malcolm.  I believe his data were also published in the Accidental Deliberations blog.

Nicky, second choice has to be read together with firmness of intention, as someone pointed out to me recently.  I might have a second choice, but stand no chance of switching to it.  Someone else might have no second choice, but is grouchy and doesn't want to vote for their own party, and thus stays home.  It's been put to me that second choice is not consistent across firmness of intention, and not correlated.

Thus, if you know what the second choice is of all (for eg) Liberals, but a quarter of them don't vote Liberal, you still don't necessarily know what that quarter of them did (whether they switched or stayed home).

There are two other ways to infer what you want: (i) look at what they did last time (although your baseline this time would be the people who *didn't* do it last time, so that's another problem), and (ii) look at what the switchers told the Canadian Election Study.


Topic locked