Hello all,

While this blog has been inactive for months (hey, it was summer!), I haven't stopped doing projections. I'm currently working on the new model which would be a lot more robust to extrapolation (since it seems both the next federal and Quebec provincial elections will take us to unknown territories). However, I don't currently have time to cover the Ontarian election. Sorry for that, but I really wouldn't be able to cover it the way I like to do it.

So see you soon!

Bryan

Ok so I'm back from Mexico (it was really nice, thanks for asking). I looked at the projections and here are some comments.

1) Polls were underestimating the Conservatives.

While the polls were mostly right for the other parties, they clearly underestimated the Conservatives. If you take the last couple of polls published during the week-end, they all put the CPC around 35-37%. Given that these polls had a sample size of 2000-3000 most of the time (Ekos, Angus, Leger), it means the actual Tory score is out of the margin of error. Numbers can be even worse at the province-level. For instance in Ontario, you never had the Tories at 45% like they did on Election Day. Pollsters will (and already have in some cases like Ekos) have to revise how they survey people I think.

2) Atlantic

Using the actual percentages for this regions (which are, respectively for the CPC, LPC and NDP 37.9%, 29.3% and 29.5%), you would get 9 mistakes. Most are coming from the tricky NF-L. However, it seems that my assumption that some of the ABC effect was gone was right. Indeed, the Tories increased more in NF-L that in the rest of the Atlantic: they increased from 16.6% to 28.4%. But the effect wasn't 50% gone, more like 25% then. With that being said, weird things happened anyway. The only riding won by Harper was Labrador and unless I wasn't aware of a new star-candidate, it was definitely not the most likely to switch. In the rest of the Atlantic, the model performed quite well.

3) Quebec

The big orange wave. It's funny, with my final projections, Iw as the harshest on the Bloc (as compared to 308 or DS) and I was a little bit worried to underestimate this party... Boy I actually overestimated them. Even by using the actual percentages, I would still have the Bloc with 10 seats. I would have the Liberals at 12, but winning 6 out of 7 close races. Therefore a more adjusted projections would be 9-10. The same for the Tories with 9 seats but 5 out of 6 clsoe races, leading to adjusted results of 7. By the way, if you actually compare the close races projected to be won by the Liberals over the NDP to the actual results, you see that the model was actually quite close most of the time, instead that the NDP won these ridings slightly. Having the correct "call" is important, but being accurate is as well.

The model would have worked quite well in Quebec everywhere except in Gaspésie where the NDP won almost everything while the model was still giving most of the ridings to the Bloc. It's because historically, the Bloc was more resilient there and the NDP's swing was clearly decreased there. With that being said, th Bloc ending up with only 4 seats despite finishing second in the popular vote with as much as 23% is puzzling. I would actually be worried to have my model correctly predicting that! lol I mean, it seems the Bloc was caught in a wave AND was freaking unlucky on top of that.

4) Ontario

Ok here I need to mention something first. The following comparisons are made using a simplified model that I had. As for Quebec, I had other models ready for Ontario in case all parties would enter extrapolation-territories. Since the final polls were showing the NDP around 25%, Liberals around 27% and, most importantly, the Conservatives slightly down compared to 2008, I decided to use the full model, and not the simplified, more extrapolation-friendly one. BUT, if the polls had shown the CPP at 45%, I would have switched to the other model. People reading this blog regularly know that I have many specifications. After all, I did switch for Quebec after the orange wave started.

So anyway, no matter which model I would have used, my projections would have have been mostly ok in Ontario, except in the GTA. There is no way to reconcile historical trends and what we saw last Monday. In particular, while the provincial swings explain how the Tories won ridings in the suburbs of Toronto, they shouldn't have won seats in Toronto. For instance, the seat of Ignatieff should have remained a Liberal one. Historically, the Tories swing was decreased in Toronto. Look at the following table:

		04 to 06				06 to 08
	CPC	LPC	NDP		CPC	LPC	NDP
east905	3.3%	-4.0%	1.0%		6.8%	-10.1%	-0.7%
west905	4.1%	-3.9%	0.2%		3.8%	-5.5%	-0.8%
Tor. east	4.0%	-3.1%	0.5%		2.5%	-4.7%	-1.3%
Tor. West	3.3%	-3.8%	1.0%		3.9%	-5.4%	-1.0%
prov. swing	3.6%	-4.8%	1.3%		4.1%	-6.1%	-1.2%

As you can see, traditionally, the Tories were increasing a little bit less in this region while the Liberals were holding up better (except in the east 905 where the Tories already had all but one seats since 2008). So ridings liks the one from Ignatieff where the difference between the Tories and the Liberals was more than 11-points were not likely to switch when the provincial swing for the Tories was +5 and -8.5 for the Liberals. You have a lot of ridings in this situations where the provincial swings were actually amplified for both parties! What it means is that there was a major change occuring in the GTA during this election. Maybe it actually all makes sense. I mean, yes traditionally the LPC was holding up better, but now that the Liberals were already low everywhere but in the GTA, it kind of makes sense that any further drop province-wide would have to come from the GTA! I'll think about it while building the next model. In any way, I was very clear that my model was probably not appropriate for extrapolation purposes. For this, simplistic model assuming an uniform swing (or a proportional one) as better suited.

5) What to do next?

My objectives has always been to turn such a model into a published paper. Now that we have more data (and in particular, I now have variation for the NDP, a party that used to move very little and was tricky to estimate), I will be able to test various specifications more in details. I couldn't do that before because, in order to estimate the coefficients, I need more elections. Indeed, with only 3 (2004 to 2008) and thus 2 swings, I hadn't enough data points. Now, I can use data from 2004 to 2008 to see how I could have explained 2011 better. One thing I'll have to try to make the model more extrapolation-friendly. With a majority government, I have more time to do that! I'll keep you posted with the updates.

Doing seat projections is always tricky. You can go simple and be right sometimes, but sometimes you are wrong. In 2008, for the Quebec election, my model was the only one to predict 7 ADQ MLAs instead of the 2-3 predicted everywhere else. Why? Because I had regional coefficients. But when the regional patterns change, like this time in Quebec (Gaspésie) or in the GTA, a model based in historical patterns would necessarily be wrong. Seat projection is not an exact science. However, I still think there is a place for a scientific, complicated and transparent model like mine.

I'll also be back for the next Quebec election, around 2012 I guess. I don't know yet if I'll have time to build a model for Ontario or BC.

Ok so Harper won his majority. Why were the predictions wrong?

Well, if you excude the Atlantic where I would have to look more in-depth into it, it's mostly the fact that the Tory vote was underestimated by most polls. This is especially true in Ontario. If you enter the actual percentages in the model, you get results much closer to the actual ones. I understand this is a lame excuse but it at least shows that the conversion % to seats wasn't that wrong. And it also shows that it's important to provide a simulator like I do. It allows people to use the percentages they believe are right (i.e: correcting the polls).

As for the predictions, I was the hardest model on the Bloc, so yea I guess... For Quebec, surprisingly, except in Gaspésie and a little bit in Montreal (Mostly Mtl-East), the model wasn't that bad for the NDP.

The collapse of the Liberals is way worse than predited in the polls. This is crazy.

I'll look at the details results next week, when I'm back. In the mean time, congratulations to DemocraticSpace who was the only model to predict a majority even though it was a smaller one. On the other hand, DS, was more off than me for the Bloc and NDP.

Thanks to Louis-Benoit L'Italien-Brunot, here is a map of the final projections. It's very nice.

By the way, I'm slow for answering comments because I'm currently in... Mexico! I left yesterday. Yes yes I know, so unprofessional for a seat projecter lol But I couldn't leave later because I start teaching next Monday and my girlfriend couldn't leave earlier... So we voted in advanced polling and we booked a resort with wifi. Hey, if NDP candidates can do that (and even get a chance to be elected), I sure have the right to do the same! lol

So I'll reply today. I have to say, no matter the results, it was a pleasure running this site and having all your comments and suggestions. I'll do this for sure during the next elections.

Here we are, the final projections! So let's answer the big question first: Stephen Harper and the Conservatives are projected to win this election, but only with another minority. So the big wish of Harper to finally get his majority is not projected to happen. It could still happen (look at the total safe+100% all races for the Tories), but it will all come down to Ontario. It's way more likely to be another minority. Thus the official call of this site.

The second big news is the NDP finishing second, thanks to a big surge in Quebec as well as, to a lesser extent, in the rest of Canada. Jack Layton is projected to become the leader of the opposition surely. Will he be the PM within the next 3 months? Maybe, it all depends on two things. First, how will Harper react to another minority? Will he try to reach out to the opposition or not? Also, will the Liberals support the NDP or the Tories? I would say that a big factor in this equation is whether NDP+LPC have the majority. But all this is beyond the basic objective of this website and will not be resolved until weeks.

For the Liberals and the Bloc, it's like a race for the biggest collapse. For the Grits and Michael Ignatieff, this election could well be remembered as the equivalent of 93 for the Tories (although, I'm not saying the Liberals will only elect two MPs)! They have no chance to win this election and are even out of reach of the second place.

As for the Bloc, for the first time since this party exists, it will almost surely finish second in Quebec! And not only in the number of votes as in 2000 (when the Liberals finished first), but also in the number of seats. This is historic. In the worst case scenario, Gilles Duceppe could even lose his seat and/or be the leader of only 10 MPs. Please notive that the projection of 14 seats is probably too harsh. Indeed, the Bloc is projected to win only 4 out of 20 close races. It's too low, the Bloc should probably be projected around 19 seats. And Liberals and CPC should be a little bit lower in la Belle Province.

So let's go region by region.

1. Atlantic Canada.

It was a source of a lot of potential gains for the Tories at the beginning of the campaign. However, recent regional polls, as well as some polls conducted in specific ridings in NF-L, tend to show that the Tories won't make many gains there. Remember that my model take into account the fact that the regional swing for the each party has two components: the ABC effect and an actual regional swing. That could be very important to predict the numbers of seats won by each party outside of NF-L. Indeed, imagine everything is the same except that the ABC is gone. That would result in a regional lead of 6-points for the Tories! But if you don't adjust for the fact that this increase is entirely in NF-L, you'll overestimate the Tories everywhere else, in particular in IPE.

In this region, the party who is actually likely to make gains is the NDP which was projected to finish first in many last minute polls. Finishing first in the Atlantic won't likely give Jack Layton the majority of the seats in this region (mostly because of the large majorities a lot of Liberals in incumbent had in 2008), but would boost the numbers of MPs from 4 to potentially 8.

2. Quebec.

Ok let's be clear here: this province will likely be completely crazy. I worked very hard on my model for this province, trying many specifications, in order to be as correct as possible in the context of a possible historic orange wave. At the end, my projections are favourable to the NDP but less than a pure uniform swing model. If you compare my projections to the ones from Ekos for instance (which uses, I'm sure, a uniform swing), you'll see the Bloc and Liberals higher and the NDP a little lower. While I completely believe the NDP will finish first in this province, I think it will fall short of a total sweep where the Bloc would fall to 10 seats or less. However, no matter the specification of the model used, the number of seats won by the NDP vary a lot depending on whether this party gets 37% or 40% of the votes (and alternatively, whether the Bloc gets 25 or 28%). So it could really go either way, but I choose the medium one. You can see the high volatility in the 95% CI.

3. Ontario.

This is the key of a possible Tory majority. Actually, it is really the only place where it can happen. As I said, the Atlantic is unlikely to provide Harper with enough seats, while the CPC will probably be happy to lose only 3 seats in Quebec! In Ontario, the potential gains are mostly in the GTA, in particular in the suburb of Toronto (in the west 905). However, the Conservatives have been polled around 41% (+2 compared to 2008) consistently during the entire campaign. This provincial swing would probably not be enough to secure enough gains. Yes the CPC might shift its support to the GTA, but that would then mean they would be down somewhere else in the province. Given the surge of the NDP, that could well result in some losses. So the overall net gains will, in all likelihood, not be enough to reach a majority (the Tories basically need to win 65-70 seats in this province). One way it could happen is if the NDP "only" reaches 23% and the Liberals fall to 28%. That scenario would lead to a lot of vote splitting and give some seats to the Tories.

4. Prairies/Alberta.

I haven't covered these regions very much during this campaign. But I had a good reason: we haven't seen much change in the polls. The Conservatives will surely win a large majority of the seats over there. Will they make any gains? Maybe some riding, but the overall results should be, at best, very similar to 2008. The NDP could even make gains there. This party will also likely keep its only Alberta riding.

5. British Columbia.

The Tories already won 22 of the 36 ridings in 2008, so they don't have that many potential gains. With the NDP polled close second in some recent polls, the only hope for the Tories is to steal some seats from the collapsing Liberals. For instance Vancouver South where the Liberals incumbent won by only 21 votes last time. However, the Conservatives candidate got into a lot of troubles during the campaign. On a personal note ( I live in Vancouver Kingsway but I drive a lot in Van South), I would say I saw more Liberals sign this time around than in 08. So I think the Liberals will lose some seats, but not all of them. In any ways, Harper just can't really count on BC to get a majority, not with the NDP so high.

Also, Elizabeth May is projected to lose her riding but it will be a close race. I actually wouldn' be surprised if she won, mostly thanks to a heavy campaign from the Green and a lot of Liberals voters switching. Her "boost" might be even bigger than estimated using the 2006 by-election and the Nova-Scotia run of 2008 (I "boost" her by 20-points in the model, more than any other candidate).

Here is the pdf. Enjoy the election!

Here it is, more details to come later today, with the latest Ekos poll. ou'll have a riding-by-riding pdf with the min/max of each party in each riding.

But in the mean time, you have the official call of this site: Tories minority, NDP official opposition.

Here is the pdf.

Some people argued that the NDP will suffer the same fate as the Lib-Demm in the last UK elections. In particular, because NDP supporters are less commited and more likely to switch their vote, they will end up performing under what the polls predicted. It was true most of the campaign.

However, look at this graph from the latest Ekos poll. The NDP voters are actually more likely to vote now, as compared with the Liberals. We find similar evidence in the latest Angus-Reid poll. So, another effect of the orange wave... And another reason to worry for the Liberals.

By the way, I'm busy making the final projections that should be up Sunday morning, and for sure Sunday evening.

Today, I wrote about how strategic voting is useless in my opinion. It can matter in maybe 2-3 ridings, but that's it. For the other ones, you would need like 40% of the voters to do the same and this is very unrealistic. This is my opinion though, I know some of the commenters here strongly believe in the impact of strategic voting.

Some readers yesterday rightfully pointed out the weird projections in Guelph where the model has the NDP closely winning against the Tories. While I acknowledge the fact it would require a major swing for the NDP to win this riding, I have to mention one technicality about the model.

In Ontario, my model (with all the variables) is able to explain around 70% of the riding-level swing for the CPC and LPC. This is not as good as Quebec but still pretty good (remember, the only actual source of information is the provincial swing). It means I would have made around 8-11 mistakes during the last two elections. However, the NDP is tricky because this party experienced almost no provincial swing. Indeed, the party of Jack Layton moved from 18.1% in 04, to 19.4% in 06 and back to 18.2% in 08. It's very small and not a good source of variation (which is required for the model). I can only explain around 30% of the variation. I'm not trying to find excuses, I'm just explaining clearly where the model works best and where it works less well.

When this is the case, I usually switch to a uniform swing model for the NDP (it is also the case in the Prairies). A linear swing model is les sophisticated but has been proven as providing an overall good approximation of the actual results (this is for instance what Ekos is using). It's more "extrapolation-friendly". Nevertheless, if you do that, the NDP would gain only one seat in Ontario, despite a provincial swing of 5-points (currently; wait two more days and if the trend goes on, it's gonna be 10-points lol).

So I'm left with two choices: I can go safe but have the NDP gaining almost nothing, or I can go with the model and have more gains. Since I believe the NDP will indeed gain more than one riding, I'm taking the chance to miss-identify those ridings. So yes, maybe it's not gonna be Guelph, maybe the 5-points swing will be concentrated in another region or riding, but the overall results for Ontario should be the NDP at 21-23 seats, not 18. It makes more sense to me.

Again, for the final projections I'll carefully look at various models and decide which one is best suited. But for now on, I'll stick to the original model and give Guelph to the NDP. You have to admit that it's not the most unlikely riding to switch! lol

Note to my English readers: since I firmly believe in a bilingual Canada and since this post is about Quebec, I decided to write it in french. You can use Google translate if you really have to, or better, try to read it anyway. Posts in french will remain the exception.

Nous avons droit à huit sondages dans des comtés de la région de Québec et du Saguenay-Lac-Saint-Jean. Nous pouvons ainsi les comparer aux projections. Bien sûr, ces sondages par comté sont toujours moins fiables. Je ne ferai pas une comparaison comté par comté, mais par région en général.

Dans la région de la ville de Québec, les projections semblent marcher assez bien. Le vote Bloc semble un peu sous-estimé. Il semblerait que ce parti mange des votes aux Conservateurs. Mais en règle général, le modèle a le NPD projeté correctement. C'est une bonne nouvelle.

Dans l'autre région, le modèle ne semble simplement pas fonctionner. Ne nous voilons pas la face, les deux sondages dans Jonquière-Alma et Roberval-Lac-Saint-Jean montrent une bonne progression du NPD, alors que mon modèle prévoit une très faible hausse dans ce coin de la province. Je dois m'avouer très surpris, surtout pour le comté de Blackburn. En 2008, le NPD n'avait récolté que 4.88% des votes! Bien en-dessous de la moyenne provinciale. Alors de voir le NPD sondé à 32% est assez étrange vu qu'historiquement, le "swing" provincial du NPD était réduit dans Jonquière-Alma. Mais bon, voilà ce qui arrive lors de vagues et de changements historiques! Cependant, je me souviens de sondages plaçant Blackburn perdant en 2008 et au final, il avait remporté la victoire facilement.

Finalement, dans Chicoutimi-Le-Fjord, le NPD n'est pas dans la course (telle que projeté par le modèle) mais les Conservateurs semblent bien plus bas que prévus. Si cela est vrai (ainsi que dans la région de Québec), il y a une chance que le PCC perde beaucoup de sièges le 2 mai. Un autre obstacle pour l'obtention de la majorité. Par ailleurs, le fait que le Bloc soit en meilleure posture que prévue à Québec avait déjà été remarqué ici.

Est-ce que je devrais manuellement modifier le modèle, en particulier dans le Nord-Est du Québec? Peut-être, mais je ne pense pas le faire. Je suis ok si mon modèle se trompe dans quelques comtés après tout, même quelques régions au Québec!

Harris-Decima finally released a new poll. They were pretty much the only pollster which didn't publish a poll since the NDP's surge. So I updated the projections with this new HD, one new Nanos and one new Ekos (by the way, we can see that voting intentions are becoming stable in Ekos, which is a really good news for the projections!).

As you can see, the NDP is now projected solidly in second place. This is the first time it happens when using an average of most recent polls. As for the Bloc, be carefull as the current projections have this party winning only 6 out of 17 close races. But the trend is definitely there for this party and it's starting to look ugly.

If we actually get these results, the NDP couldn't simply rely on the support of the Liberals in order to govern. On the other hand, Harper would need "only" the Bloc. Speaking of the Conservatives, the majority is still technically possible but it's becoming more and more difficult. Rifht now, they would basically need to win 95% of the close races they are involved in. I personally don't think they will secure a majority anymore. If anything, if you look at the trend in Ekos, AR or Nanos, you see the NDP rising way too high for that. Especially since the Liberals will not finish third with only 37 seats like the NDP last time. So I think the new parliament will have an opposition at least as important as last time. What does it mean for Harper? Well I don't know. The current projections would probably mean he remains PM and tries to govern. But if the CPC's numbers start to fall a little bit more and Harper actually loses seats (along with the NDP+LPC having the majority), I think we would be in for a lot of instability. These are all speculations of course. By the way, I believe that people wouldn't mind so much having the NDP forming the government in the event that the Tories would fall quickly. I think the difference with the 2008 coalition is that, at that time, the Conservatives did win the election and the Liberals were the clear losers. But this time around, the moral winner would probably be the NDP. This poll from Ekos kinda confirms my thoughts.

Riding-by-riding projections are here.

Let me know what you think: if Harper gets a (even reduced) minority and NDP+LPC have the majority, what kind of government do we have within the first three months?

A friend of mine just let me know about these two polls conducted in the ridings of Lévis-Bellechasse and Lotbinière-Chute-de-la-Chaudière. They both show the NDP rising but still 2nd or 3rd. However, the polls were conducted between April 12th and 17th, so before the main surge.

So in order to see if my model was correct, we need to compare the projections made with polls released during the same period. I found back the pdf for the projection of April 20th. The provincial percentages were:

CPC: 19.9%

LPC: 19%

NDP: 22.1%

Bloc: 33.8%

Using these, here are the projections for the two ridings.

Lévis-Bellechasse
	Poll	Proj
CPC	37.7%	45.5%
LPC	5.9%	8.5%
NDP	22.9%	22.3%
Green	6.4%	2.9%
Bloc	25.5%	20.8%
Lotbinière-Chutes-de-la-Chaudière
	Poll	Proj
CPC	35.4%	46.9%
LPC	11.5%	5.9%
NDP	23.3%	24.7%
Green	1.1%	2.8%
Bloc	28.7%	19.8%

The Bloc was a little bit underestimated for the second riding, but overall I'm pretty satisfied (especially since the margins of errors are big for these two polls) The NDP was spot on in particular!

Bloc supporters shouldn't be too happy though, as I've said, these polls and projections were valid for before the big surge. If you look at the current projections, you see the Bloc in third and way behind.

Today, we've got a new poll from Forum Research. With a sample size of more than 3000 (I'm starting to really like all these polls with large samples), they show... that the Tories only have a 3-points lead over the NDP nationally! Completely crazy but kind of consistent with the last polls from Ekos and Angus-Reid. So before talking projections, I would like to explain why I think the surge is real and why I don't believe that the NDP will suffer the same fate as the Lib-Dem did in the 2010 UK election. For those who don't know, the Lib-Dem got a big boost after the (first ever in this country) tv debate and was polled to be around 30% all campaign long. But then, on election day, they only increased their share of votes from 23 to 24% and actually lost 5 seats! So here are the reasons:

1. I believe in polls.

Obviously I wouldn't spend my time studying every single poll published if I thought polls were wrong! While I acknowledge the limits and problems of polls, I still trust them. In particular, when I see that every single poll shows the same thing, I will not start tweaking the vote intentions because the NDP voters might not show up to actually cast their ballots. So, I trust polls and we have enough of them to get a pretty accurate idea of the voting intentions.

2. The polls were NOT that wrong in the UK.

Check this pdf (I'll refer to it again later in this post) on page 6, you see that poll did indeed overestimate the Lib-Dem, but not by 10-points. But the Lib-Dem was not, in average, projected to gain 10-points, or 25 like this is currently the case in Quebec for the NDP. So yes, it is possible the NDP is currently being overestimated, but don't believe the NDP will end up at 19% on May 2nd or at 15% in Quebec.

3. The surge has come (so far) from Quebec.

Recent polls start showing a NDP surge outside of Quebec (in the Atlantic and in Ontario in particular) but it's nothing compared to the surge in Quebec. And because we know that Quebec can work by waves, I trust this surge more. Specifically, we've seen that more than once in this province. In 2006, Harper started the campaign at 5% and ended up at 25% with 11 MPs. The surge happened during the Christmas break. In 2007, the ADQ started with 4 MLAs and 15% in the polls. They were supposed to be doomed after this election. At the end? The ADQ rose all campaign long, up to 25% and actually got as much as 30% on Election Day! They got 39 MLAs and fell 5 seats short to be the government! So it is well possible to see the NDP climbing in the mid 30's.

4. The vote commitment of NDP supporters is softer but also has a lot of potential with the second choices.

Check the pdf and you'll see that the Lib-Dem voters were less commited with as much as 34% saying they might change their mind. In Canada, currently, the Tories have the highest level of commitment according to Angus-Reid. If the commitment is above 80% for the CPC' it's only around 70% for both the Liberals and NDP. But at the same time, we get that 80% of Bloc supporters are sure to vote for the Bloc. 3 weeks ago, you would also have got the same level of commitment from the Bloc. Yet, 3 weeks later, the Bloc has fallen from 35% to 25%... On top of that, second choices from Ekos show the NDP is the main second choice of every party's supporters, in particular the Bloc and Liberals (as much as 53%!). Therefore, among the 30% of un-sure Liberal voters, you could well have a lot of them switching to the NDP. On the other hand, the NDP's second choices are more diversified. Therefore, the NDP might lose votes to the Tories, Green and Liberals, while the Grits would switch mostly to the NDP. The net effect would be a pretty stable NDP.

5. The Lib-Dem didn't have the edge on any policy issues, the NDP does.

If you still look at the pdf document, you see that the Lib-Dem were high only because of their leader and his performance during the (first) debate. They weren't seen as having the best policies for any big issues. It is different here in Canada. First of all, I believe that the debat had a weirdest potential impact in the UK since it was the first time such a debate was broadcasted on tv. Secondly, Jack Layton is of course not seen as the best to handle the economy, but he has the edge for health care, as shown here. Also, if you look at the leadership index from Nanos for instance, you'll see that Layton has almost always outperformed Ignatieff. This wasn't the case for Nick Clegg and the Lib-Dem.

So all that to say that I don't think the NDP will suffer a big disapointment on May 2nd. Yes, they might and probably will be below some of the crazy scores we've seen in some polls, but they will not be at 19%.

Speaking of crazy numbers, if I was to use only the Forum poll, that would give me:

CPC: 130

LPC: 62

NDP: 107

Green: 1 (yes May would win)

Bloc: 8

Looks crazy doesn't it? Well, this is only one poll but as I've said before, the surge is still occuring (adding the Forum's numbers to the latest projections would give 70 seats to the Liberals and 69 to the NDP, +15 if we compare to the projections from april 26th! So one more day and the NDP will be projected to finish second). I could get very similar projections by using the last polls from Ekos, Angus-Reid and Forum, which would combine for a 10k sample size...

A short post where I thought I would let you know of this poll conducted in the riding of Saanish-Gulf Islands where Elizabeth May is running. She is projected to lead 45% to 38%, Howeber, this is a poll with only 398 respondents and conducted by a firm that I never heard of. If you apply the margins of errors, they are both statistically tied.

It's possible that May is indeed leading though, but that would mean that she would benefit from a huge boost, because the current polls don't put the Green that high in BC (or anywhere in Canada actually). I'm already giving May a boost of 20-points, taken mostly from the NDP. So I don't think I'll change that and give us an even bigger boost, even though she's is currently projected to lose this riding in my projections. If you missed it, here is where I explain how I estimated the Elizabeth May effect.

No matter the polls or projections, this will likely provide us with one last close race to follow during election night. Although, depending on the outcome, we might be more concerned with what type of government we'll get.

As a polls and political junkie, I'm of course very excited by the recent turn of events we've seen during this campaign. However, I have to admit that making seat projections for this campaign will be difficult. First of all, polls give us a snapshot of the situation, but with 2-5 days of delay. This delay can clearly be seen by comparing polls from Ekos to Nanos for instance. Because Nanos uses a rolling panel, it takes longer to see the surge occuring. For projections, by using recent polls, there is the risk of being behind or outdated already. It is especially the case in Quebec where each new poll seems to give 1 more point to the NDP and one less to the Bloc. Right now, by simply updating the list of polls I'm using to calculate the average, I'm basically in a situation where the NDP gains 2-3 seats every day in Quebec.

So let's try to only use the four most recent polls: last one from Nanos, last two from Ekos and the recently released Angus-Reid. Ekos and AR have in common to place the NDP way ahead in Quebec and strong second nationally. They all place the NDP first in Quebec though. These four polls provide an overall sample size of above 8000, which is quite good. Here are the results:
And if it wasn't for Nanos who has the NDP lower in Ontario, the NDP could even reach above 90 seats. Notice how this scenario would lead to another incredibly unstable parliament where LPC+NPD would fall just short of a majority (by one seat actually if you elect the Speaker from the other party).

Ekos and AR also show a NDP surge outside of Quebec. In Ontario for instance, it's very interesting as the NDP is currently taking votes from both the Tories and the Grits. So again, NDP's surge doesn't mean Conservatives majority, not at all. We'll have to wait and see if it's confirmed elsewhere. But it seems more and more likely that the Liberals could well finish third, especially since they starting dipping below 30% in Ontario. For the Tories, if they can keep their support in this province around 38-39%, the NDP might split some seats for them. However, this is true only if the NDP grabs 2-3 points from the Grits and reach 21-22%. As soon as they go above that, every point switched from the LPC to the NDP leads to stable seat projections. For instance, try to input AR's numbers only and you'll see that the Tories would lose a lot of seats in Ontario.
So again, I said it many times on this blog but I'm not a big believer of the vote splitting. Those who say that are forgetting how many ridings are a race between CPC and NPD.

But you can see how difficult making projections can be right now. Simply compare these projections to the ones published yesterday, using other polls as well. If you believe the surge is mostly over, then by simply adding new polls this week, we should be ok. But if it isn't, we'll likely have the wrong percentages and therefore the wrong seat projections. On top of that, the swings are so large and weird that they push the model to its limit. The fact is that we never saw the NDP around 28% nationally or above 30% in Quebec. So my model, based heavily on estimation, might be in trouble lol rest assure though that for my final projections, I'll first compare my model to other ones. I have many excel files with many types of models (uniform, proportional, etc). So I can always see if one model predicts something very different and look into the details. But right now, I think the biggest challenge isn't that much to convert percentages into seats, but rather to have an idea of what the percentages will be. Hopefully things will settle down soon and people will spend the week-end watching the Royal wedding and not changing their minds!

By the way, I'd encourage everyone to publish, (by comments) their own projections (using the simulator under the tab "make your own projections"), based on the percentages you think will happen on May 2nd. That would make an interesting thread.

These are probably my last projections until the final ones, which should be published Saturday or Sunday. The big story is of course the constant and continuous rise of the NDP that is now projected at 55 seats, including 13 in Quebec. So let's talk a little bit more about la Belle Province.

Some of you would probably think I'm too generous with Jack Layton in Quebec or that I'm using the 2-3 polls placing the NDP above 30% in this province. Well, you would be wrong. These projections are based on 10 polls, all released and conducted within the past 2 weeks. I have two Nanos (I don't use more Nanos because I don't want to bias my average with too many polls from the same pollster), two Ekos (including the crazy one from yesterday, with a sample size of 3000), one Angus-Reid, one Forum, one Leger, one Ipsos and one from Environics. Also, for Quebec only, I have the Crop one. Among these polls, only one, yes one, has the NDP below 20% (Harris-Decima and its probably because it's a 2-weeks poll). I actually have 6 polls where the NDP is first in Quebec. So all that to say that I'm using quite a lot of recent polls. Therefore, it's only normal for the upward trend for this party to be reflected in terms of seats. Also, the NDP is currently on the edge of sweeping Quebec. Indeed, they are involved (or competitive) in 23 ridings in Quebec, of which they only win 6 (the Bloc is 14 out of 20). It's a fairly low convertion ratio, but I'm kinda ok with it as the NDP will still have to get its vote out on May 2nd. However, if the trend goes on this week, the NDP will soon be projected to be around 35% in Quebec and would win a lot more seats. For instance, if I was to use only the poll from Ekos published yesterday, I would get the NDP at 50 seats in Quebec and 99 nationally! And remember, it is a poll with 3000 sample size, therefore we have quite accurate estimates in each province. The orange wave will be THE thing to look for on May 2nd. Will it happen or will it be a burst like the Lib-Dem during the latest UK elections? I don't know. The fact the wave is happening in Quebec makes me believe it will indeed happen.

By the way, if the NDP does finish first in Quebec, with around 35% of the votes, this party has a real chance at finishing second nationally. If the Conservatives fall short of a majority (or even lose some seats), we could have a very unstable parliament. The Liberals would have to decide who they want to support. I'm guessing by that time, they woud also be looking for a new leader and Bob Rae looks to be the frontrunner (a former leader of the NDP in Ontario...). Could this mean that Harper would fall quickly and be replaced by a NDP government? And would it be a NDP-only or a NDP-led coalition? Not sure Jack Layton would feel very generous to a party who said no twice to a coalition (under Ignatieff), but a new leader could fix that. It's all political fiction of course, but the fact is: it's is possible! In particular, if NDP+LPC=majority without the Bloc, a NDP government could well become reality. You don't even need to run crazy simulations, just use a poll with 3000 interviews...

For the Conservatives, it's quite stable and quiet. They are mostly unaffected by the NDP's surge in Quebec. And since the NDP is rising only marginally elsewhere (especially in Ontario), the Tories are still projected around 145 seats. It's quite amazing how stable they are actually. For instance in Ontario, they have been projected at 41% for the last 3 weeks. Every time a poll shows up where the CPC is higher than that, there is another poll placing them below this number (at some point, I even double checked if my excel spreadsheet was woriking properly as I would not see any change for the CPC in Ontario!). A majority is still possible though, but they need to gain many seats in Ontario (in the GTA in particular). Not sure if they can really grab enough seats there though, especially since they don't enjoy a large provincial swing. If they were to shift their support to the GTA (and thus gain seats there), that would necessarily mean some losses elsewhere in Ontario. As I've said before, they also need the Atlantic, but recent polls (especially In NF-L) don't show this region to be a source of many seats for them.

For the Liberals and Michael Ignatieff, things are ugly. In Quebec they are down as some federalist are switching to the NDP. In Ontario, while they were very stable around 34% (like the Tories), they have started slipping to 32% recently. And they must pray for the NDP's surge to be contained in Quebec. While Ignatieff did pretty good during the first two weeks, the impact of its policies announcements are fading away. On the other hand, his less than stellar performance at the English debate still has a negative impact on its numbers. Just look at the traking polls from Nanos and you see that the LPC is clearly down. While the initial plan was to gain some seats and prevent a Tory majority, the current plan must be to "sauver les meubles" and not finish 3rd. Look at the list of polls. Every non-Nanos one published since April 20 show the same thing: the LPC behind the NDP. As for Nanos, the NDP is increasing and getting really close.

Nothing to say about the Green. The only riding to follow on May 2nd is the one of Elizabeth May. They are down in the polls as compared to 2008.

Finally, the Bloc is maybe close to an epic collapse. The NDP could well succeed where the Liberals and the Conservatives failed: finish first in Quebec, both in terms of votes and seats. We are still a little bit short of that, but the trend is not good for the party of Duceppe. And I'm not sure bringing back Jacques Parizeau is the smartest move ever. My guess is that the Bloc is now in a panic mode where the objective is to gather all sovereignist voters (which is around 30-35% currently). The problems? First of all, some separatists don't get out to vote for a federal election. Secondly, some of them just think it isn't the time and place to decide Quebec's future fate in the federation (and they would be right), so they feel free to support federalist parties, such as the NDP. But the Bloc knows that if they fall below 30%, they will enter the "losing zone" where each point lost translates into many seats gone.

The riding-by-riding projections are available here.